, | i |L i|| );^ ' < "\ \ 'iiiliiili .111 Wmm fiilii Hlmliiii!: 1 ' : :)il ■ i;»' J ; 1' . i niiiiiii' il!HMin|i«llt||nij|lHi|!| m ■I liisiiijilili liiilil ^BiiSI Hiiifflliiiiiiiilili!;- ■iiiiHIi^ w§m aiiiiiiiliijiiilliilii ;iii :,l- ill M iifci -^""- ^iillliiiiinifijiiisitlliiliiili, . . LIBRARY . . Connecticut Agricultural College. VOL lA.X.RX CLASS NO .4?.....v? >-' COST ^HN" DATE Zr^-i ... 3 . 19/4, Digitized by the Internet Archive in 2009 with funding from • Boston Library Consortium IVIember Libraries http://www.archive.org/details/geologyoflakesupOOvanh UNITED STATES GEOLOGICAL SURVEY GEORGE OTIS SMITH, Director THE GEOLOGY OF THE LAKE SUPERIOR REGION BY t CHARLES RICHARD VAN HISE AND CHARLES KENNETH LEITH 5SC£/MLHij DtPARTMENT RECEIVED JAN 2 1C6? Wilbui Cross Library Univeisity ci Connecticut WASHINGTON GOVERNMENT PRINTING OFFICE 1911 M 1 ^X. CONTENTS. Page. Chapter I. Introduction 29 Outline of monogra]5h 29 Acknowledgments •. 30 Geography 30 Topography 33 Relief '. 33 Drainage 33 Chapter II. History of Lake Superior mining 35 The Keweenaw copper district of Michigan 35 Copper mining on Isle Eoyal and elsewhere 37 Marquette iron district 38 Menominee iron district 39 Crystal Falls, Florence, and Iron River iron districts 39 Gogebic iron district 40 Vermilion iron district 40 Mesabi iron district 41 Accounts of the district before its opening 41 Opening and development 42 Cuyuna iron district 44 Baraboo iron district ., 45 Less important developments 45 Clinton iron ores of Dodge County, Wis 45 Paleozoic iron ores in western Wisconsin .■ 45 Iron ores of the north .''hore of I^ake Superior 45 Silver mining on the north shore of Lake Superior 46 Lake Superior gold mining 46 General remarks 47 Industrial changes 47 Smelting 47 Influence of physiography on industrial development 48 Production of iron ore 49 Chapter III. History of geologic work in the Lake Superior region 70 General statement 70 Work of individuals 70 Growth of geologic knowledge 72 Bibliography 73 Michigan 74 Northern Wisconsin 77 Minnesota 78 Ontario 81 Lake Superior region (general) 83 Chapter IV. Physical geography of the Lake Superior region, by Lawrence Martin 85 Topographic provinces 85 The Lake Superior highlands 85 Topographic development 85 The broad uplands - - . . 89 Position, relief, and sky line 89 Relation of original and present topography 89 Monadnocks 90 Valleys in the peneplain 90 ■ 5 6 CONTENTS. Chapter IV. Physical geography of the Lake SuiJerior region, }>y I^awrence Martin — Continued. Page. The Lake Superior highlands — Continued. The broad uplands — Continued . Soil and glacial topogra])h y 91 Description of dif trict.s in detail 91 Gabbro ])lateau 91 St. Louis plain 92 Vermilion district 94 Rainy Lake and Lake of the Woods district 92 Hunters Island and Thunder Bay region 94 Region north of Lake Superior 95 Region northeast of Lake Superior 95 Michipicoten district 95 Region north of Sault Ste. Marie 95 Marquette district 96 Menominee district 96 Crystal Falls district 96 Keweenaw Point 97 Northern ^Wisconsin 97 Central Wisconsin 98 Northeastern Wisconsin 98 Linear monadnocks and other ridges 98 General description 98 Keweenawan monoclinal ridges 99 General statement , 99 Northeastern Minnesota 99 Isle Royal and Michipicoten Island 99 Keweenaw Point and northern Michigan and Wisconsin 100 Keweenawan mesas 100 Huronian monoclinal ridges and valleys 102 Gunflint Lake district 102 Penokee Range 102 Giants Range 103 Marquette district 105 Menominee district 106 Crystal Falls district 107 North-central Wisconsin 107 Northwestern Wisconsin 108 The lowland plains 108 Area 108 Character and structiu-e 108 Denudation 109 The belted plain 109 The Minnesota lowlands 110 The basin of Lake Superior 110 General character and origin ' 110 Description of escarpments 112 Duluth escarpment 112 Keweenaw escarpment 115 Escarpment of northern Wisconsin (Superior escarpment) 115 Isle Royal escarpment 115 Age of escarpments 116 Bearing of escarpments on age of peneplain 116 Chapter V. The Vermilion iron district of Minnesota 118 Location, area, and general geologic succession 118 Topography 119 Archean system 119 Keewatin series : 119 Ely greenstone 119 Distribution 119 Appearance and structiu-e 119 Mineral constituents 120 Clastic rocks 121 CONTENTS. 7 Chapter V. — The Vermilion iron district of Minnesota — Continued. Page. Archean system — Continued. Keewatin series — Continued. Ely greenstone — Continued. Acidic flows 121 Intrusive rocks 121 Extension of Ely greenstone beyond district 122 Soudan format ion 122 Distribution 122 Deformation 123 Lithology 124 Origin 126 Relations of Ely greenstone and Soudan formation 126 Laurentian series 128 Porphyry 128 Granite of Basswood Lake 128 Algonkian system 129 Huronian series 129 Lower-middle Huronian .• 129 General statement 129 Ogishke conglomerate 129 Distribution 129 Deformation 129 Lithology 130 Greenstone conglomerate 130 Granite conglomerate 130 Porphyry conglomerate 131 Chert and jasper conglomerate 131 Common Ogishke rock 131 Metamorphism 131 Relations to adjacent formations 132 Thickness 132 Agawa formation 132 Knife Lake slate 132 General statement 132 Lithology 133 Microscopic character '... 134 Deformation 135 Relation to adjacent formations 135 Thickness 135 Intrusive rocks 135 Upper Huronian (Animikie group) and Keweenawan series 136 Theironoresof the Vermilion district, Minnesota, by the authors and W. J. Mead 137 Distribution, structure, and relations 137 Chemical composition 139 Mineral composition of the ores and cherts 140 Physical characteristics of Vermilion ores 140 Texture 140 Density 141 Porosity 141 Cubic contents 141 Secondary concentration of Vermilion ores 141 Precedent conditions 141 Mineralogical and chemical changes 142 Sequence of secondary alterations and development of textures 142 Volume change in Ely ore 142 Distribution of phosphorus , 143 Chapter VI. The pre-Animikie iron districts of Ontario 144 Lake of the Woods and Rainy Lake district 144 Introductory statement 144 Archean system 144 Keewatin series 144 Laurentian series 145 8 CONTENTS. Chaptek VI. The prc-Animikie iron districts of Ontario — Continued. Page. Lake of the Woods and Rainy Lake district.— Continued. Algonkian sy.-ftoin 146 Iluronian series 14(5 Steep Rock Lake district 147 General geology 147 Iron ores 149 Atikokan district 149 Kaministikwaa and Matawdn district : 149 Michipicoten district 150 Geography and topography 150 Succession 150 Archcan system 151 Keewatin series 151 Gros Cap greenstone 151 Distribution 151 Petrographic character 151 Wawa tuff 151 Distribution 151 Petrographic character 151 Structure and thickness 1.52 Helen formation 152 Distribution 152 Structiu'e and thickness 153 Petrographic character 1,53 Relations to other formations 153 Eleanor slate 154 Laurentian series 154 Algonkian system 154 Huronian series 154 * Lower-middle Iliuonian 154 Dore conglomerate 154 Distribution, topography, and structure 154 Petrographic character 154 Thickness 1.55 Relations to underlying rocks 155 Michipicoten extensions 155 The iron ores of the Michipicoten district, by the authors and W. J. Mead 156 General statement 156 Chemical composition 156 Mineral composition 156 Physical characteristics 157 Color and texture 157 Density 157 Porosity 157 Cubic feet per ton 157 Secondary concentration of the Michipicoten ores 157 Chapter VII. The Mesabi iron district of Minnesota 159 General description 159 Archean system or "Basement Complex" 160 Distribution 160 Kinds of rocks 160 Structure 161 Algonkian system. . .'. 161 Iluronian series 161 Lower-middle Huronian 161 Distribution ^ 161 Graywackes and slates 161 Conglomerates 162 Giants Range gi'anite 162 Relations of Giants Range granite to the lower-middle Htu-onian sediments and of both to other rocks 162 Structure and thickness 163 Conditions of deposition : 163 CONTENTS. 9 Chapter VII. The Mesabi inm district of Minnesota — Continued. Algonkian system — Continued . Huronian series — Continued. Upper Hm-onian ( Aniniikie group) Igg General character and extent 163 Pokegaraa quartzite 1(;4 Biwabik formation Ig4 Distribution Ig4 Kinds of rocks Ig5 Greenali'te rocks 165 Ferruginous, amphibolitic, sideritic, and calcareous cherts 168 Siliceous, ferruginous, and amphibolitic slates 170 Paint rock 171 Sideritic and calcareous rocks 171 Conglomerates and quartzites 171 Thickness ; 171 Alteration by the intrusion of Keweenawan granite and gabbro 171 Virginia slate 172 Distribution 172 Slate 173 Cordierite homstone resulting from the alteration of the Virginia slate by the Duluth gabbro. 173 Relations to the Biwabik formation 174 Structure 174 Thickness I74 Structm-e of the upper Huronian ( Animikie group) 175 Relations of the upi)er Huronian (Animikie group) to other series 176 Conditions of deposition of the upper Huronian (Animikie group) 176 Keweenawan series 177 Duluth gabbro I77 Diabase 177 Embarrass gianite 178 Cretaceous rocks 178 Distribution and character 178 Fossils 179 Pleistocene glacial deposits I79 The iron ores of the Mesabi district, by tlie authors and W. J. Mead 179 Distribution, structure and relations I79 Chemical composition of ferruginous cherts and ores 180 Analyses ISO Representation by means of triangular diagram 182 Mineralogical composition of ferruginous cherts and ores 183 Physical characteristics of the ores 183 Texture 183 Density 184 Porosity 184 Cubic contents 185 Magnetic phases of the iron-bearing formation 185 Occurrence 185 Chemical composition 185 Secondary concentration of Mesabi ores 186 Structural conditions 186 Original character of tlie iron-bearing formation 186 Alteration of sideritic or greenalitic chert to ferruginous chert (taconite ) 187 Chemical change 187 Mineral change 187 Volume change 187 Development of porosity 187 Alteration of ferruginous cherts (taconite) to ore 188 Volume changes 188 Method of expressing volume changes by triangular diagram 189 Data used in triangle 189 Consideration of the triangular diagram 190 Alterations of associated rocks contemporaneous with secondary alteration of the iron-bearing for- mation 191 10 CONTENTS. CuAiTKR \'II. The Mcsalii iron dit-trift of Minnesota — Continued. Page. The iron ores of I lie Mesabi district, liy the authors and W. J. Mead— Continued. Pliosplionis in Me.s;il)i ores 192 Dislrilnition in the iron-bearing formation '. 192 Secondary concentration of pliosphorus 194 Explanation of phosphorus in the paint rock •. . 19.5 Phosphorus in the amphibole-inagnetite phases of I lie iron-bearing formation 195 Minerals containing phosphorus 195 Detrital ores in the Cretaceous rocks 190 Sef[uence of ore concentration in the Mesain district 197 Chapter VI 11. The G\inflint Lake, Pigeon Point, and Animikie iron districts of Minnesota and Ontario 198 Gunllint Lake district 198 Geography 198 Succession of rocks 198 Algonkian system 198 Huronian series 198 Upper Huronian (Animikie group) 198 General description 198 Gunflint formation 199 Distribution ' 199 Structure 199 Potrographic character 200 Contact metamorphism 200 Thickness 200 Eove slate 200 Distribution 200 Structure 201 Petrographic character 201 Contact metamorphism 201 Thickness 201 Keweenawan series 201 Duluth gabbro. 201 Logan sills 202 Relations of the Keweenawan rocks to one another and to adjacent formations 202 Geologic relations 202 Topography as related to geology 20,3 The iron ores of the Gunflint Lake district 203 Chemical composition 204 Physical characteristics 204 Pigeon Point district 204 Animikie or Loon Lake district of Ontario 205 Location and general succession 205 Archean system 205 Algonkian system 205 Huronian series 205 Lower-middle Huronian 205 Kinds of rocks 205 Intrusives 206 Upper Huronian (Animikie group i 206 General description 206 Iron-bearing formation 206 Conglomerate 206 Lower iron-bearing member 207 Interbedded slate 207 Upper iron-bearing member 207 Upper black slate 207 Keweenawan series 207 General description 207 Logan sills 208 Structural features 208 General topographic features in their relations to geology 208 Westward extension of the Animikie district 209 CONTENTS. 11 Chapter VIII. The Gunflint Lake, Pigeon Point, and Animikie iron districts of Minnesota and Ontario — Con. Page. Animikie or Loon Lake district of Ontario — Continued. The iron ores of tlie Animikie district of Ontarii i 209 Occurrence 209 Character of the ore 210 Secondary concentration of tlie Animikie ores 210 Structural conditions 210 Original character of the iron-bearing formation 210 Nature of alterations 210 Sequence of ore concentration 210 Chapter IX. The Cuyuna iron district of Minnesota and its extensions to Carlton and Cloquet, and the Minne- sota River valley of southwestern Minnesota 211 Cuyuna iron district and extensions to Carlton and Cloquet 211 Geography and topography 211 Succession of rocks 211 Algonkian system 212 Huronian series 212 Upper Huronian (Animikie group) 212 ■' General statement 212 ' Distribution and structure 212 Lithology and metamorphism 213 Correlation 213 Keweenawan series (?) 215 Cretaceous rocks 215 Quaternary system 216 Pleistocene glacial deposits 216 The iron ores of the Cuyuna district, by the authors and Carl Zapffe 216 Distribution, structure, and relations 216 Character of the ores ? 219 General appearance 2*19 Chemical composition 220 Mineralogical composition 221 Texture 223 Secondary concentration of Cuyuna ores 223 Structural conditions 223 Original character of the Deerwood iron-bearing member 223 Mineralogical and chemical changes 223 Phosphorus in Cuyuna ores 224 Minnesota River valley of southwestern Minnesota 224 Chapter X. The Penokee-Gogebic iron district of Michigan and Wisconsin 225 Location, succession of rocks, and topography 225 Archean system 226 General statement 226 Keewatin series 226 Laurentian series 226 Relations of Keewatin and Laurentian series • 227 Algonkian system 227 Huronian series 227 Lower Huronian 227 Sunday quartzite 227 Lithology and distribution 227 Relations to adjacent formations 228 Bad River limestone 228 Distribution 228 Lithology 228 Metamorphism 228 Relations to adjacent formations 228 Upper Huronian (Animikie group) 229 General statement 229 Palms formation 229 Distribution 229 Lithology 229 Relations to adjacent formations 230 12 CONTENTS. Chapter X. The Penokee-Gogebic iron district of Michigan and Wisconein — Continued. Page. Algonkian system — Continued. Huronian scries — Continued. Upper Ilitfonian (Aniraikie group) — Continued. Ironwood formation : 230 Distribution. 230 Lithology 231 Relations to adjacent formations 232 Tyler slate 232 Distribution 232 Lithology 232 Metamorphism 232 Relations to adjacwil formations 233 Upper Huronian (Animikie group) of the eastern area 233 Keweenawan series 234 General description 234 Relations to adjacent series 234 Cambrian sand.^tone 235 The iron ores of the Penokee-Gogebic district, by the authors and \V. J. Mead 235 Distribution, structure, and relations 235 Chemical composition of the ferruginous cherts and ores 238 Mineralogical composition of the ferruginous cherts and ores 240 Physical characteristics 240 General appearance -40 Density 240 Porosity 241 Cubic contents 241 Texture 241 Magnetitic ores 24 1 Secondary concentration of Gogebic ores 242 Structural conditions 242 Original character of the iron-bearing formation 243 Alteration of cherty iron carbonate to ferruginous chert 243 Chemical change 243 Mineral change 243 Volume change 243 Development of porosity 243 Alteration of ferruginous chert to ore 244 Triangular diagram illustrating secondary concentration of Gogebic ores : 246 Alteration of rocks associated with ores during their secondary concentration 240 Occurrence of phosphorus in the iron-bearing formation 247 Phosphorus content 247 Minerals containing phosphorus 248 Behavior of phosphorus during secondary concentration 249 Sequence of ore concentration in the Gogebic district 2.50 Chapteh XI. The Marquette iron district of Michigan, including the Swanzy, Dead River, and Perch Lake areas 251 Marquette district 251 Introduction -• 251 Location, succession, and general structure 251 Archean system 253 Northern area 254 Keewatin series 254 Laurentian series 255 Southern area 255 Isolated areas of Archean rocks 256 Algonkian system 256 Huronian series 2.5G Lower Huronian 256 Mesnard quartzite 2.56 Name and distribution 256 Lithology 256 Metamorphism 257 CONTENTS. • 13 Chapter XI. The Marquette iron district of Michigan, etc.— Continued. Page. Marquette district— Continued. Algonliian system— Continued. Huronian series — Continued. Lower Huronian — Continued. Mesnard quartzite — Continued. Relations to adjacent formations 257 Thickness ^^^ Kona dolomite Name and distribution ~^^ Lithology ~^^ Metamorphism ""_ Relations to adjacent formations '-^^ Thickness -^* We we slate -^^ Distribution 258 Lithology 2^^ Metamorphism '''^^ Relations to adjacent formations 259 Thickness 259 Middle Huronian 2o9 Ajibik quartzite ■ Name and distribution "^^ Deformation " Lithology'. 260 Metamorphism """ Relations to adjacent formations - ■ ■ • 260 Thickness 261 Siamo slate "" Name and distribution 2G1 Deformation 261 Lithology 261 Metamorphism 261 Relations to adjacent formations 262 Thickness 262 Negaunee formation -"2 Name and distribution 262 Deformation -^'- Lithology, including metamorphism 263 Relations to adjacent formations 264 Thickness 264 Intrusive and eruptive rocks 264 Upper Huronian ( Animikie group) 265 Goodrich quartzite -"^ Distribution and structure 265 Lithology, including metamorphism 265 Relations to adjacent formations 265 Thickness 265 Bijiki schist 266 Name and distribution 266 Lithology, including metamorphism 266 Relations to adjacent rocks 266 Thickness 267 Michiganime slate " ' Name, distribution, and correlation 267 Deformation -"' ' Lithology 267 Metamorphism -"' . Relations to adjacent formations -"''' Thickness 268 Clarksburg formation -"^ Distribution 268 Lithology 268 14 • CONTENTS. Chapter XI. The Marquetto iron flistrii-l of Michigan, etc. — Continued. Page. Marquette district — Continued. Algonkian system^Continued. Huronian series — Continued. Upper Huronian (Animikie group) — Continued. C'lark.slmrg formation — Continued. Relations to adjacent formations 268 Thickness 268 Intru.'ive igneous rocks 268 Cambrian .• 301 Felch Mountain district 302 Location, structure, and general succession 302 Archean system 302 Laurentian series 302 Algonkian system 302 Huronian series ; 302 Lower Huronian 302 Sturgeon quartzite _. 302 Randville dolomite .' 302 Upper Huronian (Animikie group) 303 Felch schist 303 Vulcan formation 303 Keweenawan series (?) 304 Intrusive rocks 304 Paleozoic sandstone and limestone 304 Correlation 304 Laurentian series 304 Lower Huronian 30.5 16 CONTENTS. CiiAi'iKH XII. The Crystal Falls, Sturgeon, Felch Mountain, Caluiiiel, and Iron River districts, etc. — Conta. Page. Fclcli Mdunlain di.-Jtrict — Conlinur-d. Correlation — Continued. Upper Huronian (Animikie group) 305 Keweenavvaii series (?) .• 30.5 Calumet district 306 Location and general succession 30C Arehean system 306 Laurentian series 306 Algonkian system 306 Huronian series 306 Lower Huronian 306 Sturgeon ([uartzite 306 Randville dolomite 306 Upper Huronian (Animikie group) 307 Felch schL^^t 307 Vulcan formation 307 Michigamme slate 307 Paleozoic limestone and sandstone 307 Correlation 307 Iron River distrirt, by R. C. Allen 308 Location and extent 308 Topography and drainage 308 ( 'haracter of the glacial drift 309 General succession 309 Arehean (?) system 309 Keewatin series (?) 309 Algonkian system 310 Huronian series 310 Lower Huronian 310 Saunders formation 310 Distribution 310 Lithologic characters 310 Structure 311 Thickness 311 Relations to adjacent formations 311 Upper Huronian (Animikie group) 311 Michigamme slate 311 Distribution and general characters 311 General structure 312 Vulcan iron-bearing member 313 Distribution and exposures ■ ■ ■ 313 Relations to Michigamme slate 313 Thickness and structure 314 Lithologic characters 314 Distribution and local structure 315 Local magnetism in the Vulcan iron-bearing member 317 Intrusive and extrusive rocks in the upper Huronian (Animikie group) 318 Relations of upper Huronian (^nimikie group) to luiderlying rocks 318 Ordovician rocks 319 Florence (Commonwealth) iron district of Wisconsin 320 Location and general succession 320 Algonkian system 321 Hiu'onian series 321 Upper Huronian (Animikie group) 321 Michigamme slate 321 General character and distribution 321 Vulcan iron-bearing member , 321 Intrusive and extrusive greenstones and green schists 322 Quinnesec schist 322 Intrusive and extrusive greenstones and green schists other than Quinnesec 322 Granite and gneiss intrusives 323 Paleozoic sandstone 323 Quaternary deposits 323 CONTENTS. . 17 Chapter XII. The Crystal Falls, Sturgeon, Felch Mountain, Calumet, and Iron River districts, etc. — Contd. Page. The iron ores of the Crystal Falls, Iron River, and Florence districts, by the authors and \V. J. Mead 323 Distribution, structiu-e, and relations 323 Chemical composition : 324 Mineral composition • ■ 325 Physical characteristics .• 325 Secondary concentration 320 Structural conditions 326 Chemical and mineralogical changes 326 Time of concentration 326 The iron ores of the Felch Mountain and Calumet districts, by the authors and W.J. Mead 326 Felch Mountain district 327 Cahmiet district 327 Secondary concentration of the Felch Mountain and Calumet ores 328 Structural conditions 328 Chemical and mineralogical changes 328 Chapter XIII. The Menominee u'on district of Michigan 329 Location and extent 329 Topography 329 Succession of formations 329 Archean system 330 Laurentian series and unseparated Keewatin 330 Algonkian system 331 General character and limits 331 Huronian series 332 Lower Huronian 332 Succession and distribution 332 Sturgeon quartzite 332 Distribution 332 Lithology 332 Deformation 332 • Relations to adjacent formations 332 Thickness 333 Randville dolomite 333 Distribution 333 Lithology 333 Deformation 334 Relations to adjacent formations 334 Thickness • 334 Middle Huronian 334 Upper Huronian (Animikie group) 335 Vulcan formation , 335 Subdivision into members 335 Distribution '. 336 Traders iron-bearing member 337 Brier slate member 337 Curry iron-bearing member 337 Deformation 338 Relations between the members of the Vulcan formation and the ilichigamme slate 338 Thickness 339 Michigamme ("Hanbury ") slate 340 Distribution 340 Name 340 Lithology 340 Defor^iation 341 . Thickness 342 Relations of Upper Huronian to underlying rocks ._ 342 Relations between Vulcan formation and the lower Huronian 342 Relations between Michigamme ("Hanbury") slate and the middle or lower Huronian... 343 Igneous rocks in the Algonkian 344 Quinnesec schist 344 Green schists at Fourfoot Falls 345 47517°— VOL 52—11 2 18 . CONTENTS. Chapter XIII. The Menominee iron district of Michigan — Continued. Page. Paleozoic rocks 345 Cambrian system 346 Lake Superior sandstone 346 Lithology 346 Relations to adjacent formations 346 Cambro-Ordovician 346 Hermansville limestone 346 The iron ores of the Menominee district, by the authors and W. J. Mead 346 Distribution, structure, and relations 346 Chemical composition of the ores 350 Average iron content of the iron-bearing formation 351 Mineral composition of the ores 351 Physical characteristics of the ores 352 Iron ore at base of Cambrian sandstone 353 Secondary concentration of the Menominee ores 353 Structural conditions ■ 353 Mineralogical and chemical changes 354 Sequence of ore concentration in the Menominee district 354 Chapter XIV. North-central Wisconsin and outlying pre-Cambrian areas of central Wisconsin • 355 Northern Wisconsin in general 355 Wausau district 355 Location, area, and general geologic succession 355 Archean (?) system 356 Algonkian system 356 Huronian series 356 Middle Huronian (?) 356 Rocks intrusive in middle Huronian (?) and Archean (?) 357 Upper Huronian (?) 357 Cambrian system 357 Barron, Rusk, and Sawyer counties 357 Vicinity of Lakewood 358 Necedah, North Bluff, and Black River areas 358 Baraboo iron district 359 Location and general geologic succession 359 Archean system 360 Laurentian series 360 Algonkian system 361 Huronian series • 361 Middle Huronian (?) 361 Baraboo quartzite 361 Seeley slate 361 Freedom dolomite 361 Upper Huronian (?) 361 Paleozoic sediments 361 Quaternary deposits 362 The iron ores of the Baraboo district, by the authors and W. J. Mead 362 Occurrence 362 Chemical composition 362 Mineralogical character 363 Physical character 363 Secondary concentration 363 Structural conditions 363 Original character of the iron-bearing member 363 Mineralogical and chemical changes ^ 364 Waterloo quartzite area 364 Fox River valley 365 Chapter XV. The Keweenawan series 366 General characteristics 366 Distribution 366 Succession 366 CONTENTS. 19 Chapter XV. The Keweenawan series — Continued. Page. Black and Nipigou bays and Lake Nipigon 3G7 Lower Keweenawan 367 Middle Keweenawan 368 Black and Nipigon bays and adjacent islands 368 Lake Nipigon 368 Relations of the Keweenawan of Black and Nipigon bays to other rocks 369 Northern Minnesota 370 The Keweenawan area 370 Lower Keweenawan 370 Middle Keweenawan 371 Effusive rocks 371 Intrusive rocks 372 A. Basic rocks 372 ■^ Duluth laccolith 372 Area and character 372 Relations to other formations •. 372 The Beaver Bay and other laccoliths and sills 373 Anorthosites 374 ' Basic dikes 374 Acidic rocks ' 374 Keweenawan rocks in the Cuyuna district of north-central Minnesota 375 Thickness of the Keweenawan of Minnesota 375 Northern Wisconsin and extension into Minnesota 376 Distribution 376 Structiire 376 Lower Keweenawan 376 Middle Keweenawan 377 Upper Keweenawan 378 Relations of the Keweenawan to other series 378 Keweenawan granites of Florence County, northeastern Wisconsin 379 Northern Michigan 380 Distribution 380 Keweenaw Point 380 Succession and correlation 380 Lower and middle Keweenawan of Keweenaw Point 381 Order of extrusion 381 Presence of basic intrusive rocks 381 Acidic intrusive rocks 382 Nature and source of detrital material 382 Variations in thickness of sedimentary beds 382 Faults 383 Upper Keweenawan 383 Relations to Cambrian rocks 384 Main area west of Keweenaw Point, including Black River and the Porcupine Mountains 384 The South Range 385 Rocks of possible Keweenawan age in outlying areas 386 Thickness of the Keweenawan of Michigan 386 Eagle River section 386 Portage Lake section 387 Black River section 388 Relations of the Keweenawan of Michigan to underlying and overlying formations 388 Isle Royal 389 Michipicoten Island 390 East coast of Lake Superior 391 General consideration of the Keweenawan series 393 Lower Keweenawan ■■ 393 Middle Keweenawan 394 Igneous rocks 394 Varieties 394 Review of nomenclature of Keweenawan igneous rocks, by A. N. Winchell 395 The grain of Keweenawan igneous rocks — the practical use of observations 407 The extrusive masses 408 20 CONTENTS. Chapter XV. The Keweenawan series — Continued. Page. General consideration of the Keweenawan series — Continued. Middle Keweenawan — Continued. Igneous rocks — Continued. The intrusive masses 410 Source of lavas '"1 Sedimentary rocks 412 Source and nature of material 412 Extent of sediments 413 Upper Keweenawan 413 Relations to underlying series 414 Relations to overljdng series 415 Conditions of deposition 416 Thickness of the Keweenawan rocks 418 Areas of Keweenawan rocks 419 Volume of Keweenawan rocks 419 Length of Keweenawan time 420 Jointing and faulting 420 The Lake Superior synclinal basin 421 Metamorphism '■ 423 R6sum6 of Keweenawan history 424 Chapter XVL The Pleistocene, by Lawrence Martin 427 The glacial epoch 427 Plan of presentation 427 Ice advances 427 Driftless Area :■ = 429 Retreating ice - - -. ^"^ Contrasted general effects of glaciation - 430 Destructive work of the glaciers trj 430 Removal of weathered rock ^,..- 430 Striae and roches moutonn^es - - 431 Broadened and deepened valleys 431 Glacial rock basins 431 Transporting work of glaciers 432 Constructive work of glaciers 433 Ground moraine 433 Drumlins 433 Eskers *134 Terminal moraines 434 Kames 435 Recessional and interlobate moraines 435 Drainage of drift-covered areas 435 Differences between younger and older drift 435 Effect of nunatak stages on distribution of drift 436 Variation of deposits with slopes 436 Outwash deposits 437 Pitted plains ■ 438 Loess - 438 Valley lakes due to variation in stream load 438 Distribution of glacial drift 439 Marginal lakes 441 Glacial Lake Agassiz 442 Marginal glacial lakes 442 Lake Nemadji 443 Lake Duluth 444 Intermediate glacial lakes 445 Lake Algonquin 446 Nipi.ssiiig Great Lakes 447 Effect of tilting on glacial lakes 448 Present iiosition of raised beaches 449 Glacial-lake deposits 452 The four Pleistocene provinces 453 Grounds for distinction 453 CONTENTS. 21 Chapter XVI. The Pleistocene, by Lawrence Martin — Continued. Page. The four Pleistocene provinces — Continued. Drif tless Area 454 Area of older drift 454 Area of last drift 454 Areas of glacial-lake deposits 454 Postglacial modifications 455 Modifications on the land 455 Modifications in and around the Great Lakes 456 Summary of the Pleistocene history 459 Chapter XVI I . The iron ores of the Lake Superior region, by the authors and W.J. Mead 460 Horizons of iron-bearing formations 460 General description of ores of the Lake Superior pre-Cambrian sedimentary iron-bearing formations 461 Introduction 461 Kinds of rocks in the iron-bearing formations 461 Chemical composition of the iron-bearing formations 462 Ratio of ore to rock in the iron-bearing formations 462 Structural features of ore bodies 462 Shape and size of the ore bodies 475 Topographic relations of the ore bodies 476 Outcrops of the ore bodies 476 Chemical compo.sition of the ores 477 Mineralogy of the ores 479 Physical characteristics of the ore 480 General character 480 Cubic contents of ore 481 Range and determination 481 Use of the diagram 482 Construction of the diagram 482 Effect of porosity 482 Effect of moisture 483 Moisture of saturation 433 Excess of moisture handled in mining 434 Exploration for iron ore 434 Magnetism of the Lake Superior iron ores and iron-bearing formations 486 Manganiferous iron ores 433 Iron-ore reserves 433 Data available for estimates 433 Availability of ores 433 Reserves of ore at present available 439 Estimates 43g Life of ore reserves at present available 499 Reserves available for the future 49]^ Estimates 492 Comparison of Lake Superior reserves with other reserves of the United States 492 Lowering of grade now discernible 493 Effect of increased use of low-grade ores 494 Comparison with principal foreign ores 495 Tra,usportation 4g5 Mine to boat 495 Docks 496 Boats 497 Dock to furnace 497 Total cost of transportation 497 Methods of mining 497 Rates of royalty and value of ore in the ground 499 Origin of the ores of the Lake Superior pre-Cambrian sedimentary iron-bearing formations 499 Outline of discussion 499 The iron ores are chiefly altered parts of sedimentary rocks -. 500 Conditions of sedimentation 5OO Iron-bearing formations mainly chemical sediments 5OO Order of deposition of the iron-bearing sediments 5OX Are the iron-bearing formations terrestrial or subaqueous sediments? 5OI 22 CONTENTS. Chapter XVII. The iron ores of the Lake Superior region, by the authors and W. J. Mead — Continued. page. Origin of the ores of the Lake Superior pre-Cambrian sedimentary iron-bearing formations — Continued. Conditions of sedimentation — Continued. Pog and lagoon origin of part of tlie iron-bearing rocks 502 Hypothesis of bog and lagoon origin not applicable to the main masses of the iron-bearing sediments. . 502 Hypothesis of glauconilic origin not. applicable 503 Iron-bearing sediments not laterite deposits 503 Iron-bearing sediments not characteristic transported deposits of ordinary ero-sion cycles .503 As.-!Ociation of iron-bearing sediments with rontemporaneotis eruptive rocks ,506 Association of iron-bearing sediments and eruptive rocks outside of the Lake Superior region 508 Significance of ellipsoidal structure of eruptive rocks in relation to origin of the ores 510 Eruptive rocks associated \yith iron-bearing sediment;* of Lake Superior region carry abundant iron. 512 Genetic relations of upper Huronian slate to associated eruptive rocks 513 Main mass of iron-bearing sediments probably derived from associated eruptive rocks 513 Direct contributions of iron salts in hot solutions from the magma 513 Contribution of iron salts from crystallized igneous rocks in meteoric waters 514 Contribution of iron salts by reaction of hot igneous rocks with sea water 515 Conclusion as to derivation of materials for the iron-bearing formations 516 Variations of iron-bearing formations with different eruptive rocks and different conditions of deposition 516 Chemistry of original deposition of the iron-bearing formations 518 Natiu'e of the problem ; 518 Formation of iron carbonate and limonite 519 Nature of carbonate precipitate 520 Precipitation of greenalite 521 Processes 521 Nature of greenalite precipitate 522 Source of alkaline silicates necessary to produce greenalite 525 Reactions betwaen greenalite and iron carbonate, or carbon dioxide 526 Source of carbon dioxide for reactions with greenalite 527 Deposition of hematite, magnetite, and silica directly from hot solutions 527 Deposition of iron sulphide 527 Correlation of laboratory and field observations 527 Secondary concentration of the ores 529 General statements 529 Chemical and mineralogical changes involved in concentration of the ore under surface condi- tions 529 Outline of alterations 529 Oxidation and hydration of the greenalite and siderite producing ferruginous chert 530 Alteration of ferruginous chert to ore by the leaching of silica, with or without secondary intro- duction of iron 537 Processes involved 537 Conditions favorable to leaching of silica 538 Solution of silica favored by alkaline character of waters ^. 538 Transfer of iron in solution 539 Secondary concentration of the ores characteristic of weathering 539 Mechanical concentration and erosion of iron ores 540 General character of mi ne waters 540 Localization of the ores controlled by special structural and topographic features 544 Quantitative study of secondary concentration 545 Alterations of iron-bearing formations by igneous intrusions 546 Ores affected 546 Possible contributions from igneous rocks 546 Temperature at which contact alterations were effected 549 Character of iron-bearing formations at the time of intrusions of igneous rocks 549 Chemistry of alterations 550 Banding of amphibole-raagnetite rocks 551 Recrystallization of quartz 552 High sulphur content of amphibole-magnetite rocks 552 Secondary iron carbonate locally developed at igneous contacts 552 Contact alterations not favorable to concentration of ore deposits 552 SiU'face alterations of amphibole-magnetite rocks 553 Summary of alterations of iron-beaiing formations by igneous intrusions 554 CONTENTS. 23 Chapter XVII. The iron ores of the Lake Superior region, by the authors and W. J. Mead — Continued. Page. Origin of the ores of the Lake Superior pre-Cambrian sedimentary iron-bearing formations — Continued. Alteration of iron-bearing formations by rock flowage 554 Cause of varying degree of hydration of the Lake Superior ores 555 Sequence of ore concentration 557 Origin of manganiferous iron ores 5(j0 Part of the metamorphic cycle illustrated by the Lake Superior iron ores of sedimentary type 500 Titaniferous magnetites of northern Minnesota 561 Magnetites of possible pegmatitic origin 562 Brown ores and hematites associated with Paleozoic and Pleistocene deposits in Wisconsin 562 Ores in the Potsdam 562 Brown ores in "Lower Magnesian " limestone 562 Geology and topography 565 Oilman brown-ore deposit 565 Cady brown-ore deposit 565 Origin of Spring Valley brown-ore deposits 566 Postglacial brown ores 566 Clinton iron ores of Dodge County, Wis 567 Occurrence and character 567 Origin of the Clinton iron ores 568 Summary statement of theory of origin of the Lake Superior iron ores 568 Other theories of the origin of the Lake Superior pre-Cambrian iron ores 569 Genetic classification of the principal iron ores of the world 571 Chapter XVIII. The copper ores of the Lake Superior region, by the authors, assisted by Edward Steidtmann. 573 The copper deposits of Keweenaw Point 573 General account 573 Transverse veins of Eagle River district 575 Dipping veins of Ontonagon district 576 Amygdaloid deposits 576 Copper in conglomerates 577 Composition of copper-mine waters .579 Copper in Keweenawan rocks in parts of the Lake Superior region other than Keweenaw Point 580 Origin of the copper ores 580 Common origin of the several types of deposits 580 Previous views of nature of copper-depositing solutions and source of copper 580 Outline of hypothesis of origin of copper ores presented in the following pages 581 Association of ores and igneous rocks 581 Ore deposition limited mainly to middle Keweenawan time 581 Deposition of the copper accomplished l)y hot solutions 582 Nature of gangue minerals 582 Nature of wall-rock alterations ; 582 Paragenesis of copper and gangue minerals 585 Contrast with present work of meteoric solutions : 585 Source of thermal solutions 586 Three hypotheses 586 Were the thermal solutions derived from extrusive or from intrusive rocks? 587 Significance of sulphides of copper in the intrusives and lower effusi^■es 588 Conclusions as to source of copper-bearing solutions 588 Chemistry of deposition of copper ores 589 Cause of diminution of richness with increasing depth 591 Relation of copper ores to other ores of the Keweenawan 591 . Chapter XIX. The silver and gold ores of the Lake Superior region 593 Silver ores 593 Production T 593 Silver Islet 593 General account of silver in the Animikie group 594 Origin of silver ores in the Animikie group 595 Gold ores 595 Chapter XX. General geology 597 Introduction 697 Principles of correlation 597 General character and correlation of the Archean 599 Keewatin series 599 24 CONTENTS. Chapter XX. General geology — Continued. Page. General character and correlation of the Archean — Continued. l.aurentian scries 600 General statements concernini; the Archean system 601 General statements concerning the Algonkian system 602 Character and subdivisions 602 Northern Huronian subprovince 602 Lower middle Huronian 602 Lithology and succession 602 Igneous rocks 603 Conditions of deposition 603 Correlation 603 Upper Huronian (Animikie group) 604 Lithology and succession .' .604 Igneous rocks 604 Conditions of deposition 605 Correlation 605 Southern Huronian subprovince 605 Lower Huronian 605 Lithology and succession 605 Igneous rocks 605 Conditions of deposition 606 Correlation 606 Middle Huronian 607 Lithology and succession 607 Igneous rocks 607 Conditions of deposition 607 Correlation . i 608 Upper Huronian (Animikie group) 608 Lithology and succession 608 Igneous rocks 609 Conditions of deposition 609 Correlation 609 General remarks concerning the upper Huronian ( Animikie group) of the Lake Superior region 610 Character 610 Conditions of deposition of the upper Huronian (Animikie group) 612 Keweenawan series 614 Lithology and succession 614 Igneous rocks 615 Conditions of deposition 615 Correlation 615 Paleozoic rocks • • - ■ 615 Cretaceous rocks 616 Pleistocene deposits 617 Pre-Cambrian volcanism 61" Pre-Cambrian life 617 Unconformities 617 Unconformity lietween the Archean and lower Huronian 617 Unconformity between the lower and middle Huronian 618 Unconformity at the base of the upper Huronian ( Animikie group i 619 Unconformity at the base of the Keweenawan 619 Unconformity at the base of the Cambrian 619 Deformation and metamorphism 620 General conditions , 620 Principal elements of structure 621 The Lake Superior basin 622 R6sum6 of history 623 Index 627 ILLUSTRATIONS. Page. Plate I. Geologic map of the Lake Superior region, wilh sections In pocket. II. Relief map of the Lake Superior region, showing the larger topographic features 86 III. .4, Pre-Oambrian peneplain in Ontario, near Michipicoten; B, Jasper Peak, near Tower, Minn. ... 88 IV. A, Topographic map of Rib Hill, Wis.; B, Typical monoclinal ridge topography, U\e Royal, Mich.. 90 V. A, The Duluth escarpment and even upland of peneplain on Duluth gabbro in Minnesota; B, Lake shore escarpment of Archean schists and Iluroniauquartzite near Marquette, Mich. .' 112 VI. Geologic map of the Vermilion iron-bearing district, Minn 118 VII. A, Ellipsoidal parting in Ely greenstone; B, Ellipsoidally parted Ely greenstone, showing spheru- litic development ^ VIII. Geologic map of the Mesabi iron-bearing district, Minn In pocket. IX. Sharp folding of beds of iron-bearing Biwabik formation in Mesabi district, Minn.; A, Hawkins mine; B, Monroe mine ^"^ X. Typical cross section through iron-bearing Biwabik formation, Mesabi district, Minn., from drill records ■.■■'.■■' XI. A, Panoramic view of the Mountain Iron open-pit mine, Mesabi district, Minn.; B, Panoramic view of the Shenango iron mine, Mesabi district, Minn 180 XII. Geologic map of Pigeon Point, Minn • -O'l XIII. Geologic map of the Animikie iron-bearing district, north of Thunder Bay, Ontario 206 XIV. Map of central Minnesota, including Cuyuna district 212 XV. Map of part of the Cuyuna iron district of Minnesota, showing magnetic belts 212 XVI. Geologic map of the Penokee-Gogebic district 226 XVII. Geologic map of the Marquette iron-bearing district, Mich In pocket. XVIII. Map of Carp River fault, sees. 4, 5, and 6, T. 47 N., R. 25 W., Mich 252 XIX. Detailed map of quartzite ridges of Teal Lake, showing faulting and unconformity of Ajibik and Mesnard formations -^'* XX. Geologic map of Dead River area, Mich 286 XXI. Map of Perch Lake district, Mich., showing distribution of outcrops In pocket. XXII. Geologic map of the Crystal Falls district, Mich., including a portion of the Marquette district. In pocket. XXIII. Geologic map of the Calumet district, Mich 306 XXIV. Geologic map of the Iron River district, Mich -' In pocket. XXV. Geologic map of the Florence iron district, Wis In pocket. XXVI. Geologic map of the Menominee iron district, Mich In pocket. XXVII. Vertical north-south cross sections through the Norway-Aragon area, Menominee district, Mich., illustrating geologic structure 346 XXVIII. Geologic map of the Keweenaw Point copper district, Mich - . 380 XXIX. A, Hanging valley near Helen mine, Michipicoten; B, Lake clay overlying stony glacial till in Mountain Iron open pit, Mesabi range, Minn 43- XXX. .4, Terminal-moraine and outwash-plain topography in glaciated area of western Wisconsin; B, Glaciated valley of Portage Lake on Keweenaw Point, Mich,, with hanging valley of Huron Creek. 434 XXXI. A, Characteristic Driftless Area topography in northern Wisconsin; B, Characteristic muskeg and ground-moraine topography in glaciated area of Minnesota 436 XXXII. Jaspilite from Marquette district, Mich 464 XXXIII. A, Folded and brecciated jaspilite of the Soudan formation, Vermilion district, Minn.; B, Hema- titic chert from Negaunee, Marquette district, Mich 466 XXXIV. Ferruginous chert and slate of iron-bearing Biwabik formation, Mesabi district, Minn 468 XXXV. .4, Amphibole-magnetite chert from Republic, Mich.; B, Sideritic magnetite-grunerite schist from Marquette district, Mich 4/0 XXXVI. .4, Jaspery filling in amygdules from ellipsoidal basalt of the Crystal Falls district, Mich.; B, Cherty siderite from Marquette district, Mich.; C, Cherty siderite from Penokee district, Mich 472 XXXVII. Greenalite rock from Mesabi district, Minn 474 25 26 ILLUSTRATIONS. Page. Plate XXXVIII. Characteristic specimens of iron ores 480 XXXIX. CharacteriHtic specimens of iron ores 480 XL. Diaf^ram showing relation of density, porosity, and moisture to cubic feet per ton 480 XLI. yl. Ore dock.s at Two Harbors, Minn.; B, Excavations at Stevenson, Minn 496 XLII. Photomicrosiraphs of natural and artificial greenalite granules, cherty siderite, and concre- t ionary ferruginous chert 524 XLIII. Photomicrographs of greenalite granules 532 XLIV. Photomicrographs of ferruginous chert showing later stages of the alteration of greenalite granules 534 XLV. Photomicrographs of granules and concretionary structtires in Clinton iron ores 536 XLVI. A, Ore and jasper conglomerate from Marquette district, Mich.; B, Ferruginous chert from Marquette district, Mich 542 XLV II. Photomicrographs of ferruginous and amphiboli tic chert of iron-bearing Biwabik formation near contact with Duluth gabbro 548 XLVIII . Ferruginous chert or jasper, of possible pegmatitic origin, in basalt 564 XLIX. Map showing location of copper-bearing lodes and mines on Keweenaw Point 574 Figure 1. Key map showing location of Lake Superior region 31 2. Sketch map of the Lake Superior region, showing iron districts, shipping ports, and transportation lines - 32 3. Diagram showing annual production of iron ore in Lake Superior region since the opening of the region. 49 4. Generalized topographic map of the Lake Superior region 87 5. The topographic pro\ances of the Lake Superior region, with some subdi^dsions of the peneplain 88 6. True-scale cross section of Keweenawan monoclinal ridges near the end of Keweenaw Point 99 7. Hypothetical cross section showing relation of secondary lowlands, mesas, monoclinal ridges, etc., to peneplain - 101 8. Graben or rift valley of western Lake Superior, showing escarpments on either side and peneplain above 112 9. The drainage of the St. Louis and Mississippi headwaters before the stream captures along the Duluth escarpment 113 10. The drainage of the St. Louis and Mississippi headwaters at present, after stream captures and diversions 113 11. Structure profile in northern Wisconsin, showing the south edge of the peneplain on the pre- Cambrian rocks and the northern part of the belted plain of the Paleozoic 116 12. Diagram to illustrate folding of "drag" type, common in the Vermilion and other ranges 123 13. Section across jasper belt in sees. 13 and 14, T. 62 N., R. 13 W., Vermilion iron range, Minn 123 14. Transverse sections of Chandler, Pioneer, Zenith, Sibley, and Savoy mines, Vermilion district, Minn 138 15. Diagram illustrating volume changes involved in the alteration of jasper to ore at Ely, Minn 142 16. North-south cross section of an ore deposit on the Mesabi range near Hibbing, Minn 180 17. Triangular diagram showing composition of various phases of Mesabi ores and ferruginous cherts 182 18. Section through iron-bearing Biwabik formation transverse to the range, showing nature of circula- tion of water and its relations to confining strata 186 19. Dia"-ram showing volume changes observed in the alteration of ferruginous chert to ore 188 20. Graphic representation of the changes involved in the alteration of greenalite rock to ferruginous chert (taconite) and ore 189 21. Triangular diagram representing volume composition of the various phases of ferruginous cherts and iron ores of the Mesabi district 190 22. Diagram showing relation of phosphorus to degree of hydration in Mesabi ores 192 23. Diagram showing relative amounts of phosphorus and lime in Mesabi ores 196 24. Cross section of iron-bearing Gunflint formation east of Paulson mine, Gunflint district, Minn 199 25. Plan and cross section of the iron-ore deposit in sec. 12, T. 43 N., R. 32 \V., Crow Wing County, Minn. 218 26. Triangular diagram showing mineralogical composition of various phases of iron ores and ferruginous cherts of the Cuyuna district, Minn ■ 2_1 27. Triangular diagram showing volume composition of various phases of iron ores and ferruginous cherts of the Cuyuna district, Minn ""- 28. Cross section showing the occurrence of ore in pitching troughs formed by dikes and quartzite foot- wall, in the Gogebic district -36 29. Ore depo.sits of the Penokee-Gogebic district ; 237 30. Triangular diagram showing chemical composition of various jihases of Gogebic ores and ferruginous cherts - ■ ^39 31. Diagrammatic; rei)re.sentation of the changes involved in the alteration of cherty iron carbonate to ferruginous chert and ore, Gogebic district "^ ILLUSTRATIONS. 27 Page. FiGtJRE 32. Triangular diagram showing volume composition of the ferruginous cherts and iron ores of the Gogebic range 245 33. Diagram showing relation of phosjihorus to degree of hydration in Gogebic ores 248 34. Diagram showing relative amounts of phosphorus and lime in Gogebic ores 249 3.5. Idealized north-south section through the Marquette district, sho^ving abnormal type of synclinorium. 2.53 36. Ore deposits of the Marquette district .• 270 37. Graphic representation of the volume composition of the principal phases of the iron-bearing Negaunee formation 276 38. Triangular diagram showing the volume composition of the several grades of ore mined in the Mar- quette district in 1906 277 39. Diagram showing relation of phosphorus to degree of hydration in Marquette ores 280 40. Diagram showing relative amounts of phosphorus and lime in Marquette ores 282 41. Outcrop map of Swanzy district, Mich 284 42. Geologic map of west end of Marquette district, Mich 289 43. Sketch map to show general relations of iron-bearing rocks, principally upper Huronian, in Crystal t- Falls, Iron River, Florence, and Menominee districts 292 I 44. Section showing roughly the succession of beds in the Vulcan iron-bearing member near Atkinson, in the Iron River district, Mich 318 45. Geologic map and cross section of Iron Hill, Menominee district, showing relations of lower and mid- dle Hiu-onian 3^5 46. Horizontal section of the Aragon mine at the first level, Menominee district, Mich 347 47. Horizontal section of the Aragon mine at the eighth level, Menominee district, Mich 348 48. Vertical north-south cross section through Burnt shaft. West Vulcan mine, Menominee district, Mich. . 349 49. Sketch to show pitch of a drag fold in a monoclinal succession 350 50. Triangular diagi-am representing the volume composition of the various grades of ore mined in the Menominee, Crystal Falls, and neighboring districts in 1907 352 51. Sketch map showing occurrence of quartzites of Huronian age in Tps. 33 and 34 N., Rs. 15, 16, and 17 E., Wis ■ 358 52. Sketch map showing occurrence of Huronian quartzite near Necedah, Wis. 358 53. Sketch map showing Baraboo, Fox River valley, Necedah, Waushara, and Waterloo pre-Cambrian areas of south-central Wisconsin 359 54. Generalized cross section extending north and south across the Baraboo district 360 55. Vertical section of Illinois mine 364 56. Section on south cliff of Great Palisades, Minnesota coast " 371 57. Sketch showing unconformable contact between Keweenawan diabase porphyry and Cambrian sand- stone at Taylors Falls, Minn 379 58. Diagrammatic section illustrating the assigned change of attitude of a series of beds, like the Kewee- nawan, from an original depositional inclination to a more highly inclined attitude 419 59 Map of the Lake Superior basin, designed to show the structure and e.xtent of the Keweenawan trough 422 60. Sketch map showing the glaciation of the Lake Superior region, giving names of lobes and probable general directions of ice flow 428 61. Sketch showing the glacial cirque, the rock basins, and the hanging valley near the Helen mine, Michipicoten 432 62. Sketch showing the origin of the drift deposits overlying the ore in the Mesabi iron range 443 63. Glacial Lake Nemadji 444 64. Glacial Lake Duluth 445 65. Hypothetical intermediate stage vrith the expansion of glacial Lake Chicago and the later stage of glacial Lake Duluth 446 66. Glacial Lake Algonquin 447 67. Part of Nipissdng Great Lakes 448 68. Sketch map shoiving Driftless Area and regions of older drift, last drift, and lake deposits 453 69. St. Louis Ri\-er at the stage when it cut its valley and emptied directly into Lake Nipissing 456 70. The present St. Louis River, which has been converted into an estuary by post-Nipissing tilting 457 71. Triangular diagram showing chemical composition of all grades of iron ore mined in the Lake Supe- rior region in 1906 478 72. Textures of Lake Superior iron ores as shown by screening tests 481 73. Diagram showing relation between estimated ore reserves of the Lake Superior region and rate of pro- duction 490 74. Diagram representing decline in grade of Lake Superior iron ore since 1889 493 75. Cross section of Keweenaw Point near Calumet, showing copper lodes in conglomerates and amyg- daloids ^"4 76. Triangular diagi-am comparing the amomita of undecomposed silicates, quartz, and residual weathered products in different kinds of muds, shales, and weathered rocks 612 THE GEOLOGY OF THE LAKE SUPERIOR REGION. By C. R. Van Hise and C. K. Leitii. CHAPTER I. INTRODUCTION. OUTLINE OF MONOGRAPH. The Lake Superior rejjion is a part of the southern margin of tlie great pre-Cambrian shield of northern North America. It is bordered and overlapped on the south by Paleozoic rocks of the iEssissippi Valley and on the southwest by Cretaceous deposits. The pre-Cambrian rocks . of the area, which may be divided into a considerable number of lithologic and time units, contain the great iron and copper deposits by which the region is most widely known. The great development of the mineral industry in this region has afforded the geologist unusual opportunity for study, as it has not only made the region more accessible but has justified larger expenditures for geologic study than would otherwise have been made. Tliis fortunate combination of a field containing an exceptionally full record of a little-known part of the geologic column \vith the means of studying it has warranted tlie study of the pre-Cambrian with a degree of detail that has been practicable in but few other significant pre-Cambrian regions. Geologic surveys of various parts of the Lake Superior region have been conducted under national, state, and private supervision almost without interruption since the early part of the nineteenth century, especially since the opening of the mining industry in the middle of the century. The later reports have naturally been more adequate than the earlier ones, because they have included the results of the earlier work and have gained the advantage derived from the greater accessibility of the district. The reports thus far issued have dealt with small parts of the region or with certain phases of its general geology. State and private surveys have neces- sarily worked within jirescribed areas, so that notwithstanding the multiplicity of reports certain parts of the region have not yet been adetpiately covered. It has been the proper func- tion of the United States Geological Survey to make detailed surveys designed to accomplish the uniform treatment and correlation of the several ore-bearing districts, and finally to publish a monographic report on the region as a whole. Work under a general plan for these surveys was begun in the early eighties under the direction of Prof. R. D. Irving, whose monograph on the copper-bearing rocks of Lake Superior " appeared in 1883, though it was partly prepared at an earher date, while he was connected with the Wisconsin Geological Survey. The develop- ment of this plan has since been continuous. Until 1888 the work was in charge of Professor Ir\ang: since that time it has been under the direction of Dr. Charles R. Van Hise, the senior author of this monograph. Detailed monographs on the live leading iron ranges have been published and also papers covering different phases of the general geology of the region. This monograph represents the first attempt to give a connected account of the geology of the Lake Superior region as a whole, with special reference to the iron and copper bearing for- mations. Attention is dii'ected primarily to general features of correlation of the formations, o Mon. U. S. Geol. Survey, vol. 5, 1883. 29 30 GEOLOGY OF THE LAKE SUPERIOR REGION. to the geologic history of the region, and to tlio origin of the iron and copper ores. In addition, brief chapters are presented on several parts of the district which had not yet been reported on by the United States Geological Survey. No attemj)t is made to give details. For these tlie reader is referred to the pul)lications of the United States Geological Survey and of state geological surveys and to other sources specified in ai)j)ropriate places in this volume. Tiiough tills monograph may be regarded as completing a stage in the jjrogress of the geologic survey of tlie region, and lience may be considered final in one sense, it may also properly be regarded as only the first of a series of general studies of tlie tlistrict. The area is so large and the record is so complex that this monograph will accomplish its purjjose if it discloses the elements of some of the major j)rol)lcms of the region and affords a basis for a better-directed attack on them than has heretofore been possible. Future monographs will untloubtedly be written on each of the many phases of subjects that are barely touched upon in this monograph, such, for instance, as the petrography and consanguinity of the igneous rocks of different periods, the conditions of sedimentation of various series, the relations of volcanism to ore deposition, and the correlation of major and minor structural features of the Lake Sujjerior region with one another antl with the various structural features of Xortli America. Besides, certain areas not yet fully reported on will require detailed monographic description. It is hoped that the work of the United States Geological Survey in the Lake Sujierior region may be continued along the lines indicated. Parts of the region have been studied at different times by men occup3'ing different view- points. Some areas which have recently become commercially prominent have not yet been adcciuately studied in detail. Finally, mining, drilling, and various public and private surveys are so rapidly extending the knowleclge of the geology of the region that it is practically impos- sible at the present time to write a monograph that will not require modification in some par- ticulars almost before it comes from the press. Because of these facts this work shows inequal- ities and inadequacies of treatment for different parts of the region and for different phases of the subject. It is hoped, however, that the monograph will be measured by the advance it represents over previous available knowledge and especially by its attempt to bring out sig- nificant general features of the geology not heretofore discussed, and not by its deficiencies, of which the writers have a lively appreciation. The parts of the report written partly or wholly by others than the authors bear the names of the writers. It will be understood that any chapter or section for which no names are given has been written by C. R. Van Hise and C. K. Leith. ACKNOWLEDGMENTS. The completion of tliis monograph and the detailed studies leading up to it have been facilitated by the cordial cooperation of the mining men of the region. To attempt to mention the names of all who have gone out of their way to render aid in these studies would involve the publication of a list including the greater number of local mining men, and even from such a list some names would probably be inadvertently omitted. Especially valuable has been the infor- mation furnished by the Ohver Iron Mining Company (United States Steel Corporation), which has a most highly developed and efficient engineering and geologic staff. Valuable aid has been given by state and provincial surveys and by the Minnesota tax commission. To all these men and organizations we express our indebtedness and thanks. We are indebted to Messrs. W. J. Mead, Lawrence Martin, Alexander N. Winchell, A. C. Lane, R. C. Allen, and Edward Steidtmann for sections of this report bearing their names, and to numerous other men mentioned in the report who have contributed in ilifferent ways. Not the least of our indebtedness is to Mr. A. C. Deming for efficient clerical service. GEOGRAPHY. The Lake Superior region comprises parts of Michigan, Wisconsin, Minnesota, and Ontario- adjacent to Lake Superior. (See figs. 1 and 2.) The accoiii])anying general geologic map- INTRODUCTION. 31 (PI. I, in pocket) covers the area between parallels 44° and 49° north and meridians 84° and 95 west, comprising ai)proximately 1,81,000 s(|uare miles — an area almost equal to that of the sLx New England States and New York, New Jersey, Pennsylvania, and Maryland, or that of Sweden and Belgium. v^-O KENTUCKYy-^j^^j^i^ 'OKLAHOMA I ) ''" 200 300 WILES FiGiTRE 1. — Key map showing location of Lake Superior region. The region includes several ore-bearing districts of comparatively small area — the Ke- weenaw copper-bearing district of Keweenaw Point, Michigan, about 1,350 square miles; the Marquette iron-bearing district of Michigan, extending westward from the city of Marquette 82 GEOLOGY OF THE LAKE SUPERIOR REGION. on tlic hike slioro, about .330 s(|uare miles; the Menominee iron-hearir^s^ district, e.\tendin<^ from Iron Mountain in Michigan eastward alon» Menominee River, a The general topography of this lake has been reviewed by M. W, Harrington (Nat. Geog. Mag. vol. 7, 1S9G, pp. 111-120). who hasalso studied the currents in the Great Lakes in detail (Bulletin B, Weather Bur., U. S. Dept. Agr., 1895). c The physical geography of a part of this region was described in its larger aspects in 1S50 by Foster and Whitney, Report on the geology and topography of a portion of the Lake Superior land district in Michigan, vol. 1, pp. lS-83. liSeliermerhom, L. Y., Am. Jour. Sci., 3dser., vol. 33, 1887, p. 282. • 47517°— VOL 52—11 S 34 GEOLOGY OF THE LAKE SUPERIOR REGION. A number of short streams, such as Manistique, White, and Escanaba rivers, flow south- ward into Lake Michigan and Green Bay. Menominee River, which forms the Michigan- Wisconsin boundary, flows southeastward into Green Bay, receiving as tributaries the Paint and the IMichigamme. Peshtigo and Wolf rivers drain northeastern Wisconsin. A number of small streams drain the northeastern part of the Lower Peninsula of Micliigan. Another large part of the Lake Superior region in Wisconsin and Minnesota is tributary to the Mississippi and so to the Gulf of Mexico. The principal tributaries in tliis area are Wisconsin, St. Croix, Black, Chippewa, Swan, and Prairie rivers. A third large part of the Lake Superior region, in northern Minnesota and western Ontario, is tributary to Lake Winnipeg, and hence to Nelson River and Hudson Bay. This system com- prises the numerous large lakes occupying a large portion of the area of northern Minnesota and western Ontario, including Lakes Rainy and Vermilion and Lake of the Woods. The divide between the St. Lawrence and the Mississippi drainage systems extends from Portage in central Wisconsin, between Wisconsin and Fox rivers, north to the Wisconsin- Michigan boundary (fig. 4, p. 87) thence northwest and west into Minnesota, and thence north between upper Mississippi River and St. Louis River to the Giants Range. The Giants Range, extending east-northeast across the northern part of Minnesota, separates the Mississippi and the St. Lawrence systems on the southwest and southeast, respectively, from the Nelson River and Hudson Bay system on the north. The areas of these three large drainage systems within the Lake. Superior region are as follows: St. Lawrence, 107,000 square miles; Mississippi, 52,000; Hudson Bay, 22,000. As a whole the drainage of the Lake Superior region is very imperfect. The rxumerous lakes, swamps, waterfaUs, and rapids are features of an immature drainage. CHAPTER II. HISTORY OF LAKE SUPERIOR MINING. THE KEWEENAW COPPER DISTRICT OF MICHIGAN (1844).° The existence of copper was known to the Chippewa Indians met in the Lake Superior region bj^ the earliest explorers. They exhibited crude ornaments of native copper but seemed to make no further use of their knowledge. There is evidence that mining was carried on at a far earlier period. ■^Tiether the mining was done by ancestors of the aboriginal tribes discovered in possession of the Lake district by the earliest white explorers, or by some antecedent people of higher civilization, is a point that archaeologists and ethnologists are still arguing. Whatever may have been the derivation or fate of that prehistoric race of copper miners vaguely termed "mound builders," it is certain that they enjoyed at least a rudimentary civilization and were suc- cessful metallurgists, for they possessed the art of tempering copper. Weapons for the chase and war and domestic utensils of good finish and style and highly tempered are dug from mounds and found in sand dunes along the southern shore of Lake Superior from time to time. 6 The existence of native copper on Keweenaw Point was reported by La Garde in 1636, by the Jesuit missionaries in the "Relations," extending from 1632 to 1672, by Baron Le Houtan in 1689, by P. de CharlevoLx in 1721, and by Jonathan Carver in 1765. The report of Captain Carver led to the formation of a mining company which actually mined copper ore in 1761 and 1762, but \vithout commercial success. In 1771 Alexander Henry, an Englishman, began mining operations, but he desisted in 1774. The copper ores were noted in 1819 by H. L. School- craft and in 1823 by Major Long, both of them conducting explorations for the Government. The first systematic survey and study of the copper ores was made by Douglass Houghton for the first Micliigan Geological Survey. In 1830, in company with Gen. Lewis Cass, he first visited the copper region, and some years later began combined geologic and topographic surveying, for which, by considerable effort, he had procured support from the Michigan legislature. His first report was published in 1841. Previous stories of mineral wealth on the southern shore of Lake Superior had been too vague and confused to interest capitalists sufficiently to venture their money in attempts at mining in a country which was then much farther from the centers of wealth and population than is Cape Nome to-day, measured by time and transportation facilities. This apathy was dispelled by Dr. Houghton's first report, which was clear and concise and bore upon its face the stamp of truth. He told the world that vast stores of copper existed upon the southern shore of Lake Superior. Pressure was brought to bear upon the Federal Government, and in 1843 an arrangement was concluded with Dr. Houghton by which he was to combine a linear survey for the United States with a topographical and geographical survey he was then making for the State of Michigan. It was necessary that the linear survey be made before mining locations could be granted by the federal authorities, as there were no boundaries other than those of nature before that time. The work was begun in 1844, and during that and the following year rapid progress was made. Dr. Houghton's career was brought to an untimely end by his accidental drowning in Keweenaw Bay in the late fall of 1845, but his work was then so far advanced that it was taken up and pushed to early completion by competent successors." The first actual copper mining at Lake Superior was done in 1844, and the first product secured was a few tons of oxide ore — not native copper — taken from a fissure vein near Copper Harbor, Keweenaw County, by the Pittsburg and Lake Superior Mining Company, which later developed the Cliff mine, nearly 20 miles to the southwest. The Minnesota mine, in Ontonagon County, was opened shortly after. <* The subsequent history of the copper district is one of continuous rapid growth with only minor fluctuations. o In the following history of the Keweenaw copper district the authors have drawn freely on the excellent brief account of early conditions in the Copper handbook, by Horace J, Stevens. (■Stevens, H, J., Copper handbook, vol. 6, 1906, p, 14, tidem, vol, 2, 1902. pp, 16-17, rfldem, vol, 0, 1906, p, 17. 35 36 GEOLOGY OF THE LAKE SUPERIOR REGION. The following table of annual iiroduclion sliows, in amount ami in percentage, tin' relation of Lake Superior shipments to those of the United States: Anmuil production of Lake Superior copper district, compared with annual production of United States, 1850 to 1907. "^ Lake Superior Lake Superior Lake Su >erior or Michigan dis- or Miciiigau dis- or Mictiigau dis- | trict trict. trict Year. United States. Year. United States. Year. United States. Per- Per- Per- Amount. cent age. Amount. cent- age. Amount. cent- age. Long tons. Lomi tons. LoTUi Ions. Long tons. Long tons. Long tons. 1850 MO 572 88 1871 13.000 11.942 91 1892 154,018 54.999 30 1851 900 779 86 1872 12.500 10.961 87 1X93 147.0.3:! 50,270 34 1852 1,100 792 72 1873 15.500 13.433 86 1894 158.120 51,031 32 1853 2.0OO 1,297 65 1874 17, .500 15, ,327 87 1895 109.917 57,7.37 34 1854 2.250 1,819 81 1875 18,000 16,089 89 1896 205,384 63, 418 31 1855 3.000 2,593 86 1876 19,000 17,085 89 1897 220,571 «i,706 29 1856 4,000 3.666 91 1877 21,000 17, 422 83 1898 2.35.050 06,056 28 1857 4.800 4.255 88 1S78 21,500 17,719 82 1899 253.870 fo,(103 20 1858 5.500 4,088 74 1879 23, OOO 19, 129 83 1900 269,111 63.461 24 1859 6.300 3.985 63 I8.S0 27.000 22, 204 82 1901 268,522 09,501 26 1860 7,200 5,388 74 1881 32,000 24,363 76 1902 294.297 76. 050 26 1861 7,500 6,713 89 1S.S2 40, 467 25,439 62 1903 311.582 86,848 27 1862 9.000 6,005 67 1883 51,574 26. 663 51 1904 .362. 739 93,001 26 1863 8,500 5,797 68 1884 64, 708 30, 961 47 1905 402,704 102.874 25 1864 8,000 5,576 69 1885 74,052 32,209 43 1900 409. 414 102.514 25 1865 8,500 6,410 75 1S86 70,4.30 36, 124 51 1907 386. 655 96, 480 25 1866 8,900 6,l38 69 1887 81,017 33.941 42 1908 420.953 99, 408 23 1867 10,000 7,824 78 1888 101,054 38.604 38 1909 502, 425 103,290 20. S 1868 11, (»0 9,346 80 1889 101,239 39.364 38 1910 493,705 99,545 20 1869 12,500 11, .886 95 1890 115,966 45.273 39 1870 12,600 10,992 87 1891 126,839 50,992 40 c Stevens, H. J., op. cit., vol. 9, 1909, p. 1594. Production for 1909 and 1910 from Engineering and Mining Journal, For many years the district held first place as a producer of copper ore in the United States, and in total production it is stUl first; but in 1887 and later years, except 1891, its annual ship- ments have been surpassed by those of the Butte district of Montana and since 1904 by the copper districts of Arizona. The deposits first to be developed were the transverse fissure veins, rich m mass copper, cutting across the strike of the beds in the Eagle Harbor region, at the northeast end of the district. The Cliff mine was discovered by Charles T. Jackson in 1845. Production contmued in this district until 1895. It is now inactive but has been newly explored with a view to a reojienmg. Next to be developed were the vein or mass-copper deposits following the trend of the Keweenaw beds in Ontonagon County, at the southwestern end of the district. The presence of copper in this district was known for many years, but systematic mining was not started until a few years after the Eagle River district was opened. The principal mines were the Min- nesota (now the Michigan), the National, and the Mass. The Minnesota was discovered in 1847 by S. O. Knapp, through surface indentations of ancient workings. In one of these was found a mass of copper weighing 6 tons, together with rotted timbers on wMch it had been supported. The first shipment from tints mine was made in 1848, and for fourteen years 70 per cent of the ore was "mass." The opening of the Minnesota mine was followed by that of the National, Mass, and other mines. The district is still actively producing, but prmcipally from the ain^g- daloidal beds, mass copper at present (1908) constituting only about 25 per cept of the ore produced. The am3-g(ialoid deposits of the central part of the district were the next to receive atten- tion. The first of these deposits was discovered, in 1848, on the present Pewabic location, and the second on the Isle Royal location. The Quincy had been opened in 1847 on a transveise vein, but the Quincy ain3-gdaloid was not found until 1S5(), the same year that the main "Pewabic" bed was found. During 1856 the Quincy proihiced 13,462 pounds of cojiper. liut it did not become profita])le until ISOO. In 1877 tlie Osceola amygdaloid was discoveretl, and that 3'ear the Osceola mine pi-oduced 2,744,777 pounds of coj)per. The 'Wolvei'ine was opened before 1890 but was not profitable until 1S97. The Atlantic niuie was openeil in 1872. The HISTORY OF LAKE SUPERIOR MINING. 37 richest amygdaloid bed in the district is tire "Baltic," whicii was ffrst f)r()ved valuable by the Baltic mine in 1S97, and a few years later was discovered on the Champion location. The amygdaloid deposits are now the most numerous, and in 1907 produced 73.1 per cent of the total copper ore of the district, of which about 75.5 per cent came from Houghton County. A larger proportion of the production will come from the amj^gdaloids in the future. The last of the inincipal types of deposits to be discovered were those in the Allouez con- glomerate and the Calumet and Hecla conglomerate. Both conglomerates were discovered by E. J. Hulbert and associates. The Allouez conglomerate was found, in 1S59, at tlie site of the Allouez mine, and was worked for a short time, but soon proved to be unj)roductive, in this locality at least. Later it was found to be productive farther south, on the Boston and Albany location, later the Peninsula and now the Franklin Junior. The Allouez conglomerate has jdelded but little profit. The site of the Calumet and Hecla was bought by Hulbert in 1860, the evidence being a number of copper-bearing conglomerate bowlders and a few depressions, sucii as in other parts of the district were found to indicate ancient workings. In 1S69 Hulbert and liis associates returned to the spot and dug through an amygdaloid into the conglomerate bed. The Calumet and Hecla paid their first dividends in 1869 and 1870. Up to January, 1910, ths dividends of the Calumet and Hecla have aggregated .1110,550,000 on a capital of .S2, 500,000. The table below shows the relation in percentage of the annual production of the Calumet and Hecla mine, from 1867 to 1908, to the amiual production of the Alichigan district for the same period. Percentage of total Michigan copper production produced by the Calumet and Hecla mine, 1867 to 1908." 1867. 1868. 1869. 1870. 1871. 1872. 1873. 1874. 187.5. 1876. 1877. 1878. 1879. 1880. 7 5 24 4 46 57. 61 66. 62 5 5S 5 5 59. 56. 5 60. 63. 5 61. 63. 5 1881 57.5 1882 56.0 1883 52.5 1884.. 58.0 1885... . 65.'5 1886 62.5 1887 60.5 1888 58.0 1889 55 1890 . . 59. 1891 56. 5 1892 4fi. 1893 53 5 1894 53.5 1895. 1896. 1897. 1898. 1899. 1900. 1901. 1902. 1903. 1904. 1905. 1906. 1907. 1908. 61.5 63.0 58.0 58.5- 61.0 54.5 53.5 47.6 39.8 38.7 41.2 43.6 37.9 36.6 The onh" other mine now operating on the conglomerates is the Tamarack, opened m 1881. COPPER illNING ON ISLE ROYAL AND ELSEWHERE. Isle Royal is unusually rich m interesting evidences of prehistoric copper mining. The first minmg of historical record was begun soon after the opening of Keweenaw Point, in 1844, culminating in 1847 and 1848 and wanmg in 1855, when the island was again without perma- nent inhabitants. Another brief period of development, from 1871 to 1883, resulted in the opening on the island of the Saginaw and Minong mines, with a combined production of less than 10,000 tons of copper. Since the nineties exploration has been going on intermittently, but without success. No mines are operating at the present time. The ores are essentially the same as those of Keweenaw Point. As mined they were low grade, probably less than 1 per cent. They occur principally in fissure veins m the traps. The copper-bearing formation has been found elsewhere in the Lake Superior region, but the copper-mining industry has practicallj" not extended beyond Keweeiraw Point and Isle Royal. The southwestern extension of the Keweenaw district in Wisconsin and Minnesota is a Calculated from data in Stevens's Copper hand book, vol. 9, 1909. 38 GEOLOGY OF THE LAKE SUPERIOR REGION. being extensively exploted and opened for mining, l)ut thus far the production has not been important. As the copper-producing area has been restricted to that of the early discoveries and as the co]iper-mining industry has developed evenly, it is unnecessary' for our purposes to follow its lustory in greater detail. , IVLiRQUETTE IRON DISTRICT (1848). Iron was first discovered in 1844 near the site of Negaunee by the Government linear surveying party in charge of William A. Burt, himself under the direction of Douglass Houghton. The Michigan legislature having failed in 1843 to renew appropriations for the Michigan Survey, Dr. Houghton had turned to the Federal Government and had succeeded in procuruig an additional allowance per mile for geologic work in connection with the linear survey of the Upper Peninsula, which liad already been begiui, and he iiimself took the contract for the linear survey in order to have the direction of the work. In 1848 iron ore was mined by the Jackson Association (subsequently the Jackson Iron Companj^) and carried by team to a Catalan forge which they had constructed near Carp River. The project was not commercially successful and was closed in 1850. The ilarquette Iron Com- pany opened the Cleveland mine near the present town of Ishpeming in 1849 and carted its ore to a forge at Marquette. This also was a financial failure and was discontinued in 1854. In 1850 and again m 1852 a few tons of ore were shipped from the district to Pemisylvania for trial in Pennsylvania furnaces. The openmg in 1855 of the ship canal along St. Marys River, connecting Lake Huron and Lake Superior, was followed in 1856 by the first regular shipments of iron ore from the Marquette district to the lower lakes, amoimtmg to 6,.343 tons. Up to that time the local forges had consumed about 25,000 tons of ore. The completion in 1857 of the Iron Mountain Railway (later the Houghton, Marquette and Ontonagon Railway and ulti- mately the Duluth, South Shore and Atlantic Railway) between Marquette and the mines gave easy transit to the lake, and 22,876 tons were shipped in 1858 and five times that amoimt in 1860. From 1855 to 1862 transportation facilities were so far improved as to make it possible to get ore out, but the mines had not yet been really brought into relation with the iron market. Therefore the companies met with no real success whether they tried to make iron themselves or to send their ore down to the furnaces of Ohio and Pennsyl- vania. The Lehigh Valley, and not Pittsburg, was still the iron center of the United States. The war suddenly changed the whole outlook. A great demand sprang up for all kinds of iron goods, and both mining and iron making on the Upper Peninsula received a strong impetus. Shipments increased from 49,000 tons in 1861 to five times that amount in 1864, while the companies made fabulous profits. * * * The year 1865 marked a slight retrogression, but the eight years following saw a wonderful growth, the boom in iron and steel reflecting the rapid industrial develop- ment of the country, and from 1870 to 1873 registering its speculative excitement. * * * In 1863 but three mines shipped ore; in 1864, five; in 1865, seven; in 1866, nine; 1868 added tour more mines, 1870 three more, while in 1872 the table of shipments increases the total number of mines by 11 to 29, and in 1873 no less than 40 are represented. The total shipments of 1866 were just below 300,000 tons; those of 1873 almost exactly foiu- times that amount." The opening of the Republic, Michigamme, and Spurr mines in 1872 practically completed the area of the Marquette district as known at present, though a few discoveries of importance have been made within the area since that time. Exploration is still vigorous. The field for deep exploration opened by recent discoveries is a large one. The necessary increase in means of shipment was made by the building of the Chicago and Northwestern Railway from Negaunee to Escanaba and bj^ increase in the capacit}' of the docks already built at Marquette. As a result of the panic of 1873 — development work ceased, production fell off almost 25 per cent in 1874 and yet further in 1875, and the number of mines reporting shipments declined from 40 in 1873 to 33 in 1874 and to 29 in 1875. The working force of those that continued operations was largely reduced, and only five mines showed a larger output in 1874 than in 1873. & a Mussey, H. R., Combination In the mining industry: Studies in history, economics, and public law, Columbia Univ., vol. 23, No. 3, 1905, pp. 55,57, 59. I> Idem, p. 73. HISTORY OF LAKE SUPERIOR MINING. 39 Returning prosperity brought an increase in shipments of 80 per cent between 1878 and 1882, and tlie number of producing mines increased from 29 in 1875 and 1877 to 48 in 1882. The foUowmg year saw a considerable depression because of overproduction, but thenceforth the production showed a general increase until 1891, with a minor depression in 1885. The years 1891 and 1893 saw another falling off in production, the latter contemporaneous with the general panic of 1893. From that time to the present there has been a general increase in production, with shght recessions in 1904 and 1906. The Lake Superior and Ishpeming Railway was constructed in 1896 to carry the ores of the Cleveland-Chffs Iron Company from the Ish- peming district to Lake Superior. The table of production of iron ore from the Marquette range (pp. 51-60) summarizes the development of the district. The Swanzy district, southeast of the Marquette district proper, is reached by the Chicago and Northwestern Railway; its production is usually credited to the Marquette district. The district was first explored in 1869, and the Smith (later the Cheshire and Princeton) mine was opened in 1871. Systematic exploration by drilling, begun in 1902 by the Cleveland-Cliffs Company, greatly extended the ore reserves and determined the probable limits of the district. A largely increased production may be looked for. MENOMINEE IRON DISTRICT (1872). The Marquette district had been the sole producer of iron ore in the Lake Superior region for nearly thirty years when its first competitor, the Menominee district, entered the field. The first practical discovery of iron ore in Menominee County was made by the brothers Thomas and Bartley Breen some time previous to 1867, though the veteran explorer S. C. Smith claims to have been and probably was aware of its existence in that section as early as 1855, in which year he traversed what he called a new range, south and east from Lake Michigamme to Escanaba, locating what is now the estate of the Republic Iron Company on the way. The first practical work in the way of development was done by N. P. Hulst for the Milwaukee Iron Company at the Breen and Vulcan mines in 1872, and by John L. Buell at the Quinnesec the following year." The existence of ore in shipping quantity had been demonstrated in 1874, but the distance of the district from the Great Lakes and the lack of facilities for shipment prevented its further development until the extension of the Menominee branch of the Cliicago and Northwestern Railroad from Escanaba to Quinnesec. This was carried through to Iron Mountain in 1880, and thence northwest to Iron River and the Gogebic range. The Chicago, Milwaukee and St. Paul Railway entered the district in 1886 and the Wisconsin and Michigan Railway in 1903. When shipment had once started it increased much more rapidly than that of the Marquette district. The first year's output of 10,405 tons jumped to 95,221 tons the following year, to 269,609 tons the third year, and to 592,086 tons the fourth year, and reached the million mark in 1882. In 1901, 1902, and 1903 the Menominee surpassed the Marquette range in shipment, but for the most part in later j'ears it has been producmg about the same amount yearly as the Mar- quette district. Its total slupment to the end of 1909 is 71,212,121 tons as compared with a total of 91,838,558 tons from the Marquette district. The table (pp. 61-65) includes shipments from the outlying Florence, Iron River, and Crystal Falls districts to the northwest. CRYSTAL FALLS, FLORENCE, AND IRON RIVER IRON DISTRICTS (1880). The Crystal Falls, Florence, and Iron River districts may be regarded as northwesterly outliers of the Menominee range, and they are included in it in tables of production. For a number of years after the opening of the Menominee range prospectors worked in various places, among others in the vicinity of Crystal Falls, seeking to follow the iron range west of the Menominee River. As a result of this endeavor, the deposits at Florence, Wis., and then those farther north and west at Crystal Falls, Mich., were in turn located. It was not until 1881 that sufficient exploratory work had been done at Crystal Falls to warrant a belief in the future of this iron-bearing area. In April, 1882, the Chicago and Northwestern Railway completed o Swineford, A. P., .\nnual review of the iron mining and other industries of the Upper Peninsula for the year ending December 31, 1881; Mar- quette, 1882, p. 119. 40 GEOLOGY OF THE LAIvE SUPERIOR REGION. its branch to Crystal Falls, and the shipment of ore began. The Amasa deposits were not exploited to any great extent until the year 1888, when the Chicago and Northwestern Railway built a branch from Crystal Falls to Amasa. The Chicago, Milwaukee and St. Paul Railway in 1893 completed a line from Channing to Sidnaw, which runs through Amasa." These districts have as a whole (hn-cloped slowly as compared witli tlio other principal iron districts of the Lake Superior country, partly because of the slightly lower grade of many of the ore bodies and partly because of the lack of exposure, making exploration difricult and costly. Consequently large areas remain to be tested underground. The increasing demand for iron ore of the lower grades has brought about a revival of exploration in this area during the last few years. This is one of the most promising fields of exploration yet remain- ing ui the Lake Superior region, and the next few years are likely to see large developments. GOGEBIC IRON DISTRICT (1884). The Gogebic range of Michigan and its extension, the Penokee range of Wisconsin, some- times referred to together as the Penokee-Gogebic district, were long known to explorers and had been mapped by the geologists of the Michigan and Wisconsin surveys prior to their opening in 1884. The first recorded notice of their discovery appears on the plats of the township surveys. It is remarkable that subsequent discoveries have been restricted to the areas first determined by the geologic mappmg. Early exploration was largely confined to the weU- exposed magnetic portions of the formation at the west end of the range, which have been less productive than the central, less well exposed portions of the iron-bearing formation. In 1884 the first shipment of 1,022 tons was made from the Colby mine to Marquette. In the following year the shipment reached 119,860 tons, owing to good transportation facilities and to the remarkable speculation wliich in 1886 and 1887 led to the formation of mining companies in this district with a nominal capital exceeding $1,000,000,000. The inevitable collapse in the fall of 1887 took the savings of smaller investors and many mines were closed down, but the stronger companies weathered the storm and in spite of the speculative failure the production of ore steadily increased until 1890, when for a period of several years the shipments reflected the depressed and unstable conditions which affected the Lake Sui)erior region as a whole. In the autumn of 1885 the Milwaukee, Lake Shore and Western Rail- way (subsequently part of the Chicago and Northwestern) was finished froni the mines to Ashland. The Wisconsin Central Railway crossed the range at Penokee Gap in 1873, connecting with Ashland, and in 1887 extended a branch to the center of the district. The Duluth, South Shore and Atlantic Railway already paralleled the range on the north at the time of its dis- covery and afforded easy connection with the lake. VERMILION IRON DISTRICT (1885). J. M. Clements describes the opening of the Vermilion district, in Minnesota, as follows:'' The first mention of the occurrence of iron ore in the Vermilion district was made by J. G. Norwood, who obser^-ed it during his explorations in 1850 and published a statement concerning it in the report accompanjdng that of D. D. Owen.c The iron he observed is that which occurs near Gunfiint Lake, at the extreme ea.st end of the district, and which geologically belongs vnth the ores of the Mesabi range. In this part of the Vermilion district the ores have never been exploited to any extent and are at present of little commercial importance. Interest in what is now known as the Vermilion iron-bearing district was aroused in the .sixties by the reported occurrence of gold in the \-icinity of Vermilion Lake. There was considerable excitement tor several years and a small rush to the district. Shafts were sunk and stamp mills were erected, the machinery ha\ing been jiacked in from Duluth over the Vermilion tr^il. A town site was laid out near Pike River, at the southwest extremity of \'er- milion Lake, and some buildings were erected. In all a good deal of money was fruitlessly expended, as no gold deposits of any importance were found. a Clements, J. M., and Smyth, H. L., The Crystal Falls iron-bearing district of Michigan: Mon. V. S. Geol. Survey, vol. 30, 1899, p. 175. kClcnients, J. M., The Vermilion iron-bearing districi of .Minnesota: Mon. U. S. Geol. Survey, vol. 45, 1903, pp. 213-215. c A report of the geological survey of Wisconsin, Iowa, and Minnesota, 1852, p. 417. HISTORY OF LAKE SUPERIOR MINING. 41 Some time after this, in 1875, the first exploration for iron ore in this district was taken up by Mr. George R. Stuntz, accompanied by Mr. John Mailman, who began to prospect the iron formation and iron ore exposed on Lee Hill, southwest of the Bay of Vermilion Lake, which is now known as Stuntz Bay, named after Mr. Stuntz. The ore deposits on Soudan Hill were then discovered. In 1880 Prof. A. H. Chester examined the Vermilion Lake iron f>>rmation for private parties and Mr. Bailey Willis studied it for the Census Office. Systematic and extensive efforts were made in the late seventies and the early eighties to develop the iron ores. By this time the Minnesota Iron Company had been organized and all of the properties which at that time were known to contain ore and great stretches of country which were in the continuation of the ore range had been purchased, the company owning over 20,000 acres of land on the Vermilion range proper and in the vicinity of the good harbor on Lake Superior known now as Two Harbors. On August 1, 1884, the Duluth and Iron Range Railroad was completed from Two Harbors to Tower, near Vermilion Lake. This road was 72 miles long. At a later date it was connected with Duluth, 25 miles away. During the first year (1884) 62,124 tons of ore were shipped, some of this having come from the stock piles which had been growing during the years of development preceding the opening of the railroad. Prospectors were busy in the years prior to the opening of the railroad in prospecting the district to the east of Tower, and in 1883 outcrops of ore were found by Mr. H. R. Harvey in sec. 27, T. 63 N., R. 12 W.., near the present town of Ely. The body of iron ore indicated by these outcrops was further tested in 1885-6 and led to the opening up of the great deposits at Ely on which are now working the Chandler, Pioneer, Zenith, Sibley, and Savoy mines. During 1888 there were shipped from the Chandler mine 54,612 tons of high-grade ore. •• From this time on the development of the range was rapid and steady, as is shown by the annual increase in the shipments of ore. The Vermilion range was thus opened at about the same time as the Gogebic range, but its mines, in contrast to those of the Gogebic, were from the start in the hands of a strong company, which controlled the railroad and prevented active competition. To quote from Mussey : A comparison of the output of the two ranges by years discloses an interesting contrast between centralized control backed by adequate capital in the Vermilion district and competitive exploitation based on small undertakings and insufficient funds in the Gogebic district. The Gogebic district, which was not really opened up till 1885, in the second year following produced more than a million tons; the A'ermilion, though opened a year earlier, did not reach the million mark till 1892, when the Ciogebic produced almost three millions, only to fall off to less than half that amount the next year. Production on the Gogebic moves upward by leaps and starts, one season rising to excess, the next sinking back to deficiency; the output of the Vermilion, on the other hand, climbs with a regularity that is surprising, when one considers the variable conditions of the market in which it had to be sold.'' MESABI IRON DISTRICT (1891). ACCOUNTS OF THE DISTRICT BEFORE ITS OPENING. In penetratmg the vast wilderness north and west of the Great Lakes country, the early explorers were compelled for the most part to stick close to the waterways, for the nature of the country made travel for long distances exceedingly arduous by any other method than 'canoeing. Three of tlie canoe routes to the country northwest of Lake Superior cross the Giants or Mesabi Range* and its eastward continuation. Mississippi River and its tributaries, Prairie and Swan rivers, touch the western portion of the district. Embarrass Lake, tributary to St. Louis River, and thence to Lake Superior and the St. LawTcnce, crosses the Giants Range near its east-central portion. Gunfiint Lake, one of a chain of lakes tributary to Rainy River and Nelson River and thence to Hudson Bay, lies far to the east, on a continuation of what is now known as the Mesabi district. Hence the first published references to the Mesabi district concern the parts of the district immediately adjacent to these canoe routes. Brief descriptions of Pokegaana Falls on Mississippi River and of adjacent areas were made by Z. M. Pike in 1810, by James Allen and Henry R. Schoolcraft in 1832, and by J. N. Nicollet in 1841. In 1841 also Nicollet published his map of the hydrograpliic basin of the upper Mississippi, on which the Giants or Mesabi Range, called "Missabay Heights," was for the first time delineated, a Mussey, H. R., Combination in the mining industry: Studies in history, economics, and public law, Columbia Univ., vol. 23, No. 3, pp. 90-91. t The name "Mesabi" has been variously spelled and applied with various limits to the ridges of this district, and the use of the same term to denote the iron-l>earing district as such has added to the confusion. The spelling "Mesabi" has been adopted by the United States Geo- graphic Board. It has become usual, for the sake of clearness, to speak of the main topographic feature as the Giants Range. In this report the terms are definitely distinguished, Mesabi range being applied only to the iron-bearing district that occupies a linear belt of low sloping land at the base of the Giants Range. 42 GEOLOGY OF THE LAKE SUPERIOR REGION. by hachures, although very imperfectly. In 18.52 J. G. Norwood reported the occurrence of iron-bearinj^ rocks at Gunflint Lake and mentioned fjranite and gneiss seen in crossing the range at Embarrass Lake. In 1866 Charles 'V\liittles<^y reported on explorations made m northern Minnesota durmg the years 1848, 1859, and 1864. He mentioned Pokegama Falls and made vague reference to the granitic rocks of the range. "Mesal)i Range" was used in an indefinite way to cover what are now known as the Giants and Vermilion ranges. In 1866, also, Henry H. Fames, the first state geologist of Minnesota, reported granite and gneiss seen on a trip across the range at Emliarrass Lake. In describing the ranges of the northern part of the State, including the "Missabi Wasju," he stated that they appear to be traversed by metal- bearing veins. Presumably, however, this statement refers mainly to the Vermilion range. In a second report, published the same year, Mr. Fames is more explicit, and, referring to the general elevated area of the northern part of the State, including tlie Giants Range, states: "In this region are found also immense bodies of the ores of iron, both magnetic and hematitic, occurring in dikes and associated with tlie rock in which it is found; in some of these formations iron enters so largely into its composition as to affect the magnetic needle." Pokegama Falls and Prairie River Falls were visited, and at the latter place the presence of "iron ore " was noted. Tlu'se reports of Fames contain the first references to iron ore in the Mesabi district proper, although iron-bearing rocks had been noted by Norwood in 1852 at Gunflint Lake. From this time on desultory exploration work was done in certain portions of the district. It was confined for the most part to the area west of Birch Lake, in Rs. 12, 13, and 14 W., and to the vicinity of Prairie River. No published accounts of the earlier portion of this explora- tory work are to be found. The first examination of the Giants Range by a mining expert with particular reference to the occurrence of iron ore in workable deposits, noted in print, was made in 1875 by A. H. Chester, of Hamilton College, New York. Reaching the Giants Range at Embarrass Lake, he worked eastward toward Birch Lake. In his report (published in 1884) he called attention to the magnetic character of the iron m this area and to the fact that the alternating iron laj^ers are not thick or continuous. The percentage 44.68 was given as a fair average of iron in the rocks of this part of the district. In general, one gathers the impression that he was not favor- ably impressed with the economic prospects of this area. Between the time of Chester's exam- ination of the range, in 1875, and the publication of his report, in 1884, N. H. Winchell, state geologist of Minnesota, briefly noticed the Mesabi district in two of his reports. In 1879 he told of the occurrence of iron ore in R. 14 W. and published analyses. In 1881 he told of a trip from Embarrass Lake east to range 14 and noted the magnetic character of the iron-bearing formation in range 14, as well as its similarity to the formation at Gunflint Lake. Indeed, the iron-bearing formation in range 14 was called the "Gunflint beds." In 1883 Irving called the iron-bearing rock series in the Mesabi district Animikie, a term which had been applied to similar rocks at Thunder Bay and westward to Gunflint Lake, and correlated the Animikie rocks with the original Huronian rocks of the north shore of Lake Huron and with the iron-bearmg forma- tion and associated rocks of the Penokee-Gogebic iron range of Michigan and Wisconsin. From this time on the term Animikie is much used in the literature on the Mesabi range to designate the iron-bearing formation and associated rocks. In 1884, in the same volume in which Chester's report was published, N. H. WincheU discussed the age of the Mesabi rocks, assignmg them to the "Taconic," then regarded as Lower Cambrian, and, following Irving, correlated them with the iron-bearing rocks of the Penokee-Gogebic district. In the late eighties a number of other reports on the district were issued by the Minnesota Survey, but they contain no important points not noted in reports above cited. This brmgs us to the openmg of the district for minmg. OPENING AND DEVELOPMENT. Since the late sixties there had been more or less exploration, particularly along the eastern portion of the district, from Embarrass Lake to Birch Lake, and the presence of iron-bearmg rocks had been recognized and discussed in the reports mentioned above. However, not a single HISTORY OF LAKE SUPERIOR MINING. 43 deposit of iron ore of such size and character as to warrant mining had been revealed. In fact, the range had been "turned down" by many mining men who had examined it. This was largely because of the fact that they confined their attention principally to the eastern, magnetic end of the range, where exposures of the iron-bearing formation are numerous. Even up to the present time no ore has been fovmd there in quantity. Yet the impression was gradually develop- ing that iron ore in large quantity was to be found in this district, and a few prospectors were working diligently. Among the more persistent of the Mesabi range explorers were the Merritts — Lon Merritt, Alfred :\Ierritt, L. J. Merritt, C. C. Merritt, T. B. Merritt, A. R. Merritt, J. E. Merritt, and W. J. Merritt — of Duluth, Minn. Their faith in the range was the first to be rewarded. On November 16, 1890, one of their test-pit crews, in charge of J. A. Nichols, of Duluth, struck iron ore in the NW. i sec. .3, T. 58 N., R. 18 W., just north of what is now known as the Mountam Iron mine. This was followed in 1891 by the discovery of ore in the area now covered by the Biwabik and Cincinnati mines. John McCaskill, an explorer, observed iron ore clinging to the roots of an upturned tree on what is now the Biwabik property. Test pitting by the Meiritts, in charge of W. J. Merritt, led to the discovery of the Biwabik in August, 1891. The Cincinnati mine was opened the same fall. The Hale, Kanawha, and Canton mines were opened in the spring of 1892. The discovery of ore near the sites of the present towns of Virginia, Eveleth, McKinley, and Hibbing followed in rapid succession. The excitement followmg the fu'st discovery of ore at Mountain Iron was greatly augmented by each succeeding find, and in 1891 and 1892 there was the inevitable rush of explorers. Up to October, 1892, there were two railways touching the range— the Duluth and Iron Range, crossing the range at Mesaba station on its way to the Vermilion range, and the old Duluth and Winnipeg (now the Great Northern), reaching the range at Grand Rapids. Both these places were far removed from the exploring centers. Most of the explorers went through Mesaba station. Reaching this place by rail, they were compelled to travel 12 to 50 miles to the west along "tote roads" which were all but impassable. The time, money, and energy needed to conduct even modest explorations at this time can be appreciated only by those who have experienced the difficulties of inland travel in the Lake Superior region away from railways. The stories of this "totmg" period contain the usual records of misfortunes, lucky strikes, and enterprise incidental to a mining boom. The railways were not long in getting into the field. In October, 1892, two lines were put in operation. The Duluth, Missabe and Northern Railway was built to connect the Mountain Iron mine with the old Duluth and Winnipeg Railway (now the Eastern Railway of Minnesota, a part of the Great Northern system) at Stony Brook Junction, and later was extended to Duluth. Almost immediately after the connection with Mountain Iron a branch was sent out to Biwabik. About the same time the Duluth and Iron Range Railroad sent out a branch from its main line to the group of mines at Biwabik. Very soon thereafter both railways got into Virginia. Hib- bing was reached by the Duluth, Mssabe and Northern in 1893. Eveleth was reached by the Duluth and Iron Range in 1894 and by the Duluth, Missabe and Northern very soon thereafter. The Mississippi and Northern (Eastern Railway of Minnesota) about the same time projected a spur from Swan River to the Hibbing district. With the advent of railways the development of the range went on by leaps and bounds. This marvelous development has continued to the present time. The only considerable check occurred during the period of general financial depression wliich the country underwent in 1894, 1895, and 1896. Almost an untouched wilderness m 1890, the district is to-day the greatest producer of iron ore in the world. The rapidity of the development of the nuning industry of the district, carrying with it all the prosperity of the range, can not be better told than by the table of sliipments from the district (pp. 65-68). The development of the Mesabi range eastward toward the magnetic portions of the iron- bearing formation has been less satisfactory than that to the west. A small amount of ore 44 GEOLOGY OF THE LAKE SUPERIOR REGION. was opened up at the Spring mine, formerly the site of the Mailman mine, leading to the ron- struction of a spur railway, and minor discoveries not yet exploited have been reported frmii places farther east. Also certain ore deposits have been developed in the vicinity of the town of Mesaba, near the Iron Range Railway track. The last-named depo-sits mark about the eastern limit of the principal mining operations. The most noteworthy developments of tiie district in late years have been the exploration and exploitation of the ores in the western part of the district, wliich, because of their content of loose quartz grains, giving them the name "sandy taconite," were long regarded as worthless. As a result of elaborate experiments in washing tests it was found possible to utihze these ores, and mining operations are now being conducted and planned on an enormous scale. Since 1900 several to^\^as have sprung up in what was before a wilderness. The town of Coleraine, built by fiat of the OUver Iron Alining Company, is an example of what may be accomphshed in a short time by large capital intelligently expended by a single group of indi\nduals working on a uniform plan. The railways have followed up and made possible much of the development of the western Mesabi district. It is reached by spurs from both the Duluth, ilissabe and Northern and the Great Northern railways, leaving the main lines south of Hibbing. , Still more recent has been the extension of the district by exploration for 12 miles or more west of Pokegama Lake, near Mississippi River. The ores have been found to be lean but probabty merchantable. The iron-bearing formation pinches out at the southwest end of the district, the overl^ang slate coming into contact with the underlying quartzite. This part of the district, together wdth magnetic belts farther west, particularly the one running through Leech Lake, the east end of which comes within 12 miles of the Mesabi district, affords interesting possibiUties for exploration, which will be adequately undertaken. CUTUNA IRON DISTRICT (1903). The development of the Cuyuna, the newest of the Lake Superior iron districts, in the same geologic group as the Mesabi district, is unique in a way. The other iron ranges of the Lake Superior region were all discovered through more or less conspicuous surface indications of ore bodies. Outcrops of the ore or of iron-bearing rocks existed. There are no rock out- crops in the Cuyuna district, the drift mantle being SO to 3-50 feet thick, and the first tliscovery of magnetic iron-bearing rocks in tliis region was made with the dip needle by Cuyler Adams, about 1895. The dip needle was the sole factor used in the subsecpient tracing of the ore formations b}" Cuyler Adams and afterward by others, preparatory to drilling, from the time of the first discovery of magnetic iron-bearing formations until 1907, when more or less indiscriminate drilling began. The first drilhng was done hi 1903 at a point just south of Deerwood, ^liim., by Cuyler Adams, and has continued in greatly increasing amount to the present time, some 2,000 drill holes and two shafts having been put down, resulting in the discovery of a number of ore deposits. (See pp. 216-219.) The distribution of the ore bodies and the limits of the district are yet very imperfectly known. Extension of magnetic surveys to the west and north have shown isolated magnetic belts at several places, some of them beyond the western boundary of Minnesota. The ilistribution of some of these belts is showm on the general map. Underground exploration of these belts has just begun. The next few years wiU see rapid exploration of the Cuyuna range and the coun- try to the north and west. For some time before the drilling began, geologists had suspected the existence of iron- bearing formation in the CuAnma district. The general geologic map of Minnesota, published by the Minnesota Cieological and Natural History Survey in 1901, showed this area as occupied by a .southwestern extension of the slates and cjuartzites of the Mesabi district. In 1903 C. K. Leith published a sketch showing tlie hypothetical extension to the southwest of the iron- HISTORY OF LAKE SUPERIOR MINING. 45 bearing formation of the Mesabi district through the since chscovered Cuyuna district. A similar view of the geologic possibilities was held by W. N. Merriam, geologist for the United States Steel Corporation. The Northern Pacific Railway extends tliroughout the length of the Cuyuna district and affords easy access to the ores. It also runs near some of the magnetic belts west and north- west of the Cuyuna district. The Minneapolis, St. Paul and Sault Ste. Marie Railway passes the district on the southeast and in 1910 completed a spur into the district. For both rail- ways the lake port ^dll be Superior.' BARABOO IRON DISTRICT (190.3). The discovery of ore in the outlying and relativel}'- small Baraboo district, in Wisconsin, ■was not made until 190.3. The quartzite ranges here conspicuously exposed had long been recognized as Iluronian, and suggestion had been made that iron-bearing rocks might be asso- ciated with them. In fact, for several years the Cliicago and Northwestern Railway had quar- ried small amounts of paint rock' within a few feet of what is now known as the Illinois mine. Because of the covering of Cambrian sandstone and glacial deposits the ore deposits them- selves escaped detection until drilhng was, in 1900, begun by W. G. La Rue in the vicinity of the Illinois mine near North Freedom. Since that time, as a result of almost uninterrupted exploration, ore deposits have been found at various places in the Baraboo syncline. Only three shafts have been sunk and ore has been slupped from only one, the Illinois mine. The development of the district has not been rapid because of the relatively low grade of ore, the considerable cost of mining, and the great expense of deep drilling, although these factors have been partly offset by lower freight rates to Cliicago. Both mining and exploration in the Baraboo district are in their infancy. LESS IIMPORTANT DEVELOPMENTS. CLINTON IRON ORES OF DODGE COUNTY, WIS. (1849). There is no record of the fu-st discovery of the Clinton iron ores in Dodge County, Wis., for they are exposed at the surface in accessible country. Ore was first mined from them in 1849. The ores have been partly used in local charcoal furnaces at Mayville and Iron Ridge and partly sliipped to ^Milwaukee, Cliicago, and adjacent points. Because of their low percentage of iron, liigh phosphorus, and moderate quantity, they have not figured largely in Lake Superior production. PALEOZOIC IRON ORES IN WESTERN WISCONSIN (1857). Small hematite deposits scattered through the driftless portion of the Cambrian sandstone area north of Wisconsin River, in Wisconsin, were opened up about the time of the discovery of the Marquette district. In 1857 a charcoal furnace was built at Ironton, in Sauk County, to use these ores. Another was built at Cazenovia, in Richland County, in 1876, and torn down in 1879. None of these ores has been mined since 1880. Records of production are not avail- able, but before 1873 about 25,000 tons of ore was mined from these deposits. Farther north, at Spring Valley, in Pierce County, Wis., brown-ore deposits dissociated ■wdth Ordovician limestone were opened about 1890, and a charcoal furnace was built to use these ores in 1893. At a later period coke supplanted charcoal as a fuel. IRON ORES OF THE NORTH SHORE OF LAKE SUPERIOR (1900). Since the opening of the Lake Superior region for mming the north shore has been more or less explored and a considerable number of iron-bearmg belts have been located in the territorv extending from the Lake of the Woods beyond Michipicoten. Only three ore bodies have been found. The best knowTi of these is the Helen ore body, which was discovered in 1897 ui the 46 GEOLOGY OF THE LAKE SUPERIOR REGION. Michipicoten district, on tlie northeast side of Lake Superior. This district was cortnected with Lake Superior l)y tlie buildmjx of the Algoma Central and Hudson Bay Railway (12 miles) in 1899 luul bof^an shii)nieMt in 1900. Discovery of the Helen mine led to rather vigorous exploration in tne many kno\ui iron- bearing belts in the immediatel,v adjacent territory, m some places by drilling, but without conspicuous success. A small body of ore was found iit the Josephine mine, a few miles north- east of tlie Helen mine. The Atikokan ore furiuues, but these lake ports are outside of the region. In 1908" there were in operation a coke furnace at JXduth, Minn., tliree in Wisconsin outside of Milwaukee, five char- coal furnaces in the Upper Peninsula of ^Michigan, three furnaces in the northern part of the Lower J'eninsula of Michigan, and the steel plant at Sault Ste. Marie, Ontario. The largest plant yet projected for the local use of iron is to be buUt for steel making in West Duluth by the L'nited States Steel Corporation; it may be in operation in 1912. In recent years there has been an attempt to recover by-products from the charcoal burned, the first notable project being the Cleveland-Cliffs furnace at Presque Isle, in the Marf(uette district. This plant is most elaborately eciuijiped for the recover}-, as by-products, of wood alcohol and creosote. The Lake Superior Iron and Chemical Company, at Ashland, Wis., also has a well-equipped by-product plant. The Zenith furnace at Duluth has been rebuilt on a large scale to recover by-products from coke. At present it is supplying gas to the city of Duluth. The steel plant now planned at Duluth by the United States Steel Corporation will utUize the gases as fuel. With increase of population directly tributary to the Great Lakes it is very likely that the local smelting of the ores will increase. The depletion of the timber \vill prubabl}- compel mcreased use of coke instead of charcoal. Peat, which is found locally in large quantities, may be considered as a possible fuel for the future. INFLUENCE OF PHYSIOGRAPHY ON INDUSTRIAL, DEVELOPMENT. One of the principal relations between the physiography and history of the industrial Lake Sui)erior region seems sufficiently distinct to be summarized in a few words. The early stages of development were closely controlled by conditions of accessibility. The early explor- ers, traders, and prospectors were confined to the lake and river shores and to country easily accessible from them. Wlien mining and lumbering began there was also a distinct localization of these mdustries in accessi])le jjlaces. With the growth of theindustiy and the introduction of railways the influence of physiography on the local distribution of activity gradually became less marked, until at present this distribution is but little affected by the configuration of the surface and tlrainage. The situation of ore deposits has of course localized the mining development. Favorable conditions of access, though advantage has been taken of them, have been subordinate factore. An iron deposit would be utilized whether it was in a swamp or on a mountain, whether easily accessible or not. In other words, increased demand for the raw materials of the Lake Superior region, due to general commercial conditions and the westward movement and increase of population, has gradually overridden and more or less obliterated the natural phj'siographic channels of development. The relation of the Great Lakes to cheapness of water transportation and of the simple topography of the region to ease of railway construction to any mineral-proilucing district continues to be an important physiographic influence and one that is unusual in a mining district. <• Map of the United States showing location of lilast furnaces in 19Wf, compiled by W. T. Thom from Swank's Iron and steel works directory tor 1908: Mineral liesources U. S. for 1908, pt. 1, V. S. Geol. Survey, 1910, PI. II. HISTORY OF LAKE SUPERIOR MINING. 49 PRODUCTION OF IRON ORE. The production of iron ore from the several producing ranges of the Lake Superior region since their opening is given in the following table, compiled principally from the Iron Trade Review. The figures refer mainly to shipments rather than to production. The figures of the United States Geological Survey do not go back far enough for the purposes of this table. The facts of the table are graphically expressed in figure 3. 45 ,000,000 40,000,000 35 ,000,000 30,000,000 (0 O 25,000,000 (D Z o -■ 20,000,000 15,000,000 5 ,000,000 1550 1855 1660 1865 1870 1875 1880 I6S5 1890 YEAR5 FiGUKE 3.— Diagram showing annual production of Iron ore in Lake Superior region since the opening of the region. Table of Lake Superior iron-ore shipments from the earliest shipment to date.<^ Gogebic Range. [Gross tons.] Name of mine. lSS-1. 1SS5. 1886. 1887. 1888. 1889. 1890. 1891. 1892. Ada. (Included in Ironton.) Anvil 10,075 175,563 1.369 159. 252 16, 101 1,799 21,721 29,763 24,676 174, 183 47,000 257,915 45.690 435,949 73 267,439 42,090 6.741 74,015 231,896 Atlantic 5,422 94,553 4,788 179,937 199, 865 246,695 83,554 319,482 8,880 2,697 40,639 53, 267 56,542 80,486 152,878 46,574 131,896 130,833 119,676 CastiJe Colby c 1.022 84,302 257,432 258,518 285,880 136,833 193,038 1,497 23,794 21, 150 9,619 69,968 21,754 13,907 6,778 10.055 8,515 1,997 Geneva o Figures for 1893-1909, inclusive, from Supplement to the Iron Trade Review, vol. 46, No. 9. March 3,1910. Figures for previous years compiled from the annual tables published by the Iron Trade Review and from "Annual review of the iron mining and other industries of the Upper Penin- sula for the year ending December, 1880," by A. P. Swineford. b Under Norrie group after 1904. c Includes Tildeu prior to 1891. 47517°— VOL 52—11 4 50 GEOLOGY OF THE LAKE SUPERIOR REGION. Table of Lake Superior iron-ore shipments from the earliest shipment to dale — Continued. Oogeblc Range— Continued. [Gross tons. J Name of mine. 18S4. 1885. 1886. 1887. 1888. 1889. 1890. 1891. 1892. 5,468 19,734 61,714 53,918 28,721 30,000 103,169 76,545 51,551 52,000 63,903 110,368 22,383 15,759 1,506 4,283 Iinju'iial. (See Federal.) 161,635 9,950 551 18,424 2,249 Iron King. (See Newport.) 24,762 8,635 6,247 300 3,944 18,497 52, 179 1,228 2,882 10,144 64,779 Mik'iiio Montreal f Section 331 23,013 20,184 4,105 124,844 13, 714 17,979 43,989 75,660 23,217 237,254 30, 475 19,906 1,414 38,015 69,145 1,313 412. 196 5,412 49,976 9,725 26,087 116,094 36,987 143,691 71,488 108,684 105,606 73,409 New Davis. (See Davis.) 165,962 15,419 674,394 13,354 116,376 35,245 574 906,728 1,005 172,060 50,004 758,572 985,216 6,711 Pabst *> 1,103 130,226 32,227 113,245 102,382 Pike 16,388 45,000 3,058 9,472 11,694 913 Section 33. (See Montreal.) | 2,912 1,405 10,963 18, 137 6,010 64,902 28.41S 56,046 Tilden c 233,356 10,780 12,764 2,387 10,683 1,878 Vaughn. (See Aurora.) 14,576 37,210 97 53,242 Wisconsin. (See Davis.) Yflle ^\Vc<;t Colbvl 1,022 119,860 753,369 1,324,878 1,437,096 2,008,394 2,847,810 1,839,574 2,971,991 Name of mine. 1893. 1894. 1895. 1896. 1897. 1898. 1899. 1900. 1901. Ada. (Included in Ironton.) 13,297 83,020 68,064 126,096 70, 989 245,883 57,483 91, 149 60,727 187, 169 5,037 123,208 38,058 133,076 1.101 66,067 111,625 50,307 166, 122 154, 615 19,964 170,309 232.961 135,955 193,111 286.399 190. 13.5 179,028 203, 152 223.747 Brotherton 18,905 28,578 47,148 47, 156 40,567 52,349 50,490 38,821 46,186 37,308 73. 198 43,162 78,858 62,524 89,804 125,496 103.109 179,374 Castile 504 48.492 633 32.572 3,569 59,346 15,210 31,385 32,616 22,921 152,875 103,239 5,029 23,475 10,253 26,105 18,329 4,544 7,964 1,015 1,255 986 7,728 54,664 10,358 21,475 Imperial. (See Federal.) Iron Ilelt . 23,976 45,109 148,228 81,351 96,763 58,418 105,934 43,883 Trnii Chief No 2 Iron King. (See Newport.) 7,977 25,047 33.893 1,651 1,265 19,988 9,604 11,782 332 10,324 153,307 263,711 7.844 1.090 107,524 217,201 34,140 Mikado 4,788 138,882 157,821 11.397 191, 106 150,979 91.846 Montreal (Section 33) New Davis. (See Davis.) 34,299 109,718 46,037 150,392 131,531 142,369 270,776 196,953 72,945 190,448 472,062 3,930 104.510 2,058 621,608 2,4.37 206,074 37,911 738,480 329,068 604,281 700,990 714,069 666,389 660,965 219.960 46,905 68,984 114,108 13,185 220.496 207, 153 120 223.891 175,925 263.869 154,705 239,242 139,658 198. 6S6 7.603 Pike 3.434 6,346 1 21.788 Section 33. (See Montreal.) 1 12,196 10, 102 15,691 11,819 r, 1 1,950 20.970 418,188 22,876 135, 118 34.323 209,077 89.441 250,205 45,815 270,890 i2,526 500,830 74.097 481.909 89.997 Tildcnc 287,203 446,670 o Includes Aurora after 1904 and Pabst afti^r 1901. ftUniier Norrio ^roup after 1901. eUnder Colby prior to 1891. d linder Norrie group after 1904. ' Includes Tilden prior to 1891. HISTORY OF LAKE SUPERIOR MINING. 51 Table of Lake Superior iron-ore shipments from the earliest shipment to date — Continued. Gogebic Range — Continued. [Gross tons-l Name of mine. 1894. 1895. 1896. 1897. 1899. 1900. 1901. Trimble Tylers Forks Upson Valley Vaiigim. (See Aurora.) Windsor (now Gary) Wisconsin. (See Davis.) Yale (West Colby) 2,474 11,438 1,329,385 1,809,468 2,547,976 488 2,875,295 841 12.836 2,938,155 Name of mine. Ada. (Included in Ironton.) Anvil Ashland Atlantic Aurora « Bessemer Blue Jacket ' Brotherton Cary (and Superior) Castile Chicago , Colbyi) Davis ( W'isconsin) Eureka Federal First National Geneva Germania ( Harmony) Hennepin ". Imperial. (See Federal.) Iron Belt Iron Chief Iron Chief No. 2 Iron King. (See Newport.) Ironton Jack Pot Kakagon (now Cary) Meteor (Comet) Mikado Montreal (Section 33) New Davis. (See Davis.) Newport Nimikon (now Cary) Norrie group f Ottawa (Odanah) Pabstd Palms Pence Pike Puritan (Ruby) Section 33. (See Montreal.) Shores Sparta Sunday Lake Tilden'e Trimble Tylers Forks Upson Valley Vaughn. (See Aurora.) Windsor (now Cary ) Wisconsin. (See Davis.) Yale (West Colby) 1902. 135, 502 301,824 190, 213 402,981 63,256 136,896 44,625 22,526 31,. WO 20, 502 36,383 79,121 8,555 102 19,117 98.834 136,354 141,571 1,080,032 26, 141 32,113 "'6.' 343 144,630 468, 672 11,065 26,043 11,309 274, 138 148, 385 356, 365 94,986 89,221 22,965 54,915 734 7,108 2,240 862 26,353 16,875 31,709 6, 150 108, 709 93, 139 790,346 87,929 60,800 115 91.383 211,534 46,211 45, 595 344, 102 77, 224 212,920 84,870 01,860 81,141 11,225 23,364 23,197 6,638 59, 587 26,611 163,021 618, 638 30, 420 53,718 50, 625 204,681 1905. 82, 118 409, 131 208,039 137,351 146, 414 S3. 736 3,160 2,973 2,589 140, 740 107,854 1, 627, 128 21,980 13,963 'iiiiei 79, 209 188, 104 1900. 79, 493 341,841 97, 689 147.281 216. 992 2,108 113,001 9.436 6,768 3,227 106, 168 154, 043 139,202 1,245,!>97 57,219 5,622 'i7,'934 86,879 169,697 56,667 3,664,929 2,912,708 2,398,287 3,705.207 3.643,514 3.037,102 2.699.866 4.088,067 1907. 39, 496 298,056 91,759 104.224 209,407 6, 157 17,. 347 190. 9(« 163,891 169, 763 1,109,085 46,424 24,922 101,899 312,490 38,010 1909. 35, 937 2.59,611 41.4I!6 22,927 269. 612 124,846 96, 776 96,3.18 103,090 224,251 26, 982 68.305 'i22,'324 170,095 'ii5,'6()2" 2,508 152 44,560 277,694 80,617 177.006 99, 195 191,611 773,243 33,893 977, 054 100, 223 22, 174 111,130 111.184 93.712 154. 506 14,874 71,4.58 Total. 706, 962 5,. 387, 166 1,547,123 3,961,683 20,889 1,799 1,762,498 2,289,618 36, 247 68,727 2.450,347 103, 961 462,134 36,443 1,997 7. 108 422,239 259,733 1,186,602 12. 199 551 848,986- 99,090 71,904 216,307 997, 0&5 2,861.252 5,845,039 28,035 17, 744, 658 481,359 2,360,583 1,284,489 40, 566 98, 732 109,572 55.808. 4.862 1,306,975 5. 088. 635 25. 931 10.683 11,375 1,878 148,905 373, 173 60, 896, 457 Marquette Range. Name of mine. Years un- known. 1854. 1855. 1856. 1857. 1868. 1859. 1800. 1801. 1862. Bessemer. (See Lillie.) Beaufort ( Ohio ) a Under Norrie group after 1904. t> Includes Tilden prior to 1891. c Includes Aurora after 1904 and Pabst after 1901. d Under Norrie group after 1901. f Under Colbv prior to 1891. / Under Iron Clifis, 1890-1895; under Cleveland-Cliffs group after 1895. 52 GEOLOGY OF THE LAKE SUPEKIOIt REGION. Tabic of Lake Superior iron-ore shipments from the earliest shipment to date — Continued. Marquette Range— Continued. [Gross tons.] Name of mine. Years un- known. 1854. 1855. 185C. 1857. 1858. 1859. 1860. 1861. 1802. Blue. (See Queen group.) T> . j/Mitohell -^,. Braastad|\^.i,„l,^„p;;;;;;;;--_'";;;;;;; Breitung Uematite No. 2 ■^.. BulTaloi Cambria ... . Cheshire. (See Princeton.) Chester. (See Rolling Mill.) Cleveland &. . . 3.000 1,44& 6,343 13,204 7,909 15, 787 40,091 11,793 40 364 Cleveland Hematite. (Included under Cleveland.) Curry Dalliba (Phenix) Detroit Dexter Dey East New York Edison Edwards. (See Samson.) Erie Etna Fitch Fosterd Foxdale Gibson Goodrich . . Grand Rapids ( Davis) Green Bay. (See Bay State.) Hartford Home (P. and L. S.) (now Volunteer),.. Tmpprial Indiana. (See Bay State.) Iron Cliffs « 30,000 12,442 10,309 28,377 41.295 . i2,9i9 46,096 Keystone. (See East Champion.) Lake Angeline Lake Superior, 4,658 24,668 33,015 25,195 37,709 Llllie Maas Manganese (Negaunee) . . . Mary Charlotte Mesahi's Friend Miphic;iTnTnp e Miller Milwaukee Mitchell Moore Negaunee New York (York). . North Champion. (See Hortense.) ' Nortnwest Ogden 1 Palmer (Cascade). (See Volunteer ) Pioneer i Pittsburg and Lake Angeline. (See under Lake Angeline.) Piatt Portland Prince of Wales <» Queeno a Under Queen ^roup after 1890. 6 UndcT ricvclaii.l-ClilTs group after \HXi. c Iiicliidos Clovrhmd nfti'r 1S.S3; incUidos nanuim. Foster. Iron Clifls, Uichigamme, and Salisbiuy after 1895. drndcr Iron CliiTs, 1S91-1S;W: under rieveland-Clifls group alter 1S95. < Under Clcveland-riills group after 1895. / Under Winthrop after 1892. HISTORY OF LAKE SUPERIOR MINING. 53 Table of Lake Superior iron-ore shipments from the earliest shipment to date — Coiitiuued. Marquette Range — Continued. [Gross tons.] Name of mine. Years un- known. 1S54. 1855. 1856. 1857. 1858. 1859. 18l». 1861. 1862. Queen group o Republic Richards Riverside Roiling MiU Saginaw Sam Mitchell. (See Mitchell.) Schadt Section 12 Smith. (See Prmcetou.) South Buffalo c Spurr Star West (Wheat) ' St. Lawrence. (See Nonpareil.) Sterling. (See American.) Taylor Teal Lake. (See Cambria. ) Titan Washington ^ Webster < West Republic Wheeling Wheat. (See Star West.) 30,000 3,000 1,449 6,343 25,646 22,876 08,83? 114.401 49,909 124, 169 Name of mine. 1863. 1864. 1866. 1866. 1867. 1868. 1869. 1870. 1871. 1872. American (Sterling) Austin Bamum « 14, 385 33, 484 44,793 45, 939 38 381 Bay State Bessemer. (See Lillie.) Bessie Beaufort ( Ohio) Blue. (See Queen group.) T> *- J (Mitchell 197 BraastadVi„jjj^„p 3, 409 11.088 14,239 Breitung llematite No. 2 Butfalo c Cambria 6,255 21,635 73,161 67,588 68,408 Cheshire. (See Princeton.) Chester. (See RolUng Mill.) Cleveland / 40, S42 . - 44, 959 33, 355 42,680 75,864 102, 112 100, 133 132,884 142,058 151, 724 Cleveland llematite. (Included under Cleveland.) roliiTTiIiia. (Klmnani ClUTV Dalliba ( Phenix) Detroit Dexter Dey East Champion East New York Edison Edwards. (See Samson). Empire Erie -■> Etna Fitch 6,000 14,540 23, 458 13,532 18,684 Foxdaie . Goodrich Oreen Bay. (See Bay State.) Hartford "Includes Buffalo, Prince of Wales, Queen, and South Buffalo after 1890. 6 Under Iron Cliffs, 1891-1895; under Cleveland-Cliffs group after 1895. c Under Queen group after 1890. d Prior to 1890. see Braastad: includes Marquette after 1892. « Under Iron Cliffs, 1S:10-1S95; imdor Cleveland-CUas group after 1895. / Under Cleveland-ClilTs group after 18.83. s Includes Cleveland after 1SS3; includes Bamum, Foster, Iron Cliffs, Michigamme, and Salisbury after 1895. 54 GEOLOGY OF THE LAKE SUPERIOR REGION. Table of Lake Superior iron-ore shipments from the earliest shipment to date — Continued. Marquette Range— Continued. [Gross tons. J Name of mine. 1803. 18G4. 1865. 1860. 1867. 1868. 1869. 1870. 1871. 1872. 1,160 4,782 15,150 25,440 35,757 58,402 79,702 48,725 38,841 Indiana. (See Bay State.) 77,237 83,905 19,500 86,763 65,505 20,151 50,201 92,287 24,073 08,002 127, 491 46,607 119,935 130,524 27,051 105,745 125,908 35,432 131,343 127,642 .53, 407 100,582 132,297 33,645 158,047 119,910 Keystone, (See East Champion.) 35,221 78,970 185,070 Lillie 4,866 15,942 24,153 Marv Chirlotte 141 8,000 12,214 33,761 43,302 43,665 71,456 94,S09 1,809 70,381 2,921 68,950 9,925 North Champion. (See Hortense.) . Pendill Palmer (Cascade). (See Volunteer.) Pittsburg and Lake Angeline. (See under Lake Angeline.) Piatt Portland 13,445 1 ■ 11,025 Rolling Mill 236 6,772 18,503 Salisbury e ... 545 SamMitcheU. (See MitcheU.) 2,843 4,928 17,360 19, 151 24,232 26,437 28,380 Schadt Section 12 Smith. (See Princeton.) South Buffalo c Star West ('^Tieat) Sterling. (See American.) Teal Lake. (See Cambria.) 4,171 39,495 West Republic Wiiithrop / Wheat. (See Star West.) 203,055 243,127 186,208 278,796 443,567 491,454 617,444 830,934 779,607 893,169 a Under Cleveland-riiffs group after 1895. bUnder Winthrop after IS92. c Under Queen group after 1890. d Includes Uullalo. I'rince of Wales, Queen, and South Buffalo after 1890. « Under Iron Cliffs, 1891-1895: under Cleveland-Clifls group after 1895. / Prior to 1890, see Braastad; includes Marquette after 1892. HISTORY OF LAKE SUPERIOR MINING. 55 Table of Lake Superior iron-ore shipments from the earliest shipment to date — Continued. Marquette Range— Continued. [Gross tons.] Name of mine. 1873. 1874. 1875. 1876. 1877. 1878. 1879. 1880. 1881. 1882. 797 4,702 8,006 Amps 48,076 41,403 43,209 37,6,32 8,583 37,909 26,680 24.015 3,336 24.522 2,208 27,883 583 41,778 1,236 Bay State... Bessemer. (See Lillie.) Bessie Beaufort (Ohio) 5,532 18,245 Blue. (See Queen group.) 6,478 13,279 45,247 14,824 21,146 43, 630 ■D * J (Mitchell. 8,658 33,456 7, .549 7,549 , 5, 696 27,236 3,898 12,549 4,259 23,740 11,131 26,595 33,396 Braastadj^^.j^j^^^p ■-;;•;;;;;;;;;■;;;;;; 7,502 23,005 Butialo & Camliria 2.610 47,097 6,329 66,002 10,083 70,883 3,754 73, 464 6,724 94,027 949 131,167 6,958 112,401 2,415 212,748 19,246 145, 427 5,531 198,569 64, .545 72,782 56,877 1.59,009 Cheshire. (See Princeton.) Chester. (See Rolling MUl.) Chicago Cleveland c... 133,265 105,858 129,881 140,393 152, 188 152,737 206, 120 Cleveland Hematite. (Included under Cleveland.) 21,065 35,088 8,059 6,663 11,158 12,066 Dalliba(Phenix)... 10,986 44,836 Detroit 5,402 Dey 10,426 5,227 3,346 7,715 14, 495 5,401 4,029 10,217 3,408 4,002 Edison Edwards. (See Samson.) Erie 2,731 Etna.. . Fitch 18, 107 4,719 847 i25 4,804 1,122 3,011 11,648 Foxdale Gibson Goodrich /6,.338 503 7,547 3,992 11,131 10,245 9,998 Green Bay. (See Bay State.) Hartford Hortense (North Champion) Home ( P. and L. S. ) (now Volunteer) . . . Himiboldt (Washington). 21,498 38,014 1.362 27,890 1,225 23,921 492 18,204 285 14,726 9,642 3,333 16, 545 20, 302 43,463 Indiana. (See Bay State.) IronClifls? Iron Moimtain Jackson 130, 131 43,933 158,078 105,600 31,526 114,074 90,568 26, 370 129,339 98, 480 22.5.39 111,766 5,945 17,276 80,340 19,112 127.349 10, 127 19,691 83,121 28, 161 109,674 8, 506 30, 180 103,219 25,321 173,938 22,380 28,962 i26,626 14,928 204,094 18,347 31,200 118,939 18,0150 262.235 16, 748 28,051 90, 830 Keystone. (See East Champion.) Lake Angeline... 14.326 296,509 Lillie 27,494 Lucy (McComber) 38,969 2,642 10,407 40,406 Maas Manganese (Negaunee) Mesabi's Friend 29,107 45,294 44,763 70,074 28,238 58,622 56,970 52,766 57,272 43, 712 Miller Milwaukee 941 13, 142 31,635 40,891 Mitchell . National . . 4,191 33,310 29,351 24,833 23,366 Negaunee Construction Works 1,177 New York (York) 70,882 6,629 77,017 70, 103 987 58,863 556 55,581 3,307 21,903 4.547 57,528 2,609 58,512 2,192 50,074 56,806 New York Hematite 2,105 North Republic 9,998 18,880 Pendill... . . ... 4,000 12, 549 3,959 13,686 9,987 Palmer (Cascade). (See Volunteer.) Pittsburg and Lake Angeline. (See under Lake Angeline.) a Under Iron Cliffs, 1890-1895; under Cleveland-Cliffs group after 1.S95. ' Under Qucon group after 1890. c Under Clevoland-Clifls group after 1883. d Includes Cleveland after 18.83; includes Barnum, Foster, Iron Cliffs, Michigamme, and Salisbiu-y after 1895. ' Under Iron Cliffs, 1S91-1.S95; under Cleveland-Cliffs group after 1895. / Includes shipments for prior years. e Under Cleveland-Cliffs group after 1895. » Under Winthrop after 1892. 56 GEOLOGY OF THE LAKE SUPERIOR REGION. Table of Lake Superior iron-ore shipments from the earliest shipment to date — Continued. Marquette Range— Continued. [Gross tons.] Name of in [nc. 1873. 1874. 1875. 1876. 1877. 1878. 1879. 1880. 1881. 1882. Piatt Portland Princeton (Swauzey or Cheshire) 9,328 187 225 8,434 16,924 17,985 13,202 15,011 31,498 105,453 122,639 119,726 120.095 165,836 176,221 135,231 235,387 233,786 235,109 Rolling Mill 11,319 37,138 11,023 38,968 16. 643 45.486 6,730 2,849 37.806 55.318 4.571 12,804 53,265 56.979 20. 510 19,330 38, 121 44.005 37.869 10, 419 30. 773 54.097 52.155 10,351 10,039 43,396 39,293 5,455 15, 172 35,059 21,457 1,668 30.793 43,690 4,584 163 16,276 42,243 Sam Mitchell. (See Mitchell.) 12,421 Schadt 5,027 330 13,243 3,287 Smith. (Sec Princeton.) 31.933 1,091 42.068 2,139 23,094 20,276 22,801 2,225 1,409 851 2,746 9,040 8,873 Star West fWheatl 3,323 9,554 St. Lawrence. (See Nonpareil.) Sterling. (See vVmerican.) 1,110 10,559 15,146 Teal Lake. (See Cambria.) 1,778 28,920 18, 198 4,071 15,324 20,211 4.704 24. 141 38,596 39,276 41,456 4,443 7,354 27,865 1,777 Wheeling Wheat. (See Star West.) 1,158,249 919,257 889,477 1,006,785 1,010,494 1,023.083 1,130,019 1,384,010 1,579,834 1,829,394 Name of mine. 1883. 1884. 1885. 1886. 1887. 1888. 1889. 1890. 1891. 1892. American (Sterling) 3,618 2,916 1,483 13,099 20,032 21,000 21,604 15,076 62.752 631 69,408 47,458 52,975 16,123 10,211 12,835 Bessemer. (See Lillie.) 847 Beaufort (Ohio) 18,976 20, 190 18,360 2,218 17. 166 17,354 12,829 Blue. (See Queen group.) 7,017 58,743 16,419 74,067 4,091 86,789 Braastad|^/j'^^{{^^;j^-p 50,143 73,144 53,913 155,341 10,860 58,784 137,593 24,686 41,130 146,330 30.801 57.861 174,680 50.919 72.780 215,098 100.464 80.359 223,442 Cambria. . 47.508 104,960 117 218.219 59,742 210, 180 50,796 173,915 34. (»2 133.413 41.549 109.978 Che^shire. (See Princeton.) Chester. (See Rolling Mill.) Cleveland TTematite. (Included under Cleveland.) 225,674 218,757 203,664 207,441 184.316 274,048 331,713 221,788 310,907 714 Curry 16,671 Dalliba ( Phenix) i.687 12.314 4,878 1.605 26,099 Detroit. 3.809 16.202 2.709 19.125 750 39,400 18.500 1,821 10,112 3,895 6,080 9,130 5.448 13.000 Bey.. 5,039 2.697 29,739 893 East New York 13.094 36,431 50,293 35,175 Edison Edwards. (See Samson.) Empire.. . .... Erie 5. 405 1.091 Fitch 16.550 21,949 15.093 Foster c... 10.029 9,675 9,643 Foxdale a Under Queen group after 1890. ftlnclu'ios liulTalo, Prince of Wales. Queen, and South Buffalo after 1896. cUndor Iron C'lilTs. 1891-1895; under Cieveland-ClitTs group after 1895. d Prior to IKOO. see Iira;istiid: includes Marquette after 1S92. e Under Iron cliiVs. Iviii is'i.i; inider Clevcland-Clitfs group after IS9.'). / Under Cli'veluiid-Clilfs croiii) afler IKti. a Includes Cleveland after 1S83; includes Bamum, Foster, Iron Cliffs, Michiganune, and Salisbury after 1895. HISTORY OF LAKE SUPERIOR MINING. 57 Table of Lake Superior iron-ore shipments from the earliest shipment to date — Continued. Marquette Range— Continued. [Gross tons.] Name of mine. 1883. 1884. 1885. 1886. 1887. 1888. 1889. 1890. 1891. 1892. 1,515 12,142 2,700 Grand Kapids (Davis) 1,200 11,611 20,058 566 7,757 26,426 9,362 22,823 Green Bay. (See Bay State.) Hartford 5.678 886 5,685 16,246 ' TTnTTibnlHt. ( Wn.shinfjr.nn) . , , 31.866 23,763 11,766 20,207 19,873 11,655 15,866 23,259 38,460 188,776 19,879 18,552 278,270 4,571 7,194 Indiana. (See Bay State.) Iron Cliffs a . . . 87,346 393 109,906 191, 120 302, 909 23,041 12, 139 78.520 134,616 289,395 .Tnrt^nn 71,278 27, 259 200, 799 4,614 14,678 83,251 86,922 204,796 2, 683 68,657 111,051 226,040 708 89,370 131,731 267, 622 3,957 101,909 223,600 240, 225 .32,692 22,276 128,891 229,070 288,784 33,916 32,982 124,682 261,680 318,321 31,812 43,483 92,979 241,605 308,831 19, 551 27,683 92,507 Keystone. (See East Champion.) Lake Angeline... 287,517 366, 715 T;illip 29.005 26,326 lU^ins 397 1,484 3,111 1,367 5,229 20, 441 7,060 70, 128 23, (,92 16,802 9,555 Minhigammeo 42,533 25,935 12,373 48,790 58,726 36,448 66,999 80,777 23, 169 1,894 Miller" 805 25,991 38,465 46,693 8,823 50. 490 8,411 48,908 540 52,727 24,763 Mitchell 21,178 13.987 Negaunee 5,259 45,304 78,318 76,488 64,218 85,846 Negaimee Construction Works 10.394 1,517 43 1.077 1.094 5,128 12,844 2, 422 11,220 New York Hematite. North Champion. (See Hortense.) 289 ll,9i;i 1,436 Northwest.. 1,687 2,200 3,553 12,605 1,594 18,249 10,072 Pendill 318 Palmer (Cascade). (See Vcflunteer.) 5,140 1,203 9,060 Pittsburg and Lake Angeline. (See un- der Lake Angeline.) Piatt .... 2,676 Portland 32, -115 Princeton (Swanzey or Cheshire) 13,730 3,557 8,328 2,842 7, .301 29, 403 491 66, 122 Queen c. . . 5,527 109.217 479. 509 191,127 379.719 Republic. . 152,565 277,757 250,835 241,161 220,624 87 1,374 235,062 21,030 287.390 22, 122 220,0(,5 3,915 167,991 5,022 402 3,712 6,783 Rolling Mill . . 1,528 9,108 17,028 15,700 1,820 946 26, 629 1,334 3,437 4,403 1,058 4,320 29, 503 51,667 1,133 48,304 74,947 4,512 72, 449 2,796 85,798 1,218 Sam Mitchell. (See Mitcheli.) .^flm^nn (Arf^lp) 600 Schadt " Smith. (See Princeton.) 4,964 24,706 69,359 146,383 Spurr 9,067 6,625 752 15,867 Star West (Wheat). 6,824 9,200 17, 538 4,987 7,997 15,141 4,412 St. Law-rence. (See Nonpareil.) Sterling. (See American.) 6,155 13, 128 19,414 Teal Lake. (See Cambria.) 19,411 11,748 23,340 5,679 13,865 24,034 16,003 47,486 2,846 56,321 60, 1.56 141,524 92,699 127, 130 934 19, 623 4,585 4,098 6,229 10,558 10,756 2,054 12,872 3,335 74 448 1,510 19, 679 30,734 2,777 12,700 5,887 6,383 9.861 2,074 Winthrop / 109,576 122,042 191,658 Wheat. (See Star West.) 1,305,425 1,558,034 1,430,422 1,627,380 1,851,634 !l, 923,727 2,642,813 2,993,664 2,512,242 2,666,856 a Under Cleveland-Cliffs group after 1895. 6 Under Winthrop after 1,S92. c Under Queen group after 1890. ^Includes Buffalo, Prince of Wales. Queen, and South Buffalo after 1890. e Under Iron Cliffs, 1891-1895; under Cleveland-Cliffs group after 1895. / Prior to 1890, see Braastad; includes Marquette after 1892. 58 GEOLOGY OF THE LAKE SUPERIOR REGION. Table of Lake Superior iron-ore shipments from the earliest shipment to daJf-Continued. Marquette Range— Continued. [Gross tons.] Name of mine. American (Sterling) Ames .\ list in ; ll;irnilin a * W.w Slate I!.',,i'iiier. (See LtUie.) lic'.iufoi't (Oliio) liluo. (See Queen group.) Boston , .(Mitchell nraastaci^^-inthrop Brc'iiiinp Hematite No. 2 liiillalo'' C'aiubria Champion • Clioshire. (See Princeton.) Chester. (See Rolling MUl.) Chiiaso Cleveland "^ -.--.'•J j'" Clevfhiii.l Hematite. (Included under Cli'velunrl.) Cleveland l-clifls group i Cohiinliia ( Kloman) Currv Dallllia (Phenlx) Detroit Dexter East Champion East New York Edison Edwards. (See Samson.) Empire Erie Etna Fitch Foster « Foxdale Gibson Goodrich Grand Rapids (Davis) Green Bay. (See Bay State.) Hartford Hortense (North Champion) Home (P. and L. S.) (now Volunteer). Humboldt ( Washington) Imperial ■ Indlaua. (See Bay State.) Iron Clifls / 1894. 1,103 30, 445 61,648 218, 105 1896. 5,195 . 47, 218 42,788 143,706 41,656 100,398 587 95,086 113,375 1897. 7,833 21,740 911 352 6,513 Iron Mountain Jackson : ■ ■ • • Keystone (See East Champion.) Lake .\ngeline Lake Superior Llllie Lucy (McComber) Maas Magnetic (stock pile) Manganese (Negaimee) Marquette Mary Charlotte Mesabi's Friend Mlchigamme / Miller Milwaukee Mitchell Moore National Negaunee Negaunee Construction Works New York (York) New York Hematite North Champion. (See Hortense.) North Kepublic Nonpareil (St. Lawrence) Northwest Norwood Ogden Pascoe PendlU Palmer Palmer (Cascade). (See Volunteer.) 130,812 51,009 351,973 329,010 68. 861 21,964 221,153 513,119 13,752 18,903 12,073 6,764 940 253,760 935 69,732 '25,'666 32,288 355,453 344, 758 78,388 1,610 132,581 ' 21,' 487 259,042 42,186 313,555 342, 439 54,285 5,503 3,214 67 1,532 '2,'297 110,648 141,728 718, 408 1,154 102,623 163, 190 869,482 1899. 124,930 215,074 1,011,048 80,710 342,251 469, 576 107,532 79, 102 489.685 376. 761 112.781 10,033 1900. 1901. 80,432 113,743 881,021 27,987 90,882 175,394 1,041 182, 169 55,012 460, 333 686, 563 211,023 11,846 23,235 88,230 464,988 682, 595 196,200 62, 321 31,714 389,128 709, 143 114,990 4,338 68,907 99,026 860,484 31,696 4,647 1902. 5,007 59,781 63,976 205,721 1,104,864 38,761 7,440 38,271 481,574 635,642 98,788 191,330 195,573 "'6,' 642' 4,648 '126,' 829 ■3,' 327 15,449 304,125 832,796 79,919 37,655 '234,Vi3 204,286 Pioneer Pittsburg and Lake .\ngeline. (See un- der Lake Angeline.) a Under Iron ClilTs. 1s;iii-1h;i.^,: under Clevelatid-Cliffs group after 1896. 6 Under Queen gruiip aflrr IS'ill. SKdSae^la;^,! ai^if Is^S; 'IJSu^S-Barnum, Foster, Iron Cli.K Michigamme, and Salisbury after IS95. « Under Iron CiiUs, ISM ls.i.5; under Clcveland-Cliils group after 189o. / Under Cieveland-Cliils group after 1895. g Under Winthrop after 1S92. HISTORY OF LAKE SUPERIOR MINING. 59 Table of Lake Superior iron-ore shipments from the earliest shipment to date — Continued. Marquette Kange— Continued. [Gross tons.] Name of mine. 1893. 1894. 1895. 1896. 1897. 1898. 1899. 1900. 1901. 1902. Piatt 5,448 41,226 13, 198 11,296 Portland Primrose 6,040 Prince of Walesa Princeton (Swanzey or Clteshire) 19,096 •■■•' 6,593 25,247 55,802 75,037 67,051 118,048 Quartz Queenu 120,673 64, 195 232, 469 105,719 204,957 1 323,057 242,293 124,342 61,022 140,312 342,978 137,085 398,298 130, 126 400,845 104, 604 418, 044 157,646 Republic Reduction Co Ricliards 6,887 4,630 1,088 24,464 4,613 51,303 54^181 50,041 Riverside 43 Rolling Mm 3,975 22,685 22,815 24,874 Saginaw Salisbury c SamMltclieU. (See MitclieU.) Samson ( A rgyle) Scliadt 1,261 Section 12 Smitli. (See Princeton.) Soutli Buffalo u Spurr Star West ( Wtieat) 5,550 51,207 9,658 942 6,716 15,987 St. LawTence. (See Nonpareil.) Sterling. (See American.) Steplienson Tavlor Teal Lalie. (See Cambria.) Titan Volunteer (see also Home) 69,561 26,946 32,672 53,216 l',617 29,983 47,578 32,736 Waslilngton Webster 20,797 West Republic Wlieeling 180,071 134,365 119,120 150,496 106,894 122,592 171,318 148,945 109 129,496 Wlieat. (See Star West.) 1,835,893 2,060,260 2,097,838 2,604,221 1 1 2,715,035 3, 125, 039 3,757,010 3,457,522 3,245,346 3,868,025 Name of mine. 1903. 1904. 1905. 1906. 1907. 1908. 1909. Total. American (Sterling) 419 13,764 23, 222 90,001 240,339 ''98 Ames Austin 195, 950 111,229 125,858 433,037 801 851 Barnum e Bay State 16,037 Bessemer. (See Liliie.) Bessie 29,718 134.648 21.S79 38,306 1.646 25. 781 78,029 01,035 72,987 Blue. (See Queen group.) Boston 02 542 Ti t. J 1 Mitcliell Braastadj^^j^^^^gp 831 445 Breitung Hematite No. 2 7,8.54 9,809 38,671 59, 667 55,849 129,673 301 583 Buflaloa 017 73U Cambria ... 41,168 74,238 84,852 174 81,791 64,680 40,628 115,007 i 35. 145 107,577 85,977 313 136,815 11,199 9 037 717 4,394,335 9,012 2, son, 298 15,239,906 94,813 111,(371 5Q 114 Cliesltire. (See Princeton.) Chester. (See Rolling Mill.) Cleveland / Cleveland Hematite. (Included under Cleveland.) 810,845 743.263 1,288,416 1.330.944 1,030,928 438,379 877,433 Columbia ( Kloman) Dalliba (Plienix) 140,841 Dexter 118 512 Dey 2,709 East Cliampion 7tt 002 East New York 22,523 7,299 33.095 Edison .. 893 Edwards. (See Samson.) Empire 40,565 53,637 108,993 203 095 Erie .- 8.136 Etna... 1,0*11 Fitch 31,817 Fosterc 171,893 Foxdale 5.053 3,429 3,303 31.447 Cibson 16,357 a Under Queen group nftor 1,890. 6 Includes [iuil'iilo. I'riiicc of Wales. Queen, and South Buffalo after 1890. cUniier Iron I'lilf^, ]v,ll-]S95; under Cleveland-CliUs group after 1895. d Prior to 1S90, see Braastad: includes Marquette after 1.S92. e Under Iron Clills, 1S90-1,S95; under Clevcland-CliUs group after 1895. /Under Cleveland-Cliffs group after I8.S:!. e Includes Cleveland after 1883; includes Barnum, Foster, Iron Clifls, Michigamme, and Salisbury after 1895. 60 GEOLOGY OF THE LAKE SUPERIOR REGION. Table of Lake Superior iron-ore ahipmcnlsfrom the larliial sldpmenl to date — Continued. Marquette Range— Continuird. [Gross tons.) Name of mine. 1903. 1904. 1905. 1906. 1907. 1908. 1909. Total. 49.754 110,736 Green Bay. (See Bay State.) Hartford ' 20, 085 179,980 322,209 364,801 328,161 278,366 250,680 1,760,951 30,574 26,022 713,961 727 1,661 6,076 55,756 48,231 115,478 376, Ol Indiana. (See Bay State.) 1,700,537 393 5,409 310,950 604,829 77,454 33, 180 374.183 727,378 9,868 5,066 269, 116 635,671 32, 781 85 61,345 283.373 674,066 80.545 11.060 280,298 349.435 61,708 1,672 159, 197 3,885,513 Keystone. (See East Champion.) 262,480 590. 339 63,209 220,410 261,955 8,632 1.115 29,030 8,285,400 14,931,563 Lillie 1,743,490 519,031 32,378 220,611 292 292 6.359 152.907 34,303 4S.8S5 221,738 257.088 155,633 99,104 240,433 1,057, IM 16.043 880,362 375, 451 MitrhplI 11,539 29.319 25,828 68.131 150.216 224,665 i45,i32 239,554 253,448 196, 170 232,219 312,217 3,61.2,127 12,708 Npw York- fYork") 1,123,071 37,587 North Champion. (See Hortense.) ^ 289 23,395 1,687 5,753 986 59,806 45,993 13, 131 14, 172 Palmer (Cascade). (See Volunteer.) 15,409 Pittsburg and Lake Angeline. (See un- der Lake Anceline.) 73,844 Portland 79,652 79,652 , 6,040 32,415 Princeton ( S wanzey or Cheshire) 84,223 76, 461 129,079 166,894 177,863 36,033 42,934 1,271,761 491 180,866 Queen group rf 254,658 155, 415 311,479 124,506 263,377 150,699 221.096 177,220 309.917 170,554 104,098 67,999 237,509 176,575 5,315,998 6,193,469 47, 174 8,261 55,593 68,134 86,129 89,563 35, 156 60,994 102,566 688,455 16, 160 Rolling Mill 6,786 28,766 49,204 52,147 133,139 578,916 451,424 686,411 Sam Mitchell. (See MVtcheli.) 267,805 Schadt 1.261 21,887 Smith. (See Trinceton.) 245.412 165.244 Star West fWheatl 204, 649 St. LawTence. (See Nonpareil.) 39,869 64,075 39,869 Sterling. (See American.) Stephenson 6,305 52,588 ■ 122,968 .32.970 Teal Lake. (See Cambria.) Titan 90, .371 7,395 71,870 100,281 38,544 10,022 1,393.175 20,625 44,716 65.341 Webster 34.905 133.077 50.870 10..i55 WinthroD f 72,433 1,759.115 Wheat. (See Star West.) 3,040,245 2,843,703 4,215,572 4,057,187 4,388,073 2,414,632 4,256,172 91,83S,55S a Under Clcveland-Clifls Rroup after 1895. tUiKlor Winl.hrop nftor isaii. c Under Queen group after 1890. d Inclnde"! Buffalo. Prince of Wales, Queen, and South Buflalo afler 1890. el'nder Iron ClilTs, 1891-1S95: under Clevelami riifTs p-oup after 1895. / Prior to 1S90, sec Braastad; mcludcs Marquette after 1S92. HISTORY OF LAKE SUPERIOR MINING. 61 Table of Lake Superior iron-ore shipments from the earliest shipment to date — Continued. Uenomlnee Range. [Gross Ions.] Xame ot mine. 1S77. 1878. 1S79. 1880. 1881. 1882. I8,s:i. 1884. 1885. Alpha Antoinc (Clitt'ord) Aragon Armenia Bauer Baltic Berkshire Beta Breen 5,812 4,796 1,463 5,359 Brier Hill 10,593 4,388 Bristol (Claire) Calumet 5,847 29,239 3,627 Caspian 34,566 134,521 247,506 265,830 290,972 Chatham Clifford : Columbia " 15,948 115,862 4,3,34 21,493 6,774 34,622 r^imninnwpnlth 9,643 30,856 97,410 11,816 42,947 Cornell Crystal Falls 1,341 Cuff Candy 12, 803 46, 168 21,851 14,368 17,534 12,644 13,374 18,287 3,676 22,675 3,410 10, 079 24,099 608 4,897 49,897 9,880 Cvclopsa 6,028 Delphic Dober 6 T)iinTi Eleanor (Appleton) Emmett 12,397 22,474 31,136 . 648 Fairbanks c 8,045 160,165 455 40,232 Florence 14, 143 100, 501 Fogarty Forest Genesee (Ethel) Gibson Great Western 587 22,826 20,710 Groveland Half and Half Hemlock Hersel Hiawatha Hilltop HoUister Hope 4, 280 29,115 4,362 100,369 636 52,684 2,739 56,693 Iron River d James Keel Ridge 11,496 19,611 23,425 6,033 Kimball Lament (Monitor) Lee Peck e Lincoln Loretta Ludington / 8,816 3,374 52,152 102,632 101,165 124, 194 Manganate Mansfield 3,477 18,677 18,187 McDonald Metropolitan 23,854 36,643 27,577 Michigan Exploration Co Millie (Hewitt).. 4,362 9,500 7,516 7,927 4,627 Monongahela Munro 2,480 29 221 7^202 114,836 5,973 37,620 10,004 71,710 11,652 Northwestern 7,276 73,519 198, 165 137,077 165,547 6,515 Penn Iron Mining Co. » Perry 3,138 Pewabic (see also Walpole) Quinnesec 25,925 41,954 52,436 43,711 44,240 21,676 16,996 14 110 Riverton (see also Dober and Iron River) h 13,465 49,196 60,406 73,648 76,514 38, 120 18,020 Selden Sheridan Shelden & Shafer (Union). (See Columbia.) South Mastodon 798 23,089 10,856 Sturgeon River Tobin Verona a Under Penn Iron Mining Co. after 1892. bUndor Riverton aficr 1900. c Included in Paint Hivrr after 1S93. d Under Riverton after 1892. e Cherry Valley ore. / Included in Chapin after 1S94. g Includes Curry. Cyclops, Norway, and Vulcan prior to 1S93. ft Includes Iron River after 1S92; includes Dober after 190U. 62 GEOLOGY OF THE LAKE SUPERIOR REGION. Table of Lake Superior iron-ore shipments from the earliest shipment to date — Continued. Menominee Range — Continued. [Gross tons.] Name of mine. 1877. 1878. 1879. 1880. 1881. 1882. 1883. 1884. 1885. V ulcan " 4,593 38,799 56,975 86,976 8o,2?4 94,042 79,874 101,722 124,125 Walpoleb 6,198 15,292 8,344 10,405 95,221 269,609 592,088 739,635 1,136,018 1,(M7,415 895,634 690,435 Name of mine. 1886. 1887. 1888. 1889. 1890. 1891. 1892. . 1893. 1884. 1 Antoine fCliffordl 1,745 50,275 46,609 26,649 96,829 167,948 127,901 138,209 ^ Baltic jj 1,585 1,226 1,400 ' Brier Hill 57,352 9,612 * Chapin (see also Ludington) 198,871 336, 128 290,871 518,990 742,843 488,749 660,052 489,134 235,895 Columbia . 14,282 51.189 4,566 2,377 56,609 2,064 10,936 61,818 11,385 108,515 60, 133 116,786 70,770 134,982 57,682 249,113 22,426 151,291 10,300 174,921 Cornel! 3,974 CuH . _ 5,376 14,693 28, 722 6,101 72, 162 7,361 100,681 10,599 125,773 1,697 37,189 17,648 14,297 2,272 24,677 118,096 151,828 156,963 162,721 133,666 4,377 58,590 5,618 24,538 8,210 79,399 142,585 196, 269 218,570 48,806 48,246 9,634 2,726 Genesee (Ethel) « 16,357 87,487 22,267 23,239 21,860 38,454 72,546 62,464 1,049 67 58,197 35,531 661 Half and Half 5,961 8,347 1,496 17,072 872 600 8,801 2,183 65,459 Hemlock 11,323 955 Hiawatha 1,683 Hilltop HoUister 2,020 1,057 1,021 15,543 Hope 2,275 5,854 78,591 83,018 110,000 179,238 155,458 59,345 1,176 5,997 3,298 Kimball 12,348 31,139 26,226 42,819 2,844 26,019 13,777 2,600 Lincoln 1,813 8,757 8,131 109 Loretta 55,983 74,454 101,653 61,883 116,297 97,355 6,844 18,303 66,526 141,303 15,777 354 Mansfield 49,836 45,370 69, 2,i9 9,150 69. 558 23,485 41,640 48,792 51,463 63,511 McDonalil Metropolitan 6,393 9,070 3,490 Michigan Exploration Co 505 77 Millie (Hewitt) 5,517 1,163 11,124 12,274 39,232 5,889 6,780 13,062 Miinro 5,400 30,460 5,744 3,441 13,200 Northwestern 93,878 13,933 95,726 10,240 87,260 12,506 68.044 32,700 61,717 62,654 4,089 45,435 44,767 18,390 Paint Uivertseealso Fairbanks). . Penn Iron Mininc Co. i 280,450 i75,274 a Under Penn Iron Mininc Co. after 1892. dlnclu'leil in I'l'wabii- uficr 1,S91. c Under Kiverion after I'.iimi. i Included in Paint River after 1893. « Includes shipments lor prior years. / Under Riverton after 1892. ff Cherry \';\IIey ore. * Included in C'hapin after I.'!94. » Includes Curry, Cyclops, Norway, and \'uk-un prior to 1S93. HISTORY OF LAKE SUPERIOR MINING. 63 Table of Lake Superior iron-ore shipments from the earliest shipment to date — Continued. Uenomlnee Range— Conlinucd. [Gross tons.] Name of mine. 1886. 1887. 1888. 1889. 1890. 1891. 1892. 1893. 1894. Peny 28,991 64, .507 115,273 165,745 304,010 13,442 6,585 2,249 Riverton (see also Dober and Iron River) a ... 12,853 790 10,834 1,302 1^684 13, 354 11,971 1,102 4,005 595 1,476 7,137 45,745 2,234 Shelden & Shafer (Union). (See Columbia.) 2,722 1,018 3,589 6,829 7,800 4,775 Vivian Vulcan b ' 143, 930 305,03ii 1,740 r29,541 900 153,900 9,614 104,996 2,940 78,967 3,895 179,904 25,635 34,418 12,699 44,400 3,705 1 880,006 1,193,343 1,191,101 1,796,754 2,282,237 1,824,619 2,377,856 1,466,197 1,137,949 Name of mine. 1895. 1896. 1897. 1898. 1899. 1900. 1901. 1902. Antoine fCllfford) . . . 27.931 183, 296 2,045 110,821 95,809 98,847 149,694 104,510 295,821 93,025 337,807 119,940 404,645 63,429 477,212 18,750 110,993 646,203 100, 864 Baker Baltic 17,326 64,664 Beta Brier Hill ... Bristol (Claire) 80,915 51,639 36,593 129,035 Chapin (see also Ludington) 618,589 420,318 643,402 724,768 940,513 929,937 929,701 966,812 Clifford 70,867 208,880 87,202 93,707 24.623 98,283 14,199 250,687 126.290 117,295 97,531 63,342 19,963 77,799 186,798 112,704 Cr v^tal Falls 13,037 44,526 95,210 128,233 147, 346 20,210 100,902 197, 770 38,209 141, 148 230,614 195,555 Cuff 3,395 41,942 76,877 178,800 183,052 Delnhic 52 5,009 49,381 '°:l^ 49,203 90,885 2, 107 47,081 31,062 2,816 Fairbanks « 22,820 35,136 37,594 93,663 74,235 36, 756 15,395 130, 798 Forest 14,455 Gibson 14,643 33,851 43,316 98,550 123,361 11,444 42,470 7,699 Half and Half 949 94,646 96,032 69,865 110,209 72,413 149,966 123,331 Hersel 1,201 11,008 6,410 20,355 2,503 74,596 3,496 i 3,373 1 . Keel Ridse i9,44i 4,900 Kimball 1 67,652 31,323 47, 267 43,622 , 64,824 72.959 61,219 19,727 54,985 7,747 53,160 34,334 64, 104 68,447 128,300 1 37,182 1 60,739 86,607 90,155 74, 113 31,181 33, 733 60 McDonald " Includes Iron River after 1S92; includes Dober after 1900. ^ Under Pfnn Iron Mining Co. after 1S92. f Included in Pewabic after 1891. d Under Riverton after 1900. e Included in Paint River after 1S93. / Under Riverton after 1892. g Cherry Valley ore. ft Included in Chapin after 1894. 64 GEOLOGY OF THE LAKE SUPERIOR REGION. Table of Lake Superior iron-ore shipments from the earliest shipment to dale — Continued. Menominee Range— Continued. [Gross tons.] Name of mine. 1895. 1896. 1897. 1898. 1899. 1900. 1901. 1902. 1 1.071 10,924 216 10,374 53,272 25,935 Rlillii- < Hewitt). 21,815 17,430 15, 194 14,922 12,133 2,397 1 1 • 1,324 1,316 197,606 10,383 290,622 179,917 237,886 223,713 229,651 358,126 273,443 Pewabic (see also Walpole) 262,551 761 273,587 279,855 305,072 530, 129 11,050 2,202 374,043 25,967 71,004 507, 786 06,383 119,860 530,291 62,531 Eiverton (see also Dober and Iron River) <^ 215,850 2,161 Selden 16,754 3,419 146 31,104 8,063 Shelden A: Shafer (Union). (See Columbia.) ■ Tobin 18,957 11,475 55,238 5,143 43,245 40,384 13 661 1,923,798 1,560,467 1,937,013 2,522,265 3,301,052 3,261,221 3,619,053 4,812,509 Name of mine. 1903. 1904. 1905. 1906. 1907. 1908. 1909. Total. Alpha 1,370 107,886 522,035 31,901 1,370 Antrjinp CClitlordl 81,164 374, 944 16,577 138,395 423,098 195, 855 431,000 27,882 100,996 441,636 36,665 l,a53,792 Aragon 226,354 246,984 5,836,279 311,608 45,003 174,426 34,295 4o,«)3 Baltic 123,236 151,114 133,246 186,495 1 189, 119 129,037 3,440 1,10s, 663 37,735 Beta 1 :::::;:;:;:::::::::;::;.:::; 4,211 16,625 21,004 I 20,366 75,425 Brier Hill I 14,981 246,581 132,420 210,388 298,031 15, 773 80,875 943,425 345,676 51,646 13S.S67 855,308 14,883 190,300 15,222 102,628 391,620 45,826 396,825 2,18.5,367 121,354 2,088 704,051 4,242 541,324 10,248 902,628 189,023 587, 647 68,730 103,626 527.971 Chapin (see also Ludington) Chatham 16,182.416 129, 439 Clifford 103, 626 27,883 8,085 942,703 5,051 1,617 6,346 50,787 2,511,784 49,302 Crystal Falls Cuff 117,096 180,983 152,255 111,871 114,158 296 986 1,735,251 58,419 iii,85i 1,410 5,512 844,889 416,928 286,093 Dclnhic 33,770 65,192 5,365 21,051 1,819 91,476 3,121 141,992 1,677 8,829 193,3% 1,521,871 18,719 66,655 8,500 95,877 153,452 233,858 169,459 178,905 7,949 140,354 32,560 231,191 77,356 2,718,019 Fogarty 117,Sll5 Forest 11,988 132,380 11,9SS 61,694 77,370 80,971 38,984 6S,,5S5 36.246 112,747 24,933 471,4.S9 4,548 124,246 9,123 57,151 101), 751 1,294 68, sis 4,737 191,265 311,218 234,492 13,913 1,872,228 Groveland 74,092 Half and Half. 7,524 96.072 79,420 136,232 124,450 106,437 117,181 83,834 112,481 1,589.818 Hersel ».t5 53,828 38,288 9,704 20 7,820 138,190 136,739 4SS.6I2 Hilltop 20, 229 6,371 10,671 25,842 46,;ts2 7,339 . 2s,.i.10 17.S71 904,,i,s7 2, .360 59,760 90,851 152.971 Keel Kidco ] 93, 101 Kimball 1 16,224 16,224 a Under Penn Iron Minfne To. aripr 1892. t> Includes Curry, ryclops, Norway, and Vulcan prior to 1893. « Includes IronRiv'i'r iificr 1n:i2: iucludes Dobor after 1900. d Included in I'ewabic alter IS'.il. e lender Riverton after 1900. / Included in Paint Kiver after 1893. e Under Hiverton after 1892. HISTORY OF LAKE SUPERIOR MINING. 65 Table of Lake Superior iron-ore shipments from the earliest shipment to rfai Manganate Mansfield 51,440 79,163 38,584 183, 532 44,633 118,713 Mastodon McDonald 1,144 Metropolitan Michigan E.vploration Co 58,088 146 36,815 39,819 18,091 603 3,322 Millie ( Hewitt) 40,860 6,913 8,739 10,887 Monongaliela Munro 32,332 9,080 92,183 91,238 47,454 91,792 46,834 53, 778 27,773 306 23,241 Nanatmo Northwestern 17,280 Norway c Paint River (see also Fairbanks).. 9,863 343,543 ii,257 141,948 11,973 423,244 28,321 496,582 75,805 381,128 Perm Iron Mining Co. t^ 176,211 428,004 4,837,348 3,138 6,917,700 502,903 1,141,098 501,985 2,092 116.299 8,203 39,350 19,404 1,394,737 130,973 405,412 1 668 654 Perrv Pewabic (see also Walpole) Quinnesec 489, 175 49,708 97,633 372,791 33 81,543 633,413 493,891 457, 796 365,341 465,453 3,147 171,200 19,994 Riverton (see also Dober and Iron River) «.. 82,611 161,704 21,017 90,358 26,080 47,073 38,069 Saginaw (Perkins) Selden Sheridan Sheldon & Shafer (Union). (See Columbia.) South Mastodon Stephenson Sturgeon River Tobin 45,386 50,910 12, 122 113,669 20,202 81,354 166,529 235,867 237, 781 161,642 359,668 Verona Vivian 90,426 122,577 48, 493 10,056 Vulcan c Walpole/ 19,089 375,385 161,425 12, 135 10,926 47,583 92,632 70,094 154,150 Youngs town Zimmprman, 1,832 10,303 3,749,567 3,074,848 4,495,451 5,109,088 4, 964, 728 2,079,156 4,875,385 71,212,121 Mesabl Range. Name of mine. 1892. 1893. 1894. 1895. 1896. 1897. 1898. 1899. 1900. 1901. 59,141 234,562 170,738 390,800 720,474 777,346 829,118 .Adriatic Agnew 14,903 64,218 41,300 Albany Alberta .\uburn 108,210 "96,' 048" 376,970 "'247,069" 131,478 '"242," .56.5' 176,263 '427,404' 235.030 "383,'i86' 385.992 '553,' 836' 263, 692 ' '924,"868' 427,510 '416,074 Bessemer Biwabik...^ . .... "isi.'soo' Brav Brunt Burt Canisteo Canton 24.416 213,853 359,020 10,261 99,498 Cass Chisholm 34, 573 26,372 17,187 57,324 32,912 246 Clark 63,071 199, 506 15,627 66, 137 7,213 22,003 60,798 80, 494 152,947 278,416 Crosby Cyprus Day 18, 651 1,975 Diamond Duluth 37.026 112,155 564 166,435 9,647 128,587 121,707 150,024 224,630 Elba Euclid Fayal 136,601 248,645 642, 939 575, 933 1,072,257 1,252,504 1 656 973 Forest Fowler Franklin 46,617 223,399 280,423 231,080 30, 128 200.400 (iO,000 168,524 39,299 Franz Genoa., 17, 136 309,514 279,077 276. 559 253,651 332,022 Gilbert o Cherry Valley ore. !> Included in Chapin after 1894. <: Under Penn Iron Mining Co. after 1892. 47517°— VOL 52—11 5 d Includes Curry, Cyclops. Norway, and Vulcan prior to 1893. ' Includes Iron River after 1892; includes Dober after 1900. / Included in Pewabic after 1891. 66 GEOLOGY OF THE LAKE SUPElilOR REGION. Table of Lake Superior iron-orr shipments from the earliest shipment to date — Continued. Uesabl Range— Continued. [Gross loiis.l Name o! mine. 1892. 1893. 1894. 1895. 1896. 1897. 1898. 1899. 1900. 1901. Glen Grant ■. Hanna Hartley Hawkins Hector (Hale) 3,616 24,167 31,004 70,006 13.728 18,«07 32,901 30 929 Higpins No. 2 Hobart HoUanil i Hull Hull-Rust Jordon Kellogg Kinnev LaBelle.. . Lake Superior group 58,123 67,659 259, 912 135,404 154,320 284,023 594,761 Larkin (Tpsoraj Laura '. Leonard.. . ' Longyear. . Mahoning.. . 117,8S4 167,245 519,892 .520. 751 750,341 28,615 911,021 65,340 705, 872 126,299 Mariska ■ Miller 13.858 2,140 Minorca Monica . Monroe Morris Mountain Iron (and Rath and Aetna) . . . 4,245 121,463 573,440 371,274 159,744 773,538 650,955 1,137,970 1,001,324 1,058,100 Pearce Penobscot. . ... 11,933 29,652 85,619 146, Ml 221,080 Pettit Pillsburv 99,691 106,487 57,847 101.032 41,905 120.723 Rol)erts". 18, 614 42, 756 Rust Sauntrv- Alpena. .. . . . 53,004 1)8, 560 328,739 Sellers 47,433 153,037 112,765 174,867 56,280 34,918 Seville Sharon ... 56,810 Sliver.... 66,722 226,156 237,143 202,144 156,426 Spring 5,628 47,700 96,280 12,215 1,621 101,675 279,515 st.ciair ;:": r St Paul Stephens 56,031 666,273 Sweenev Syracuse Troy 8,297 93,109 Utica . Virginia group 123,015 544,954 622,712 955, 739 749,499 560, »>8 293,651 417,473 5,420 Webb. . 3,046 11,249 12,357 18.238 Wills Yates 1 4,245 613.620 .793,052 2,781,587 2,882,079 1,275,809 4,613,766 B,626.3S4 ,7,809,535 9.004.890 HISTORY OF LAKE SUPERIOR MINING. 67 Table of Lake Superior iron-ore shipments from the earliest shipment to date — Continued. Mesabi Range— Continued. [Gross tons.] Name of mine. 1902. 1903. 1904. 1905. 1906. 1907. 1908. 1909. Total. Adams 1,242,923 1,109,750 940, 105 1,140,984 1,238,350 3,294 103, 2(X) 9,057 366,371 1,136,513 70, 187 149,084 765.. 592 108, li9 164, 486 1,829,372 107,317 151,536 12,585,828 288,927 923,851 207, 650 45,582 24,829 108,847 23,933 109,608 90,435 912 153,433 44,651 28,439 241,186 Albany 437,521 31,032 120,332 64,860 51, 143 35,747 368,057 1,731,036 82 175 Alberta Alexander 15,073 60,547 231, 699 38,283 2,143,028 756,853 9,121,509 Bessemer 80, 303 647,614 112,630 1,092,987 131,791 807,374 78,012 803, 750 120,350 365,781 227,767 642,821 65,514 14, 212 1,660,101 85,505 623, 127 807,511 Bray ... 65,514 269,l>i4 75,401 1,376,875 178,935 1,501,272 5,454 636 1,460,998 2,760 Burt 1,860,462 7,859,698 93,719 713,048 241,343 1,946,993 152,075 2,942,375 16,987 2,201,854 636, 176 678 192 Canton Cass 50, 155 168,831 29,554 130, 732 .59,552 231, 296 965 358,091 1,300 146,901 66,961 379, 156 1,373 274,394 36, 121 258,793 6,309 319,983 Chisholm 200,029 228,386 4,790 334,594 314,697 4,637 484,512 Clark 350,799 300,492 266,873 Commodore. . . 65,833 59,292 20, 436 34,043 249 30,131 263, 401 100,606 115,373 162,533 192,144 477,203 172, 326 227.365 349, 853 260,948 116,069 77,674 152,084 154,868 115,745 409, 148 135,366 183,470 159,038 107,685 Crosby Croxton 18,594 100,297 121,818 107, 781 348 .244,343 84,530 130,228 235,351 1,075.759 1,278,034 319,453 171 Cynrus 106,516 Diamond 171 93, 120 134,488 150,220 207,454 150,053 93,616 149,819 123,425 142, 172 125,724 158,336 255,580 149,185 147,916 150,501 224,202 82,627 1,879,357 6,304 99,892 51,393 1,737,233 Elba 1, 668, 853 Euclid 82, 627 Faval 1,919,172 1,460,601 975, 102 85,280 1,358,922 99,785 1,634,8.53 41,647 1,878,812 4.840 34.014 30.921; 907 108,610 100,178 205,426 1,439,879 2.420 21.511 8,246 18,132,550 Forest Fowler 155 417 111,085 92,019 65,528 62,884 244,150 66, 935 11,0C,8 179,468 1,712,008 145. 069 70.210 281,081 Genoa 399,719 303,700 2,985 287 Gilbert 336,927 272, 142 783, 683 396,591 1 220 788 Glen 23,875 51,946 171,705 18,928 280, 412 44,413 287, 835 49,227 279,424 1,917.410 164,514 Grant Hanna 238,873 ""3i6,'783' 30, 726 322,604 238 873 Hartley 334, 646 270.984 65, 952 173, 439 7,339 16,908 8,068 157,366 2,900.493 254,329 99,812 61,996 65,462 248, 246 390,108 1,646,523 418,336 1,111,146 8,314 270 864 Hawldns 5,892 54,289 107,905 99, 0.55 202, 070 4,990 238,598 294,588 .37.221 341,319 975 95,472 Hector (Hale) Higgins No. 2 35,286 158, 484 1,682 163,020 2, 926, 083 151,071 18.313 118,529 31,331 176,510 391,157 400. 907 Hull 233,065 282,592 1,690.311 190,971 84,715 110,708 836 043 Hull-Rust 3,039,911 162,510 10, 477 13,754 165,468 287,431 7,464 27,216 10,557,398 877 767 17,562 50, 215 61,109 . . 013 317 Jordon 147,931 190,024 97,474 185,854 925,330 32,352 6,225 89, 161 57,691 145,989 796,349 7,464 473, 668 La Btlle 70, 753 766,311 48.298 1,226,066 89,554 1,415,884 78,597 50,466 56,146 51,638 4,963,469 94 722 Larkin (Tesora) 12,001 175,670 138,001 308.989 254,308 367,192 22,040 301,522 149,410 301.368 l.W,316 297,870 14, 030 79' 313 176, 726 289,490 46,661 366,543 178,110 653, 162 6,857 303,066 53,335 79,286 200,163 10.591 279.399 81,604 105,170 3,778 228,536 151,952 153,822 221 197,192 27,207 352.004 297,011 275.777 16,778 1,277.745 Laura . . . 16,453 28,784 768 970 2,263,496 858 095 Leonard Lincoln 87,908 22,788 379, 219 2,144,263 121 391 17,706 1,564.332 82,065 137 113, .521 279,463 1.399 611.592 93,072 30,226 89,981 1,561,893 92,356 77,690 109, 086 1,038, (H5 222,640 1,009.446 11,675 706,325 66,641 1,011,661 139,853 1,274,232 115,763 12,531,132 1,044,325 Malta lOS 053 107,244 234,071 2^0 7(>5 Miller.. 118,520 224,321 525 80,330 119,439 277,119 1 133 484 16,523 900, 41)3 557 315 35,499 115,886 121,739 117,653 155,541 92,715 154,601 128, 870 119,164 210,291 7,614 147,621 1,831,187 Mohawk 7,014 628,899 7,316,409 279,396 Monroe 13,730 1,070.937 60,725 2,495.0,S9 188,508 310,839 1,809,743 64,073 2,563,111 228,451 156,809 2,076, 3,S8 34,935 1,973,519 153,770 19,172 621 71,645 528,154 1,571 206, 098 160,249 .35,571 1,617,772 49.409 1,348,714 33.012 1,168,855 Mountain Iron (and Rath and Aetna ) . . . Myers 17,198 871 193. 698 11.940 59,389 914 736 31 112 Onondat^a 30,887 90 797 54,884 50, 204 235 66,862 242,830 68,RS3 706,071 59,029 496,830 1,040,265 190,154 997,065 700,140 1 IGS Pearson 68,683 209,531 1,615 Perkins... 59,029 83,548 Pettit 17,278 238.122 28,972 52.700 229,133 27,088 140,239 161,924 82.757 33.646 30,074 489,718 57,140 59,889 Pillsbury. Rust 272,114 284,617 213,355 227,079 249,837 Scranton 1, 168 207,990 193,428 251,631 261,501 241,031 155,000 354,780 026. 169 23,585 2 870 890 Seville 23,585 68 GEOLOGY OF THE LAKE SUPERIOR REGION. Table of Lake Superior iron-ore shipments from the earliest shipment to date — Continued. Mesabl Range— Continued [Gross tons.] Name of mine. 1902. 1903. 1904. 1905. 1906. 1907. 1908. 1909. Total. 224,520 48,199 329,535 2, .328 605 Shenango. . 51,712 213,097 383,717 387,093 401,887 49,291 831.099 256,073 Sliver .305,361 Sparta . 227,444 40,458 59,692 27,777 235 1,241,197 15,257 610,457 20.510 430,633 w 35, 773 SDruce (Cloauet) . 543,203 587,153 6,148 589,319 26,748 606,295 61,792 674,602 579,903 5,166,199 94,688 137.430 st.ciair ':::::;::::::::::::: 24,230 113,200 87,0,'i5 1,014,582 367,764 1.428,614 454 819 1,4.34,681 1,652,021 1,041.500 20,984 1,142.977 137,207 516. 770 182,352 7,579 1,030,742 243,049 9,984. 191 583,. 'J92 7,579 5,509 5,509 256,384 86,520 S8, 136 87,584 174,309 146,849 20, 691 268,281 64,820 5,674 6,766 165,604 17,685 190,903 100,730 61,825 .304,864 90,090 1,015,717 158,692 113, 334 35,267 174,6.33 40,283 20,937 57, 194 21.310 661,329 853 765 15.099 91.496 156,180 12,759 489.824 103,622 9,009 399 877 Utica 120,697 185,944 201.480 113,305 1,843,450 60,966 1,3' .3, 649 Victoria 2S9,.525 Virginia Rroup 5,131 5,866 5.395 402,224 8.218 097 2*'*6 424 Webb 71,235 19,610 369 783 Williams (Nortli Cincinnati) . 97 84'' Wills 12, 158 4,550 3,440 84,614 5, .362 45, 790 20.148 Winnifred 39,179 81.686 53.179 3,415 265,289 94,867 210, 726 15,453 61,341 86, 308 84, 446 365. 102 Yates 58,174 079, 038 145. 689 13,342,840 12,892,542 12,156,008 20,158,699 23,819,029 27,495,708 17,257,3.50 28,176,281 195,703,424 Vermilion Range. Name of mine. 1884. 1885. 1886. 1887. 1888. 1889. 1890. 1891. 1892. 1 54,612 306,220 3,144 336.002 12,012 373.969 3,079 651,655 2,651 Sibley 62,124 225,484 304,396 394,252 457.341 535,318 532,000 517,570 498. 353 Zenith 14 991 62, 124 225,484 304,396 394,252 511,953 844,682 880,014 894,618 1,167,650 Name of mine. 1893. 1894. 1895. • 1896. 1897. 1898. 1899. 1900. 1901. 435.930 558.050 605.024 40.054 471.545 149.073 4.38.365 207. 103 715.919 123. 183 80?, 359 339.897 81.022 5.169 457.732 79.323 644.801 460.794 170.446 4,670 325.020 60.089 627.379 678 310 212,008 Sibley Soudan (Minnesota) 370.303 14.388 390,463 432,760 448.707 18.765 592. 196 40,817 426,040 208 284 Zenith 60 082 820,621 948,513 1,077,838 1,088,090 1,278,481 1,265,142 1.771,502 1.655,820 1.786.063 Name of mine. 1902. 1903. 1904. 1905. 1906. 1907. 1908. 1909. Total. Chandler 645,786 673,863 243.9.37 78,304 275, 168 167,205 460,548 696.736 169.616 113.595 175. 114 161.091 422.162 505.432 74.866 122.783 70.713 86,557 365,739 663,682 91,775 251, 170 205,002 109,818 318,990 766,853 106,9.33 271,496 146,503 181,580 245.684 830,700 43.320 226,8.35 102,977 236,751 50.639 477,606 82.521 127,544 53,070 50,264 9.537 378 477.226 83.167 151.009 74.862. 321,951 6.991.297 1.359.611 Siblev 1.352.575 Soudan (Minnesota) 8 2S1 752 Zenith 1.602.672 2,084,263 1,676,699 1,282,513 1.677,186 1,792,355 1,685,267 841.544 1.108.215 29.125.285 Miscellaneous (In WIsconslni. Name of mine. 1892. 1893. 1894. 1895. 1896. 1897. 1898. 1899. 1900. 1901. Illinois Maj^ille 9,044 7,925 10,511 16,472 13,144 10,546 18, 151 ig.Tsi 20.986 22.400 9,044 7,925 10, .511 16,472 13,144 10,546 18,151 19,731 20,986 22.400 HISTORY OF LAKE SUPERIOR MINING. 69 Table of Lake Superior iron-ore shipments from the earliest shipment to date — Continued. Miscellaneous (In Wisconsin)— Continued. [Gross Ions.] Name of mine. 1902. 1903. 1904. 1905. 1906. 1907. 1908. 1909. Total. Illinois 47,922 19, 558 26,562 71,413 39,978 20,610 67, 118 61,624 1.5,847 72,180 3.966 19,644 51, lOS 309,741 158,994 411,892 17,913 18,836 15,955 66,804 23,338 71,. 541 23,338 36,749 94,042 132,001 144,589 95,790 122,449 82,759 880,627 Sununaiy. Years unknown. 1854. 1855. 1856. 1857. 1858. 1859. 1860. 1861. 1862. 30,000 3,666 1,449 6,343 25,646 22,876 68,832 114, 401 49,909 124, 169 Grand total-... 30,000 3,000 1,449 6,343 25,646 22,876 68,832 114,401 49,909 124, 169 1863. 1864. 1865. 1866. 1867. 1868. 1869. 1870. 1871. 1872. 203,055 243, 127 186,208 278,796 443,567 491, 454 617,444 830,934 779,607 893, 169 203,055 243,127 186,208 278,796 443,567 491,454 617,444 830,934 779,607 893.169 1 1873. 1874. 1875. 1876. 1877. 1878. 1879. 1880. 1881. 1882. 1,158,249 919,257 889,477 1,006,785 1,010,494 10,405 1,023,083 95,221 1,130,019 269,609 1,384,010 592,086 1,579,834 739,635 1,829,394 1, 136, 018 1,158,249 919,257 889,477 1,006,785 1,020,899 1,118,304 1,399,628 1,976,096 2,319,469 2, 965, 412 1883. 1884. 1885. 1886. 1 1887. 1 1888. 1889. 1890. 1891. 1892. 1,022 1,558,034 895,634 1 119,860 430,422 690,435 1, 753, 369 627,380 880,006 1,324,878 1,851,634 1,193,343 1,437,096 1,923,727 1,191,101 2,008,394 2,642,813 1,796,754 2,847,810 2,993,664 2,282,237 1,839,574 2,512,242 1,824,619 2,971,991 1,305,425 1,047,415 2,666,856 2,277,856 4,245 62,124 225,484 304,396 394,252 511,953 844,682 880,014 894,618 1,167,650 9,044 Grand total .. . 2,352,840 2,516,814 2,466,201 3,565,151 4, 764, 107 5,063,877 7,292,643 9,003,725 7,071,053 9,097,642 1893. 1894. 1895. 1896. 1897. 1898. 1899. 1900. 1901. 1,329,385 1,835,893 1,466,197 613,020 820, 621 7,923 1,809,468 2,060,200 1,137,949 1,793,052 948,513 10,511 2, 547, 976 2,097,838 1,923,798 2.781,587 1,077,838 16,472 1,799,971 2,604,221 1,. 560, 467 2,882,079 1,088,090 13, 144 2, 258, 236 2,715,035 1,937,013 4, 275, .809 1,278,481 10,546 2, 498, 461 3,125,039 2,522,265 4,013,700 1,205,142 18,151 2, 795, 866 3,757,010 3,301,062 0,626,384 1,771,502 19,731 2, 875, 295 3,457,522 3,261,221 7,809,535 1, 655, 820 20,986 2, 938, 155 3,245,346 3,619,053 9,004,890 1,786,063 Miscellaneous (in Wisconsin)... 22,400 Grand total 6,073,641 7, 759, 753 10,445,509 9,947,972 12,475,120 14,042,824 18,271,535 19,080,379 20,615,907 1902. 1903. 1904. 1905. 1906. 1907. 1908. 1009. Total. 3,654,929 3,868,025 4,612,509 13,342,840 2,084,263 23,338 2,912,708 3,040,245 3,749,567 12,892,542 1,676,699 36, 749 2,398,287 2,843,703 3.074,848 12,156,008 1,282,513 94,042 3,705,207 4,215,572 4,495.451 20, 158, 699 1,677,186 3,643,514 4,057,187 5,109,088 23,819,029 1,792,355 3,637,102 4,388,073 4,904,728 27, 495, 708 1,685,267 2,699,850 2,414,632 2,679,156 17,257,350 .841,. 544 4,088,057 4,256,172 4,875,385 28,176,281 1,108,215 82,759 60,896,457 Marquette range 91,838,558 71,212,121 Mesabi range 19.5,703,424 29,125,285 Miscellaneous (in Wisconsin). . . 132,001 144, 589 95,790 122, 449 880,627 Grand total 27,585,904 24, 308, 510 21,849,401 34,384,116 1 38.565.762 42,266.668 ! 26.014.987 42,586.869 449,656,472 CHAPTER III. HISTORY OF GEOLOGIC WORK IN THE LAKE SUPERIOR REGION. GENERAL STATEMENT. The Lake Superior region is .among the first in which detailed study and mapping of tlie ancient crystalline complex have been extended over large^^ areas; it has had special attention })ecause of the magnitude of the mining industry and the commercial importance in mining of a correct understanding of geologic structure. Without the mines, expenditure for geologic work upon so large a scale would scarcely have been undertaken in a district so inaccessible. The increase of Ivnowledge concerning the geology of the region has closely followed the development of mming. The earlier geologic work in the Lake Superior region was of a most general nature and was necessarily confined to the shores of Lake Superior and to parts immediately accessible from canoe routes tributary to Lake Superior. The great distances and the difficulties of travel made detailed mapping impracticable over large areas in the interior. Numerous important observa- tions were made which have subsequently been found to be of value, but these were in the main fragmentary. Detailed geologic work has been for the most part confined to the ore-bearing areas and was not begun until these areas had been located or opened for mining. WORK OF INDIVIDUALS. On the Canadian shore of Lake Superior and in adjacent territory the geologic work has been of a somewhat general nature except in one or two localities. This is so largely because no ore-bearing districts have been discovered in this part of the region of sufficient commercial importance to warrant large expenditures for geologic work. The geologists who have contrib- uted most to the loiowledge of this portion of the district are Bigsby (1825, 1852, 1854), Bayfield (1829, 1845), Logan (1847, 1852, 1863), Murray (1847, 1863), Macfarlane (1866, 1868, 1869, 1879), Robert Bell (1870, 1872-1878, 1883, 1890), Selwyn (1873, 1883, 1885, 1890), G. M. Dawson (1875), Lawson (1886, 1888, 1890, 1891, 1893, 1896), H. L. Smj^th (1891), Pumpelly (1891), W. II. C. Smith (1892, 1893), Coleman (1895-1902, 1906, 1907, 1909), Willmott (1898, 1901, 1902), Van Hise (1898, 1900), Mclnnes (1899, 1902, 1903), Parlis (1898, 1902, 1903), Clements (1900), Miller (1903), W. N. Smith (1905), Burwash (1905), J. M. Bell (1905), and Moore (1907, 1909). All were in the service of the Canadian government or of the Canadian Geological Survey except Coleman, Willmott, J. M. Bell, Burwash, and Moore, who represented the Ontario Bureau of Mines, and Pumpelly, II. L. Smyth, Van Hise, Clements, and W. N. Smith, American geologists. The principal detailed mapping has been that in the Lake of the Woods and Rainy Lake district by Lawson (1886-1888), that in the Steep Rock Lake region by Pumpelly and Smyth (1891), and that in the Michipicoten iron district by Coleman, Willmott (1898), Burwash (1905), and J. y[. Bell (1905). Closely related is the extremely important work of Logan and Murray (1863) in the original Iluronian district east of Lake Superior and north of Lake Huron. In the United States portion of the Lake Superior region early general observations were made by exj)lorers sent out by the United States Government. Schoolcraft visited the south shore of Lake Superior and ascended St. Louis River (1821, 1854). Owen (1847, 1851, 1852) visited particularly the west end of Lake Superior and the upper Mississippi and its tribu- taries. Norwood (1S52) ascended Montreal and St. Louis rivers. Wiiittlesey (1852, 1S76) explored nortliern Wisconsin and northern Mijinesota. Whitne}' (1854, 1856, 1857) visited 70 HISTORY OF GEOLOGIC WOEK IN THE REGION. 71 nearly all parts of tho I^ake Superior shore. Houghton (lcS4U-lS41) made general observa- tions on the Lake Superior region as a whole. However, much the larger part of the early geologic exploration was confined to the regions now known as the Marquette iron and Keweenaw copper districts, the extension of the Keweenaw district into the Gogebic district, and adjacent parts of the Upper Peninsula. The first important detailed report on the Keweenaw copper district was that of Douglass Houghton, of the Michigan Geological Survey, in 1S41, based on work done several years before. This report led directly to the opening of the Keweenaw copper district. He was followed by Whitney (1847-1850), Foster (1848, 1850), Jackson (1849, 1S50), and Agassiz (1850, 1867). Subsequent geologic work on Keweenaw Point of great importance was that of Brooks and Pumpelly (1872, 187.3), Marvine (187.3), Rominger (1873), and others, for the Michigan Geological Survey. Field study leading to the preparation of a monograph on the copper-bearing rocks of Lake Superior was begun by Irving prior to 1880 for the Wisconsin Geological Survey and completed in 1882 for the United States Geological Survey. This volume " has remained the standard reference book on the district to the present time, though contributions of much value have been made by Hubbard, Lane, Seaman, and others. The extension of Houghton's work in the copper district and that of Burt, his assistant, led directly to the discovery and opening of the Manjuette iron-bearing district in 1848. The important early geologic work in this district was done by Burt (1850), Foster and Wliitney (1851), Kimball (1865), and Credner (1869),' all in the service of the United States Government. Later followed the important contributions of the geologists of the Michigan Geological Survey — Brooks (1873, 1876), Wright (1879, 1880), Rominger (1873, 1881), and others. Wadsworth's contributions to the geology of the Marquette and Keweenaw districts (1880, 1881, 1S84, 1890, 1891) have been the subject of much controversy. After the opening of the Keweenaw and Marquette districts geologic mapping began to be extended to the south and west through the Upper Peninsula of Michigan and northern Wisconsin. Particularly noteworthy are the reports of the Michigan Geological Survey on the general geology of the Upper Peninsula of Michigan, but particularly of the Manjuette, Menominee, and Gogebic districts, by Brooks (1873, 1876), Wright (1879, 1880), Rominger (1881, 1895), and Alexander Winchell (1888). The Menominee range in its Wisconsin extension was reported on by Wright (1880) and Brooks (1880) for the Wisconsm Survey, and Fulton (1888). The Penokee district and adjacent territory in northern Wisconsin was described by the geologists of the Wisconsin Survey —Lapham (1860), Whittlesey (1863), Irving (1874, 1877, 1880), Sweet (1876), Chamberlin (1878), and Wright (1880). Early general observations in northern Wisconsm were contributed by Percival (1856), Daniels (1858), Lapham (1860), Hall (1861-62), Irving (1872-1874, 1877, 1878, 1880, 1882, 1883), Murrish (1873), Eaton (1873), Wright (1873), Chamberlm (1877, 1878, 1880, 1882, 1883), Strong (1880), Sweet (1880, 1882), and Van Hise (1884). The detailed geologic work by the United States Geological Survey leadmg up to the prep- aration of the series of monographs on the iron-bearing districts of Michigan and Wisconsin was begun in the Gogebic district by R. D. Irving and C. R. Van Hise in 1884. On the comple- tion of work there detailed work was taken up in the Marquette district, 1888 to 1895, by Van Hise, Bayley, Merriam, Smyth, and others, and a monographic report ^ was issued in 1895; similar work was done in the Crystal Falls district from 1893 to 1898 by Van Hise, Bayley, Clements, Smyth, Merriam, and others, and a monograph'' was issued in 1899; and the Menomi- nee district was examined by Van Hise, Baylej^, Clements, Weidman, and others, and a mono- graph'* was issued in 1904. Since the completion of the work in the Menominee district in 1900 the United States Geological Survey has been devoting its attention to Mhuaesota, although a small amount of general work has been done in Michigan and Wisconsin. While the United States Geological Survey has been mapping the districts of the Upper Peninsula, the Michigan Geological Survey has given relatively less attention to this area than it had previously, . o Mon. U. S. Geol. Survey, vol. 5, 1SS3. c Idem, vol. 36, 1899, 512 pp., 53 pis. * Idem, vol. 28, 1S95, 008 pp., 35 pis., and atlas. d Idem, vol. 40, 1904, 513 pp., 43 pis. 72 GEOLOGY OF THE LAKE SUPERIOR REGION. but durinf,' this poriod it lias issued important rp])()rts on tho districts of Keweenaw Point, Por- cupine JMountains, and Isle Royal l)y Hubbard, Lane, Wri> Idem, vol. 43, 1903, 316 pp., 33 pis. HISTORY OF GEOLOGIC WORK IN THE REGION. 73 to make closer approximations to actual conditions. It also illustrates well the fact, sometimes lost sight of, that a geologic map represents an approximation to the truth, limited in its accu- racy and adequacy by the general stage of advancement of the science, and perhaps falling short of tliis limit if the map maker does not fairly represent that advance. The n'aps published with tliis monograph are closer approxmaations to the truth than the maps previously pub- hshed. These maps in turn will be superseded by better approximations as facts accumulate and geologic knowledge advances. It is hoped that the user of these maps will measure them by their advance over preexistmg maps rather than by the distance they fall short of the ideally perfect map. In the geologic literature on the Lake Superior region a progressive change may be noted from the fragmentary descriptions of earlier writers to more elaborate descriptions accompanied by attempts at stratigraphic and structural classification and the development of better prin- ciples for that purpose, and in turn a change to better understanding of the principles of corre- lation of the rocks, based on better knowledge of these rocks and of the conditions of the forma- tion of rocks of tliis kind. The work on ore deposits similarly began with fragmentary descrip- tions, followed by fuller descriptions and attempts at lithologic and structural classification, then by hj^^otheses on the origin of the ore, which gradually gave way to accepted theories based on qualitative evidence. The present monograph is believed to mark a further devel- opment in the same direction by transferring the theories of origin of the ore more largely from a qualitative to a quantitative basis. Mention of names in connection with the general tendencies outlined above would lead to endless detail, but the tendencies may be noted in terms of years and organizations. Before 1870 the geologic work was fragmentary, descriptive, and as a whole unorganized, though work of exceptional merit was done by individuals. The period from 1870 to 18S0 was marked by the better organized efforts of the Michigan and Wisconsin geological surveys, with corresponding improvements m the organization of geologic knowledge of the parts of the Lake Superior region studied, affording the first real contribution to the stratigraphic and structural geology of the region. Then the kinds of geologic work really began which are now followed in the Lake Superior region. In the early eighties the United States Geological Survey took up the study of the district, its first reports being based largely on information previously gathered by Irving and other members of the Wisconsin and other State geological surveys. Since its entrance into the region the United States Geological Survey has studied the problem more continuously than the state surveys, over a larger area, and with a uniform plan, with the result that its publications since the early eighties mark the principal steps in the advance- ment of knowledge of the region. This is said without disparagement of contemporaneous work by the Michigan, Wisconsui, Minnesota, Ontario, and Canadian surveys, which have issued reports on different phases of the problem, but for reasons mentioned above these reports for the most part have been more limited in their scope than those of the LTnitetl States Geological Survey. In recent years the Wisconsin Geological Survey has again taken up the mapping of the crystalline rocks of northern Wisconsin with thorouglmess and with good results. The Michigan Geological Survey also has now taken up work in the Upper Peninsula of Michigan, on the iron-bearing district of Iron River and on the copper-bearing series, which is rapidly advancing our knowledge. It is to be hoped that all local organizations will continue to develop. Even though they do, however, there will still be need for attention to the region b\' the United States Geological Survey, because its field of work is broader and it is in better position to take up general correlation and structural problems common to the district. BIBLIOGRAPHY. The following bibliography comprises references to literature on the geology of the region arranged first by districts and then by date. Reports on districts and mines that do not refer primarily to the geology are not here included. Also no reference is made in the following list to literature dealing with the physical geography or with the Pleistocene geology of this region. iUl references to these subjects will be found as footnotes in Chapters IV and XVI. 74 GEOLOGY OF THE LAKE SI^PERIOK REGION. MICHIGAN. Third annual report of the Geological Survey of Michigan, by Douglass Houghton, State of Michigan, House of Representatives, No. 8, pp. 1-33. Fourth annual report of the state geologist, Douglass Houghton. Idem, No. 27, 184 pp. See also Metalliferous veins of the Northern Peninsula of Michigan, by Douglass Houghton. Am. Jour. Sci., 1st ser., vol. 41, 1841, pp. 183-18G. Geology of Porters Island and Copper Harbor, by John Locke. Trans. Am. Phil. Soc, vol. 9, 1844, pp. 311-312, with maps. Mineralogy and geology of Lake Superior, by 11. D. Rogers. Proc. Boston Soc. Nat. HLst., vol. 2, 1840, pp. 124-125. Report of observations made in the survey of the Upper Peninsula of Michigan, by John Locke. Senate Docs., 1st sess. 30th Cong., 1847, vol. 2, No. 2, pp. 183-199. Report of J. D. Whitney. Idem, pp. 221-230. Report of J. D. Whitney. Senate Docs., 2d sess. 30th Cong., 1848^9, vol. 2, No. 2, pp. 1.54-1,59. Report of J. W. Foster. Idem, pp. 159-163. The Lake Superior copper and iron district, by J. D. Whitney. Proc. Boston Soc. Nat. Hist., vol. 3, 1849, pp. 210-212. On the geological structure of Keweenaw Point, by Charles T. Jackson. Proc. Am. Assoc. Adv. Sci., 1849, 2d meetmg, pp. 288-301. Report on the geological and mineralogical survey of the mineral lands of the United States in the State of Michigan, by Charles T. Jackson. Senate Docs., 1st sess. 31st Cong., 1849-50, vol. 3, No. 1, pp. 371-935, with 14 maps. Contains reports by Messrs. Jackson, Dickenson, Mclntyre, Barnes, Locke, Foster and \Miitney, Whitney, Gibbs, ^\^litney, jr., Hill and Foster, Foster, Burt, Hubbard. United States geological survey of public lands in Michigan; field notes, by John Locke. Idem, pp. 572-587. Synopsis of the explorations of the geological corps in the Lake Superior land district in the Northern Peninsula of Michigan, by J. W. Foster and J. D. Whitney. Idem, pp. 605-626, with 4 maps. Notes on the topography, soil, geology, etc., of the district between Portage Lake and the Ontonagon, by J. D. "WTiitney. Idem, pp. 649-666. ' Report of J. D. "WTiitney. Idem, pp. 705-711. Report of J. W. Foster. Idem, pp. 766-772. Notes on the geology and topography of the country adjacent to Lakes Superior and Michigan, in the Chippewa land district, by J. W. Foster. Idem, pp. 773-786. To].)(>graphy and geology of the survey with reference to mines and minerals of a district of township lines south of Lake Superior, by William A. Burt. Idem, pp. 811-832, with a geologic map opposite p. 880. General observations upon the geology and topography of the district south of Lake Superior, subdi\'ided in 1845 under direction of Douglass Houghton; deputy surveyor, Bela Hubbard. Idem, pp. 833-842. Geological report of the survey "with reference to mines and minerals" of a district of township lines in the State of Michigan, in the year 1846, and tabular statement of specimens collected. Idem, pp. 842-S82, with a geologic map. Report on the geology and topography of the Lake Superior land district; part 1, copper lands, by J. W. Foster and J. D. WTiitney. Executive Docs., 1st sess. 31st Cong., 1849-50, vol. 9, No. 69, 224 pp., with map. Report on the geology and topography of the Lake Superior land district; part 2, The iron region, by J. W. Foster and J. D. Whitney. Senate Docs., special sess. 32d Cong., 1851, vol. 3, No. 4, 406 pp., with maps. Sec also Aperfu de I'ensemble des terrains Siluriens du Lac Sup&ieur, by J. W. Foster and J. D. AMiitncy. Bull. Soc. geol. France, vol. 2, 18.50, pp. 89-100. On the Azoic system as developed in the Lake Superior laud district, by J. W. Foster and J. D. Whitney. Proc. Am. Assoc. Adv. Sci., 1851, 5th meeting, pp. 4-7. On the age of the sandstone of Lake Superior, with a description of the phenomena of the association of igneous rocks, by J. W. Foster and J. D. Whitney. Idem, pp. 22-38. Geology, mineralogy, and topography of the lands around Lake Superior, by Charles T. Jackson. Senate Docs., 1st sess. 32d Cong., 1851-52, vol. 11, pp. 232-244. Age of the Lake Superior sandstone, by Charles T. Jackson. Proc. Boston Soc. Nat. Hist., vol. 7, 1860, pp. 396-398. Age of the sandstone, by William B. Rogers. Idem, pp. 394-395. First biennial rei)ort of the progress of the geological survey of Michigan, by Alexander Winchell. Lansing, 1861, 339 pp. Some contributions to a knowledge of the constitution of the copper ranges of Lake Superior, by C. P. Williams and J. F. Blandy. Am. Jour. Sci., 2d ser., vol. 34, 1862, pp. 112-120. On the iron ores of Marquette, Michigan, by J. P. Kimball. Idem, vol. 39, 1865, pp. 290-303. On the position of the sandstone of the southern slope of a portion of Keweenaw Point, Lake Superior, by Alexander Agassiz. Proc. Boston Soc. Nat. Hist., vol. 11, 1867, pp. 244-246. Die vorsilurischen Gebilde der "Obcm Halbinsel von Michigan" in Xord-Amerika, by Hermann Credner. Zeitschr. Deutch. geol. Gesell., vol. 21, 1869, pp. 516-554. See al.so Die Gliederung der eozoischen (vorsilurischen) HISTORY OF GEOLOGIC WORK IN THE REGION. 75 Formationsgruppe Nord-Amcrikas, by Hermann Credner. Zeitschr. pesammtcn Naturwissenschaften, vol. 32, Giebel, 1868, pp. 353-405. On the age of the copper-bearing rocks of Lake Superior, by T. B. Brookn and R. Pumpelly. Am. Jour. Sci., 3d ser., vol. 3, 1872, pp. 428^32. Iron-bearing rocks, by T. B. Brooks. Geol. Survey Michigan, vol. 1, pt. 1, 1869-1873, 319 pp., with maps. Copper-bearing rocks, by R. Pumpelly. Idem, pt. 2, pp. 1^6„62-94, with maps. Copper-bearing rocks, by A. R. Marvine. Idem, pt. 2, pp. 47-61, 95-140. Paleozoic rocks, by C. Rominger. Idem, pt. 3, 102 pp. Observations on the Ontonagon silver-mining district and the slate ciuarries of Huron Bay, by C. Rominger. Geol. Survey Michigan, vol. 3, pt. 1, 1876, pp. 151-166. On the youngest Huronian rocks south of Lake Superior and the age of the copper-bearing series, by T. B. Brooks. Am. Jour. Sci., 3d ser., vol. 11, 1876, pp. 206-211. Classified list of rocks observed in the Huronian series south of Lake Superior, by T. B. Brooks. Idem, vol. 12, pp. 194-204. Metasomatic development of the copper-bearing rocks of Lake Superior, by Raphael Pumpelly. Proc. Am. Acad. Arts Sci., vol. 13, 1878, pp. 253-309. First annual report of the commissioner of mineral statistics of the State of Michigan for 1877-1878, by Charles E. Wright. Marquette, 1879, 229 pp. Notes on the iron and copper districts of Lake Superior, by M. E. Wadsworth. Bull. Mus. Comp. Zool. Harvard Coll., whole ser., vol. 7; geol. ser., vol. 1, No. 1, 157 pp. See also On the origin of the hon ores of the Marquette dis- trict, Lake Superior. Proc. Boston Soc. Nat. Hist., vol. 20, 1878-1880, pp. 470-479. On the age of the copper-bearing rocks of Lake Superior (abstract). Proc. Am. Assoc. Adv. Sci., 29th meeting, pp. 429-4.30. On the relation of the "Keweenawan series" to the Eastern sandstone in the vicinity of Torch Lake, Michigan. Proc. Boston Soc. Nat. Hist., vol. 23, 1884-1888, pp. 172-180; Science, vol. 1, 1883, pp. 248-249, 307. Upper Peninsula, by C. Rominger. Geol. Survey Michigan, vol. 4, 1881, pp. 1-248, with a geologic map. Geological report on the Upper Peninsula of Michigan, exhibiting the progress of work from 1881 to 1884, by C. Rominger. Geol. Survey Michigan, vol. 5, pt. 1, 1895, 179 pp. On a supposed fossil from the copper-bearing rocks of Lake Superior, by M. E. ^^'adsworth. Proc. Boston Soc. Nat. Hist., vol. 23, 1884-1888, pp. 208-212. Observations on the junction l>etween the Eastern sandstone and the Keweenaw series on Keweenaw Point, Lake Superior, by R. D. Irving and T. C. Chamberlin. Bull. U. S. Geol. Survey No. 23, 1885, 124 pp., 17 pi. Mode of deposition of the iron ores of the Menominee range, Michigan, by John Fulton. Trans. Am. Inst. Min. Eng., vol. 16, 1887, pp. 525-536. Report of N. H. Winchell. Sixteenth Ann. Rept. Geol. and Nat. Hist. Survey Minnesota, for 1887, pp. 13-129. Report of Alexander Winchell. Idem, pp. 133-391. Unpublished field notes made by W. N. Merriam in the summers of 1888 and 1889. Unpublished field notes made by C. R. Van Hise in the summer of 1890. The greenstone schist areas of the Menominee and Marquette regions of Michigan, by George Huntington Williams. Bull. U. S. Geol. Survey No. 62, 1890, pp. 31-238, with 16 pis. and maps. See also Some examples of dynamic meta- morphism of the ancient eruptive rocks on the south shore of Lake Superior. Proc. Am. Assoc. Adv. Sci., 36th meeting, 1888, pp. 225-226. A sketch of the geology of the Marquette and Keweenawan district, by M. E. \\'adsworth. Along the south shore of Lake Superior, by Julian Ralph. 1st ed., 1890, pp. 63-82. Explanatoi-y and historical note, by R. D. Irving. Bull. U. S. Geol. Survey No. 62, 1890, pp. 1-30. The Penokee iron-bearing series of Michigan and Wisconsin, by R. D. Irving and C. R. Van Hise. Mon. U. S. Geol. Survey, vol. 19, 1892, 534 pp., with plates and maps. See also Tenth Ann. Rept. U. S. Geol. Survey, for 1888-89, 1890, pp. 341-507, with 23 pis. and maps. A sketch of the geology of the Marquette and Keweenawan district, by M. E. \\'adsworth. Along the south shore of Lake Superior, by Julian Ralph. 2d ed., 1891, pp. 75-99. On the relations of the Eastern sandstone of Keweenaw Point to the Lower Silurian limestone, by M. E. Wadsworth. Am. Jour. Sci., 3d ser., vol. 42, 1891, pp. 170-171 (communicated). The South Trap range of the Keweenawan series, by M. E. Wadsworth. Idem, pp. 417^19. Unpublished field notes made by Raphael Pumpelly and C. R. Van Hise in the summers of 1891 and 1892. The Huronian volcanics south of Lake Superior, by C. R. Van Hise. Bull. Geol. Soc. America, vol. 4, 1892, pp. 435-436. Microscopic characters of rocks and minerals, by A. C. Lane. Rept. State Board Geol. Survey Michigan for 1891-92, Lansing, 1893, pp. 176-183. A sketch of the geology of the iron, gold, and copper districts of Michigan, by M. E. Wadsworth. Idem, pp. 75-174. See also Ann. Repts. 1888-1892, pp. 38-73. Subdivisions of the Azoic or Archean in northern Michigan, by M. E. Wadsworth. Am. Jour. Sci., vol. 45, 1893, pp. 72-73. The succession in the Marquette iron district of Michigan, by C. R. ^'an Hise. Bull. Geol. Soc. America, vol. 5, 1893, pp. 5-6. 76 GEOLOGY OF THE LAKE SUPERIOR REGION. The geology of that portion of the Menominee range east of Menominee River, by Nclscm P. Ilulst. Proc. Lake Superior Min. Inst., March, 1893, pp. 19-29. A pontact between the Lower Iluronian and the underlying graiiile in the Ropublir trough, near Republic, Mich., by n. L. Smyth. Jour. Geology, vol. 1, Xo. 3, 1893. pp. 2G8-274. Two new geological cross sections of Keweenaw Point, by L. L. Hubbard. Proc. Lake Superior Min. Inst., vol. 2, 1894, pp. 79-96. Chai-acter of folds in the Marquette iron district, by C. R. Van Hise. Proc. Am. Assoc. Adv. Sci., 42d meeting, 1894, p. 171 (abstract). Relations of the Lower Menominee and Lower Marquette series of Michigan (preliminary), by II. L. Smyth. Am. Jour. Sci., 3d ser., vol. 47, 1894, pp. 216-223. The quartzite tongue at Republic, Mich., by H. L. Smyth. Jour. Geology, vol. 2, 1894, pp. 680-691. The relation of the \ein at the Central mine, Keweenaw Point, to the Kearsarge conglomerate, by L. L. Hubbard. Proc. Lake Superior Min. Inst., vol. 3, 1895, pp. 74-83. The Marquette iron range of Michigan, by G. A. Xewett. Idem. pp. 87-108. With geologic map. The volcanics of the Michigamme district of Michigan (preliminary), l)y J. Morgan Clements. Jour. Geology, vol. 3, 1895, pp. 802-822. A central Wisconsin base-level, by C. R. Van Hise. Science, new ser., vol. 4, 1896, pp. 57-59. See also A northern Michigan base-level. Idem, pp. 217-220. Organic markings in Lake Superior iron ores, by W. S. Gresley. Science, new ser., vol. 3, 1896, ])p. 622-623; Trans. Am. Inst. Min. Eng., vol. 26, 1897, pp. 527-534. The Marquette iron-bearing district of Michigan, by C. R. Van Hise and W. S. Bayley; with a chapter on the Republic trough, by H. L. Smyth. Men. U. S. Geol. Survey, vol. 28, 1897, 608 pp. With atlas of 39 plates. Pre- iminary report on same district published in Fifteenth Ann. Rept. U. S. Geol. Survey, 1895, pp. 477-650. The origin and mode of occurrence of the Lake Superior copper deposits, by M. E. Wadsworth. Trans. Am. Inst. MLn. Eng., vol. 27, 1898, pp. 669-696. Some dike features of the Gogebic iron range, by C. M. Ross. Idem, pp. 556-563. Geological report on Isle Royale, Michigan, by A. C. Lane. Geol. Survey Michigan, vol. 6, pt. 1, 1898, 281 pp. With geologic map. Keweenaw Point, with particular reference to the felsites and their associated rocks, by L. L. Hubbard. Geol. Survey Michigan, vol. 6, pt. 2, 1898, 155 pp. With plates. Unpublished notes by Prof. A. E. Seaman and thesis on the Gogebic district, by W. J. Sutton, Michigan College of Mines. The Crystal Falls iron-bearing district of Michigan, by J. Morgan Clements and H. L. Smyth, with a chapter on the Sturgeon River tongue, by W. S. Bayley, and an introduction by C. R. Van Hise. Mon. V. S. Geol. Survey, vol. 36, 1899. With geologic maps. Geology of the Mineral range, by A. E. Seaman. First Ann. Rept. Copper-Mining Industry of Lake Superior, 1899, pp. 49-60. Note sur la region cupriffere de Textremit^ nordest de la peninsula de Keweenaw (Lac Superieur), par Louis Duparc. Archives sci. phys. et nat., vol. 10, 1900, p. 21. The Menominee special folio, by Charles R. Van Hise and ^\■. S. Bayley. Geologic Atlas U. S., folio 62, U. S. Geol. Survey, 1900. Unpublished notes by Prof. A. E. Seaman made for Michigan Geological Survey, Michigan College of Mines, and United States Cieological Survey. See also unpublished maps prepared for Michigan exhibit at St. Louis exposition, 1904. Unpublished thesis by W. O. Hotchkiss, Geol. Dept. Univ. Wisconsin, 1903. Report of special committee on the Lake Superior region to Frank D. Adams, Robert Bell, C. Willard Hayes, and Charles R. Van Hise, general committee on the relations of the Canadian and the United States geological sur\-eys, 1904. Jour. Geology, vol. 13, 1905, pp. 89-104. The special committee consisted of Frank D. Adams, Robert Bell, C. K. Leith, C. R. Van Hise. There were present by invitation W. G. Miller, A. C. Lane, and for parts of the trip A. E. Seaman, W. N. Merriam, J. U. Sebenius, and W. N. Smith. Maps of the Marquette, Menominee, and Gogebic districts, Michigan, prepared by A. E. Seaman for the St. Louis exposition, 1904. Unpublished. The Menominee ii-on-bearing district of Michigan, by W. S. Bayley. Mon. U. S. Geol. Survey, vol. 46, 1904, 513 pp. The geology of some of the lands in the Upper Peninsula, by R. S. Rose. Proc. Lake Superior Min. Inst., 1904, pp. 88-102. Unpublished notes of field work done in 1905, by G. W. Corey and C. F. Bowen. Black River work, by A. C. Lane. Ann. Rept. Geol. Survey Michigan for 1904, 1905, pp. 158-162. Report of progress in the Porcupines, by F. E. Wright. Ann. Rept. Geol. Survey Michigan for 1903, 1905, pp. 33^4. Also Preliminary geological map of the Porcupine Mountains and vicinity, by F. E. Wright and A. C. Lane. Ann. Rept. Geol. Survey Michigan for 1908, 1909, pi. 1. The geology of Keweenaw Point -a brief description, by A. C. Lane. Proc. Lake Superior Min. Inst., vol. 12, 1907, pp. 81-104. HISTORY OF GEOLOGIC WORK IN THE REGION. 77 Notes on the geological section of Michigan; part 1, the pre-Ordovician, by A. C. Lane and A. E. Seaman. Jour. Geology, vol. 15, 1907, pp. 680-695. Also notes on the geological section of Michigan, part 2, from the St. Peter sand- stone up, by A. C. Lane. Jour. Geology, vol. 18, 1910, pp. 393^29. A geological section from Bessemer down Black River, by W. C. Gordon and Alfred C. Lane. Ann. Rept. Geol. Survey Michigan for 1906, 1907, pp. 396-507. Unpublished geologic maps of Menominee and Florence districts, Michigan and Wisconsin, prepared for Oliver Iron Mining Company by W. N. Merriam. Unpublished maps and report on geology of Crystal Falls, Menominee, and Iron River districts, Michigan, prepared during commercial surveys, by C. K. Leith, R. C. Allen, and others. Report on geology of Iron River district of Michigan, by R. C. Allen. Michigan Geo!, and Biol. Survey, pub. 3, 1910, 151 pp., with geologic map. The intrusive rocks of Mount Bohemia, Michigan, by F. E. Wright. Ann. Rept. Michigan Geol. Survey for 1908, 1909, pp. 361-397. NORTHERN WISCONSIN. Report of a geological reconnaissance of the Chippewa land district of Wisconsin, etc., by David D. (.)wen. Senate Docs., 1st sess. 30th Cong., 1848, vol. 7, No. 57, 72 pp. Preliminary report containing outlines of the progress of the geological survey of Wisconsin and Iowa up to October 11, 1847, by Da\-id Dale Owen. Senate Docs., 1st sess. 30th Cong., 1847, vol. 2, No. 2, pp. 160-173. Description of part of Wisconsin south of Lake Superior, by Charles WTiittlesey. Report of a geological survey of Wisconsin, Iowa, and Minnesota, 1852, pp. 419-470. The Penokee iron range, by Increase A. Lapham. Trans. Wisconsin State Agr. Soc, vol. 5, 1858-59, pp. 391^00, with map. See also Report to the directors of the Wisconsin and Lake Superior Mining and Smelting Company, in the Penokee iron range of Lake Superior, with reports and statistics showing its mineral wealth and prospects, charter and organization of the Wisconsin and Lake Superior Mining and Smelting Company, Milwaukee, 1860, pp. 22-37. Geological report of the State of Wisconsin, by James Hall. Report of the Superintendent of the Geological SiuT^ey (1861), exhibiting the progress of the work, 52 pp. Physical geography and general geology, by James Hall. Report on the geological survey of the State of Wisconsin, vol. 1, 1862, pp. 1-72. The Penokee mineral range, Wisconsin, by Charles Whittlesey. Proc. Boston Soc. Nat. Hist., vol. 9, 1863, pp 235-244. On some points in the geology of northern Wiscon.sin, by R. D. Irving. Trans. WiscoiLsin Acad. Sci., vol. 2, 1873-74 pp. 107-119. See also On the age of the copper-bearing rocks of Lake Superior, and on the westward continuation of the Lake Superior synclinal. Am. Jour. Sci., 3d ser., vol. 8, 1874, pp. 46-56. Ann. Rept. Progress and Results of Wisconsin Geol. Survey for 1876, pp. 17-25. Report of progress and results for 1874. Geology of Wisconsin, vol. 2, pp. 46-49. Notes on the geology of northern Wisconsin, by E. T. Sweet. Trans. Wisconsin Acad. Sci., 1875-76, vol. 3, pp 40-55. Note on the age of the crystalline rocks of Wisconsin, by R. D. Irving. Am. Jour. Sci., 3d ser., vol. 13, 1877, pp, 307-309. Report of progress and results for the year 1875, by O. W. Wight. Geology of Wisconsin, vol. 2, 1873-1877, pp, 67-89. Geology of central Wisconsin, by R. D. Irving. Idem, pp. 409-636, with 2 atlas maps. On the geology of northern Wisconsin, by R. D. Irving. Ann. Rept. Wisconsin Geol. Sm-vey for 1877, pp. 17-25. Report on the eastern part of the Penokee range, by T. C. Chamberlin. Idem, pp. 25-29. General geology of the Lake Superior region, by R. D. Irving. Geology of Wisconsin, vol. 3, 1880, pp. 1-24. Geol- ogy of the eastern Lake Superior district. Idem, pp. 51-238, with 6 atlas maps. Mineral resources of Wisconsin. Trans. Am. Inst. Min. Eng., vol. 8, 1880, pp. 478-508, with map. Note on the stratigraphy of the Huronian series of northern Wisconsin, and on the equivalency of the Huronian of the Marquette and Penokee districts. Am. Jour. Sci., 3d ser., vol. 17, 1879, pp. 393-398. Huronian series west of Penokee Gap, by C. E. Wright. Geology of Wisconsin, vol. 3, 1880, pp. 241-301, with an atlas map. Geology of the western Lake Superior district, by E. T. Sweet. Idem, pp. 303-362, with an atlas map. Geology of the upper St. Croix district, by T. (". Chamberlin and Moses Strong. Idem, pp. 363-428, with 2 atlas maps. Geology of the Menominee region, by T. B. Brooks. Idem, pp. 430-599, with 3 atlas maps. Geology of the Menominee iron region (economic resources, lithology, and westerly and southerly extension), by Charles E. Wright. Idem, pp. 666-734. The quartzitea of Barron and Chippewa counties, by Moses Strong, E. T. Sweet, F. H. Brotherton, and T. C. Chamberlin. Geology of Wisconsin, vol. 4, 1873-1879, pp. 573-581. Geology of the upper Flambeau Valley, by F. H. King. Idem, pp. 583-615. Crystalline rocks of the Wisconsin Valley, by R. D. Irving and C. R. Van Hise. Idem, pp. 623-714. 78 GEOLOGY OF THE LAKE SUPERIOR REGION. General geology (of Wisconsin), by T. C. Chaniberlin. Geology of Wiscontiin, vol. 1, 1SS3, pp. 3-:50O, ■n-ith an atlaa map. Lithologry of Wiscon.sin, by R. D. Irving. Idem, pp. 340-361. Transition from the copper-bearing series to the Potsdam, by L. C. Wooster. Am. Jour. Sci., 3d ser., vol. 27, 1884, pp. 463^65. Geology of the St. Croix Dalles, by C. P. lierkey. Am. Geologist, vol. 20, 1897, pp. 345-383; vol. 21, 1898, pp. 139-155, 270-294. Preliminary report on copper-bearing rocks in Douglas County, Wis., by U. S. Grant. Hull. Wisconsin Geol. and Nat. Hist. Survey No. (i, 1901. The pre-Potsdam Peneplain of the pre-Cambrian of north-central Wisconsin, by S. Weidman. Jour. Geology, vol. 11, 1903, pp. 289-313. Unpublished thesis Univ. Wisconsin, 1905. The geology of north-central W'isconsin, by S. W'eidman. Bull. Wisconsin Cieol. and Nat. Hist. Survey No. 16, 1907. Summary fiu-nished by author in 1905. MINNESOTA. Account of a journey to the Coteau des Prairies, viilh a description of the red pipestone cpiarry and granite bowlders found there, by George Catlin. Am. Jour. Sci., 1st ser., vol. 38, pp. 138-146. Report of J. G. Norwood. Senate Docs., 1st sess. 30th Cong., 1847, vol. 2, No. 2, pp. 73-134. Description of the geology of middle and western Minnesota, including the country adjacent to the northwest and part of the southwest shore of Lake Superior; illustrated by numerous general and local sections, woodcuts, and a map, by J. G. Norwood. Report of a geological survey of Wisconsin, Iowa, and Minnesota, 1852, pp. 209-418. Report of the State geologist on the metalliferous region bordering on Lake Superior, by Henry II. Eames. St. Paul, 1866, 23 pp. Geological reconnaissance of the northern, middle, and other counties of Minnesota, by Henry II. Eames. St. Paul, 1866, 58 pp. Notes upon the geology of some portions of Minnesota, from St. Paul to the western part of the State, by James Hall. Trans. Am. Philos. Soc, new ser., vol. 13, 1869, pp. 329-340. Report on the geological survey of the State of Iowa, containing results of examinations and observations made within the years 1866, 1867, 1868, and 1869, by Charles A. WTiite. Des Moines, 1870, vol. 1, 391 pp.; vol. 2, 443 pp. First Ann. Rept. Geol. and Nat. Hist. Survey of Minnesota, by N. H. Winchell, 1873, 129 pp. Thegeology of the Minnesota Valley, by N. II. Winchell. Second Ann. Rept. Geol. and Nat. Hist. Survey Minne- sota, 1874, pp. 127-212. Ueber die krystallinischen gesteine von Minnesota in Nord-.\merika, by A. Streng and J. H. Kloos. Leonhard's Jahrbuch, 1877, pp. 31, 113, 225. Translated by N. H. Winchell in Eleventh Ann. Rept. Geol. and Nat. Hist. Survey Minnesota, 1883, pp. 30-85. Sixth Ann. Rept. Geol. and Nat. Hist. Survey Minnesota, for 1877, by N. H. WincheU, 226 pp. Sketch of the work of the season of 1878, by N. H. Winchell. Seventh Ann. Rept. Geol. and Nat. Hist. Survey Minnesota, for 1878, pp. 9-25. The cupriferous series at Duluth, by N. H. Winchell. Eighth Ann. Rept. Geol. and Nat. Hist. Survey Minnesota, for 1879, pp. 22-26. ' Preliminary report on the geology of central and western Minnesota, by Warren Upham. Idem, pp. 70-125. Report of Prof. C. W. Hall. Idem, pp. 126-138. Preliminary list of rocks, by N. H. Winchell. Ninth Ann. Rept. Geol. and Nat. Hist. Sur\-ey Minnesota, for 1880, pp. 10-114. The cupriferous series in Minnesota, by N. H. Winchell. Proc. Am. Assoc. Adv. Sci., 29th meeting, 1881, pp. 422-425. See also Ninth Ann. Rept. Geol. and Nat. Hist. Survey Minnesota, for 1880, pp. 38.5-387. Preliminary list of rocks, by N. H. Winchell. Tenth Ann. Rept. Geol. and Nat. Hist. Survey Minnesota, for 1881, pp. 9-122. Notes on rock outcrops in central Minnesota, by Warren Upham. Eleventh Ann. Rept. Geol. and Nat. Hist. Survey Minnesota, for 1882, pp. 86-136. The iron region of northern Minnesota, by Albert H. Chester. Idem, pp. 154-107. Note on the age of the rocks of the Mesabi and Vermilion iron district, by N . H. Winchell. Idem, pp. 168-170. See also Proc. Am. Assoc. Adv. Sci., 1884, 33d meeting, pp. 363-379. The geology of Minnesota, by N. 11. Winchell and Warren Upham. Final Rept. Geol. and Nat. Hist. Survey Minnesota, voL 1, 1884, 695 pp.; vol. 2, 1888, 697 pp. Notes of a trip across the Mesabi range to Vermilion Lake, by N. II. Winchell. Thirteenth Ann. Rept. Geol. and Nat. Hist. Survey Minnesota, for 1884, pp. 20-24. The crystalline rocks of Minnesota, by N. H. Winchell. Idem, pp. 36-38. The crystalline rocks of the Northwest, by N. H. Winchell. Idem, i)p. 124-140. Report of a trip on the upper Mississippi and to Vermilion Lake, by Bailey Willis. Tenth Census, vol. 15, 1886, pp 457^67. HISTORY OF GEOLOGIC WORK IN THE REGION. 79 Report of geological observations made in northeastern Minnesota during the season of 188G, by Alexander Winchell. Fifteenth Ann. Rept. Geol. and Nat. Hist. Survey Minnesota, fur 1886, pp. 5-207. Geological report of N. H. Winchell. Idem, pp. 209-399, with a map. Report of N. H. Winchell. Sixteenth Ann. Rept. Geol. and Nat. Hist. Survey Minnesota, for 1887, pp. 13-129. Report of Alexander Winchell. Idem, pp. 133-391. See also The unconformities of the Animikie in Minnesota. Am. Geologist, vol. 1, 1888, pp. 14-24. Two systems confounded in the Huronian. Idem, vol. 3, 1889, pp. 212-214, 339-340. Systematic results of a field study of the Archean rocks of the Northwest. Proc. Am. Assoc. Adv. Sci., 37th meeting, 1889, p. 205. The geological position of the Ogishke conglomerate. Idem, 38th meeting, 1890, pp. 234-235. Report of H. V. Winchell. Sixteenth Ann. Rept. Geol. and Nat. Hist. Survey Minnesota, for 1887, pp. 39-5-462, with map. The distribution of the granites of the Northwestern States and their general lithologic characters, by C. W. Hall. Proc. Am. Assoc. Adv. Sci., 37th meeting, 1889, pp. 189-190. Report of N. H. Winchell. Seventeenth Ann. Rept. Geol. and Nat. Hist. Survey Minnesota, for 1888, pp. 5-74. See also The Animikie black slates and quartzites and the Ogishke conglomerate of Minnesota, the equivalent of the "Original Huronian." Am. Geologist, vol. 1, 1888, pp. 11-14. Methods of stratigraphy in studying the Huronian. Idem, vol. 4, 1889, 342-357. Report of field observations made during the season of 1888 in the iron regions of Minnesota, by H. \'. \\'inchell. Seventeenth Ann. Rept. Geol. and Nat. Hist. Survey Minnesota, for 1888, pp. 77-145. See also The diabasic schists containing the jaspilite beds of northeastern Minnesota. Am. Geologist, vol. 3, 1889, pp. 18-22. Report of geological observations made in northeastern Minnesota during the summer of 1888, by U. S. Grant. Seventeenth Ann. Rept. Geol. and Nat. Hist. Survey Minnesota, for 1888, pp. 149-215. Conglomerates inclosed in gneissic terranes, by Alexander Winchell. Am. Geologist, vol. 3, 1889, pp. 153-165, 256-262. Some thoughts on eruptive rocks, with special reference to those of Minnesota, by N. II. \\'inchell. Proc. Am. Assoc. Adv. Sci., 37th meeting, 1888, pp. 212-221. The Stillwater, Minn., deep well, by A. D. Meads. Am. Geologist, vol. 3, 1889, pp. 341-342. ^ On a possible chemical origin of the iron ores of the Keewatin in Minnesota, by N. II. and H. V. \\'inchell. Idem, vol. 4, 1889, pp. 291-300, 382-386. Also Proc. Am. Assoc. Adv. Sci., 38th meeting, pp. 235-242. Some results of Archean studies, by Alexander Winchell. Bull. Geol. Soc. America, vol. 1, 1890, pp. 357-394. The Taconic iron ores of Minnesota and of western New England, by N. II. and H. V. Winchell. Am. Geologist, vol. 6, 1890, pp. 263-274. Record of field observations in 1888 and 1889, by N. H. Winchell. Eighteenth Ann. Rept. Geol. and Nat. Hist. Survey Minnesota, for 1889, pp. 7-47. The iron ores of Minnesota, by N. H. and H. V. Winchell. Bull. Geol. and Nat. Hist. Survey Minnesota No. 6, 1891, pp. 430, with a geologic map. Geological age of the Saganaga syenite, by Horace V. Winchell. Am. Jour. Sci., 3d ser., vol. 41, 1891, pp. 386-390. Notes on the petrography and geology of the Akeley Lake region, in northeastern Minnesota, by W. S. Bayley. Nineteenth Ann. Rept. Geol. and Nat. Hist. Survey Mitmesota, for 1890, pp. 193-210. The stratigraphic position of the Ogishke conglomerate of northeastern Minnesota, by U. S. Grant. Am. Geologist, vol. 10, 1892, pp. 4-10. Paleozoic formations of southeastern Minnesota, by C. W. Hall and F. W. Sardeson. Bull. Geol. Soc. America, vol. 3, 1892, pp. 331-368. The basic massive rocks of the Lake Superior region, by W. S. Bayley. Jour. Geology, vol. 1, 1893, pp. 433—456, 587-596, 688-716; vol. 2, 1894, pp. 814-825; vol. 3, 1895, pp. 1-20. The crystalline rocks, by N. H. Winchell. Twentieth Ann. Rept. Geol. and Nat. Hist. Survey Minnesota, for 1891, 1893, pp. 1-28. Anorthosites of the Minnesota shore of Lake Superior, by A. C. Lawson. Bull. Geol. and Nat. Hist. Survey Min- nesota No. 8, 1893, pp. 1-23. The geology of Kekequabic Lake, in northeastern Minnesota, with special reference to an augite-soda granite, by U. S. Grant; thesis accepted for degree of Ph. D. in Johns Hopkins University, 1893. Twenty-first Ann. Rept. Geol. and Nat. Hist. Survey Minnesota, for 1892, 1893, pp. 5-58. With geologic map and plates. The eruptive and sedimentary rocks on Pigeon Point, Minnesota, and their contact phenomena, by W. S. Bayley. Bull. U. S. Geol. Survey No. 109, 1893, with maps and plates. Field observations on certain granitic areas in northeastern Minnesota, by V. S. Grant. Twentieth Ann. Rept. Geol. and Nat. Hist. Survey Minnesota, 1893, pp. 3.5-110. Sketch of the coastal topography of the north side of Lake Superior, with special reference to the abandoned strands of Lake Warren, by A. C. Lawson. Idem, pp. 181-289. Actinolite-magnetite schists from the Mesabi ir(.in range, in northeastern Minnesota, by W. S. Bayley. Am. Jour. Sci., 3d ser., vol. 46, 1893, pp. 176-180. The Mesabi iron range, by H. V. Winchell. Twentieth Ann. Rept. Geol. and Nat. Hist. Survey Minnesota, for 1891, 1893, pp. 111-180. Preliminary report of field work during 1893 in northeastern Minnesota, by U. S. Grant. Twenty-second Ann. Rept. Geol. and Nat. Hist. Survey Minnesota, pt. 4, 1894, pp. 67-78. 80 GEOLOGY OF THE LAKE SUPERIOR REGION. Notes on the geology of Itasca County, Minn., by G. E. Culver. Idem, pt. 8, 1894, pp. 97-114. Preliminary report of field work during 1893 in northeastern Minnesota, by A. H. Elftman. Idem, pt. 12, 1894, pp. 141-180. The stratigraphic position of the Thompson slates, by J. E. Spurr. Am.. lour. Sci., Sdser., vol.48, 1894, pp. 1.59-1G.5. The iron-bearing rocks of the Mesiibi range in Minnesota, by J. Edward Spurr. Bull. Geol. and Nat. Hist. Survey Minnesota No. 10, 1894, 2(i8 pp., with geologic maps. The origin of the Archean greenstones, by N. H. Winchell. Twenty-third Ann. Repl. Geol. and Nat. Hist. Survey Minnesota, for 1894, pt. 2, 1895, pp. 4-3.5. Preliminary report on the Rainy Lake gold region, by II. V. ^^'inchell and U. S. Grant. Idem, pp. 36-105. The iron ranges of Minnesota, by H. V. Winchell. Proc. Lake Superior Mining Inst., vol. 3, 1895, pp. 11-32. Notes upon the bedded and banded structures of the gabbro and upon an area of troctolyte, by A. H. Elftman. Twenty-third Ann. Kept. (iool. and Nat. Hist. Survey Minnesota, for 1894, 1895, pt. 12, pp. 224-230. The geological structure of the western part of the Vermilion range, Minnesota, by H. L. Smyth and J. Ralph Finlay. Trans. Am. Inst. Min. Engineers, vol. 25, 1895, pp. 595-605. The Koochiching granite, by Alexander Winchell. Am. Geologist, vol. 20, 1897, pp. 293-299. Some new features in the geology of northeastern Minnesota, by N. H. Winchell. Idem, pp. 41-51. The origin of the Archean igneous rocks, by N. H. Winchell. Proc. Am. Assoc. Adv. Sci., vol. 47, 1898, pp. 303- 304 (abstract); Am. Geologist, vol. 22, 1898, pp. 299-310. Some resemblances between the Archean of Minnesota and of Finland, by N. H. Winchell. Am. Geologist, vol. 21, 1898, pp. 222-229. The significance of the fragmental eruptive debris at Taylors Falls, Minn., by N. H. Winchell. Am. Geologist, vol.22, 1898, pp. 72-78. The oldest known rock, by N. II. Winchell. Proc. Am. Assoc. Adv. Sci., vol. 47, 1898, pp. 302-303 (abstract). Sketch of the geology of the eastern end of the Mesabi iron range in Minnesota, by L'. S. Grant. Engineers' Year Book L'niv. Minnesota, 1898, pp. 49-62. With sketch map. The geology of Minnesota, by N. H. Winchell, V. S. Grant, James E. Todd, Warren I'pham, and H. V. Winchell. Final Rept. Geol. and Nat. Hist. Survey Minnesota, vol. 4, 1899, pp. 630. With 31 geologic plates. Structural geology of Minnesota, by N. H. Winchell, assisted by U. S. Grant. Idem, vol. 5, 1900, pp. 1-SO, 972-1000. Vol. 4 contains an account of detailed field work in northeastern Minnesota, with incidental discussion of general problems. The area is treated by counties and smaller arbitrary geographic divisions, in the description of which several men have taken part. This manner of treatment leads to repetition in the discussion of the general geologic features, and in many cases it is extremely ditBcult to correlate the facts recorded in the different sections. Vol. 5 contains an account of the general structural geology of the State based on the detailed work described in vol. 4. Grant's views, as in- dicated in the detailed descriptions of special areas, in some cases differ somewhat widely from those of Winchell. The gneisses, gabbro schists, and associated rocks of southwestern Minnesota, by C. W. Hall. Bull. V . S. Geol. Survey No. 157, 1899, 160 pp. With geologic maps. Mineralogical and petrographic study of the gabbroid rocks of Minnesota, and more particularly of the plagioclas- tites, by Alexander Winchell. Am. Geologist, vol. 26, 1900, pp. 153-162. With geologic sketch map of northeastern Minnesota. L^npublished field notes, summer of 1900, by C. R. Van Hise and J. Morgan Clements. Final Rept. Geol. and Nat. Hist. Survey Minnesota, vol. 6, 1900-1901. (N. H. Winchell.) Keewatin area of eastern and central Minnesota, by C. W. Hall. Bull. Geol. Soc. America, vol. 12, 1901, pp. 343-370, pis. 29-32. Keweenawan area of eastern Minnesota, by C. W. Hall. Idem, pp. 313-342, pis. 27-28. Sketch of the iron ores of Minnesota, by N. H. Winchell. Am. Geologist, vol. 29, 1902, pp. 154-162. The Mesabi iron-bearing district of Minnesota, by C. K. Leith. Mon. U. S. Geol. Survey, vol. 43, 1903, 316 pp. The Vermilion iron-bearing district of Minnesota, by J. Morgan Clements. Idem, vol. 45, 1903, 463 pp. Some results of the late Minnesota Geological Survey, by N. H. Winchell. Am. Geologist, vol. 32, 1903, pp. 246-253. The geology of the Cuyuna iron range, Minnesota, by C. K. Leith. Econ. Geology, vol. 2, 1907, pp. 145-152. The Cuyuna iron district of Minnesota, by Carl Zapffe. Unpublished bachelor's thesis L^niv. Wisconsin, 1907. See also The Cuyuna iron-ore district of Minnesota, by Carl Zapffe. Supplement to the Brainerd (Minn.) Tribune, Sept. 2, 1910, pp. 32-35, with map. The iron-ore deposits of the Ely trough, Vermilion range, Minnesota, by C. E. Abbott. Proc. Lake Superior Min. Inst, (for 1906), vol. 12, 1907, pp. 116-142. Geological history of the Redstone quartzite, by Frederick W. Sardeson. Bull. Geol. Soc. America, vol. 19, 1908, pp. 221-242. Contribution to the petrography of the Keweenawan (mth geologic ma]3'l, by Frank F. Grout. Jour. Geology, vol. 18, 1910, pp. 633-657. The iron formation of the Cuyuna range, by F. S. Adams. Econ. Geology, vol. 5, 1910, ])p. 729-740; vol. 6, 1911, pp. 60-70, 156-180. HISTORY OF GEOLOGIC WORK IN THE REGION. 81 ONTARIO. Notes on the geography and geology of Lake Superior, by John J. Bigsby. Quart. Jour. Sci., Lit. and Arts, voL 18, 1825, pp. 1-34, 222-269, with map. Outlines of the geology of Lake Superior, by H. W. Bayfield. Trans. Lit. and IILst. Soc. Quebec, vol. 1, 1829, pp. 1^3. On the junction of the Transition and Primary rocks of Canada and Labrador, by Captain Bayfield. Quart. Jour. GeoL Soc. London, voL 1, 1845, pp. 450-459. On the geology and economic minerals of Lake Superior, by W. E. Logan. Rept. Prog. Geol. Survey of Canada for 1846-47, pp. 8-34. On the geology of the Kaministiquia and Michipicoten rivers, by Alexander Murray. Idem, pp. 47-57. On the age of the copper-bearing rocks of Lakes Superior and Huron, and various facts relating to the physical structure of Canada, by W. E. Logan. Eept. Brit. Assoc. Adv. Sci., 21st meetmg, 1851, pp. 59-62, Trans.; Am. Jour. Sci., 2d ser., vol. 14, 1852, pp. 224-229. On the geology of the Lake of the Woods, south Hudson Bay, by Dr. J. J. Bigsby. Quart. Jour. Geol. Soc. London, vol. 8, 1852, pp. 400-406. With a geologic map of the Lake of the Woods. On the physical geography, geology, and commercial resources of Lake Superior, by John J. Bigsby. Edinburgh New Phil. Jour., vol. 53, 1852, pp. 55-62. On the geology of Ramy Lake, south Hudson Bay, by Dr. J. J. Bigsby. Quart. Jour. Geol. Soc. London, vol. 10, 1854, pp. 215-222. With a geologic map of Rainy Lake. On the geological structure and mineral deposits of the promontory of Mamainse, Lake Superior, by John W. Dawson. Canadian Naturalist and Geologist, vol. 2, 1857, pp. 1-12, with a section. Report of progress of the Geological Survey of Canada from its commencement to 1863, by W. E. Logan, 1863, 983 pp., with an atlas. On the Laurentian, Huronian, and upper copper-bearing rocks of Lake Superior, by Thomas Macfarlane. Rept. Prog. Geol. Survey Canada, 1863-1866, pp. 115-164. On the geological formations of Lake Superior, by Thomas Macfarlane. Canadian Naturalist, 2d ser., vol. 3, 1806-1868, pp. 177-202, 241-256. On the geology and silver ore of Woods Location, Thunder Cape, Lake Superior, by Thomas Macfarlane. Cana- dian Naturalist, 2d ser., vol. 4, pp. 37^8, 459^63, with a map. On the geology of the northwest coast of Lake Superior and the Nipigon district, by Robert Bell. Rept. Prog. Geol. Survey Canada, 1866-1869, pp. 313-364, with a topographic sketch map. Report on the country north of Lake Superior, between the Nipigon and Michipicoten rivers, by Robert Bell. Idem, 1870-71, pp. 322-351. Report on the country between Lake Superior and the Albany River, by Robert Bell. Idem, 1871-72, pp. 101-114. Notes of a geological reconnaissance from Lake Superior to Fort Garry, by A. R. C. Selwyn. Idem, 1872-73, pp. 8-18. On the country between Lake Superior and Winnipeg, by Robert Bell. Idem, pp. 87-111. The geognostical history of the metals, by T. Sterry Hunt. Trans. Am. Inst. Min. Eng., vol. 1, 1873, pp. 331-345; vol. 2, 1874, pp. 58-59. On the country between Red River and the South Saskatchewan, with notes on the geology of the region between Lake Superior and Red River, by Robert Bell. Rept. Prog. Geol. Survey Canada, 1873-74, pp. 66-90. Report on the geology and resources of the region in the vicinity of the Forty-ninth parallel, from the Lake of the Woods to the Rocky Mountains, by George Mercer Dawson, 387 pp., with a geologic map. The mineral region of Lake Superior, by Robert Bell. Canadian Naturalist and Geologist, 2d ser., vol. 7, 1875, pp. 49-51. On the country west of Lakes Manitoba and Winnipegosis, with notes on the geology of Lake Winnipeg, by Robert Bell. Rept. Prog. Geol. Survey Canada, 1874-75, pp. 24-56. Report on an exploration in 1875 between James Bay and Lakes Superior and Huron, by Robert Bell. Idem, 187.5-76, pp. 294-342. Report on geological researches north of Lake Huron and east of Lake Superior, by Robert Bell. Idem, 1876-77, pp. 213-220. Remarks on Canadian stratigraphy, by Thomas Macfarlane. Canadian Naturalist, 2d ser., vol. 9, 1879, pp. 91-102. Report on the geology of the Lake of the Woods and adjacent country, by Robert Bell. Rept. Prog. Geol. and Nat. Hist. Survey Canada, 1880-1882, pp. 11-15 c, with a map. On the geology of Lake Superior, by A. R. C. Selwyn. Trans. Roy. Soc. Canada, vol. 1, sec. 4, 1883, pp. 117-122. Age of the rocks of the northern shore of Lake Superior, by A. R. C. Selwyn. Science, vol. 1, 1883, p. 11. See also The copper-bearing rocks of Lake Superior. Idem, p. 221. Notes on observations, 1883, on the geology of the north shore of Lake Superior, by A. R. C. Selwyn. Trans. Roy. Soc. Canada, vol. 2, sec. 4, 1885, p. 245. 47517°— VOL 52—11 6 82 GEOLOGY OF THE LAKE SUPERIOR REGION. Report on the geology of the Lake of the Woods region, with s])ecial reference to the Keewatin (Eluronian?) belt of Archean rocks, by A. C. Lawson. Ann. Kept. Geol. and Nat. Ilist. Survey Canada for 1885, new ser., vol. 1, pp. ■5-1.51 cc, with a map. Geology and lithology of Michipicoten Bay, by C. L. Herrick, W. G. Tight, and 11. L. Jones. Bull. Denison Univ., vol. 2, 188(), pp. 120-144, with 3 plates. Thecorrelationof the Aniraikrie and Huronian rocks of Lake Superior, by Peter McKellar. Proc. and Trans. Roy. Soc. Canada, vol. 5, sec. 4, 1887, pp. 63-73. Report of the geology of the Rainy Lake region, by A. C. Lawson. Ann. Rept. Geol. and Nat. Hist. Survey Canada for 1887-88, new ser., vol. 3, pp. 1-196 f, with 2 maps and 8 plates. See also The Archean geology of the region north- west of Lake Superior. Etudes sur les schistes cristallins. Internat. Geol. Cong., London, 1888, pp. 66-88. Geology of the Rainy Lake region, with remarks on the classification of the crystalline rocks west of Lake Superior; prelim- inary note. Am. Jour. Sci., 3d ser., vol. 33, 1877, pp. 473-480. Report on mines and mining on Lake Superior, by B. D. Ingall. Ann, Rept. Geol. and Nat. Hist. Surs'ey Canada for 1887-88, new ser., vol. 3, pp. i-131 n, \vith 2 maps and 13 plates. Tracks of organic origin in rocks of the Animikie group, by A. R. C. Selwyn. Am. Jour. Sci., 3d ser., vol. 39, 1890, pp. 145-147. The internal relations and taxonomy of the Archean of central Canada, by Andrew C. Lawson. Bull. Geol. Soc. America, voL 1, 1890, pp. 175-194. Geology of Ontario, with special reference to economic minerals, by Robert Bell. Rept. Roy. Comm. on Min. Res. Ontario, Toronto, 1890, pp. 1-70. Lake Superior stratigraphy, bj' Andrew C. Lawson. Am. Geologist, vol. 7, 1891, pp. 320-327. The structural geology of Steep Rock Lake, Ontario, by Henry Lloyd Smyth. Am. Jour. Sci., 3d ser., vol. 42, 1891, pp. 317-331. Report on the geology of Hunters Island and adjacent country, by W. H. C. Smith. Ann.' Rept. Geol. Survey Canada for 1890-91, vol. 5, pt. 1, G, 1892, pp. 5-76. The Archean rocks west of Lake Superior, by W. H. C. Smith. Bull. GeoL Soc. America, vol. 4, 1893, pp. 333-348. The laccolitic sills of the northwest coast of Lake Superior, by A. C. Lawson. Bull. Cieol. and Nat. Hist. Survey Minnesota No. 8, 1893, pp. 24-48. Multiple diabase dike, by A. C. Lawson. Am. Geologist, vol. 13, 1894, pp. 293-296. Note on the Keweenawan rocks of Grand Portage Island, north coast of Lake Superior, by U. S. Grant. Idem, pp. 437-438. Gold in Ontario; its associated rocks and minerals, by A. P. Coleman. Fourth Rept. Bur. Mines Ontario, for 1894, sec. 2, Toronto, 1895, pp. 3.5-100, with 2 geologic maps of parts of the Rainy River district. The hinterland of Ontario, by T. W. Ciibson. Idem, sec. 3, pp. 124-125. The new Ontario, by Archibald Blue. Fifth Rept. Bur. Mines Ontario, for 1895-96, pp. 193-196. A second report on the gold fields of western Ontario, by A. P. Coleman. Idem, sec. 2, pp. 47-106. The anorthosites of the Rainy Lake region, by A. P. Coleman. Jour. Geology, vol. 4, 1896, pp. 907-911; Canadian Rec. Sci., vol. 7, 1897, pp. 230-235. Malignite, a family of basic plutonic orthoclase rocks rich in alkalies and lime, by Andrew C. Lawson. Bull. Dept. Geology Univ. California, vol. 1, 1896, pp. 337-362, pi. 18. Third report on the west Ontario gold region, by A. P. Coleman. Rept. Bur. Mines Ontario, vol. 6, 1897, pp. 71-124. The Michipicoten mining divL^^ion, by A. B. Willmott. Idem, vol. 7, 1898, pp. 184-206. Geology of base and meridian lines in the Rainy River district, by W. A. Parks. Idem, pp. 161-183, with geologic map. Clastic Huronian rocks of western Ontario, by A. P. Coleman. Idem, pp. 151-160; Bull. Geol. Soc. America, vol. 9, 1898, pp. 223-238. Unpublished field notes by C. R. Van Hise, 1898. The geology of the area covered by the Seine River and Lake Shebandowan map sheets, comprising portions of Rainy River and Thunder Bay districts, Ontario, by Wm. Mclnnes. Ann. Rept. Geol. Survey Canada, vol. 10, pt. H, 1899, pp. 13-51, with geologic map. Copper regions of the upper lakes, by A. P. Colejnan. Rept. Bm-. Mines Ontario, vol. 8, pt. 2, 1899, pp. 121-174. Copper and iron regions of Ontario, by A. P. Coleman. Idem, vol. 9, 1900, pp. 143-191. Upper and lower Huronian in Ontario, by A. P. Coleman. Bull. Geol. Soc. America, vol. 11, 1900, pp. 107-114. Unpublished field notes by C. R. A'an Hise and J. Morgan Clements, summer of 1900. The iron belt on Lake Nipigon, by J. W. Bain. Rept. Biu-. Mines Ontario, vol. 10, 1901, pp. 212-214. Iron ranges of the lower Huronian, by A. P. Coleman. Idem, pp. 181-212. The Michipicoten Huronian area, by A. B. Willmott. Am. Geologist, vol. 28, 1901, pp. 14-19. The Michipicoten iron range, by A. P. Coleman and A. B. Willmott. Univ. Toronto studies, geol. ser.. No. 2, 1902, 47 pp. See also Rept. Bur. Mines Ontario, 1902, pp. 152-185. Rock basins of Helen mine, Michipicoten, Canada, by A. P. Coleman. Bull. Geol. Soc. America, vol. 13, 1902, pp. 293-304. HISTORY OF GEOLOGIC WORK IN THE REGION. 83 Nepheline and other syenites near Port Coldwell, Ontario, by A. P. Coleman. Am. Jour. Sci., 4th ser., vol. 14, 1902, pp. 147-155. See also Rept. Bur. Mines Ontario, 1902, pp. 208-213. Region southeast of Lac Seul, by William Mclnnes. Summary Rept. Geol. Survey Canada for 1901-2, pp. 87-93. The country west of Nipigon Lake and River, by Alfred W. G. Wilson. Idem, pp. 94-103. The country east of Nipigon Lake and River, by W. A. Parks. Idem, pp. 103-107. Iron ranges of northwestern Ontario, by A. P. Coleman. Rept. Bur. Mines Ontario, 1902, pp. 128-151. Iron ranges of northern Ontario, by W. G. Miller. Idem, 1903, pp. 304-317. Region lying northeast of Lake Nipigon, by W. A. Parks. Summary Rept. (ieol. SlU'vey Canada for 1902-3, pp. 211-220. Region on the northwest side of Lake Nipigon, by William Mclnnes. Idem, pp. 206-211. Nepheline syenite in western Ontario, by W. G. Miller. Am. Geologist, vol. 32, 1903, pp. 182-185. Genesis of the Animikie iron range, Ontario, by F. Hille. Jour. Canadian Min. Inst., vol. 6, 1904, pp. 245-287. The Animikie or Loon Lake iron-bearing district, by W. N. Smith (in charge of a party consisting of A. W. Lewis, J. U. Warner, G. W. Crane, and R. C. Allen). Min. World, vol. 22, 1905, pp. 206-208, with geologic map. Iron ranges of Michipicoten West, by J. M. Bell. Rept. Bur. Mines Ontario, vol. 14, 1905, pt. 1, pp. 278-355, with geologic map. See also The possible granitization of acidic lower Iluronian schists on the north shore of Lake Superior. Jour. Geology, vol, 14, 1906, pp. 233-242. The geology of Michipicoten Island, by E. N. Burwash. Univ. Toronto studies, geol. ser.. No. 3, Toronto, 1905, with map. Pre-Cambrian nomenclature, by A. P. Coleman. Jour. Geology, vol. 14, 1906, pp. 60-64. The Animikie iron range, by L. P. Silver. Rept. Bur. Mines Ontario, vol. 15, 1906, pt. 1, pp. 156-172. Iron ranges east of Lake Nipigon, by A. P. Coleman. Sixteenth Ann. Rept. Bur. Mines Ontario, 1907, pt. 1, pp. 105-135. Iron ranges eaat'of Lake Nipigon, the ranges around Lake Windebegokan, by E. S. Moore. Idem, pp. 136-148. Iron ranges of Nipigon district, by A. P. Coleman. Eighteenth Ann. Rept. Bur. Mines Ontario, 1909, pt. 1, pp. 141-153. Iron range north of Round Lake, by E. S. Moore. Idem, pp. 154-162. Geology of Onaman iron range area, by E. S. Moore. Idem, pp. 196-253. The quartz diabases of the Nipissing district, Ontario, by W. H. Collins. Econ. Geology, vol 5, 1910, pp. 538-550. Diabase and granophyre of the Gowganda Lake district, Ontario, by Norman L. Bowen. Jour. Geology, vol. 18, 1910, pp. 658-674. LAKE SUPERIOR REGION (GENERAL). Narrative journal of travels through the northwestern regions of the United States, extending from Detroit through the great chain of American lakes to the sources of the Mississippi River, by Henry R. Schoolcraft. Albany, 1821, 419 pp., with map. Report of Walter Cunningham, late mineral agent on Lake Superior, January 8, 1845. Senate Docs., 2d sess. 28th Cong., 1844-^5, vol. 7, No. 98, 5 pp. Mineral report, by George N. Sanders. Idem, No. 117, pp. 3-9. Report of J. B. Campbell. Idem, vol. 11, No. 175, pp. 4-8. Report of George N. Sanders. Idem, pp. 8-14. Report of A. B. Gray. Idem, pp. 15-22. Report of A. B. Gray on mineral lands of Lake Superior. Executive Docs., 1st sess. 29th Cong., 1845^6, vol. 7, No. 211, 23 pp., with map. On the origin of the actual outlines of Lake Superior (discussion), by William B. Rogers. Proc. Am. Assoc. Adv. Sci., 1st meeting, 1848, pp. 79-80. The outlines of Lake Superior, by Louis Agassiz. Lake Superior; its physical character, vegetation, and animals compared with those of other and similar regions, by Louis Agassiz and J. Elliot Cabot, pp. 417-426. See also Proc. Am. Assoc. Adv. Sci., 1st meeting, 1848, p. 79. Abstract of an introduction to the final report of the geological siu'veys made in Wisconsin, Iowa, and Minnesota, in the years 1847, 1848, 1849, and 1850, containing a synopsis of the geological featm-es of the country, by Da%'id D. Owen. Proc. Am. Assoc. Adv. Sci., vol. 5, 1851, pp. 119-131. On the age, character, and true geological position of the Lake Superior red sandstone formation, by Da^dd D. Owen. Report of a geological survey of Wisconsin, Iowa, and Minnesota, 1852, pp. 187-193. Report of a geological survey of Wisconsin, Iowa, and Minnesota, and, incidentally, of a portion of Nebraska Territory, made under instructions from the United States Treasury Department, by David D. Owen. 1852, 638 pp. A geological map of the United States and the British Provinces of North America, with an explanatory text, geological sections, etc, by Jules Marcou, Boston, 1853, 92 pp. See also Reponse a la lettre de MM. Foster et Whit- ney sur le Lac Superieur. Bull. Soc. g^ol. France, 2d ser., vol. 8, 1851, pp. 101105.- The metallic wealth of the United States, by J. D. Wliitney. Philadelphia, 1854, 510 pp. Observations on the geology and mineralogy of the region embracing the sources of the Mississippi River, and the Great Lake basins, during the expedition of 1820, by Henry R. Schoolcraft. Summary narrative of an exploratory 84 GEOLOGY OF THE LAKE SUPERIOR REGION. expedition to the sources of tlie Mississippi River in 1820, resumed and rompleted by the discovery of its orijjin in Itasca Lalce in 1832. Philadelphia, 1854, pp. 303-362. Remarks on some points connected with the geology of the north shore of Lake Superior, by J. D. Whitney. Proc. Am. Assoc. Adv. Sci., vol. 9, 1856, pp. 204-209. On the occurrence of the ores of iron in the Azoic sy.stem, by J. D. Whitney. Idem, pp. 209-216. Remarks on the Iluronian and Laurentian systems of the Canada Geological Survey, by J. D. \\1iitney. Am. Jour. Sci., 2d ser., vol. 23, 1857, pp. 305-314. Physical geology of Lake Superior, by Charles ^\liittlesey. Proc. Am. Assoc. Adv. Sci., vol. 24, 1876, pt. 2, pp. 60-72, mth map. The copper-bearing rocks of Lake Superior, by R. D. Irving. Mon. U. S. Geol. Survey, vol. 5, 1883, 464 pp. 15 1., 29 pis. and maps. See also Third Ann. Rept. U. S. Geol. Survey, 1883, pp. 89-188, 15 pis. and maps; Science, vol. 1, 1883, pp. 140, 359, 422; Am. Jour. Sci., 3d ser., vol. 28, 1884, p. 462; vol. 29, 1885, pp. 67-68, 2-58-259, 339-340. The copper-bearing series of Lake Superior, by T. C. Chamberlin. Science, vol. 1, 1883, pp. 453—1.55. On secondary enlargements of mineral fragments in certain rocks, by R. D. Irving and C. R. Van Ilise. Bull. U. S. Geol. Survey No. 8, 1884, 56 pp., 6 pis. Di\'isibility of the Archean in the Northwest, by R. D. Irving. Am. Jour. Sci., 3d ser., vol. 29, 1885, pp. 237-249. Preliminary paper on an investigation of the Archean formations of the Northwestern States, by R. D. Irving. Fifth Ann. Rept. U. S. Geol. Survey, 1885, pp. 175-242, 10 pis. Origin of the ferruginous schists and iron ores of the Lake Superior region, by R. D. Irving. Am. Jour. Sci., 3d ser., vol. 32, 1886, pp. 255-272. Is there a Iluronian group? by R. D. Irving. Am. Jour. Sci., 3d ser., vol. 34, 1887, pp. 204-216, 249-263, 365-374. A great Primordial quartzite, by N. H. Winchell. Am. Geologist, vol. 1, 1888, pp. 173-178. See also Seventeenth Ann. Rept. Geol. and Nat. Hist. Survey Minnesota, for 1888, pp. 25-56. On the classification of the early Cambrian and pre-Cambrian formations, by R. D. Irving. Seventh Ann. Rept. U. S. Geol. Survey, 1888, pp. 365-454, with 22 pis. and maps. The iron ores of the Penokee-Gogebic series of Michigan and Wisconsin, by C. R. Van Ilise. Am. Jour. Sci., 3d ser., vol. 37, 1889, pp. 32^8, with plate. An attempt to harmonize some apparently conflicting views of Lake Superior stratigraphy, by C. R. Van Hise. Idem, vol. 41, 1891, pp. 117-137. The Norian rocks of Canada, by A. C. Lawson. Science, vol. 21, 1893, pp. 281-282. The Norian of the Northwest, by N. H. Winchell. Bull. Geol. and Nat. Hist. Survey Minnesota, No. 8, 1893, pp. iii-xxii. An historical sketch of the Lake Superior region to Cambrian time, by C. R. Van Hise. Jour. Cieology, vol. 1, 1893, pp. 113-128, with geologic map. ■ Crucial points in the geology of the Lake Superior region, by N. 11. Winchell. Am. Geologist, vol. 15, 1895, pp. 153-162, 229-234, 295-304, 356-363; vol. 16, 1895, pp. 12-20, 75-«6, 150-162, 269-274, 331-337. See also Compt. Rend. Congrfes gfol. intemat., 6th sess. (1894), 1897, pp. 273-308. Pre-Cambrian fossiliferous formations, by Charles D. Walcott. Bull. Geol. Soc. America, vol. 10, 1899, pp. 199-244. The iron-ore deposits of the Lake Superior region, by C. R. Van Hise, assisted in Mesabi and Vermilion sections by C. K. Leith and J. Morgaii Clements, respectively. Twenty-first Ann. Rept. U. S. Geol. Survey, pt. 3, 1901, pp. 305- 434, with geologic maps. Geological work in the Lake Superior region, by C. R. Van Hise. Proc. Lake Superior Min. Inst., vol. 7, 1902, pp. 62-69. The original source of the Lake Superior iron ores, by J. E. Spurr. Am. Geology, vol. 19, 1902, pp. 335-349. A comparison of the origin and development of the iron ores of the Mesabi and Gogebic iron ranges, by C. K. Leith. Proc. Lake Superior Min. Inst., vol. 7, 1902, pp. 75-81. The Eparchean interval ; a criticism of the use of the term Algonkian, byAndrew C. Lawson. Bull. Dept. Geology Univ. California, vol. 3, 1902, pp. 51-62. The Iluronian question, by A. P. Coleman. Am. Geology, vol. 29, 1902, pp. 325-334. The nomenclature of the Lake Superior formations, by A. B. Willmott. Jour. Geology, vol. 10, 1902, pp. 67-76. Report of the special committee for the Lake Superior region, by C. R. Van Hise and others. Jour. Geologj-. vol. 13, 1905, pp. 89-104; Rept. Ontario Bur. Mines, vol. 14, pt. 1, 1905, pp. 269-277; Rept. Geol. Survey Michigan for 1904, 1905, pp. 133-143. Report of the special committee for the Lake Superior region, personal comments, by A. C. Lane. Ann. Repl. Geol. Survey Michigan for 1904, 1905, pp. 143-153. See also Comment on the report of the special committee on the Lake Superior region, Jour. Geology, vol. 13, 1905, pp. 457-461. A summary of Lake Superior geology with special reference to recent studies of the iron-bearing series, by C. K. Leith. Trans. Am. Inst. Min. Eng., vol. 36, 1906, jip. 101-153, with geologic map. The movement of Lake Superior iron ores in 1909, with a map showing distribution of ores, by John Birkinbine. Advance chapter from Mineral Resources U. S. for 1909, U. S. Geol. Survey, 1910, 7 l)p. \n Algonkian basin in Hudson Bay — a comparison with the Lake Superior basin, by C. K. Leith. Ecou. Geol- ogy, vol. 5, 1910, pp. 227-240. CHAPTER IV. PHYSICAL GEOGRAPHY OF THE LAKE SUPERIOR REGION. By Lawkkxce Martin. TOPOGRAPHIC PROVINCES. The Lake Superior region as described in this report inchides three topos^rapliic provinces (fig. 5) — (1) the Lake Superior highlands, a peneplain with hilly upland and lowland subdi- visions; (2) a series of lowland plains surrounding the peneplain on the east, south, and west; and (3) the deep basin of Lake Superior embraced between parts of the liighland and the low- land. These three topographic provinces are in various stages of development and preservation, depending on the underlying rock structure, the process by which they are being modified, and the length of their period of development. The first consists essentially of Archean and Algon- kian rocks; the second of Cambrian and other early Paleozoic rocks and of Cretaceous rocks; the third is a present seat of rock deposition, and probably includes rocks of all ages represented in the other provinces, in addition to the glacial drift of the Quaternary, which also partly man- tles the rocks in the first province and almost completely buries those of the second. The peneplain liighland was worn down from former lofty mountains." Diastropliism (warp- ing, folding, and faulting) has notably modified the peneplain, tilting its borders and introducing the deep basin of Lake Superior. (See PI. II.) Subsequent deposition of early Paleozoic and Cretaceous rocks in the Lake Superior basin and about the margin of the peneplam (see fig. 5) has been followed by the exhuming of fossil topography and the production of a belted plain with alternate uplands and lowlands in the region of horizontal and gently tilted post-Algonkian rocks. Continental glaciation has slightly modified the relief and completely altered the soil and drainage of the region (Chapter XVI, pp. 427-459). THE LAKE SUPERIOR HIGHLANDS. . TOPOGRAPHIC DEVELOPMENT. The highlands about Lake Sujierior fall into two classes — (1) those underlain by coarse- grained homogeneous rocks, chiefly igneous, of both Archean and Algonkian age, and (2) those underlain by banded (both areally and structurally) alternating weak and resistant tilted rocks, chiefly sediments and lavas of Algonkian age. The areas of homogeneous igneous rocks still preserve plateaus or liigh plains of slight relief, diversified only by monadnocks and by some valleys of greater than normal depth; the areas including belts of sediments have narrow pla- teaus, monoclinal ridges, and mesas isolated among broader .intermediate lowlands. It is possible that the whole highland area was reduced to a peneplain, now represented by the plateau surfaces, the crests of some of the higher monoclinal ridges, and the tabular surfaces of tlie higher mesas, none of the adjacent lowland areas having been down-warped or down- faulted or excavated when the peneplain was most nearly perfected. c Van Hise, C. R ., Science, new ser., vol . 4, 1896, pp. 57-59 and 217-220; Weidman, Samuel, Jour. Geology, vol. 11, 1903, pp. 289-313; Wilson, A. W. G., Jour. Geology, vol. 11, 1903, pp. 015-667; Weidman, Samuel, Bull. Wisconsin Geol. and Nat. Hist. Survey No. 16, 1907, pp. 592-603 and 385-395. 85 86 GEOLOGY OF THE LAKE SUPERIOR REGION. Diastropliism during post-Algonldan time, by changing the altitude of tlie peneplain with reference to base-level, enabled demidiitioti to reattack this peneplain. Stream erosion was renewed actively along the fault escarpments, jjossibly being delayed in areas that had been submerged and buried Ijy Paleozoic sediments (p. 116). This renewal of cutting was weak or not yet active at all in regions remote from the escarpments (here also possibly being delayed in the buried and protected parts), but was strongest in the areas of banded Algonkiau rocks, especially those near the steeper slopes. In these areas of banded rocks the remnants of the original peneplain surface are small and scattered, being largest where the vertical beds resisted erosion best, smaller wliere gentle tilting made development of monoclinal ridges and interme- diate valleys possible, and of least extent where horizontal beds allowed the opening of broad lowlands with only isolated mesas, as in the Thunder Bay region, or with protruding reexposed knobs, like the Baraboo range of Wisconsin and knobs north and east of it (figs. 5.3, 54, pp. 359, 360). The lowlands developed at several points may be incipient stages of a peneplain of a later generation, developed with respect to a much lower base-level. The older penejilain surface is found at various altitudes, some of which are shown in the fol- lowing table; Altitude of different parts of the Lake Superior hir/hlands. Localitv. Average height above sea level.a Feet. Southeast of Michipicoten 1.500—1,000 Near Mic-hipkoten 'l. 2(»^1, 400 Northwest of Michipicoten ,*1, 200— 1,400 Near Heron Bay «!.10O-l,.3a0 North of Lake Superior * 900—1,050 West of Lalte Nipigon «1. 2.50— 1,500 Thunder Bay and Hunters Island region *1, 400— 1, 700 Rainy Lake and Lake of the Woods region *1.200— 1, 400 Gunfiint Lake , 1,800-2,000 VermiUon district I 1. IM>— 1 . 700 Mesabi district j 1,400—1.500 Gabbro plateau 1,400—1,700 Northern Wisconsin *I, 400— 1,500 Keweenaw Point About 1..350 Marquette dLstrict 1,400-1,600 Crystal Fails district ' 1, 400— 1, 600 Menominee district 1 1 . 200 — 1 , 400 North-central Wisconsin ' 1, .300— 1 , 500 Edge of Potsdam sandstone *About 1, 000 Highest hill. Feet. 1,700 2,120 Lowest valley. Feet. ± 1,100 1,700 2 232 i!910 1,920 2, 320 1,900 1,409 *1,950 1,900 1,370 1.940 1,072 1,547 1.300 1.400 1.400 1,400 ± i.mo 1,120 800 1,100 « Altitudes marlced with an asterisk are accurate approximations based upon railway grades, etc. All other altitudes are averaged from accurate topographic maps. It will be noted (figs. 4 and 5) that the general peneplain surface lies between 1,000 and 1,700 feet, though it is a trifle low^er locally, and rises in monadnocks to exceptional heights of a little more than 2,300 feet. The maximum relief of the peneplain proper (excluding the basin of Lake Superior) is less than 1,450 feet (900 to 2,320), and these extremes are many miles apart. The maximum local relief of any part of the peneplain at the time of its greatest perfection may be quite safely placed between 400 and 500 feet, and the average relief would be much less, perhaps 100 to 200 feet. The present differences of elevation in the peneplain remnants might be explainetl as inherited, for the writer does not conceive of peneplains as approacliing at all closely to a plane or perfectly base-leveled surface. Possibly the peneplain in the Lake Superior region when most nearly perfect stood at levels perhaps corresponding to present elevations of 1,400 feet in central Wisconsin, 1,350 feet on Keweenaw Point, 1,600 feet in northeastern Minnesota, and 1,400 feet northeast of Lake Superior in Canada, etc. Because there was upon the well- developed peneplain a series of old streams whose valleys laj- at lower levels than the low intermediate ridges and at slighth' different levels with reference to one another, the surface beveled back smoothly up the stream courses antl the Unes Final Rcpt. Geol. and Nat. Hist. Survey Minnesota, vol. 4, pp. 212, 2G5. c Idem, pp. 207, 317. > Idem, pp. 36, 278, 299. c Geology ol the Lake of the Woods region: .\nn. Kepi, r.eol. and Nat. Hist. Survey Canada for 1885, new ser., vol. 1, 1886, Rept. CC, p. 22. i Geology of the Rainy Lake region: .\nn. Rept. (Seol. and Nat. Hisl. Survey Canada for 1887-88, vol. 3, new ser., 1889, Rept. F, p. 10. t Op. cit., vol. 1, Rept. CC, pp. 15-25 and 2f.-2.S; vol. 3. Rept. F, pp. 10-20. / Geology of Hunters Island and adjacent country: Ann. Rept. Geol. Survey Canada for 1890-91, new ser.. vol. 5, 1893, Rept. O, pp. 9-11. a Geology of the area covered by the Seine Uiver and Lake Shebandowan map sheets: Ann. RepL Geol. Survey Canada for 1S97, new ser., vol. 10, 1899, Rept. U, pp. 0-10. PHYSICAL GEOGRAPHY OF THE REGION. 95 jaspilites, and iron formation and certain schists form ridges rising at most 300 feet above tlio neigliboring lakes, whose greatest depth is 280 feet; the other Arclaean rocks are "characterized by low, rouniled hills, with softened outlines." REGION NOKTII OF LAKE SUPERIOR. W. H. Collins " describes the region between Nipigon Ba}^ and Heron Bay and northward to the Height of Land as "a peneplain of rounded hills of crystaUine rocks 300 to 400 feet high, terminating abruptly along the south, " and with steeply descending streams affording excellent water power. Collins '' also describes the Archean area north of the Canadian Pacific Railway and west of Lake Nipigon as possessing "a surface of low relief and moderate altitude." Water levels vary from 1,149 to 1,382 feet. Few hills reach 250 feet in height. The sky line is exceedingly even. The area also possesses the linear topography of the Algonkian in places and the mesa topography of the Keweenawan near Lake Nipigon and to the west. REGION NORTHEAST OF LAKE SUPERIOR. J. M. Bell " has characterized the region north of Lake Superior and west of the Michipicoten district to Heron Bay as hilly, with greater ranges of relief than elsewhere in the Laurentian peneplain, witli valleys opened on weak rocks, ridges formed on resistant beds, and with monad- nocks rising above the general peneplain level on the site of the still more resistant beds. MICHIPICOTEN DISTRICT. The part of the peneplain that includes the Michipicoten district has been described as follows:'' The topography is of the rugged character usual on the north shore of Lake Superior, and Hematite Mountain, the highest point, rises 1,100 feet above the lake within a distance of 7 miles. In general the hills form steep ridges with a direction of about 70° east of north, corresponding to the strike of the schists, and traveling is difficult across the line of strike. * * * From the summit of Hematite Mountain, which is situated about in the middle of the region and rises 200 feet above any of its neighbors, there is presented more than the usual variety of surface, including long ridges of Huronian schist, rounded hills of eruptives, which sometimes rise like islands out of lacustrine plains, stretches of the hummocky surface so common in glaciated Archean districts, lake basins, rock rimmed or bordered with muskeg, rivers with lakelike stretches of dead water, tumultuous rapids over morainic bowlders and falls over rocky descents, and, finally, the splendid promontories of the shore of Lake Superior. * * * The intimate dependence of the topography on the geological history of the country is well brought out in the Michipicoten region, where the folding of the schists has determined the direction and steepness of the main ranges of hills; while bosses and irregular masses of eruptives give rise to less uniform hills associated with the ridges or standing isolated. The basis of the topography is to be found in the pre-Cambrian arrangement and the varying power of resistance to weathering and erosion shown by the different rocks; so that the prominent features may be of very ancient date, even Paleozoic. REGION NORTH OF SAULT STE. MARIE. In the upland north of Sault Ste. Marie and east of Michipicoten relief of as much as 100 to 200 feet is common. Nearer the lake and southeast of Michipicoten there are several very deep valleys, notably those of Agawa, Montreal, Batchawana, Chippewa, and Goulais rivers. Owing to the considerable rehef , some very liigh and expensive trestles will' be required where the Algoma Central and Hudson Bay Railway is to cross the first three rivers mentioned; and the building of the railway heyond Pangissin has been hindered by the necessity of high steel bridges, though the railway is graded all the way to the Michipicoten district. Such expense in railroad building in the Lake Superior region away from the lake shore is distinctly excep- tional and indicates the high degree. of the local relief. tt Summary Rept. Geol. Survey Canada for 1905-0, pp. 80-81. i> Idem, p. 103. ••Rept. Bur. Mines Ontario, vol. 14, 1305, pt. 1, pp. 281-299. d Coleman, \. P., and Willinott, A. B., The Michipicoten iron ranges: Univ. Toronto studies, Geol. ser., 1902, pp. 4-6; also Eleventh Rept. Bur. Mmes Ontario, 1902, pp. 153-154. 96 GEOLOGY OF THE LAKE SUPERIOR REGION. The areas between these deep valleys are broad aiul relatively Hat or round topped, and some of the hills "present steep slopes toward the valleys and often dropoff in impassable cliffs 100 feet or more in height. None of the hills rise much over 1,000 feet above Lake .Superior, but many reach 900 feet "" (1,500 to 1,600 feet above sea level). The surface bevels indifferently across variously durable structures of gneiss, schist, and granite in a characteristic peneplain sur- face, with the usual nionadnocks. The deep valleys resemble those of the north shore of Lake Superior, which arc crossed near their moutlis by expensive britlges and trestles of tiie Canadian Pacific Railway ; in both regions they are deep cut because of the low adjacent base-level of Lake Superior. MARQUETTE DISTRICT.'' North of ^larquette the granite area forms a monadnock group known as tlie Huron Moun- tains, rising about 1,200 to 1,350 feet above the lake.*^ The elevations were thus described by Foster and Whitney: They do not range in continuous chains, but exist in groups radiating from a common center, presenting a series of knobs rising one above another until the summit level is attained. Their outline is rounded or waving, their slope gradual. The scenery is tame and uninteresting. C. A. Davis "* writes with regard to the same region: The hills are only 150 or 200 feet above the valleys, hence the general level is relatively high and the district is a plateau, or high peneplain, rather than mountainous. The granite of the Archean south of Marquette was early described by Brooks as having an irregular topograph j', with low knobs, ridges, and cliffs.* Rominger contrasts the area south of Marquette,^ where the granites occupy lower levels than the Huronian, with the northern granite outcrops, wliich "occupy the highest elevations and constitute the most conspicuous ridges." The topography (see topographic map and structure profiles, PI. XVII, in pocket) characteristic of the Archean formations in this district has been described in greater detail by C. R. Van Hise and W. S. Bayley s as follows: Northern complex: Mona schists: Minor rugged hills, strongly glaciated. Kitchi schists: Rugged hills similar to those of Mona schist. Gneissoid granites: Rounded knobs, invariably smoothed by glaciition. Hornblende syenite: Exactly like that of granite. Southern complex: Knobs, as in northern granite areas. MENOMINEE DISTRICT.* W. S. Bayley ' has described the topography associated witli the various Archean rock series in the Menominee district (PI. XXVI, in pocket) as follows: Quinnesec schist (southern area): Rough and broken, forming deep gorges, with many ridges and elongated hills. Quinnesec schist (western area): Without distinctive peculiaiities except small rugged knobs. Granites, gneisses, and schists of northern comijlcx: Irregular rugged knolls, intensely glaciated. CRYSTAL FALLS DISTRICT..' The topography characteristic of the Archean in the Crj'stal Falls district (PI. XXII, in pocket) has been described by J. M. Clements, H. L. Smyth, and W. S. Bayley * as follows: Granite: Small rounded isolated knobs, chiefly obscured by glacial drift (gaps in granite range where resistant greenstone dikes cross) . a Coleman, A. P.. Rept. Bur. Mines Ontario, vol. 15, pt. 1, 1906, pp. 175-177. b For topography of Marquette and adjacent districts see also the chapters on these di-stricts. r Report on the geology and topography of the Lake Superior land district, ISoO, pt. 1, p. 34. It ..>,nn. Rept. Geol. Survey Michigan, 1900, p. 2(i0. e Brooks, T. B., Geol. Survey Michigan, vol. 1, 1873, pp. 72-73. / Rominger, Carl, Geol. Survey .Vlidiigan, vol. i, ISSl, p. 1.3. e The Marquette iron-hearing district of Michigan: Mon. V. S. Geol. Survey, vol. 28, 1895, pp. 152. 102, 170, 170, 191. ft See also chapter on Menominee district, where topography is discussed. ( The Menominee iron-hearing district of Michigan: Mon. U. S. Geol. Survey, vol. 46, 1904, pp. 132, 159, 16^. > See also chapter on Crystal Falls district, where topography is discussed. t The Crystal Falls Iron-bearing district of Michigan: Mon. U. S. Geol. Survey, vol. 30, 1899 (western, p. 38; eastern, pp. 329, 386. 428, and 463). PHYSICAL GEOGRAPHY OF THE REGION. 97 Archean crystallines: Mammillated with rocky knobs separated by bowl-like depressions, the hummocks and bowls being generally elongated east and west. Granites, gneisses, schists, and amphibolites of Felch Mountain district: Characteristic rough topography with east-west elongated hummocks and bowls. A topographic depression always exists along the contact of the Archean and Algonkian, usually holding a swamp or stream. Gneissoid granites and various schists of Sturgeon River tongue: Scattered bare knolls. West of the Crystal Falls, Menominee, and Marquette districts (fig. 43, p. 292; PL XXIV, in pocket) there is a general plain produced by erosion upon the homogeneous slates, in places deeply cut by streams antl partly obscured by the glacial drift. Through both slates and drift certain knobs of resistant greenstone, etc., project as eminences. KEWEENAW POINT. On Keweenaw Point the highland peninsula, generally referred to in atlases and maps as the Copper Range, has rocks vertical or very highly inclined. Erosion has thus far been unable to significantly alter the plateau " or peneplain '' which was developed on these inclined beds in the period of base-leveling. This is the case on the part of Keweenaw Point (fig. 59, p. 422) that extends southward from Gratiot River to Portage Lake, where the ridges of the eastern tip of the point, as described by Irving, merge into "one broad swell" or "a broad central ridge" which extends west as far as the Porcupine Movmtams, beyond which it resumes its continuity to the neighborhood of Bad River, Wisconsm. Upon this long, narrow plateau relief is not wanting, small monadnocks rising above the general level, which other^vise bevels indifferently across the various weak and resistant beds. This plateau surface is also diversified between Porcupine Mountams and Bad River by "rounded ridges and knobs with cliffs facmg indiffer- ently in all directions." It is still, however, essentially a peneplain, the valleys cut in it not havmg notably dissected its surface into distinctive forms like monoclinal ridges or mesas. To the northeast, at the tip of Keweenaw Pomt, there are monoclinal ridges and longitudinal valleys, replacing the former peneplain surface, above whose level monadnocks like Mounts Houghton and Bohemia still rise, the former owing its emmence to a resistant red felsite.'' In the plateau region, where the dips have prevented equally rapid dissection, the peneplain surface remains. It is marginally cut by deep gorges, to be sure, but these valleys are of mod- erate area and are not separated by monoclinal ridges or by mesas, such as occur where the dips are below 30° or nearly horizontal respectively. Minor monadnocks rise everywhere above the partty dissected peneplain. The moderate elevation on the south shore of Lake Superior known as the Porcupine Mountams "^ forms a monachiock area rising 600 to 1,421 feet above the lake and averaging 1,800 feet above sea level. The highest pomt is 2,023 feet. These mountains owe their relief to the resistant quality of a body of quartz porphyries and felsites here faulted up against the adjacent weaker beds on the south and exposed by denudation. That they form a group of monadnocks was fii'st noted by Van Hise.'' NORTHERN WISCONSIN. R. D. Irving ^ in 1878 briefly described the topography of the Archean area south of the Penokee-Gogebic range as the "elevated interior" or "interior table-land," with a gently undulatmg surface, few ledges, low granite domes, and abundant glacial lakes and swamps. In their report on the Penokee-Gogebic district Irving and Van Hise f have not specifically described the topography associated with the north edge of the peneplain within that district, but the Archean gneisses and schists there may be mferred to have characteristic knobby topog- raphy (PI. XVI, p. 226). a Brooks, T. B., Geol. Survey Michigan, vol. 1, 1873, pp. 69-70; Irving, R. D., Mon. U. S. Geol. Survey, vol. 5, 1883, pp. 164-166, 186. h Van Hise, C. R., Science, new ser., vol. 4, 1896, p. 217. c Irving, R. D., Mon. U. S. Geol. Survey, vol. S, 1883, pp. 181-182. i Idem, pp. 206-225, and geologic section 3, pi. 20. Also Wright, F. E., Ann. Rept. Geol. Survey Michigan, 1903, pp. 35-44. « The geology of the eastern Lake Superior district: Geology of Wisconsin, vol. 3, 1S7.V1S79, pp. 61-1)2, pi. 11. / The Penokee iron-bearing series of Michigan and Wisconsin: Mon. U. S. Geol. Survey, vol. 19, 1892, p. 104. 47517°— VOL 52—11 7 98 GEOLOGY OF THE LAKE SUPEKlOli liEGION. CENTRAL WISCONSIN. R. D. Irving "■ wrote as follows regarding the topography of central Wisconsin: The region of crystiilliiu' rocks (Archeau and lluronian) of north-rcntral Wisconsin, descending gradnally south- ward, has a gently undulating surface, which is, however, often broken in minor detail by low. abrupt ridges with outcropping tilted rock ledges. Weifhnan'' has described the topography associated with the Archoaii formation in north- central Wisconsin (Pis. IV, A, p. 90; XXXI, i\., p. 436) as follows: The basal group (gneiss and schists) forms a gently sloping plain, with low crystalline ledges sometimes thinly covered by sandstone, sometimes by glacial drift, but generally exposed in the ri\-er beds. The Hiironian granite and syenite form the princij)al undiversified peneplain here. - NORTHEASTERN WISCONSIN. With regard to the topography in northeastern Wisconsin, T. C. Chamberlin '^ says: The Archean surface is very irregular, and here and there knobs rise through the superincumbent formations, giving rise to isolated hills of quartzite, porphyry, and granite in the midst of the areas of lower rocks. He infers that these knobs are protruding through the Paleozoic sediments, not intrusive in them. LINEAR MONADNOCKS AND OTHER RIDGES. GENERAL DESCRIPTION. Besides the smaller monadnocks which rise above the broad uplands of the peneplain, there are numerous Imear monadnocks and elongated ridges below the peneplain level, which are related to the formations that outcrop in narrow bands, notablj' the Algonkian formations but to some extent also the Archean. A few linear monadnocks also rise above the level of the peneplain. Where the rocks are gently inclined erosion has been able to attack them more success- fully than in the areas of steeper dips, and has developed the monoclinal ridge (PI. IV, B), which has its gentler slope following the dip of the beds and its steep escarpment on the opposite side. Part of these monoclinal ridges are monadnocks, but a number are not. In the Keweenawan rocks of the Lake Superior region these monoclinal ridges are best developed in northeastern Mimiesota, on Isle Koj^al, and at the end of Keweenaw Point; among the lluronian rocks they are well developed in northern ilinnesota and southern Ontario, near Gunflint Lake, m the Penokee Range, in the Giants Range, and in all the iron districts, and as monadnocks m the peneplain (fig. 5). The origin of these monoclinal ridges as specialized forms due to differential erosion (fig. 6) upon weak and resistant strata has not been agreed to by all the workers in the Lake Superior region. N. H. WinchelH ascribed the Sawteeth Mountains of the Minnesota coast to faulting and has been followed by A. C. Lawson,*^ who ascribes the monoclinal ridges of the Animikie in southern Ontario and northern Minnesota to faulting, and by A. H. Elftmann.-' Irving,*' on the other hand, points out that the topography "is just such as is found in every region of flat-dipping hard rocks, and especially where softer layers are interleaved, as ill this case." He also cites numerous monoclinal ridges of similar type in equivalent nonfaulted rocks on eastern Keweenaw Point, in northern Wisconsin, and elsewhere, where the sawtooth shape is well developed. U. S. Grant '' writes: The numerous northward-facing cliffs suggest the iiroljability of a series of compai'ati\'cly recent east and west fault lines, along the north sides of which the strata are depressed. * * * The evidence of profound faulting in these strata, aside from the evidence of topography, is small. It seems that the present siu'face configuration could <■ Geology ot Wisconsin, vol. 2, 1873-1877, pp. 453, 462. i> Bull. Geol. and Nat. Hist. Survey Wisconsin No. IC, 1907, p. 10. c Geology of Wisconsin, vol. 2, 1S73-1S77, p. 248. d ScTcnlh .\nn. Kept. Geol. and Nat. Hist. Survey Minnesota, 1878, p. 12. c Bull. Geol. and Nat. Hist. Survey Minnesota No. 8, IS'13, p. 33; Twentieth .\nn. Rept. Geol. and Nat. Hist. Sun-cy, Minnesota, 1S91, p. 192. / Am. Geologist, vol. 21, 1898, p. 183. » Mon. U. S. Geol. Survey, vol. 5, 18&3, pp. 142-143. ft Final Hept. Geol. and Nat. Hist. Survey Minnesota, vol. 4. 1899, pp. 483, 485. PHYSICAL GEOGRAPPIY OF THE REGION. 99 have been brought about by eroaioii acting on gently inclined strata of different degrees and fissile Animikie slates being more susceptible to disintegration and erosion than the Grant subsequently proved absence of faulting in one of the "supposetl fault scarps" to the satisfaction of a number of accom- panj'ing geologists, including N. H. Winchell ami A. II. Elftmann, two of the advocates of the fault origin of these monoclinal ridges. As major faulting has never been proved to be associated with the scarps of the monoclinal ridges, as their origin by differential erosion in nonfaulted strata has been rejjeatedly shown, and as they are associated only with marked cross faults — for instance, on Isle Royal and north of Thunder Bay — the fault hypothesis for the mono- clinal ridges (sawteeth) is regarded as not warranted. Indeed, in the Vermilion monograph J. M. Clements," who discusses this type of topography, does not even mention the possibility of faulting. As the strike of the Algonkian rocks is generalh' northeast and southwest, the trend of the monoclinal ridges and of the subsec[uent valleys between is in the same direction, the longitudinal valleys that extend parallel to the strike of tlie rocks being usually broatl and persistent, whereas the transverse valleys extendmg across the strike of the rocks are narrow and irregularly arranged. T^Tiere these ridges and vallej^s are partly submerged the resulting bays are extreme!}' long, straight, and persistent, and the peninsulas and islands are in long parallel hnes, as on the coast of Isle Royal. Glaciation, acting upon tliis monoclinal-ridge topography, has pro- duced one striking series of lakes in northeastern ilinnesota; these, as well as similar lakes in other parts of the region, are due to glacial clogging of the subsecjuent axial valleys between the monoclinal ridges. KEWEENAW AN MONOCLINAL BIDGES. GENERAL STATEMENT. In northeastern Minnesota, on Isle Royal, on the end of Ke- weenaw Point, and in northern Michigan and northern Wisconsin, the monoclinal-ridge type of topography is so well developed that the name Sawteeth Mountains* has been given to these ridges on account of their resemblance to the jagged teeth of a saw when seen in profile. The same name is also applied to the Huronian mono- clinal ridges near Gunflint Lake and northward in Ontario. (See fig. 5, p. 88.) NORTHEASTERN MINNESOTA. Ridges of this sort in Minnesota, near Grand Marais, with back slopes of 5° to 10° and steep escarpments, are described by Irvmg" as forms due to differential erosion on weak and resistant beds. ISLE ROYAL AND MICHIPICOTEN ISLAND. of resistance, diabase sills. the 111 in-bedded ^•Q ^ t ^ La ;\?. i^\ s. «S' The monoclinal ridges on Isle Royal (PI. IV, B) are described by Lane."* No other information concerning the relation of the geology to the minor topography of Michipicoten Island has been obtained by the writer. a Mon. U. S. Geol. Survey, vol. 45, 190.3, pp. 400-401. i> Irving, E. D., The copper-bearing rocks of Lalie Superior: Mon. U. S. Geol. Survey, vol, o, 1SS3, fig. 1, p. 142; also flgs. 16, 26, and 29, on pp. 297, 32.1, and 320. cidem, pp. 141-143. '' Lane, .V. C, Geol. Survey Micliigan, vol. 0, 1S93-1S97, pp. 1S0-1S3. 100 GEOLOGY OF THE LAKE SUPERIOR REGION. KEWEENAW POINT AND NORTHERN MICHIGAN AND WISCONSIN. The parallel monoclinal ridges and intervening valleys near the enil of Keweenaw Point (fig. 6) were early described by Marvine" and later in some detail by Irving,'' who associated the various valleys and "parallel ridges with cliffy southern and flat northern faces" with specific gently dipping Kewcenawan beds — the valleys with weak amygdaloids and easily decom- posable diabases, the ridges with resistant melaphyres, coarse diabases, and bowlder conglom- erates — and showed the topogiaphy associated with them in various profiles.'^ In regard to the east part of Keweenaw Point, Irving'* emphasizes the relation of dip to topography: Where the dip flattens the structure comes out finely in a series of bold ridges. Toward Portage Lake, however, the dip becomes as high as 50° or more and the several ridges merge into one broad swell. This holds until the Porcupine Mountains are reached, where, although the dip angle is as high as 30°, the .structure is most beautifully illustrated in the outer ridge. « This ridge rises from the lake shore somewhat more gradually than the dip to a height of over 1,000 feet and then drops off in a bold escarpment of 400 feet into the valley of Carp Lake. This cliff/ extends nearly continuously across T. 51 N., R. 43 W., a distance of over 6 miles. The crown of the cliff is from 800 to 1,000 feet above Lake Superior and from 400 to 600 feet above the valley of Carp Lake. The base of the cliff is marked by a long slope of fragments fallen from the diabase and amygdaloid which forms its upper por- tions, but through the greater part of its length there is a perpendicular face of about 400 feet above the talus. Farther west again, as far as Bad River,ff the dips are high, often reaching 90°, and the harder rocks constitute merely rounded ridges and knobs with the cliffs facing indifferently in all directions. Beyond Bad River and all across Wisconsin to the St. CroLx the dips flatten once more, and the "sawtooth" shape in the ridges is everywhere well marked.'' This is notably true throughout Douglas County, Wis.'' U. S. Grant •' refers briefly to the surface features characteristic of the Keweenawan in Douglas County, Wis., where four belts of different topography are produced, vaiying with the part of the Keweenawan exposed, the dip, and the glacial overburden. The more resistant portions of the Keweenawan form two main ranges in northern Wisconsin because of the s^-n- clinal structure there. T. C. Chamberlin, writing as editor of the notes of the late Moses Strong, in reviewing the surface features of northwestern Wisconsin * says that the linear topography referred to and represented in ])rofiles shows sjtlendid Keweenawan monoclinal ridges. KEWEENAWAN MESAS. On the north shore of Lake Superior the tabular-mesa topograjihy (fig. .5, p. 8S) is develo]>ed in places where the Algonkian beds lie practically horizontal and weaker strata underlie more resistant beds, so that erosion has been able to open lowlands on the weak rocks and leave iso- lated highlands or ridges. Three great valleys have been opened up in the weaker beds in the upper Iluronian (Animikie group) and the Keweenawan, and two great mesa ridges have been left between these valleys. The waters of Lake Superior have subsequently risen to such a level that they occupy the floors of these valleys and form Thunder, Black, and Xipigon bays (PI. II, p. 86). Thunder Cape, the narrow end of one of the peninsulas, is a cliaracteristic bit of mesa topography, its flat top rising 1,350 feet above the level of the lake. Pie Island is another mesa of the same kind which erosion has isolated completely, the lake waters covering the valley bottoms surrounding it, and Mount McKay,' south of Mount William, is a similar a Marvine, A. R., Geol. Survey Michigan, vol. 1, 1S73, pt. 2, p. 95. 6 Mon. U. S. Geol. Survey, vol. 5, 1883, pp. 164-106. c Idem, fig. 2 on p. ITS, and pi. 18. d Idem, pp. 142-143. c Described by Foster and Whitney, pt. 1, 1850, p. 35. Shown well in topographic map by Michigan fleol. Survey, .Vnn. Kept, for 1905, fig. 3, p. 15. Just east oi the I'onupine Mount.iins, in the ISIack River region, between Bessemer and Lake Superior, the topography o( the Kewee- nawan in a typical strip from the I'enokee Range to Lake Superior is described by W. C. Gordon, who has prepared an excellent topographic map (Geol. Survey Michigan, Ann. Rept. for 1906, pp. 40S-109, 420; pi. 32). / Mon. U. S. Geol. Survey, vol. 5, 1883, p. 218. Idem, p. 143. » See also Irving, R. D., Geology of Wisconsin, vol. 3, 187.3-1879, pp. (12, 67-(iS. • Sweet, E. T., Geology of Wisconsin, vol. 3, 1S73-IS79, pp. 310-329. } Grant, U. S., Bull. Geol. and Xat. Hist. Survey Wisconsin No. 0, 1901, pp. 6-8. tocology of the upper St. Croi.\ district: Geology of Wisconsin, vot. 3, 1873-1879, pp. 367-381. iMon. U. S. Geol. Survey, vol. 5, 1883, p. 374. PHYSICAL GEOGRAPHY OF THE REGION. 101 mesa perhaps small enough to be called a butte, rising 980 feet above Lake Superior, and isolated in the broad, unsubmerged valley of Kaministikwia River. William Mclnnes "■ refers to the area of flat-lj'ing Animikie rocks near Thunder Bay as showing "table-topped hills, and escarp- ments with perpendicular faces and sharply angular outlines." A. C. Lawson, who ascribed the escarpment of the monoclinal ridges near Gunflint Lake to faultmg, has also indicated his belief that the east side of Thunder Bay, which "presents a very bold and remarkably straight cliff several hundred feet high composed of Keweenawan sandstone resting on Animikie slate, both flat-bedded and in apparent unconformity, * * * ig prob- ably originally and genetically a fault scarp."'' The writer feels inclined to ascribe this escarp- ment to subaerial denudation, partly (1) because of the insufficient evidence of larger faulting here,"^ as pointed out in the discussion of the cliffs of the monoclinal ridges (p. 99), partly (2) because denudation m the region is producing just such escarpments wliere rei^istant horizontal strata overlie weaker beds, and partly (3) because a fault scarp m this location could not pos- sibly have retained its present position and form since the latest possible date of formation unless it were protected by some lately removed mantle, as the larger possible fault scarps of the northwest coast of Lake Superior and the southeast side of Keweenaw Point seem to have been. (See pp. 112-116.) The chief reason for doubting the fault origin of the east boundary of Thunder Bay is that such an origin would imply the fault origm of the boundaries of all the mesas in this district which have escarpments that are very similar topographically and geo- logically. Because of the great complexity of block faulting that would isolate Thunder Cape and the adjacent peninsulas, as well as Pie Island and Mount McKay, etc., and the total absence of evidence of such faulting, it seems far more reasonable to ascribe these forms to the well- established cycle of forms resulting from normal subaerial denudation. North of Lake Superior, in Ontario, near Lake Nipigon and to the east and west, there seems to be a great many more mesas and valleys of exactly this kind,"* all in an area underlain by Keweenawan rocks or by upper Huronian (Animikie) slates and Logan sills, as along the Canadian Pacific Railway east of Port Arthur and especially beyond Nipigon. A. C. Lawson « writes : It is to the presence of these trap sheets (the Logan sills) that the bold and picturesque topography of Thunder Cape, Mount McKay, Pie Island, Nipigon Bay, and the many sheer-walled mesas and tilted blocks of the region is due. All these mesas apparentl}' have their present form because erosion has had more power to open up broad valleys in a region where the rocks lie practically horizontal than in adjacent regions where the rocks are more highly inclined. Three topographic types are well represented in the Keweenawan division of the Algonkian, where they seem to form a distinctly graded series (fig. 7) ratlier directh' associated with the Monoclinal Peneplain with monadnocks ^ I^i^lf? Mesas Figure 7. — Hypothetical cross section showing relation of secondary lowlands, mesas, monocllDal ridges, etc., to peneplain. dip of the constituent beds./ In exactly the same length of time precisely the same erosional agencies have been able to produce almost no effect upon the vertical and highly inclined beds (merely cutting gorges in the peneplain), to develop longitudinal valleys between monoclinal a Geology of the area covered by the Seine River and Lake Shebandowan map-sheets, comprising portions of theRainy River and Thunder Bay districts of Ontario: Geol. Survey Canada, new ser., vol. 10, 1S97, Rept. H, p. 6. b Twentieth Ann. Rept. Geol. and Nat. Hist. Survey Minnesota, 1S91, pp. 265-266. <^ Mmor faults extending in another direction with smaller possible fault escarpments are described by R. C. Allen in an unpublished thesis (1905) of the University of Wisconsin. Allen also appeals to faulting to explain the Mackenzie Valley (PI. XIII), the depression which nearly connects Thunder and Black bays and is followed by the Canadian Pacific Railway: the writer believes it to be due to normal denudation. d Collins, W. H., Summary Rept., Geol. Survey Canada, 1906, pp. 103, 105; Coleman, A. P., Rept. Bur. Mines, Ontario, vol. 16, pt. 1, 1907, pp. 107, 110. « The laccolithic sills of the northwest coast of Lake Superior: Bull, Geol. and Nat. Hist. Survey Minnesota No. S, 1893, pp. 24, 43. /Mon. U. S. Geol. Survey, vol. 5, 1S,S3, pp. 142-143, 166. 102 GEOLOGY OF THE LAKE SUPERIOR REGION. ridges on the gently inclined beds, and to advance the region where the l)e Rominger, Carl, Geol. Sur\-ey Michigan, vol. 4, 1881, pp. 1-3. ' The Marquette iron-bearing district of Michigan: Mon. U. S. Geol. Survey, vol. 28, 1895, pp. 222, 241, 257, 283-284,314, 331-.332, 410, 417, 444-445, 461, 4,88-189, 499, 572-573: also in Fifteenth .\nn. Rept. U. S. Geol. Survey, 1895. Idem, vol. 40, 1904, pp. 125-129; Menominee special folio (No. 82), Geol. .\tlas V. S., U. S. Geol. Survey, 1900. PHYSICAL GEOGRAPHY OF THE REGION. 107 Bayley" has described the toj^ography associated with the various Algonkian rock for- mations in the Menominee district in effect as follows : Sturgeon quartzite: Great bare regular bluffs with smooth tops and almost precipitous sides. Randville dolomite (northern belt): Valleys and other depressions. Randville dolomite (central belt): Usually insignificant, forming bases of hills and rarely little plateaus with small escarpments. Randville dolomite (southern belt): Conspicuous irregular cliffs or blul'fs. Vulcan iron formation: Either inconspicuous in valleys or clinging to slopes of dolomite. Ledges rarely prominent. Hanbury slate: Entirely confined to low ground, forming njinor protrusions only where the slate is locally harder. CRYSTAL FALLS DISTRICT. In the Crystal Falls district, whose physiograjih}' lias been described by J. M. Clements and H. L. Smyth,'' the adjustment of ridges and valleys to resistant and weaker structures has been somewhat similarly developed (PI. XXII, in pocket), except that here the ridges and valleys are arranged in a less simple and orderly way. The average relief is 200 feet in the western part. The Cambrian has been almost entirely removed. Furthermore, this region seems to have been, even in pre-Cambrian time, one of less relief than the Menominee region; certamly it was a region of very slight relief (called by Clements "an approximate peneplain") when the continental glacier overrode it, and as a result the glacial deposits are far more prominent and have more thoroughly obscured the preglacial topography than in any other iron district in the Lake Superior region except the Mesabi. The topography characteristic of the several Algonkian formations in the Crystal Falls dis- trict has been described by J. M. Clements, H. L. Smyth, and W. S. Bayley,*^ in effect as follows: Western part. Rand\dlle dolomite: No marked effect on topography or drainage (in depressions). Mansfield slate: Marked depressions, followed by Michigamme River. Hemlock formation: Exceedingly irregular topography; tuffs forming valleys; lava flows or intrusives forming higher ground, and resistant tuffs forming high hills. Bone Lake crystalline schists: Apparently forms knobs, but usually covered by glacial drift. Upper Huronian: In many places covered by glacial drift or by Cambrian sandstone. Shales form valleys and softly rounded hills. Graywackes and cherty rocks form more striking topography. Eastern part — Felch MounUiin. Sturgeon quartzite: Linear ridges, usually lower than those in the Archean, though locally lower than dolomite. Randville dolomite: Low, steep-sided knolls, occasionally linear ridges. Mansfield schist: No depressions; occasionally steep-sided valleys. Groveland formation: Moderately resistant, forming elevations such as Felch Mountain and Groveland Hill, 100 feet high. Upper Huronian mica schists and quartzites: Lowlands and low flat-topped ridges. Eastern part — Michigamme Mountain and Fence River: Sturgeon formation: Apparently here weak and forming lowlands; Randville dolomite underlying swamp. Mansfield formation: Indistinguishable topographically in gently rolling plain of dolomite (miniature ridges). Hemlock formation: Rough topographical details, with abrupt ridges and narrow ravines (in some parts till covered). Groveland formation: No topographic prominence except in Michigamme Mountain; in Fence River area topog- raphy less important than that of glacial drift. NORTH-CENTRAL WISCONSIN. Weidman'' has described the topography associated with the various Algonkian formations- in north-central Wisconsin (see Pis. IV, A, p. 90; XXXI, A, p. 436) in effect as follows: a Mon. U. S. Geol. Survey, vol. 46. 1904, pp. 177, 200, 291, 402. b Idem, vol. 36, 1,S99. pp. 29-36, 331-335. cidem, vol. 36, 1899 (western, pp. .50-51, 54, 73, 148. 155. 187-190; eastern, pp. 331, 398, 406. 411, 415, 423, 4.30, 431, 438, 440. 446. 471-473). d Bull. Geol. and Nat. Hist. Survey Wisconsin No. 16, 1907, pp. 42, 55, 62, 82, 88, 91, 100, 112, 118, 177, 358, 306, 371. 108 GEOLOGY OF THE LAIvE SUPERIOR REGION. Lower sedimentary series (lower Buroniant). Rib Hill quartzite: Bold knobs forming the highest land in the region in monadnocks, and jirominent because surrounding weaker granite and syenite are base-leveled. Wausau graywacke: Not prominent, forming very few low exposures. Hamburg slate: Not forming valleys lower than adjacent more resistant formations because of lack of dissection of perfected pene])lain. Powers lUuff (luartzite: Forms notable prominence 300 to 400 feet below sunoundings; smaller ridges. Quartzite at Rudolph: Low ridges and knobs. Juration City quartzite: No notable topography. Igneous intrusive formations (rhyolite series). Wausau area: Absence of sharply rugged topography, though low ledges project slightly through younger formations. Rhyolite schists of Eau Claire River: Forms striking cliffs in dells of Eau Claire River, due to joints. Rhyolite schists of Pine River: Marked gorge, a mile long, IKO feet deep, known as dells of Pine River, with sharp tributary gorges related to joints. Upper sedimentary series {middle Huronianf). Marshall Hill graywacke: Steep slopes and ledges. i^ Arpin quartzite: Low sloping land; less resistant than Powers Bluff quartzite and more resistant than adjacent granite. North Mound quartzite: Prominent mound rising above surrounding Cambrian lowland. NORTHWESTERN WISCONSIN. The Iluronian quartzites of Barron and Chippewa counties, Wis.,° form notably prominent monochnal ridges rising as much as 300 feet above the adjacent plain ant! ha\nng gentle dip slopes and steep escarpment faces with talus at the base. THE LOWLAND PLAINS. AREA. The lowland region of horizontal or gently folded post-Algonkian rocks (figs. 4 and 5, pp. 87, 88, Pis. I, in pocket ; II, p. 86) includes cliiefly rocks of Cambrian and other early Paleozoic age so generally buried beneath glacial deposits that ledges are comparatively rare tliroughout the area and the preglacial topography is partly or wholly masked. A small area of drift-covered Cretaceous, also flat lying, is found in northern Mimiesota. The lowland is made up of narrow areas on the south shore of Lake Superior, a broad belted plain in Micliigan, Llinnesota, and Wisconsin, and another plain in Miimesota. As the map (PI. I) indicates, there is a narrow strip at the west end of Lake Superior, on the south shore, and a narrow strip fringing the shore from L'Anse to Marquette. Besides tliis rather small Httoral zone, a considerable area now buried by the waters of Lake Superior is, without much doubt, covered by horizontal Paleozoic rocks. These early Paleozoic rocks cover all of the Upper Peninsula of I\Iichigan east of Marquette and overlap the highland country of northern Wisconsin and upper ilichigan, including the Archean and Algonkian areas, in a great semicircle which extends southwestward into Wiscon- sin to the vicinity of Grand Rapids and thence northwestward through Chippewa Falls, etc., to the region where the Paleozoic overlaps the Keweenawan of northern Wisconsin and sends a narrow tongue northeastward to join the horizontal Cambrian of the head of Lake Superior at Duluth. Very small patches are found on the north shore of Lake Superior. CHARACTER AND STRUCTURE. These early Paleozoic rocks consist chiefly of Upper Cambrian sandstone overlain in places by a conformable or nearly conformable series which extends upward to the Silurian in Wisconsin and to Devonian and Carboniferous in lower Michigan. North of the Archean and o Geology of Wisconsin, vol. 4, 1873-1879, pp. 575-581. PHYSICAL GEOGRAPHY OF THE REGION. 109 Algonkian of upper Wisconsin and IVIicliigan tliis Cambrian sandstone (Lake Superior sand- stone) lies essentiall}' horizontal and is probably preserved because it is downfaulted. In upper and lower iliclugan, in Wisconsin, and in Minnesota, however, there is evidence that the sedimentary rocks have been thrown into a series of broad folds — a synclinal basin in Michigan and a broad anticline in south-central Wisconsin. The Cretaceous in northern Minnesota is essentially horizontal. DENUDATION. Earth movements have left some areas of Paleozoic rocks higher than others, and as a result of the elevation and inclination of these beds eroding agencies have removed them entirely from some areas, the boimdaries of wliich have a direct relation to the broatl folding. The upper beds of the Paleozoic are almost entirely absent in northern and central Wisconsin and northwestern Micliigan (fig. 1 1, p. 116), from wliich it is inferred that, though they were once present over the whole of tliis area, they have since been removed by the active erosion which has taken place in tliis elevated region. As an evidence of the former greater distribution of the Paleozoic sediments we may refer to the isolated horizontal Cambrian beds that cap the ridges in the Menominee district east of Iron Mountain, Mich. (PI. XXVI, in pocket), and various outliers of Cambrian age, wliich form mounds rising above the general peneplain level in Portage, Wood, and Clark counties. Wis.," far north of the area of Cambrian rocks. Quite in contrast to these mounds of the border zone between the Paleozoic and pre-Cambrian in Wisconsin are the knobs of the older rocks wliich project through the thin Paleozoic edge. The knobs are inliers; the mounds are outhers. Chamberhn '' refers briefly to such knobs that protrude through the Cambrian in northeastern Wisconsin. Tlie Baraboo quartzite ridges and those at Necedah, Waterloo, etc. (figs. 53, 54, and 55, pp. 359, 360, 364), are features of the same sort. Because of their conspicuous positions as monadnocks on the pre-Cambrian peneplain they have been the first of the older rocks to emerge when the Paleozoic sediments which formerly covered their tops were eroded. THE BELTED PLAIN. The distribution of the Paleozoic sediments in a broad semicircle on the south flank of the Archean peneplain is to be explained, therefore, as a result of erosion after unequal upUft."^ The lowest bed, the Cambrian sandstone, is distributed in a curAang lowland belt around the Archean (PI. I, in pocket), with outhers scattered far back upon the Archean surface, and the overhnng Paleozoic formations are distributed in parallel curving belts, the more resistant beds standing up as highlands, the weaker beds being worn down into lowlands. A hnear series of iTunor liiglilands underlain by the " Lower Magnesian " limestone stretches southwestward in Micliigan and eastern Wisconsin (PL I), and thence northwestward in central and western Wis- consin. South and east of this is a broad valley wliich has been eroded upon the weaker members of the Ordovician, especially the Upper Ordovician (Cincinnatian) shales ami parts of the Galena and Trenton limestones. The waters of Green Bay have filled part of this great lowlantl valley, wliich extends southward, inclucUng the broad, shallow depression containing Lake Win- nebago (PI. II, p. 86). East of tliis valley there is a long, low monochnal ridge, wliich was produced by the effects of erosion on the resistant eastward-dipping Niagara limestone, and which has a steep scarp face on the northwest side and a gently dipping back slope toward Lake Michigan, diversified by minor monochnal ridges due to weak and resistant members of the Niagara. It is overlain by glacial and lake deposits. It forms an upland ridge (fig. 5, p. 88) east of Lake Winnebago and extends north in the Door Peninsula, Washington and adjacent islands of Wisconsin, and the Garden Peninsula of upper Mchigan; the scarp con- tinues first northeast, then south as the Niagara escarpment of Georgian Bay, southern Ontario, and northern New York. East of this ridge is the lowland of weak rock in wliich Lake Michigan ies and the upland of the northern part of lower Michigan. o Weidman, Samuel, Bull. Geol. and Nat. Hist. Survey Wisconsin No. 16, 1907, pp. 400, 405-407. i> Geology of Wisconsin, vol. 2, 1873-1877, p. 248. cidem, vol. 1, 1873-1879, pp. 24S-252. 110 GEOLOGY OF THE LAKE SUPERIOR REGION. The topography in the part of western Wisconsui inchidcd in this report is (Icscribeil by Moses Strong," that in central Wisconsin by R. D. Irving,* and that in eastern Wisconsin by T. C. ChaniberUn.'^ Tlip physiography of Wisconsin as a whoK> is briefly treated by G. L. Collie.'' Russell <^ has shown that in the greater part of the northern peninsula of Micliigan the wearing dowm of the gently inchned Paleozoic rocks has resulted in belts of upland and lowland of a sufficient degree of rehef to be apparent beneath the glacial deposits. The topograpliy of this region was described previously in a more general way by Douglass Houghton ■'^ and by Brooks. 3 The portion of the southern peninsula of Michigan here mapped as within the Lake Superior region has been described by Rominger '' anil by Lane.' The arrangement of the gently inclined Paleozoic rocks in curving zones has led W. M. Davis to describe Wisconsin as an ancient coastal plain, referring to the peneplained Archean area of northern Wisconsin as an oldland, the area underlain by Cambrian sandstone as an inner lowland, with a first and a second cuesta (monoclinal ridge) extending around its margin along the outcrop of the " Lower Magnesian" and the Niagara limestones respectively.-' Objec- tion has been raised to the use of the term "ancient coastal plain" on the ground that the upland area of northern Wisconsin is not known to be the old land from wliich the local Paleo- zoic sediments were derived. Though it is hence not permissible to classify Wisconsin as an ancient coastal plain, there is good warrant for describing these parts of Wisconsin and Michigan as a belted plain (fig. 5, p. 88 ; fig. 1 1 , p. 116) with upland and lowland zones sj^stematically related to the weak and resistant rocks. THE MINNESOTA LOWLANDS. In the western part of the Lake Superior region, extending into the vallevs of Red River of the North and Mississippi River, is a great lowland region, which seems to have been reduced to a peneplain in Mesozoic time, perhaps in the Cretaceous.* The Cretaceous peneplain extends into the Lake Superior region from the west and southwest and Cretaceous sediments overlap all the westward extension of the Giants Range. Just what this distribution of the Cretaceous may mean can not be said at present; but it seems probable either that sedimentation did not take place in the Lake Superior basin during the Cretaceous or else that wliile the Cretaceous base-levehng was going on over a great part of the United States the great mass of Paleozoic and perhaps later sediments were being removed from the basin of Lake Superior and the adjacent Mglilands, perhaps uncovering the several great escarpments presently to be described and producing the several lowland belts adjacent to Lake Superior and the Paleozoic areas to the south. THE BASIN OF LAKE SUPERIOR. GENERAL CHARACTER AND ORIGIN. The basin (PI. II) wliich contains the largest of the North American lakes probably includes parts of every system of rocks known to be in the region, from the Archean to the Recent. It is not known whether Paleozoic or Keweenawan rocks occupy the greater part of the basin. The Lake Superior basin is e.xcejjtional in that it is nearly surrounded by liiglilands. Going back from Lake Superior in any direction except the southeast, one soon comes to an escarp- ment, as at Diduth or on the south shore, above wliich is a distinct upland wliicli overlooks the lake basin. In some places this escarpment overlooks the waters of the lake directly (PI. V) : in others it is some distance back (PI. II and figs. 4 and 5). Moreover, this escarpment (400 to 800 feet in height) at many points descends into very d<>op water (500 to 900 feet), so that the o Ooology of Wisconsin, vol. 4, 187;i-1879, pp. 7-37. s Idem, vol. 1, 1S73, pp. 68-09. i> Idem. vol. 2, 1873-1877, pp. 453-153, S3!, 548. * Idem, vol. 3, lS7o, pp. 1-20. eldem, pp. 97-100. < Watcr-Supplj- Paper U.S. Geol. Survey No. 30, 1S99, pp. 57-.58, 9i>-91. d Bull. Am. Bur. Oeography, vol. 2, 19;)1, pp. 270-287. i Davis, W. M., Physical geography, 1S9S, pp. 13(>-137, flg. S5. c Ann. Kept. Geol. Survey Michigan for 1904, pp. 52.^).' * Leith, C. K., Kcon. Geology, vol. 2, 1907, p. 149. / Geol. Survey -Michigan, vol. 2, 1873, p. 241. PHYSICAL GEOGRAPHY OF THE REGION. Ill whole height of the suiroundinp; rim is not everywhere apparent. Some of the other Great Lakes have siicli a bounchiry on one sitle, hut none is so nearly walled in as Lake Supciior. As the submerged contours (PI. II) show, this basin has a depth of almost 1,000 feet, the deepest sounding being 163 fathoms, or 978 feet, near latitude 87° W., longitude 47° 45' N., or nearly 400 feet below sea level, without considering the possible filling of recent lake silts or glacial deposits. There is a notable depression between the pre-Cambrian of northern Wis- consin and the pre-Cambrian of Minnesota and Canada. This depression consists of a long, narrow trough trending northeast and southwest and limited on the north by the great escarp- ment wliich extends from Duluth northeastward to the mouth of Nipigon Bay, a distance of 250 miles. This trough is 25 to 70 miles TOde. Its southern boundary is Keweenaw Point and the Michigan and Wisconsin shore; at Oronto Bay, east of Ashland, there is an angular offset in passing the Apostle Islands, diminisliing the width of the lake by half. Thence the wall of the depression goes on parallel to and near the Wisconsin shore, the fault line converging west- ward toward the Duluth escarpment fault line, probably meeting it west of the head of the lakes in Minnesota. From th^ mouth of Nipigon Bay the border of the Lake Superior depression extends southeastward to Sault wSte. Marie as a high wall or escarpment of xmknowni origin. Here it is not a straight line but has great embayments and salients. On the south shore a fault escarp- ment extends southward on the east side of Keweenaw Point. The liighland border thence trends irregularly southeastward to the vicinity of Marcjuette, beyond which it extends south and a httle west of south into Wisconsin. The area between Marquette and Sault Ste. Marie on the south shore is lowland. The North American Great Lakes are situated in pairs on either side of an escarpment which faces the boundary between the resistant pre-Cambrian and the relatively weak Paleozoic rocks. In this respect they resemble the great lakes of the pre-Cambrian area of northwestern Europe. An escarpment thus situated and formed is called by Suess a glint line. Lake Superior, however, should not be included among the glint lakes, where it is classified by Suess," together with Lake Ontario, Georgian Bay, Lake Winnipeg, etc. The southeastern part of Lake Superior might be considered a glint lake because it has one early Paleozoic and one Archean shore, as was pointed out by Agassiz,** if it were not known on other- evidence to be chiefly a structural basin. In the origin of its basin, also. Lake Superior is exceptional. The other great lakes, four to the east in the United States and four to the north in Canada, lie in lowland areas where differential erosion acting upon alternate weak- and resistant beds would produce basins if aided by glacial erosion, glacial clogging, etc., though some of the basins are possibly also in part structural. Lake Michigan, for example, lies between the broad, anticlinal, southward- pitcliing fold of central Wisconsin and the basin-like syncline of central Micliigan, its location suggesting a partly structural basin, as does also the knowm warping in the basins of the other great lakes, though the structural feature is certainly of minor importance. The correspondence of the Lake ^Michigan lowland with a belt of weak strata (Silurian and Devonian), perhaps somewhat deepened by glacial erosion,"^ is probably of principal importance. The reason for the present depression. of the Lake Superior basin is somewhat doubtful, the earliest explanations being regarded as inadequate to account for certain features of it. The fact that it is a synchne (see structure section, PI. I, in pocket), first pointed out by Foster and Whitney "^ and amplified by Irving," has never been called in doubt, for there is ample proof of it. But for so old a structural basin to remain unfilled f and for it to retain abrupt boundaries which bear all the characteristics of youth are departures from the normal con- dition wluch require special explanation. » Su3ss, Ediiard, The face of the earth (Das Antlitz der Erde), translated by H. B. C. and W. J. SoUas, vol. 2, O.xford, 190G, p. 65. 6 Lake Superior, etc., 1S50, p. 420. « Chamberlin, T. C, Geology of Wisconsin, vol. 1, 1S73-1879, pp. 253-259. i Report on the Lake Superior land district, pt. 1, 1850, p. 109. c Mon. U. S. Geol. Survey, vol. 5, l.SS.^, pp. 410-418. / Barrel! f.Jour. Geology, vol. 14, laoti, p. .33.5) has computed that it would take Mississippi River only 60,000 years to completely fill Lake Superior if it flowed into that water body with its present volume and load. 112 GEOLOGY OF THE LAKE SUPERIOR REGION. The hypothesis that the present Lake Superior basin exists because of a geosyncline, as first stated, needs to be modified, therefore, bj' consideration of the possibihty of graben or rift fauhing. The amphfication of this revised hypothesis and its verification in detail remain for future work. The possibility, however, seems worth outlining here. It is thought reasonable to suppose that after the late Algonkian deformation, whose struc- tural warping produced or redeepened the major sjTichne, the basin was filled to a considerable extent by lavas and by sediments overlj-ing the Keweenawan flows. Between the close of tliis period of deposition and the beginning of the Upper Cambrian a great period of denudation produced the pre-Cambrian peneplain, whose surface of low relief beveled across the weak and resistant members of the Archean and Algonkian, the syncUnal basin perhaps being filled with the material worn away in making the peneplain or perhaps ])eing replaced bj' part of the peneplain surface. At some subsequent date, probalily also pre-Cambrian, faulting took place, producing the great escarpment which extends northeastward from Duluth and smaller nearly parallel escarpments on the south shore of the lake. These two fault fines bound what is perhaps a great graben or rift, which forms the rectangular body of northern and western Lake Superior (fig. 8). The evidence of the fault origin of these escarpments may be gathered from a detailed consideration of their characteristics. PENEPLAIN _.^ gsCARgMENT CAMBRIAN \ LAKE SUPERIOR GRABLN Sea level UPPER HURONIAN MEWEENAWAN KEWEENAWAN (ANiMmi£ group) FiGUKE 8.— Graben or rift valley of western Lake Superior.tshowing escarpment on either side and ^J^neplain above. DESCRIPTION OF ESCARPMENTS. DULUTH ESCARPMENT. Rising steeply above the waters of Lake Superior for about 600 to 800 feet at Duluth and with diminishing height toward the northeast is the Duluth escarpment (PI. II, p. 86). It has a slope at Duluth of 450 to 1,000 feet to the mile, and the steeply ascending face is 1^ to 2 miles wide (PI. V, A). Above rises the fairly level-topped gabbro plateau, wluch extends north- ward as part of the peneplain. The escarpment, wlxich bounds tliis plateau on the southeast, is remarkably simple in its outhne, with none of the irregularity which characterizes slopes long eroded by streams. Tliis simplicity of outline is shared by the gently curved escarpment of Keweenaw Point and by that of northern Wisconsin, both of which are kno\vn to foUow fault fines. Lawson has suggested that the Duluth escarpment also foUows a fault line." We have then to account for its fresh and uneroded form, for it is quite inconceivable that a fault scarp could have been produced, as tliis may have been, in pre-Paleozoic or verj' early Paleozoic time and not have been more largely altered by weathering and stream erosion. The streams of the Duluth escarpment descend very steeply to Lake Superior; few of them head more than 4 or 5 miles from Lake Superior (PL II), the greatest distance being 12 to 14 miles, in contrast with lengths of 30 to 75 miles on the north and northeast shores of Lake Superior. Many of them have as steep an average grade as 150 to 250 feet to the mile (PI. V, ^-1), the general average being 80 to 160 feet to the mile. No one of these rather tumultuous streams has cut a significantly deep valley in the face of the escarpment and most of them have only cut short gorges with small rapids and waterfalls. Quite in contrast with these steep-graded, rapidly falling streams of the escarpment are the leisurely flowing streams of the plateau surface above. The Cloquet, the upper St. Louis, and various other rivers have an average slope of about 8 or 10 feet to the nnle. It is well established that a rapidly flowing stream with a steep grade is able to deepen its vaUey rapidly and to extend its headwater area so that it encroaches upon the area drained by an adjacent a Twentieth Ann. Rept. Ocol. and Nat. Hist. Survey Minnesota, 1891, p. 192. U) 1- . m uj (A it O III in 3 N _ n *-» O rr 11 q: < < iij 7 c o 1, 2: < «; 13 O lu ■2 > (- T "-* a> zr^ U ■3 g bJ q: 5 'J T ^ ^ <• Q- D c; It n < o () ^ m < Ul2 iLl tt ir 3 1 1 0) UJ ic < J n < 7 1- < V B 1 ^1 1 Q. Ill S -1 Z Z ■' z ? l^ ^ T > ^ D 1 n fi -) O Q 1- < s CM 1 1> /• T a; 1 UJ H 13 M Q. S fr 3 U) ii < Q hi UJ I ■■J 7 T — H "^^ - ^z PHYSICAL GEOGRAPHY OF THE REGION. 113 FiGUEE i -The drainage of the St. Louis and Mississippi headwaters before the stream captures along the Duluth escarpment. leisurely flowing stream (fig. 9), capturing and diverting the latter or some portion of its head- waters. Stream captures or piracies, as they are called, of tliis Icind are common. We should expect, then, that in the course of stream development for a great length of time several of the swiftly flowing streams of the escarpment would have extended their headwaters back to the region drained by the leisurely flowing streams of the plateau surface and captured part or all of these drainage systems. The fact that many of the large streams have not done so is evidence of their youth. The largest stream in the region, however, seems to have already done just what would be expected (fig. 10), and it is natural that the largest stream should be able to do tliis first. St. Louis River, cutting back at a point near the end of the escarp- ment where it is rather low, has been able to extend its headwater region northwest- ward until it has captured the southwestward-flowing Clo- quet and the southwestward-flowing stream that forms the present headwaters of the St. Louis itself. These captured streams had been a part of the leisurely drainage system of the plateau surface, and, it seems certain, were withm the Mississippi basin (Pis. I and II). Indeed, a large valley extending southwestward from the town of Floodwood, where the St. Louis now turns abruptly to the southeast, indicates that this is probably the latest elbow of capture at which the piratical St. Louis has been able to divert to the Lake Supe- rior-St. Lawrence drainage system a large headwater tributary of Mississippi River, as it had previously diverted the Clociuet, an- other ^Mississippi head- water, or possibly one of the St. Croix. A study of similar fault scarps acted upon by stream erosion in other parts of the world indi- cates that this fault scarp has not been acted upon by erosional agencies for a great length of time. If it had been so eroded for a long period, we should find it deeply cut by valleys with outlymg knobs on the lower slopes, like the erosion escarji- ment at ^larciuette (PL V, B), and with stream captures at the upper shoulder, where the escarpment meets the plateau top. Comparison of this escarpment with the ec[ually abnipt escai'pments on the north shore of Lake Superior from Thunder Bay to Sault Ste. Marie emphasizes the freshness of the Duluth escarpment; there is a striking contrast in stream and vallej^ distribution. The north-shore 47517°— VOL 52— 11 8 Figure 10. — The drainage of the St. Louis and Mississippi headwaters at present, after stream captures and diversions. 114 GEOLOGY OF THE LAKE SUPERIOR REGION. escarpment has much lonj^er streams flowing directly to the lake from tiie north, with deep valleys everywhere cut to lake level. It is a mucii-breached wall; the Duluth escarpment is an unbroken barrier. Tlie drainage of the former proclaims greater length of time for stream dissection in the same language by which the drainage of the hitter aimounces youth. It seems possible that erosion by the Lake Superior lobe of tlui Labrador ice sheet might have so smoothed the face of this escarpment and steepened and intensified it that topography of the kind suggested wouhl be destroyed or that longer streams draining to Lake Superior would be diverted by the ice barrier and acquire new courses. Such modification may have taken place to a slight degree, but even if the maximum of glacial erosion is assumed the lack of stream diversions is c(uite unexplained, as is also the resemblance to the acknowledged fault scarp on the east side of Keweenaw Point. Along the line by which this escarpment can be discriminated as a form initially produced by faulting rather than by glacial erosion a scrutiny of the submerged continuation of the same escarpment reveals several significant facts. Fortunately the detailed soundings made by the Corps of Engineers of the United States Army in charting the Great Lakes give us detailed information (PI. II) concerning the escarpment below present lake level. First, it continues to descend at as steep or steeper angles than on the land, a depth of 400 to 600 feet being found within 2 to 3 miles from any part of the shore. The escarpment, therefore, is not merely 400 to 600 feet but 1,000 to 1,200 feet in height. Second, it extends directly across the moutlis of the several large bays (Thunder, Black, and Nipigon) at the north end of the lake, where the escarpment feature in the unsubmerged land surface is interrupted by these broad valleys, partly drowned beneath the present lake level. These are therefore hanging valleys, entermg the lake basin or the linear depression to which they are tributary at levels 400 to 600 feet " above its bottom. (See PI. II.) This submerged hanging valley condition might be explained either by glacial erosion or by faulting. The facts in favor of glacial erosion are (a) known ice flow along this coast and parallel to it; (h) probably accentuated erosive ability in this portion of the Lake Superior basin, where more rapid movement would result from the constriction of the ice between Isle Royal and the mainland; (c) the known ability of glaciers of no greater thickness and less width to erode so deeply that main valleys receive discordant tributaries (hanging vallej-s) as much as 500 to 1,000 feet above, as in Alaska, the Swiss Alps, Scotland, Norway, New Zealand, etc. Points in favor of faulting are the following: (a) The straightness of the escarpment; (h) the continuation below lake level of a topograpliic feature whose drainage and other land phenomena are inexplicable by glacial erosion alone; (c) the uniform level at which the sub- merged hanging valleys stand (Thunder Bay 22 to 23 fathoms, Black Bay 22 fathoms, Xipigon Strait 20 to 21 fathoms). Such uniformity is unusual in glacially eroded hanging valleys, where the size of the glaciers in tributary valle^^s, their width, tliickness, and eroding power, produce hanging valleys at diverse levels. Glaciers of the unecjual sizes denoted by these bays would surely have done so. (d) The varying age, character, and resistance of the rocks beveled across by this supposed fault (Cambrian sandstones, Keweenawan lavas and sediments, upper Huronian intrusives and slates, and older rocks). The escarpment therefore seems to have features inexplicable b\' glacial erosion alone, but none that do not fit the hypothesis of glacial erosion motlifying a faulted form. The exceptional depth of water just opposite the mouth of Thunder Bay (156 fathoms), making this point 936 feet deep, or more than 300 feet below sea level, and the second deepest place in tlio lake, can be readily explained by glacial scooping at just this point, for such irregularity in the bottoms of glacially eroded channels like the Norwegian and Alaskan fiords are not uncommon. The writer accordingly feels that there is a reasonable possibility that the northwest shore of Lake Superior from a point west of Duluth to St. Ignace at the north, with its direct but broadly-curving course, represents the position of a fault line. This fault scarp, with 1 ,000 feet or more of throw, may either be very recent, though several considerations lead to the belief o Bottom of Thunder Bay, 22 fathoms or 132 feet; depth of trough opposite mouth, US fathoms or 0T8 feet. PHYSICAL GEOGRAPHY OF THE REGION. 115 that this is not so, or else it may have been faulted long ago and then buried and protected so that erosion has only recently begun to attack it. Accordingly it may owe the preservation of its southwesterly portion (Minnesota shore) to protection by Cambrian or later sediments and the dissection of its northeasterly part (Ontario shore) to the earlier removal of such a protecting Cambrian mantle. Glaciation is believed to have modified this escarpment in its minor features only, as in changing a more precipitous slope to the present flaring wall and in locally deepening the depression at its base. KEWEENAW ESCAKPMENT. The escarpment of the east side of the Keweenaw Point "■ very closely resembles the Duhith escarpment in form and condition of erosion though not so high nor so steep (PI. II). A north- east-southwest trendmg escarpment borders the east side of "an elongated promontory,* not greatly dissected by erosion nor deeply undulate nor serrate in its crest line," whose flat top has been formed by the base-levehng '^ of a series of steeply dipping Keweenawan beds and whose western and northwestern sides slope more gradually to the level of Lake Superior; the east side slopes steeply to the open lake near the tip and is elsewhere separated from the lake by the low-lying flat portion imderlain by the Cambrian sandstone (PI. XXVIII, p. 380). This escarpment differs, however, fi-om the Duluth gabbro escarpment in one important respect. It is cut entirely through by stream valleys in at least two places. It is believed that the great transverse valley of Portage Lake (PI. XXX, B, p. 434) and the valley of Ontonagon River were formed before the present Lake Superior existed, by streams which were superposed on this long, narrow peninsula through a mantle of Cambrian (Lake Superior) sandstone, whose remnants are still preserved high ujjon the fault scarp near the highest part of Keweenaw Pomt.'' Irving and Chamberlin," after careful consideration of the many earlier hypotheses, reach the conclusion that the Keweenaw Point scarp is a pre-Potsdam fault modified by wave work, buried, and slightly refaulted in post-Potsdam or post-Cambrian time. (See fig. 75, p. 574.) ESCARPMENT OF NORTHERN WISCONSIN (SUPERIOR ESCARPMENT). The escarpment wliich forms the boundary of the northern highlands of Wisconsin f and overlooks the basm of Lake Superior from a point west of Duluth eastward to the Apostle Islands is a lower and more gently sloping scarp (PL II). It has the characteristics of the other two escarpments in being without topographic outliers and in having short, steeply sloping stream courses which have not extended headward much beyond the shoulder of the escarpment. Chamberlin? concludes that this escarpment of Bajrtield and Douglas counties, Wis., is a pre-Potsdam fault scarp, and Grant ^ has supported this conclusion but makes its age post- Potsdam. Like the Duluth and Keweenaw escarpments, it seems to have been protected so that its dissection has been somewhat postponed Its youth is therefore not so anomalous as W. M. Davis has suggested. * ISLE ROYAIi ESCARPMENT. On the north side of Isle Royal there is a submerged escarpment of 400 to 500 feet, suggest- ing a parallel fault here (PL II) , wliich Irving and Chamberlin ' conceived of as possibly a contin- uation of the fault of Bayfield and Douglas counties on the south shore. There is no continua- Ir^dng, R. D., and Chamberlin, T. C, Observations on the junction between the Eastern sandstone and the Keweenaw series on Keweenaw Point: Bull. U. S. Geol. Survey No. 23, 1885, pp. 12, 98-119. 6 Idem, p. 103. <■ Van Hise, C. R., Science, new ser., vol. 4, 1896, pp. 217-220. i Bull. U. S. Geol. Survey No. 23, 1885, pp. 109-110. t Idem, p. 119. f Chamberlin, T. C, Geology of Wisconsin, vol. 1, 1SS3, pp. 105-100. Grant, U. S., Bull. Geol. and Nat. Hist. Survey Wisconsin No. 6, I90I, p. 0. 9 Geology of Wisconsin, vol. 1, 1883, p. 105. » Bull. Geol. and Nat. Hist. Survey Wisconsin No. fi, 1901, pp. 17-20. * Science, new ser., vol. 15, 1902, p. 234. 1 Bull. U. S. Geo!. Survey No. 23, 1885, p. 111. 116 GEOLOGY OF THE LAKE SUPERIOR REGION. tion of this steep slope northeast or southwest of Isle Royal, wliich stands on a high base with steep descents on all sides of it, especially the northwest and southeast. If the channel north- west of Isle Royal is ascribed to block faulting, the island itself must be regarded as a land mass that stands as a horst above the deep surrounding basin because of failure to be faulted down. Isle Royal and Keweenaw Point accordingly have certain features in commcjn aside from familiar fact that the Keweenawan rocks in Isle Royal dip soutlieast and those at Keweenaw Point dip northwest. The slopes facing each other seem to be dip slopes, but of the sides facing awiiy fiom each other that of Keweenaw Point is known to be a fault Une, and that of Isle Royal may possibly be a smaller one. This structural feature, then, would be a great synchnal trough between Isle Royal and Keweenaw Point, with downfaultmg on each side. Massing of the contours in other parts of the lake (PI. II) suggests submerged escarpments east of this trough, but there is not enough information for detailed discussion. AGE OF ESCARPMENTS. For all these subparallel escarpments grouped about the west end of Lake Superior the hypothesis is advanced that they have been formed by faulting. Their later liistory may have accorded with one of two hypotheses. One supposes that they are old escarpments (pre-Cambrian) sHghtly modified by stream erosion and in places possibly developed mto sea chffs and then buried beneath Paleozoic sediments. Durmg the ensumg long period of denuda^ tion the escarpments themselves were protected from erosion by the overlyuig sediments. They were gradually uncovered and are now just in the begiiming of a cycle of erosion, wliich was postponed until their rather recent disinterment. The alternative hypothesis that these are much more recent fault scarps (post-Cretaceous or pre-Pleistocene) is supported by the evidence of slight post-Cambrian movement along two of these scarps (along which tliere was surely much greater pre-Cambrian f aultmg) and by the evidence of post-Cretaceous and of post- Pleistocene faulting in other parts of the area. The question of the date of this faulting is a large one, involving the determination of the age of the great peneplam of the area and the age of the present Lake Superior basin. BEARING OF ESCARPMENTS ON AGE OF PENEPLAIN. There are three fields for attacking the problem of the age of the peneplain in the Lake Superior region. The first is m northern Wisconsm, where the truncated siu-face of the pre- Cambrian now dips down imder the Paleozoic. The conditions here are shown m figure IL BELTED PLAIN ''^^'^,^'.'l;n.. , . CL -. - -— A '-°r ^L^!l.°f°'?. -.—„.„-_- .... HUPONIAN SERIES FiGUEE 11.— Structure profile in northern Wisconsin, showing the south edge of the peneplain on the pre-Cambrian rocks and the northern part of the belted plain of the Paleozoic. Weidman has demonstrated that h-c is a buried pre-Potsdam peneplain and mferred that a-h is its exhumed equivalent. Van Ilise previously referred to h-d as a Cretaceous i>eneplaui and to a-h as its equivalent. So far as the writer can see, evidence for decidmg conclusively between these two hypotheses is not present, though the Paleozoic outliers on the peneplain suggest that it is pre-Potsdam rather than CYctaceous. The second field of attack is in the region to the west, in Minnesota (PI. XIV, p. 212). 'Ilere the Cretaceous overlaps the peneplain. Numerous diamond-drill holes tlirough tlie glacial drift on the Cuyuna range show the Cretaceous as a thin mantle on the peneplam of pre-Cambrian rocks. Elsewhere the drift covers it deeply, but on the border of the Giants Range monadnock, in the ]\Iesabi iron range, Cretaceous outliers are found m valleys and on ridge slopes (PI. MIT, in pocket). These are marme Upper Cretaceous, so the peneplam might perfectly well be either pre-Cambrian PHYSICAL GEOGRAPHY OF THE REGION. 117 or early Cretaceous in age. If the Cretaceous cau be found in valleys in the peneplain as well as in valleys on the slopes of its monadnocks, the probability of pre-Cambrian age will be strengthened . The thirtl and most jiromising field for investigation is in tlie fault scarps themselves. The escarpments were clearly made after the great peneplain was developed, for the nearly base- leveled upland areas now extend neatly up to the edges of these steep slopes (fig. 8, p. 112) and could not have done so when the peneplain was formed. The two latest periods of great base-leveling in the area are thought to be pre-Cambrian (pre-Potsdam) and Cretaceous. The known periods of faultmg are pre-Cambrian, post-Cambrian, post-Cretaceous, and post-Pleistocene. The Lake Superior basin was surely here in pre-Pleistocene time, so the post-Pleistocene may be elimmated as a period of major faulting. The choice seems to lie between (a) regarding the peneplain as due to Cretaceous base-leveling and the escarpments as due to post-Cretaceous faulting, to which there are certain objections, and (b) regarding the peneplam as an exhumed slightly dissected pre-Cambrian surface and the escarpments as due to pre-Camorian faulting. The assimaption of protection by Paleozoic sediments is necessary in order to explain the relatively fresh fault-scarp forms,' and from this assumption naturally follows the hypothesis of the clear- ing out of the basin and exhumation of the escarpments iluring th,e Cretaceous base-leveling and the glacial period, all the later faulting beuig considered of slight amomit. There are objections to this hypothesis also, but in the mind of the writer they are of less weight. CHAPTER V. THE VERMILION IRON DISTRICT OF MINNESOTA." LOCATION, AREA, AND GENERAL GEOLOGIC SUCCESSION. 'Ihe Vcnnilion iron-bearing district lies in northeastern Minnesota, in St. Louis, Lai1 mm lit lik Pi? o S 11 Basalt extrusives Porphyry intrusives and extrusives Jasper FiGtJEE 13.— Section across jasper belt in sees. 13 and 14, T. 03 N., R. 13 W., Vermilion iron range. Minnesota, Scale, 1 inch=about 85 feet. becomes locally buckled along an axis lying in any direction in the plane of bedding. This type of folding, while leaving great local complexity", does not destroy the general attitude or trend of the bed. It is frequently possible, where these folds are present, to work out the general trend of the formation and its top and bottom — as, for instance, in sees. 13 and 14, T. 62 N., R. 13 W., Minnesota (see fig. 13) — and for other areas it will be possible by close detailed surveys to work out the stratigraphy of the Keewatin series. 124 GEOLOOY OF THE LAKE SUPERIOR REGION. 'rii(> folding, notwithstanding tlio extraordinarily l)rittle character of the rock, was accom- plished without major fracture. Frequently Sr solid belt of jasper may be seen bent back upon itself within its own radius with no sign of fracture. The deformation, therefore, was in the zone of rock flowage, and no bettor instance is knowTi to us of this kind of earth movement. Though the folding is so complex as to give isoclinal or fan-shaped folds, ordinarily the turns are round rather than acute, as they commonly are in the Menominee district. Folding without brecciation is tlie rule, but in some places the Soudan formation has been brecciated in an extraordinary manner. It is broken tlarough and through by cracks and crevices, along which minor faulting has taken place. In some places the grinding of the fractured fragments over one another has been so marked as to give them a well-rounded char- acter, and such a rock resembles a conglomerate, though it is really autoclastic. This local brecciation of the Soudan formation has been favorable to the deposition of the ores, and it may be suggested that the general absence of the brecciation is the partial explanation, at least, of the very irregular tlistribution and scarcity of the ore bodies. The VermiUon district affords excellent illustrations of complex folds, or folding in two directions at right angles, and the formation which best exliibits tliis folding is the Soudan. This is because the banding of the formation is very marked, so that the position of bedding is readily determined, and also because for the most part the rock does not take on schistosity. Schistose structure is absent partly because the minerals of the rocks are not adapted to a parallel arrangement. Furthermore, the Soudan rocks are frec[uently found in contact wdth. the Ely greenstone, and the contacts give the pitches of the cross folds. The remarkalile complex folding partly explains the distribution of the Soudan formation with reference to the Ely greenstone. As upon the major folds are superposed secondary and tertiary folds, numerous patches of the jasper are naturally found in the greenstone. More- over, because of the cross folding these patches may be very narrow at one place, widen out within a very short distance so as to make a thick formation, and again become narrow. Wlien the extraordinary complexity of this folding is understood it is only necessary to premise an erosion extending to different depths in the Soudan formation before the lower Huronian was deposited in order to see how in the greenstone there may be patches of jasper ranging from a few feet in wiilth and length to the dimensions of great continuous formation about Tower and Ely. But folding is not the only cause of the present relations, as is shown on page 126. LITHOLOGY. The iron-bearing Soudan formation comprises two classes of rocks. To all the varieties of the first the miners apply the name "jasper," although only a portion of it falls strictly under this designation. This is the dominant variety of the rock. Locally interstratified with the "jasper" or under it is an argillaceous variety, which is mainly slaty but in some places is conglomeratic. The "jaspery" phase of the Soudan formation consists of interlaminated bands of finely crystalline quartz, iron oxides, and various mixtures of the two. With these preponderating minerals are various subordinate constituents, among which amphibole is the most abundant, including actinolite, cummingtonite, and griinerite. Pyrite is also present in many places. The alternate bands of material of different color, combined with the complicated fracturing' and brecciation of the formation, make it a striking rock which alwaj's attracts the attention of the traveler, even if he is not accustomed to closely noticing rocks. The bands of material of different color vary from a fraction of an inch to several inches across. The quartzose bands havfiT various colors — nearly pure white, gray, red of various hues, including brilliant red, and black. The diirerence in the color is chiefly caused by the contained iron. Hematite, if in sufficiently fine particles, gives the brilliant red colors; magnetite and hematite in larger particles give the grays and blacks. Between the bands dominantly (|uartzose ai-e usually liands nuiiniy composed of non oxide Tliis iron oxide may be either hematite or magnetite or various intermixtures of the two- Occasionally also some limonite is present. VERMILION IRON DISTRICT. 125 The chief varieties of the "jasper" are (1) the cherty variety, (2) the black-handed variety, (3) the red-banded variety, and (4) the white-banded variety. With these are subordinate masses of (5) the carbonated variety and (0) the ore boilies. 1. The cherty variety is characterized by the presence of a predominating amount of gray cliert, the iron oxide being subordinate. The rock is there a sliglitly ferruginous well- banded chert. 2. The black-banded form of the Soudan formtvtion has dark-gray or black chert bands interlaminated with black iron-oxide bands. Tiie iron oxide is commonly in large part mag- netite. Usually associated with this magnetite are some of the amphibole minerals already mentioned. 3. In the red-banded kind the quartzose layers are stained with innumerable minute flakes of hematite, which give the rock a red color, in many places a brilliant red. The iron oxide between the red bands is ordinarily hematite, usually specular hematite. With this hema- tite may be some magnetite. This red-banded variety is a well-known jasper of the Lake Superior region, to which Wadsworth has applied the name jaspilite. 4. In the white-banded kind the quartzose bands contain comparatively little iron oxide. The iron-oxide bands between the layers of chert are generally hematite, but this hematite differs in many places from that of the jaspilite bands in that it is of the red or brown variety. With it also, in many places, there is a certain amount of limonite. 5. The banded carbonate variety, while subordinate in quantity, is important in refer- ence to the genesis of the formation. It is a gray-banded rock, the light-colored layers of which consist largely of siderite. Between this sideritic rock and the ordinary forms there are aU stages of gradation. 6. The positions of the iron-ore bodies will be fully discussed later. In the iron ores the silica is very subordinate, the place of the quartzose bands being taken by iron oxide. The iron ore is dominantly hematite. At the contact of the Soudan formatioii and Ely greenstone the cherty variety of rock is very common indeed. In many places the rock at this horizon is much brecciated and com- monly has a conglomeratic appearance, which, however^ is believed to be due to movement rather than to deposition as a conglomerate. Ordinarily this cherty variety of the formation is not more than a few feet thick. Resting upon the cherty zone in many places is the black- banded kind. Ordinarily at the top of the formation is the red-banded rock, jasper, or the white-banded kind. The succession given above prevails in many places where the formation is now thick. Where the formation is thin the red and white banded rocks extend from the top to the bottom, and as at many places the formation is rather thin it may be said that the entire Soudan forma- tion for much of that district consists of these kinds of rocks, the cherty variety and the black- banded variety not appearing. The sideritic rock is notably local in its occurrence. It is generally found close to the over- lying upper Huronian rocks. The slaty phase of the Soudan formation differs from the ordinary phases in having between the silica and iron-oxide bands so large an amount of argillaceous material as to make laminae of slate. In some places a slaty cleavage has developed in the clayey layers but does not pass through the iron-oxide bands, and this may be so even where the bands of slate are not more than one-fourth inch across. Locally the slate may be in a belt several feet thick without inter- stratified jaspery material. In some places this slate is graphitic. At a few places at the bot- tom of the Soudan formation the slate passes down into a fme-grained conglomerate or into a tuff. A microscopic examination of the argillaceous varieties of the slates shows these sedi- ments to be made up of chlorite, actinolite, epidote, sericito, sphene, quartz, carbonaceous material (graphite), and some iron oxides, in various proportions. The graphitic slates consist essentially of graphite and quartz in exceedingly fine grains and in some specimens in very small quantity. 126 GEOLOGY OF THE LAKE SUPERIOR REGION. The conglomeratic phases of the formation, when studied under tlie microscope, are founri to be substantially identical with the tuffs of the Ely greenstone. Thej^ now consist largely of actinblite, clilorite, cpidote, and cjuartz. ORIGIN. From the foregoing facts it is clear that the Soudan is a sedimentary formation, mainly of nonclastic character. This would ])crhaps be evident from the well-bedded character of the formation and especially from the iron carbonate. Also, as already indicated by the descrip- tion of the different rock varieties, certain phases of the formation have argillaceous bands between the iron-oxide bands, which are not uncommonly graphitic. Finally, it contains local conglomerates. There is reason for believing that many varieties of rock in the Soudan formation are derived from siliceous iron-bearing carbonate, precisely as similar rocks are derived from this material in other districts of the Lake Superior region. The analogy between the vSoudan formation and the Negaunee formation of the Marquette district is especially close. Substan- tially every variety of rock which is found in one district may be found in the other. A variety may be somewhat more prevalent, however, in one district than in the other; for instance, the amphibole minerals are less abundant in the Soudan formation than in the Xegaunee formation. In the absence of local specific evidence of the original character of the iron-bearing rocks in the Vermihon district it is probably not safe to put too much stress on the similarities with. other districts where the original character of the rock is certainly knowTi. One must admit the distinct possibility that the iron-bearing sediments may have been originally deposited sub- stantially as banded chert and iron oxide of the jasper type. RELATIONS OF ELY GREENSTONE AND SOUDAN FORMATION. The main mass of the Soudan formation seems to be above the Ely greenstone. In certain places it is Itnown to be in pitching troughs formed by folding, the greenstone forming the walls and bottom, as, for instance, at Ely and Soudan. Some of the jasper belts of the Vermihon district are clearly interbedded with successive basalt extrusives. Such beds, but a few feet thick, may be traced for hundreds of yards with uniform widths, even contacts, and lack of folding. Wlien tiie adjacent igneous rocks arc examined closely it is found that the sedimentary bands lie parallel to the tops and bottoms of separate flows, as marked by amygdaloidal and other surface textures, without intervening fragmental sediments. This is well illustrated in sees. 13 and 14, T. 62 \., R. 13 W., Mimiesota. (See fig. 13, p. 123.) Many of the jasper bands are associated even more closely with intrusive and extrusive porphj^ries than with the greenstones. (See p. 128.) These porphyries are found to be closely related to the extrusive basalts but on the whole to follow them and to be associated with their later phases of extrusion. This association of the iron with the later acidic phase of extrusion is also seen in the Woman River district of Ontario. Its significance is discussed on page 513. The most common contact between the Ely greenstone and the Soudan formation is perfectly sharp — indeed, knifelike in its sharpness. The rocks are as sharply separated from each other as if the Soudan formation were intersected by the greenstone by intrusion, and doubtless this is, at least in a few places, the true significance of the relations. Contacts of the kind mentioned may be seen at many places in both the west and the east end of the district. They are especially clear and numerous in liimters Island and at Jasper Lake, Birch Lake, and Emerald Lake. At each of these lakes, almost at every large outcrop of Soudan material, somewhere along the base of the formation the contact may be found. The kind of contact next most common to that just described is that in which a brecciated rock occurs between the iron-bearing Soudan formation and the Ely greenstone. This breccia ordinarily is not more than a few feet wide. In some places it mvolves only the greenstone, elsewhere the Soudan formation only, in still other places both. Thus a conglomerate-like VERMILION IRON DISTRICT. 127 rock may show fragments and matrix mainly of greenstone or almost wholly of Soudan formation, or the two intermingled. In the last case the greenstone is more likely to be the matrix and the Soudan rock to constitute the fragments. A breccia of the greenstone class is well seen on an island near the west end of Otter Track Lake. The brecciated Soudan forma- tion is well exhibited in belts of Soudan rock north of Robinson Lake, in sec. 7, T. 62 N., R. 13 W. A breccia composed of greenstone and Soudan material is seen at various places on Lee Hill. Here is a green schist matrix containing numerous fragments of red jasper, each exhibiting its banding, which lies in diverse directions. Some of these fragments are well rounded; others are subangular; many others have angular rhomboidal forms, such as are produced by shear- ing stresses. However, these fragments are not more, angular than those in a basal conglom- erate at many localities. The c^uestion may be asked whether the breccias were conglomerates before they were breccias. At present their dominant structure is doubtless that of a dynamic breccia, but it is also possible that some of them at least were originallj^ conglomerates and were subse- quently brecciated. This question, early asked, is still unanswered. Probably certain of the rocks referred to are wholly breccias, being produced by readjustment along the contact of the two formations during orogenic movements. A sharp contact of the first class might, by close folding and adjustment between the formations, produce a contact of the second class by brecciation and rounding of the fragments, thus forming a pseudoconglomerate. At contacts of a third kind is a rock which seems to be a metamorphosed mechanical sediment. As a rule, this rock varies from a few inches to several feet in thickness. It consists of alternating laj'ers of green schist or slate and light-colored, strongly siliceous, graywacke-like material. These alternations of schist and graj^wacke naturally give a remarkably sedimentary appearance; in fact, it seems as if the banding could have' been produced in no other way. The two localities which best exhibit these materials are a neck of land between two small lakes about a mile north of Moose Lake and one place on Lee Hill. At the first locality alternating bands of slate and graywacke rest against perfectly typical ellipsoidal greenstone, and interstratified with these slates and graywackes are narrow bands of jasper. These alter- nations are overlain by a broader belt of jasper. The probable interpretation of the phenomena seen here is that a few feet of mechanical sediments were deposited upon the Ely greenstone before the deposition of the nonclastic material of the Soudan formation. Moreover, it seems that there were alternations between the condition of mechanical deposition and the peculiar condition of chemical or organic deposition of the Soudan formation. The relations at Lee Hill are substantially the same, except that at this place the folding is so close that a cross cleavage cuts through the finer-grained sediments, and on account of tlus close folding and the secondary cleavage the phenomenon is more difficult to certainly interpret. However, the slate and graywacke appear to plunge under the jasper of the Soudan formation, and the explanation is with little doubt the same as for the contact noi-th of Moose Lake. A contact of a fourth kind is marked by a thin belt of greenstone conglomerate. The best localities at which this is seen are north of Robinson Lake and at the pits of the Lee mine. At the first locality, at the west end of the belt of Soudan formation, the ellipsoidal greenstone is overlain by a layer a few feet thick of greenstone conglomerate, which passes up into gray- wacke. The pebbles of this greenstone conglomerate are flattened, and it could not be said positively that the rock is not a tuff rather than a conglomerate. Finally, the Soudan and Ely formations may be separated by a thin layer of graphitic black slate, well shown on the southwest .side of vSoudan Hill. From the fact that the greater masses of the Ely greenstone were deposited before the larger masses of the Soudan formation it is believed that the great volcanic period of the Ely greenstone had practically ceased before Soudan time. However, the extremely intricate relations and apparent interstratification of the minor masses of the Soudan formation ■with the Ely greenstone and the fact that both the Ely and Soudan formations locally contain interstratified fragmental material lead to the belief that volcanic* activity had not entirely 128 GEOLOGY OF THE LAKE SUPERIOR REGION. died out in lUl parts ot tlie district at the time of tiic deposition of the earliest Soudan rocks. In consequence tiiere are interlaminations of rocks essentially belonginj^ to t lie Ely with rocks essentially belonging to the Soudan. Wliat were the physical contlitions which peiinitted the deposition of the nonmechanical Soudan formation upon the Ely greenstone with so insignificant an amount of intci-vening mechanical sediment and erosion surfaces ? If the Ely greenstone was subaerial, it is dilhcult to understand how tliis material could have got below the water without the deposition of a greater thickness of mechanical sediments than exists in the Veraiilion district. We know that such lavas are very rough in their surface expression and vary greatly in thickness, and therefore in altitude. It is impossible to believe that the sea could advance over such an area without the production somewhere of mechanical sediments of considerable thickness. The answer to this question seems to be that the eruptions of the Ely greenstone were submarine. The ellipsoidal textures are regarded as evidence of submarine flows, for reasons given on pages 510-512. The lack of erosion surfaces in the flows and the absence of fragmental mate- rial at the base of the formation itself are evidence of such an origin. If these lavas issuing from the interior of the earth were spread out below the surface of the water, after the period of volcanism had ceased and conditions became quiescent nonmechanical sediments of the iron-bearing formation might at once be deposited, provided the conditions were proper. The conditions of sedimentation are further discussed in the chapter on the origin of the iron ores. LAURENTIAN SERIES. PORPHYRY. Intrusive into the Ely greenstone and Soudan formation are various Archean felsites and porphyries in dikes and bosses. These are exceptionally well seen in the Vermilion Lake area, especialty at Stuntz Bay. As already noted, these intrusives may be in part connected with acidic flows interbedded with some of the later flows of basalt in tlie Ely greenstone. (See p. 126.) Petrographically the porphyry comprises rhyolite porphyry, feldspathic porphyry, micro- granite, granite, microgranite porphyry, and granite porphyry. In places these rocks have been metamorphosed into sericite schists and chlorite schists. There is no doubt that these rocks are older than the lower Huronian, because they yield fragments t© the Ogishke conglom- erate, but at various places their relations to the conglomerate are extremely intricate. (See p. 131.) The folding has formed breccias and pseudoconglomerates from the felsites and porphyries, which when very much mashed have been sometimes confused with the true Ogishke conglomerate. GRANITE OF BASSWOOD LAKE. The granite of Basswood Lake extends as a great continuous formation north of the Ely greenstone and the Huronian rocks from the western to the eastern end of the district, where it is locally known as the "Saganaga Lake granite." Lakes are rather numerous in this great granitic area, but they are not so numerous nor so regularly ordered as those in the Ely and Soudan formations. On the whole the granite area is one of highlands and divides between the waters running north and south. Petrographically the granite varies from hornblende and mica granite to syenite. Struc- turally it varies from massive granite through schistose granite to gneiss. Texturally it includes granites and granite porphyries. The mineral constituents arc green hornblende, ])iotite, orthoclase, quartz, and plagioclase, with accessory sphene, zircon, and iron oxide. In many places these minerals have been very much altered, so that their places are taken largely l)y secondar}^ minerals, of which chlorite is the most prominent and ejjidote, sericite. and secondary feldspar come next. There is a variation in the mineral character, hornblende being virtually absent in some specimens and abundant in others. No specimens were found in which quartz was not present, but the amount is small in some of them. VERMILION IRON DISTRICT. 129 The granite is intrusive into the Elj^ and Soudan formations. The field relations are most complex but are practically the same in all parts of the district — that is, the phenomena to be seen in passing from the other Ai'chean formations to the granite are substantially the same whether the traverse be made at Vermilion Lake, at Bumtside Lake, at Basswood Lake, or at any other point. In apjM-oach to the granite from the Ely greenstone side little stringers of quartz first appear in the greenstone, then sparse veins of feldspar, then clean-cut dikes of granite, usually of small size. With closer approach these increase in number and size until they constitute a plexus of granite dikes in the greenstone. Still farther north the greenstone and granite may be found in such confused and intricate relations as to make it difficult to say which is the more abundant. Here great knobs of granite as well as dikes occur in the greenstone masses. In the granite knobs are included fragments of the greenstone, large and small, in many places in great numbers. Farther north the granite becomes dominant and finally altogether excludes continuous masses of greenstone. If any greenstone is found it will be only in the form of included masses. In brief, the relations are like those, so clearly described by Lawson, between the batholiths of granite and the contiguous greenstones of Rainy Lake and Lake of the Woods. The granite has been spoken of as if its intrusion were a single episode. This is not sup- posed to be true. On the contrary, the relations of the different granites to one another and to the greenstones are very intricate, hence it is thought that various intnisions were separated by long intervals of time, that many of the intrusions were of themselves complex and long continued, and that, in fact, this igneous period was a complex and long-continued one. ALGONKIAN SYSTEM. HURONIAN SERIES. LOWER-MIDDLE HITRONIAN. GENERAL STATEMENT. The inferior series of Huronian rocks occupies the general position of the lower and middle Iluronian of the south shore. It will be called lower-middle Huronian, with the understanding that it may include either or both lower Huronian and n)id(lle Iluronian. The lower-middle Iluronian consists of four divisions — (1) a lower division, predominantly conglomeratic, which is most typically developed near Ogishke Muncie Lake and is known as the Ogishke conglomerate; (2) a division represented only in the eastern portion of the district, consisting of iron-bearing rocks and known as the Agawa formation; (3) a division which is predominantly a slate formation and which is called the Knife Lake slate because it is well developed and splendidly exposed on and near Knife Lake; and (4) intrusive rocks. OGISHKE CONGLOMERATE. DI.STRIBUTION. The Ogishke conglomerate extends from the western end of the district to the east end, though it varies greatlj' in thickness. In places it is a great formation; in other places it is nearly absent or is so thin that it can not be represented on the maps without a gross exaggeration. The localities at which the conglomerate can be best studied, beginning at the west, are ( 1) southeastern Vermilion Lake and especially Stuntz Bay and vicinity; (2) Moose, Snowbank, and Disappointment lakes and vicinity; (3) Ogishke Lake and the extensions of the. belt there to the southeast, northeast, and west. DEFORMATION. The Ogislike conglomerate is infolded in an extremely intricate manner with the luulcr- lying formations. This infolding is almost if not quite as complex as the infolding of the Soudan formation and the Ely greenstone already described. Owing to isoclinal folding and cross folding with steep pitches, a rock surface cutting diagonally across the plane of contact shows 47517°— VOL 52—11 9 130 GEOLOGY OF THE LAKE SUPERIOR REGION. tho most extraordinarily irregular distribution of the Ogishke and the underlying formations. Because of this it was supposed by a number of the early geologists that the Ely greenstone and the porphyry of Stuntz Bay were intrusive into tlie Ogishke conglomerate. UTHOLOOY. In general all the belts of conglomerates arc coarser below and become finer toward higher horizons. This statement is, however, only true as an average. There are places where the conglomerate is somewhat fine at the bottom, is coarser above for a certain thickness, and thence becomes finer upward. The character of the Ogishke conglomerate depends largely on the nature of the underlying formations. These formations, as already noted, are the Ely greenstone, the* Laurentian granite of Basswood Lake, the Soudan formation, and the Laurentian porphyry of Stuntz Bay. \Miere the conglomerate rests on one of these formations the material comjiosing it is mainly derived from that formation. There are four special varieties of the Ogislike conglomerate — (1) green- stone conglomerate, (2) granite conglomerate, (3) porphyry conglomerate, (4) chert and jasper conglomerate. The common kind of Ogishke conglomerate (5) represents combinations of the special phases. Greenstone conglomerate. — The Ogishke is a greenstone conglomerate at those localities where the conglomerate rests upon the Ely greenstone and other lower formations are not atlja- cent. One of the localities which exhibit this greenstone conglomerate in its typical character is the south side of Ogishke Lake and, peripheral to the Ely greenstone massifs, to the east on Frog Rock Lake. The rock is also found in equally good development on Hunters Island, at the southwest of Lake Saganaga. At these localities the greenstone conglomerate consists for the most part of very well rounded fragments of the Ely greenstone set in a matrix derived from the same source. These fragments are ordinarily of a size to make pebble conglomerates, but at some places many of them are so large as to constitute bowlder conglomerates. Between the bowlders and pebbles are smaller fragments of the same material, and between these is a finer matrLx derived from the same source. In most places upon the weathered surface the conglomerate character of tliis rock is e\'ident, but on a freslily broken surface tlie matrix and pebbles are so similar that the rock seems to be a continuous mass of greenstone. The conglomerate character is especially diificult to discover in the unbroken forests, wliere the rocks are covered witli moss and otlier vegetation. The debris, being derived from the Ely greenstone, consists of all tlie varieties of rocks shown by that formation. There are, accordingly, fragments of dense, massive greenstone, of amygdaloidal greenstone, of various kinds of ellipsoidal greenstone, etc. These rocks grade locally into rocks that may be tuft's. In certain places the conglomerate is discriminated from the tuff only by finding that the rock occupies a definite stratigrapliic zone at the base of the lower Huronian sediments. Locally discrimination is still impossible. Granite conglomerate. — The granite conglomerate occurs along the west border of Lake Saganaga. At the west side of the south arm of Cache Bay is a great bowlder conglomerate the fragments of which are directly derived from the granite. The matrLx also came almost wholly from tliis source. The exact contact of the conglomerate and granite may be seen. Tha bowlders and pebbles of the granite conglomerate are well rounded, and in every respect tliis conglomerate bears the same relations to the granite that the greenstone conglomerate does to the Ely greenstone. The granite conglomerate is associated with a peculiar variety of rock, which may be called recomposed granite. It appears tliat when the Ogislike formation was laid down the granite only locally jnelded coarse debris. For the most part it yielded the separate individual minerals of the coarse gi-anite — that is, feldspar, cjuartz, etc. As a result a clastic formation was laid down upon the granite, the particles of which were the individual minerals of the granite. Further- more, these particles were but little waterworn. The result is that when they were recemeuted VERMILION IRON DISTRICT. 131 a rock was produced which closely resembles the gi-anite. This resemblance is, indeed, so close that the rock was first mistaken by a number of geologists for the granite. This rock is exposed along the west side of Cache Bay, at Swamp Lake, at the west side of West Seagull Lake, and at intervening points. For much of tMs distance this peculiar forma- tion has a breadth of nearly half a mile. Porphyry conglomerate. — The porphyry conglomerate is confmed mainly to the area about Stuntz Bay, the debris being derived from the Laurentian porphjrry. In the past it has been known as the "Stuntz" conglomerate. In places there is a coarse bowlder conglomerate, in other places a fine conglomerate, and in still other places a graywacke composed of the individual minerals of the porphyry, so that the rock closely resembles the original porphyry. Furthermore, so similar are the bowlders and the matrix that the conglomerate itself has been confused with the brecciated porphyry. Chert and jasper conglomerate. — The chert and jasper conglomerate is found where the underlyuig formation is the Soudan. Tliis conglomerate is, however, not anywhere known to be solely composed of the Soudan material. In tliis respect this variety of rock cUffers from the varieties already described. Locally, however, the conglomerate is predominantly composed of material derived from the u-on-bearing formation. This variety of rock may be seen on Lee Hill, just north of Tower, on the Burnt Forties southeast of Vermilion Lake, and at other localities. Common OgishJce roclc. — The varieties of the Ogishke conglomerate heretofore described, each consisting largely of material from a single source, are, on the whole, rather exceptional, though the greenstone conglomerate and the porphyry conglomerate occupy considerable areas. It is natural to suppose that the Ogishke would have material derived from more than one of the previously existing formations, and ordmarily it has. Thus the normal Ogishke conglomerate consists of interrmxtures in various proportions of the materials derived from the Ely and Soudan formations, the granite of Basswood Lake, and the Laurentian porphyry, or two or more of them. Hence there is every gradation between the average form of the Ogishke conglomerate and the special forms wliich have been described. Witliin the Ogishke conglomerate, in addition to the common fragments already enumerated, there are occasional unciuestionable slate frag- ments. These are seen at various places, but are especially abundant south of Moose Lake. It is believed that the source of the fragments of tliis kind is the slate and graywacke of the Ely and Soudan formations. METAIIORPHISM. The Ogishke conglomerate varies greatly in its metamorphism. In general the processes of the change have been mainly those of metasomatism and cementation, but locally the con- glomerate is recrystallized and schistose. These phases are especially likely to be adjacent to the massive granite, greenstone, or other rock against wliich they rest. Wliere the process has gone to an extreme it is difficult to place the exact cUviding line between the original and recomposed formations. The difficulty is particularly likely to occur in reference to the green- stone conglomerate and the Ely greenstone. The extreme phase of the metamorphism of the Ogishke conglomerate results from the intru- sion of igneous rocks, and especially the Huronian Snowbank granite and the Keweenawan Dulutli gabbro. Adjacent to these intrusives the conglomerate is a conglomerate schist or gneiss, the matrix of wlucli is usually mica schist where the Huronian is of an acidic Ivind or ampliibole scliist where it is of a basic kind. The conglomerate scliist adjacent to the gabbro may be found from points east of Fay Lake to Lake Gabimiclugami. The conglomerate scliist near Snowbank Lake and Disappointment Lake has suffered the metamorphosing effect of the Snowbank gi-anite and the Duluth gabbro. The_ changes in the conglomerate are analogous to those wliich have taken place in the Knife Lake slate, which is in a similar position with reference to the granite. (See pp. 133-135.) 132 GEOLOGY OF THE LAKE SUPERIOR REGION. RELATIONS TO ADJACENT FORMATIONS. The Ogishkc conglomerate, as the foregoing description plainly shows, is unconformable with the iiiulcrh'ing formations. It may safely be inferred tiiat this unconformity is one of great magnitude. TJie evidence is of two kinds — the ciiaracter of the detritus and the structural relations. The detritus mcludes every variety of each of the formations of the Archean, including the many phases of the Ely and Soudan formations and the granite of Basswood Lake. Jn order to produce these many varieties, the Archean went tlu-ough a long and complex history of folding, intrusions, metamorphism, and erosion. As to the structural relations, the Ogishke conglomerate is here in contact with one of the underlying formations, there with another. It is therefore clear that after the Archean complex was produced it underwent deep erosion before the deposition of the Ogishke conglomerate, for some of the formations constituting the Archean were produced at great depth. Upward the Ogishke conglomerate grades into finer and finer material and passes con- formably into the Agawa formation or the Knife Lake slate. THICKNESS. The tluclcness of the Ogishke conglomerate varies greatly. It is nowhere possible to make accurate measurements, o\ving to the general absence of bedding and to the clo.se folding, but it is certain that the formation has a considerable thickness, certainly several hundred feet, and perhaps in some places more than 1,000, possibly 2,000. From this maximum thickness the foi-mation varies to a thickness of only a few feet or less, and is absent in places. AGAWA FORMATION. In the eastern part of the district, above the Ogishke conglomerate, or, where that forma- tion is absent, beneath the Knife Lake slate, is an non-bearmg formation called the Agawa. On the American side of the international boundary this formation is so thin that it can not be regarded as continuous. On the Canadian side of the boundary, especially at That Mans, Agawa, This Mans, and Other Mans lakes, the formation ranges up to 50 feet in thickness and has all the characteristic rocks of the other iron-bearing formations of the Lake Superior region, includmg ferruginous caibonate, ferruginous slate, ferruginous chert, jasper, and iron oxides. Interlaminated wath the ferruginous varieties are belts of slate. Thus the iion-l)earing forma- tion is both small and impure. There is every reason to suppose that the origin of this iron- bearing formation is similar to that of the other Lake Superior iron-bearing formations. The Agawa formation, so far as at present knowTi,has no economic importance, but it may have a geologic significance, considering that it is in the lower-middle Huronian. The only iron foiTnation at this horizon in other parts of the Lake Sujierior region is the Negaunee, and so correlation would l:)e suggested with that formation. The bearing of this suggestion on the position of the group to which the Agawa belongs is pointed out elsewhere (pp. 603-604). KNIFE LAKE SLATE. GENERAL STATEMENT. The Knife Lake slate was so named because it occurs in its typical character at Knife Lake. Nearly all the long arms of that lake lie withm the slates, and by far the greater number of the many islands and headlands are composed of them. The slates are found in two great areas, one in the western part of the district and the other in the central and eastern parts. The western area extends from the east end of Vemiilion Lake westward to parts where the rocks are covered by the Pleistocene. It occu])ios much of the shore and many of the islands of Vermilion Lake. The eastern area begins west of Long Lake and extends eastward, becoming gradually broader, and m the eastern part of the district is the most extensive formation there found. VERMILION IRON DISTRICT. 133 LITHOLOGY. The Knife Lake slate comprises the following main varieties: 1. Argillaceous slates. 2. Cherty slates. 3. Graywacke slates and graywackes. 4. Conglomerates. 5. Tuffaceous slates. 6. Micaceous (and, less commonly, amphibolitic) schists and gneisses. 7. Graj' granular rocks. There ai'e also all gradations between these varieties. The materials of different coarse- ness are in many places finely interlaminated, so that it is easy to ascertain strikes and dips. The argillaceous slates vary in color from gray to black. They are usually very dense, break with a smooth, conchoidal fracture, and have a perfect cleavage, which in a general way commonly follows the trend of the district but whose direction varies much locally, depending on the surrounding rocks, the folding, and other factors. The chertj^ slates differ from the argillaceous slates in that they contain an unusual amount of finely crystalline quartz. In many places this quartz is the dominant constituent. Between the beds of very siliceous slate in many places there are also pure bands of chert. These cherty bantls m most places appear to be secondary segregations. In many places the amount of the fhiely crystallme quartz in the separate cherty bands and in the main mass of the slate is so great as to suggest that the deposits of fine mud had mingled with it silica of organic or chemical origm. Conchoidal fractures are especially characteristic of the cherty slates. The argillaceous slates and cherty slates pass into varieties which may be called graywacke slate and graywacke. These differ but little from the finer-grained slates except that cleavage is less likely to be developed in them. Cleavage is usually present in the graywacke slates but not in the graywackes. Not uncommonly the graywackes pass into conglomerates. The fragments found in the conglomerate comprise all the varieties of material found in the Ogishke conglomerate. These, it may be recalled, are the many phases of material derived from the Archean. Indeed, there is no essential difference between these conglomerate bands antl the Ogishke conglomerate, except that the conglomerate bands of the Knife Lake slate are ordinarily fine grained and are subordinate in quantity to the slates. During Knife Lake time there was volcanic action, and close to the volcanoes, as at Lake Kekekabic, ash and larger fragments produced by explosive volcanic action are mingled with the other materials of the Knife Lake slate. These volcanic materials constitute the tuffaceous slates. Between the tuft's and the conglomerates antl slates there are all gradation varieties. Indeed, microscopic examinations show that the ashy products of the volcanoes were widely distributed and are important constituents of the varieties of the formation already described — the argillaceous and cherty slates and graywackes. The mica slates, mica schists, and mica gneisses are confuied to areas adjacent to subse- quent intrusive rocks. The most important areas are south of Tower, along Kawisliiwi River, adjacent to Snowbank, Disappointment, and Kekekabic lakes, and adjacent to the Kewee- nawan gabbro. At Snowbank Lake and near it the granite has been intruded into the slates in a most com- plex fashion, and here next to the granite the Knife Lake slate is represented by mica schists. Between the mica schists and the ordinary slates there are gradations through mica slates. Here the granite is found in numerous great dikes intersecting the Knife Lake slate. Moreover, in many places the granite injections have followed the banding of the slate so as to give close parallel mjections. In some places there are %vithin a single hand specimen several bands of granite. Also bands are found intermediate in character between the well-recognized granite and the slate. There is no doul)t that these bands ai'e due to granitization. Wlrere the injection is of the most complex kind the rock is a mica gneiss, the darker-colored bands of which are 134 GEOLOGY OF THE LAKE SUPERIOR REGION. largely the extremely metamorphosed granite. However, some material in the black bands has doubtloss been doriviMl fiom the granite and some material in the light bands has been derived from the slate. The scliists and gneisses are especially well exposed on the north side of Snowbank Lake. South of Tower, adjacent to the granite, and especially at localities near the Duluth and Iron Range liailroad, the alterations are essentiaUy the same as at Snowbank Lake, excei)t that the amphibole schists are more prominent. Also the alteration phenomena at Kekekabic Lake are in the same direction as at Snowbank Lake, but the processes have not gone so far. At Kawisliiwi River southwest of Snowbank, and at Disappoiatment and Gabimichi- gami lakes, the great gabbro mass of the Keweenawan has profoundly affected the character of the Knife Lake slate and has produced a peculiar gray granular rock wluch the Miimesota geologists have called "muscovado." These rocks differ from the slates and schists about Snowbank Lake in being almost massive. They are particularly well seen at Disappointment Lake. Between the schists north of Snowbank Lake and the granular rocks of Disai)pointment Lake there are gradations. These granular metamorphic rocks adjacent to the gabbro are regarded by Grant as the result of contact metamorpliism of the Knife Lake slate. They recrvstallized under deep-seated static conditions at high temperature and probably influenced by abundant moisture. The difference between them and the scliists and gneisses of Snowbank Lake shows how important a part orogenic movement probably had in the production of the structures of the latter rocks. The schists and gneisses of the Knife Lake slate are the joint product of djTiamic and contact action. The granular rocks wliich are adjacent to both tlie Snowbank granite and to the gabbro have doubtless undergone two periods of metamorphism, the earlier one at the time of the introduction of the Huronian Snowbank granite and a later one by the Keweenawan gabbro. At the earlier time doubtless schists and gneisses were produced under dynamic conditions which at the earlier time were transformed to granular rocks under static conditions. MICROSCOPIC CHARACTER. Clements's microscopic study shows that the rocks of the Knife Lake slate, iacluding argillaceous and cherty slates, graywacke slates, graywackes, conglomerates, and tuffs, have as recognizable primary constituents feldspar, quartz, brown mica, wliite to green and violent-brown pyi-oxene, and greeuish-browai hornblende. The clastic mineral grains very commonly have been extensively altered, and from these have been produced the following secondary namerals, which, in some places where the rocks are completely recrystalhzed, are the sole constituents: Chlorite, epidote, sericite, actinolite, massive dark-bro\\-n and green hornblende, quartz, calcite, and pyrite. The minerals between the grains in the coarser sediments are sericite, chlorite, epidote, quartz, and feldspar. These are believesoUa granite; Twenty-first Ann. Rept. GeoL anil Nat. Hist. Sun'ey Minnesota. 1S93, p. 53. VERMILION IRON DISTRICT. 137 As has been noted, the Keweenawan Duluth gabbro bounds the eastern half of the Ver- mihon district on the south. In the lower-middle Huronian and Archean rocks are numerous comparatively fresh dolerite dikes and bosses. There are also more sparingly late acidic dikes in the Archean and the Huronian. It is supposed that these fresh rocks, showing compara- tively little orogenic movement, are of Keweenawan age, although they have not been con- nected areall}^ with the greater masses of Keweenawan rock^. The metamorphosing effects of the Keweenawan gabbro upon the Archean antl Huronian have already been considered. The Keweenawan rocks themselves are discussed in Chapter X^' (])p. 866-426). THE IRON ORES OF THE VERMILION DISTRICT, MINNESOTA. By the authors and \V. J. Mead. DISTRIBUTION, STRUCTtTRE, AND RELATIONS. The iron ores of the Vermilion district occur in the Soudan formation, belonging to the Keewatin series of the Archean system. This formaticm rests upon the Ely greenstone, is in places interbedded with it, is interbedded with and intruded by acidic porphyries, and as a whole has been closely folded, with the result that the iron-bearmg formation stands with contorted and steeply inclined bedding, with steep walls and bottoms of green schist and mashed porphyry. These constitute deep, narrow, pitchmg troughs in which the ores are found. The jaspers constitute for the most part the hanging wall of the ore. The total area of the ores is but a minute fraction of that of the iron-bearing formation of the district. It is significant that notwithstanding the enormous sums of money spent in the exploration of the district no ore deposit of magnitude has been developed outside of the two principal series of deposits at Tower and Ely, which were the first discoveries in the district. One additional deposit in sec. 30, T. 63 N., R. 11 W., about 4 miles east of Elj^, has been considerably explored, leadmg up to the first shipment of ore in 1910. On Soudan Hill near Tower the structural relations of the iron-bearing formation to the green schists and mashed porphyries are so complex that it is extremely difficult to follow the ore bodies. The steeply pitching troughs branch, change their pitch, and are duplicated by parallel troughs to such an extent that m spite of the enormous amount of underground explora- tion to which the hiU has been subjected it is not certain yet that all the ore deposits have been found. The Soudan ores may have (a) "paint rock" or "soapstone" as foot wall, below which is jasper, and similar paint rock or jasper as the hanging wall; or (&) they may have jasper both as a foot and a hanging wall, and hence may lie within it and grade in all dhections into the Soudan formation. Deposits of this kind are small. The Soudan ores are mainly of the first form. They have now been found to a depth of 2,000 feet. At Ely there is a single trough of the iron-bearing formation in the greenstone, beginning as a comparatively wide body at the west and narrowing and deepening toward the east. The northeast side of the trough seems to be formed in part by lower Huronian slates or graywackes. The greenstones associated with the ores are altered to paint rock along the contacts. This trough is a comparatively simple one, but there is also a minor parallel anticline separating the Zenith ore deposits into two portions and separating the trough longitudinally into two great s\Ticlines, one between the Zenith and Pioneer mines and the other between the Zenith and Savoy mines. (See fig. 14.) Here also parts of the formation are found separated from the main mass by greenstone masses in such a manner as to make it difficult to explain them on the basis of occurrence in troughs alone. It would seem that the main mass of the ormation here has been infolded in such a manner as to give a steep monoclinal trough dipping northward, but that in addition to this main mass, which originally rested upon the greenstone, minor masses of the iron-bearing formation may be mterbedded with the greenstone, so that after the folding they would be separated from the main mass by la3-ei's of greenstone. 138 GEOLOGY OF THE LAKE SUPERIOR REGION. The deposits of Soudan Hill come to the surface near the crest at an elevation of 1,660 feet, about 150 feet above a cross valley to the east between Sf)U(l!Ui Hill and Jasper Peak. The Ely ore dcj)osits are below comparatively low-lyinj; v \ Quartz N Quartz and other and other minerals minerals / / / / Hematite / / / / / / / / Hematite Jaspe Ore Figure 15.— Diagram illustrating volume changes involved in the altera- tion of jasper to ore at Ely, Minn. From average analyses and porosity detemiinations. Comparison of tlie volume compositions of the ore ami jasper shows the removal of a large amount of silica from the jasper. In order to suliiciently reduce the silica content it is necessary that silica equivalent to 63.7 per cent of the volume of the jasper be removed. The VERMILION IRON DISTRICT. 143 ayerage porosity of the ore is approximately 22 per cent of its volume; hence the remaining space left by the removal of silica, or 41.7 per cent of the volume of the jasper, has been filled by infiltration of iron and by mechanical slumping of the ore. The relative importance of these two factors can not be definitely determined, but it is known that both have been efl'ective. The broken and brecciated condition of the ore and the drag at the jasper contacts give abun- dant evidence of slump, and secondary hematite and siderite cementing the ores indicate that a considerable amount of iron has been introduced in solution. Figure 15 illustrates the vol- ume changes above discussed. DISTRIBUTION OF PHOSPHORUS. Phosphorus and iron contents of the Vermilion ores and associatetl ri^cks are as follows: Phosphorite and iron contents of Vermilion ores and associated rocks. Phos- phorus. Relation of phos- phorus to iron. Average ore at Ely Average jasper at Ely Average ore at Soudan Average jasper at Soudan Paint rock from Pioneer mine Per cent. 65.00 28.97 65.21 38.27 16.32 Per cent. 0. 0469 .0213 .108 Per cent.. 0. 0707 .0759 .1655 .127 .778 As calculated from the figures of the Lake Superior Iron Ore Association, 89. 3 per cent of the total production of the Vermilion range in 1906 was of Bessemer grade. The lowest phos- phorus grade was Pilot lump (iron 67.22 per cent, phosphorus 0.0297 per cent), and the high- est phosphorus grade was Vermilion lump (iron 66.07 per cent, phosphorus 0.0878 per cent). The phosphorus contents of individual samples show a much greater range than the grade analyses, the ore containmg as high as 0.500 per cent of phosphorus locally. The paint rock is an altered phase of the greenstone and porphyry, consisting principally of kaoUn more or less stained with hon oxide. It is similar both m appearance and composition to the altered dike rocks of the Gogebic range. These altered igneous rocks are higher m phos- phorus than the ores (the above analysis being a typical one), owing probably to the high phosphorus m the greenstone. "Chemical maps" have been made by the chemists and engineers of the Oliver Iron M inin g Company of the mines on the Vermilion range operated by that company, the phosphorus and iron contents of the ore being entered hi the proper place directly on the nfine maps. Study of these maps fails to show any relation between the distribution of phosjihorus and the wall rocks. The only general conclusion that may be drawn is that in general the phosphorus content is lowest in the largest ore bodies and has a tendency to be liigh m the small shoots of ore away from the mam ore body. The maps show no relation between high-phosj)horus ore and i)amt rock; in fact, in several places ore running as low as 0.030 per cent of phosphorus occurs m the immediate vicinity of high-phosphorus paint rock. Owmg to the very small amomit of phosphorus even m what are termech " high-phosphorus " ores, very little is known as to its mmeral occm-rence. So far as is known, no phosphorus minerals have been identified in the ores or jaspers; hence any conclusions regarding the chemical combi- nations m which phosphorus exists are necessarily based entirely on chemical evidence. Phos- phorus is present in the ores in at least two different forms, knowTi to the Iron-ore chemists as "soluble" and "insoluble" phosphorus, part of it being easily dissolved in hydrocliloric acid and the remamder requiring ignition before it can be dissolved. Chemical analysis of the insoluble residue shows it to be an alummum phosphate. This occurrence of both soluble and insoluble phosphorus is common to ores of the other Lake Superior districts, particularly those of the Marquette range. CHAPTER VI. THE PRE-ANIMIKIE IRON DISTRICTS OF ONTARIO. LAKE OF THE WOODS A^'D RAINY LAKE DISTRICT. INTRODUCTORY STATEMENT. The Lake of the Woods and Rainy Lake distriet includes these hirge hikes and the sur- rounding lands. The district may be considered as being bounded on the south by j)arallel 48° 30', on the north by parallel 50°, on the east by meridian 92° 30', and on the west by merid- ian 95° 30'. The area which has been most closety studied is an angular one running north- west and southeast. The Canadian Survey has published detailed reports by A. C. Lawson on the Lake of the Woods " and Rainy Lake * district and one by W. H. C. Smith <^ on Hunters Island. The geology of tliis region may be said to duplicate in most essential respects, save the distribution of the formations, the geology of'the Vermilion district of Minnesota. The rocks therefore include lower-middle Huronian, Laurentian, and Keewatin. ARCHEAN SYSTEM. KEEWATIN SERIES. The series of Keewatin rocks in the district of the Lake of the Woods is that to wliich the term was first applied. Lawson's study of it, supplemented by later work of others, shows that the Keewatin series is dominantly igneous but includes subordinate amounts of sediments, precisely as in the Vermilion district. The igneous rocks comprise ancient lava flows, volcanic elastics, and contemporaneous and subsequent intrusives. They are dominantly of basic and intermediate varieties, exactly as in the Vermilion district, and among these the characteristic ellipsoidal greenstones are conspicuous. Locally felsites and quartz porphyries occur. In many areas subsequent dynamic action has gone very far, so that the rocks uncommonly have a slaty or scliistose structure. These belts of slaty and schistose rocks Lawson has separated into two divisions,'' one of which he describes as hydromicaceous schists and nacreous schists, with some associated chloritic schists and micaceous schists and included areas of altered quartz porphyry, and the other of which he calls clay slate, mica schist, and quartzite with some fine- grained gneiss. Subsequent examinations of the areas by other geologists have led to the con- clusion that large areas of these rocks are but altered facies of the ordinars^ varieties of the Keewatin igneous rocks. Thus the slates are to a large extent mashed varieties of the ellij>- soidal greenstones and tuffs. At various places the transition between the ellipsoidal green- stones and slaty varieties of rocks produced from them by metamorphism is well shown. However, there are present with the slaty and scliistose rocks of igneous origin subordinate amounts of sedimentary graywacke and slate, including small belts of ordinary black slate which are in some parts carbonaceous. There has not yet been discovered in the Lake of the Woods district any iron-bearing formation corresponding with the iron-bearing vSouilan forma- tion of the Vennilion district, and tliis is the chief cUfference between the two series. The only rocks which could possibly be regarded as a correlative of the iron-bearing Soudan formation o Geology of the Lake of the Woods region, with special reference to the Keewatin (Huronian?) belt of the Arehean rocks: Ann. Rept. Geol. and Nat. Hist. Survey Canada for 1S85. vol. 1 (new ser.), 1.S8H, Rept. CC, pp. 5-l.*)l. with map. !> Geology of the Rainy Lake region: .\nn. Rept. Geol.andNat. Hist. Survey Canada for IS87-18S8. vol. 3 (new ser.), pi. 1. 18S.S. Rept. F, pp. 1-182, with two maps. c Geology of Hunters Island and adjacent counlry: .\nn. Rept. Geol. Survey Canada for hst«HS91, vol. 5 (new ser.). pi. 1. 1S92. Rept. G» pp. 1-7G. S(H^ also The .\ri-hean rocks west of Lake Superior: Bull. Geol. Soc. .\merica. vol. 4, 1S93, pp. 3.'13-34S. d Geology of the Lake of the Woods region, p. 5G. 144 PRE-ANIMIKIE IRON DISTRICTS OF ONTARIO. 145 are very subordinate beds of limestone which occur at various phiccs. The nature of this lime- stone is represented by that at Scotty Islands, where there are narrow bands from a fraction of an inch to 2 feet wide in a schistose and banded greenstone. The layers are usually lens-shaped, and along their strike they may become narrow and pinch out. Commonly the division between the limestone and the greenstone is rather sharp. For the Lake of the Woods district Lawson " gives various sections of the Keewatin, ranging from 6,500 feet to 23,756 feet in thickness. As tliis is a volcanic series and practically all the structures are secondaiy, it may be doubted whether these figures have any real significance. In conclusion it may be well to give the statement of the International Geological Committee,'' consisting of Messrs. Frank D. Adams, Robert Bell, A. C. Lane, C. K. Leith, W. G. Miller, and Charles R. Van Hise, concerning the Keewatin of the Lake of the Woods: In the Lake of the Woods area one main section was made from Falcon Island to Rat Portage, with various traverses to the east and west of the line of section. The section was not altogether continuous, but a number of representatives of each formation mapped by Lawson were visited. We found Lawson's descripti<')ns to be substantially correct. We were unable to find any belts of undoubted sedimentary slate of considerable magnitude. At one or two localities subordinate belts of slate which appeared to be ordinary sediment and one belt of black slate which is certainly sedi- ment are found. In short, the materials which we could recognize as water-deposited sediments are small in volume. Many of the slaty phases of rocks seemed to be no more than the metamorphosed ellipsoidal greenstones and tuffs, but some of them may be altered felsite. However, we do not assert that larger areas may not be sedimentary in the sense of being deposited under water. Aside from the belts mapped as slate, there are great areas of what Lawson calls agglomerate. These belts, mapped as agglomerates, seem to us to be largely tuff deposits, but also include exten- sive areas of ellipsoidal greenstones. At a number of places, associated and interstratified with the slaty phases are narrow bands of ferruginous and siliceous dolomite. For the most part the bands are less than a foot in thickness, and no band was seen as wide as 3 feet, but the aggregate thickness of a number of bands at one locality would amount to several feet. LAURENTIAN SERIES. The Laurentian series is represented mainly by granite, sj-enite, granite gneiss, and syenite gneiss. These rocks occur in masses varying from small areas to those many miles in diame- ter. They are intrusive in the Keewatin series and comprise batholiths, bosses, dikes, and stringers. The nature of the contacts between the Laurentian and Keewatin in the Lake of the Woods area is identical with that of the contacts in the Vermilion district. Along the borders of the batholiths the Keewatin is metamorphosed into hornblende schist or gneiss, exactty as it is in the Vermilion district. Indeed, between the more metamorphosed varieties of these rocks and their less metamorphosed forms there are all gradations. Included in the great batholiths of granite are various masses of Keewatin which have generally been pro- foundly metamorphosed and in many places partly absorbed. The chemical and mineralogical compositions of the batholiths have thus, to some extent at least, been affected by the included material. Similarly the chemical and mineralogical charac- ters of the Keewatin have been affected by the material derived from the granite. Indeed, there are few better examples of endomorphic and exomorpliic effects than those furnished by this district. All these relations may be conveniently studied in the vicinity of Rat Portage. Intrusive into both the Keewatin and Laurentian are later masses of granite and also various basic rocks, including diabase, gabbro, and peridotite. Lawson's maps of the Keewatin and Laurentian in the Lake of the Woods and Rainy Lake district show certain interesting features which have here been better worked out than any- where else in the Lake Superior region. The great batholiths have a tendency to a schistose structure, which is parallel to their borders and is more marked at their exteriors than at their interiors. The Keewatin schists around the borders are in bands, the schistosity of which is rouglily parallel to the batholith boundary. Very commonly a band of Keewatin widens or narrows within a short distance or separates into two or more bands. This subdivision may go on until a band is lost in stringers in a granite mass. With many large areas of schists there appear subordinate granite batholiths, bosses, and dikes. o Geology of the Lake of the Woods region, pp. 104-112. l> Report of the special committee on the Lake Superior region, with introductory note: Jour. Geology, vol. 13, 1905, pp. 96-96. 47517°— VOL 52—11 10 146 GEOLOGY OF THE LAKE SUPERIOR REGION. . Tlio area covcrod by the Laurentian granites is muc'li jjreater tlian tliat covered by tlie Keowatin. It is certain that after the Keewatin volcanic rocks were once spread over tlie entire region, as tliey doubtless were, the}^ must have been raised in great domes, pushed aside, and jammed in between the batholithic intrusions. It is probable that the greater areas of Keewatin which once overlaid tliese batholiths have been removed by erosion anrl that the existing masses of Keewatin are but mere renmants of a great volcanic formation which once covered the entire district. It has been suggested also tliat parts of the Keewatin may have foundered and suidv in tlie granite batholiths at thi^ time of intrusion." The foregoing facts in reference to tlu! relations of tlie Laurentian and Keewatin have led Lawson '' to his subcrustal fusion theory, his idea being that the Laurentian represents the fused and recrystallized nuxsses of the Keewatin. There is no doubt tliat along the border of the batholitJiic masses a certain amount of Keewatin material has been aljsorbed, and no doubt that the Keewatin along tlie borders of the granites has derived material from them; thus there appears in some places to be an approach to chemical gradation between the two. The known facts, then, are these: The Keewatin volcanic period antedated the Laurentian. The Keewatin rocks were intruded by the various Laurentian granites and syenites, extending tlu-ough an enormous period of time. There were important exoniorphic and endomorphic effects. There is difference of opinion as to the amount of tlie Keewatin which has been absorbed by the Laurentian. Our own view tends toward conservatism in this matter. ALGONKIAN SYSTEM. HURONIAN SERIES. The Huronian rocks in this district belong to the lower-middle Huronian. They are chiefly confined to the southern part of the Rainy Lake area. From west to east their northern boundary roughly follows Rainy River, the central body of Rain}- Lake, and Seine River with its various enlargements. From this line the Huronian extends across the mternational bound- ary into Minnesota for distances not yet determined, except at a few points. Tliis is the main mass of rocks to which Lawson gave the name "Coutchiching." In the area imder discussion the rocks consist dominantly of mica schists, but there are argillaceous slates, mica slates, gray- wackes, and conglomerates at the bottom. All the evidences of imconformable relations between these rocks and the Laurentian and Keewatin series are found in many places along the contact. Where the underlj'ing rocks are Keewatin detritus is mainl}' derived from that series; where they are granite, as at Bad Vermilion Lake, the detritus is mainly derived from granite. In intervenmg areas both granite and greenstone are found. Also with these mate- rials is found detritus of other kinds, such as felsites, quartz porphyries, and gneiss. The materials include practically all the varieties of the Keewatin rocks. In places the conglom- erate passes up into a feldspathic quart zite and tliis mto a micaceous graywacke or slate. Wherever the bedding can be recognized the dip is steep to the south. The main areas of the lower-middle H\u"onian micaceous schists have been intruded by large masses of granite which maj' be especially well seen at the end of the southeast arm of Ramy Lake and in Namakon Lake along the international bouudarj-. From masses of very considerable size the intrusive granite varies to masses of much smaller size, and cutting through the mica schists are very numerous granite dikes, a large number of which are roughly parallel to the foliation. Large Ij" in consequence of the mtrusions of the granite the mam mass of the lower-middle Huronian has been transformed to a well-crystallized mica schist. As a result of these intrusions, from the end of the southeastern arm of Rainy Lake northwestward to the base of the series the rocks are less anil less metamorphosed. Possibly this grailation was one of the factors in Lawson's conclusion that the Keewatin series was higiicr and rested uncon- formably upon his "Coutcliiching," which exactly reversed the true relation. As the relation <• Daly, R. .\., The mechanics of igneous intrusion: .\ni. Jour. Sci., 4th ser., vol. 26, 190S, p. 30. 6 Geology of the Rainy Lake region, p. 131. PRE-ANIMTKIE IRON DISTRICTS OF ONTARIO. 147 of the great mica schist series to th(> Keewatin is one about \vlii( li tliere is difference of opinion, the statement of the International Geological Committee/' consisting of Messrs. Frank D. Adams, Robert Bell, A. C. Lane, C. K. Leith, W. G. Miller, and Charles R. Van Hise, who visited this district and examined the contact, is here quoted: In the Rainy Lake district the party observed the relations of the several formations along one line of section at the east end of Shoal Lake and at a number of other localities. The party is satisfied that along the line of section most closely studied the relations are clear and distinct. The Coutchiching schists form the highest formation. These are a series of micaceous schists graduating downward into green homblendic and chloritic schists, here mapped by Lawson as Keewatin, which pass into a conglomerate kiio\vn as the Shoal Lake conglomerate. This conglomerate lies upon an area of green schists and granites known as the Bad Vermilion granites. It holds numerous large well-rolled fragments of the underlying rocks, and forms the base of a sedimentary series. It is certain that in this line of section the Coutchiching is stratigraphically higher than the chloritic schists and conglomerates mapped as Keewatin. On the south side of Rat Root Bay there is also a great conglomerate belt, the dominant fragments of which consist of green schist and greenstone, but which also contain much granite. The party did not visit the main belts colored by Lawson as Keewatin on the Rainy Lake map, constituting a large part of the northern and central parts of Rainy Lake. These, however, had been visited by Van Hise in a previous year, and he regards these areas as largely similar to the green- schist areas intruded by granite at Bad Vermilion Lake, where the schists and granites are the source of the pebblee and bowlders of the conglomerate. As to the existence of areas of sediments equivalent in age to the lower-michlle Huronian in other parts of the Rainy Lake and Lake of the Woods district, no defmite statements can yet be made. It is probable, however, that close structural studies of tliese areas will disclose such sediments. Indeed, a traverse of the Rainy Lake section by the senior author led him to think that in the belt of rocks mapped as Keewatm, running from the southeastern end of Crow Lake to Manitou Lake, there are representatives of this upper series, but the area was not sufficientlv studied for its areal distribution to be given. On the other hand, it is certain that some areas which have been mapped as "Coutchiching" on the Ramy Lake sheet of the Geological Survey of Canada, and especially on adjacent sheets, are but the chloritic and hornblendic schists of the Keewatin metamorphosed by the intrusive granite. It is plain that the term "Coutchiching," if it is to have any structural significance, must be restricted to the sedimentary series of Ramy Lake, its extensions and equivalents. It must not be used as a term to cover the more schis- tose varieties of rocks of the region without reference to their stratigraphic position. As to the thickness of the so-caUed "Coutchichmg," Lawson '' gives estimates varying from 23,760 feet to 28,754 feet. These measurements, however, are clearly based on cleavage structures rather than on bedding, and close examination shows that the two do not conform; hence there is grave doubt whether the thickness of the series is more than a fraction of these estimates. It has already been indicated that in the lower-middle Hiu-onian schists ("Coutclaiching" of Lawson) there are intrusive masses of granite which have produced metamorphic effects. In addition to these granitic masses cutting all the formations of the district are later diabases, dikes, and bosses which are supposed to be of Keweenawan age. STEEP ROCK LAKE DISTRICT. OENfiRAL GEOLOGY. The Steep Rock Lake district has been described and mapped by II. L. Smyth "^ and W. N. Merriam.'* The authors have visited the district for general correlation purposes but have not studied it in detail. The following account is based principally on Merriam's work. ajour. Geology, vol. 13, 1905, p. 95. 6 Geology ol the Rainy Lake region, pp. 131-102. (■Structural geology of Steep Rock Lake, Ontario: Am. Jour. Sci., 3(i ser., vol. 42, 1891, pp. 317-331. d Private report. 148 GEOLOGY OF THE LAKE SUPERIOR REGION. The geology of the Steep Rock Lake district is similar in essential respects to that of the Vermilion and Rainy Lake districts. The succession, in descending order, is as follows: Algonkian system: Intrusive rocks. Huronian series: Lower Huronian... I iilerbedded sediments and eruptive rocks: Dark-gray slate, agglomerate, greenstones and green schists, conglomerates, and limestone, .forming part of Steep Rock "series" of Smyth, estimated by Smyth to be 5,000 feet thick. Unconformity. Archean system: Laurentian series Granites and gneisses intrusive into Kecwatin. Keewatin series Ellipsoidal greenstones and green schists containing iron formation. The lake resembles an irregular letter M, of wliich the western arm runs north and south and the eastern arm northwest and southeast. The Keewatin greenstones have a wide distribution on the south side of the lake, especially near Straw Hat Lake. They are in isolated areas surrounded and overlapped ])y the lower Huronian sediments. The principal showing of iron-bearing formation is southwest of Straw Hat Lake. It is in contact with elhpsoidal greenstone on the west side, but the relation on the cast is not Ioiowti. Lean iron ore also outcrops on the west side of the lake and in various other parts of the district. Glacial fragments of iron ore have been found on the south side of the lake opposite Mosher's Point. The Laurentian granites and gneisses are exposed principally on the north and east sides of the lake. Along the contact of the Laurentian and Keewatin in the southeastern part of the district there is a great series of hornblende schists intricately associated with both Kee- watin and Laurentian rocks. These are regarded as contact phases of the Keewatin where it is intruded by the Laurentian, similar in all respects to those of the Verniihon district. Smyth regards them as overlying the lower Huronian sediments and as passing upward into the schists of Atikokan River, which he designates as the "Atikokan series." The lower Huronian fringes the Laurentian on the southwest. Its principal exposure is on the south and west shores of the lake, but small patches of it rest against the granite on the points projecting from the east and north sides of the lake. It dips at 60° to 80° away from the Laurentian. The basal conglomerate carries fragments of various phases of Lauren- tian and Keewatin rocks. Where the conglomerate rests against the granite it is made up so largely of granitic debris and has been so metamorphosed that it is frequently difficult to deter- mine the exact contact of the granite and the sediments. According to Smyth, the succession above the conglomerate is: Lower limestone, ferruginous horizon, interbedded crystalhne traps, calcareous green scliists, upper conglomerate, greenstones and greenstone schists, agglomerate, and dark-gray clay slate. Some of the greenstones and green scliists included by Smyth in the lower Huronian are regarded by Merriam and by the authors as, at least in "part, Keewatin unconformable beneath the Huronian. According to Smyth, the Steep Rock group is folded into an eastern syncUnal, a middle anti- clinal, and a western synclinal, the latter being faulted. The axes of these folds have a liigh pitch to the south, varying from 60° to nearly 90°. Throughout the whole area is a regional cleavage which has a nearly uniform direction transverse to all the members of the Steep Rock group and also to the contact between this group and the basement complex. This has largeh* obliterated the original lamination of the sediments and is now the donunant structure. It is therefore the effect of the last force which has left its marks upon the rocks of the lake. Before this last force acted upon tlic rocks the Steep Rock group had been folded into a south- westward-tlipping monoclinal, which, under the action of the cleavage-prochicing force in a northeast and southwest direction, caused the present fluted outcrop of the formations of the Steep Rock group. That the basement complex itself yielded to tliis latter force is sliowni by the irregular outcrops of the dikes cutting it. At least three varieties of intrusives cut the Laurentian ami Keewatin and have supplied pebbles to the conglomerate at the base of the lower Huronian. Other iiitrusivcs cut the PRE-ANIMIKIE IRON DISTRICTS OF ONTARIO. 149 Keewatin, Laurentian, and lower Iluronian rocks but have been subjected to folding. Finally, a single massive dike appears to be subsequent to the latest period of folding. IRON ORES. Lean, banded iron-bearing rocks appear in the Keewatin of the Steep Rock Lake district. The principal showing is southwest of Straw Hat Lake. The rocks are in contact mth ellipsoidal greenstone on the west side, but the relation on the east is not known. Lean iron ore also outcrops on the west side of the lake and in various other parts of the district. Glacial fragments of iron ore have been found on the south side of the lake opposite Mosher's Point. Explorations southwest of Straw Hat Lake are reported to have recently disclosed an ore deposit. ATIKOKAN DISTRICT. The existence of iron ore along Atikokan River and Sabawe Lake to the east of Steep Rock Lake requires mention of the geology of tliis area. The area has not been geologically mapped in detail. As a result of visits to the district Mr. Merriam and the authors behevo the geology to be similar in all essential features to that of the Steep Rock Lake and VermiHon districts — that is, there are represented in this district Keewatin, Laurentian, and lower Huronian rocks. The ores are in the Keewatin series. The Atikokan iron ores are 3 miles north of the Canadian Northern Railway, on the north side of Atikokan River, just east of its expansion into Sabawe Lake. Here is a ridge of magnet- ite, green scliist, massive greenstone, and iron carbonate running approximately parallel to the river. The greenstone is essentially a diorite with a large proportion of hornblende. The magnetite is coarsely crystalline and dense and carries abundant ampliibole and iron pyrites and small amounts of the nickel minerals. The relations of the magnetite and greenstones are complex, as in the VermiUon district, but as a whole the greenstone seems to be intrusive into the magnetite. To the west of the main magnetite exposure iron carbonate appears in similar relations to the greenstone. So intricate are the relations of the ore to the greenstone that it is difficult to determine the true shape of the magnetite deposit from the surface outcrop. The bands are narrow, at most not more than 44 feet, and extend along the bluff for more than 400 yards. They are now being opened for mining. The ores will be roasted and used in furnaces at Port Arthur. To the west, down the river, the iron-bearing formation is exposed with similar association to greenstone and green scliist at a number of places. KAMINISTIKWIA AND MATAWIN DISTRICT. The Kaministikwia and Matawin district is characterized by lean, slightly magnetic cherts and jaspers in vertical bands and lenses, very irregular, closely associated with green schist and ellipsoidal basalt typical of the Keewatin series. Granite and quartz poii^hyry intrude the Keewatin at many places. The association of the jasper and greenstone and porphj'ries pre- sents all the problems of the Vermilion district. The principal exposures are along Kaministi- kwia River between Kamjnistikwia and Mokoman. Just north of the railway, a mile north of Mokoman, is a jasper and greenstone breccia and conglomerate. The rock here exposed has essentially the features of a breccia, but parts of it contain fragmental quartz and are truly conglomerate, suggesting that it is perhaps the basal conglomerate of the lower Iluronian. Still farther south, near Kakabeka Falls, the flat-lying iron formation and slates of the Animikie group (upper Huronian) are exposed along the river and at Kakabeka Falls. Farther west in the same township (Conmee) are more extensive outcrops of banded jasper m chert containing impure siderite. It is strongly magnetic. Farther south in Conmee Township, on the south half of lot 7 in the sixth concession, the iron range is found again with a trend of about northwest and southeast and a nearly vertical dip on a long ridge about 150 feet wide. The silica is mainly jasper, often of beautiful color, banded with magnetite, the bands often folded in complex ways, and here also there is more or less of a peculiar breccia of grained silica or jasper in a fine gray matrix. ' 150 GEOLOGY OF THE LAKE SLtPERIOR REGION. In the southeast end of lot 7 in the fifth concession there is finely banded jasper and some impure carbonate intermixed, but on lot 4 in the third concession the rock is unusually black from the presence of magnetite, and some specimens are heavy enough to make fairly good ore. Bands having a width of 1 or 2 feet appear to be nearly solid masnotite and seem rich enough to work, though a small amount of pyrite present would lower the grade of the ore. The banding varies in direction from southeast to south; and here again a conglomerate or breccia is commonly found mixed with the ore, the wliole having a length of 10 chains and a width of 135 feet. Altogether this series of iron dejjosits has been traced for about 8 miles, running parallel, it is said, to a similar range located by Tumpelly and Smyth L' miles to the southwest; and probably both are continuations of the Matawin ranges, though curvdng in a somewhat different direction."! The Matawin iron bolt e.xtends from Kaministikwia station westerly beyond Greenwater Lake. Banded magnetic and hematitic cherts and jaspers outcrop at many places on Matawin and Shebandowan rivers. West of this belt banded iron ores were seen outcropping at Copper Lake, south of Shebandowan Lake, and on the eastern shore of Greenwater Lake. Ores which probably form an extension of the same belt occur south of Moss Township, on the farther side of the gneiss area of Greenwater Lake. MICHIPICOTEN DISTRICT. The following account of the Michipicoten district is taken largely from the writings of A. P. Coleman and A. B. Willniott'' and of J. M. Bell."^ The present writers have made no detailed survey of the district, but have visited the area and agree with the essential conclusions reached by the men named. GEOGRAPHY AND TOPOGRAPHY. The Michipicoten district, on the northeast shore of Lake Superior, is about 25 miles in length from southwest to northeast, with a greatest width of about 7 miles, and runs from the mouth of Dore River, a few miles beyond Parks Lake on the northeast. It lies northwest of Michipicoten River near its entry into the bay of the same name on the northeast side of Lake Superior and shows the rugged topography so characteristic of that shore. The country rises rapidly from the lake in steep hills, often ridgelike, with the general direction of the strike of the schists about 70° east of north, and culminates in the ridge of iron-range rock just east of the Helen mine, called Hematite Hill or Mountain, which reaches a height of 1,100 feet above the lake or 1,700 feet above the sea. This is the highest point for many miles around and makes a conspicuous landmark, though other hUls reach a level of 800 or 900 feet. As Hematite Mountain is only 7 miles from Lake Superior, the rLse is rapid, and the location of the railway to the Helen mine, which is at a level of G50 feet, just at the foot of the mountain, required some skill in the choice of a route, old lake beaches and sand plains being utilized where possible. 6 SUCCESSION. The succession of formations here given is that of Coleman and WLllmott, but the names of the series are changed m accordance with the recommendation of the special committee on the Lake Superior region and the series are grouped into the Algonkian and Archean systems. "^ Algonkian system: Huronian series: Lower-middle Huronian fBasic eruptive rocks. ^ ("Upper Hiuonian" of | Acidic eruptive rocks. Coleman and Willmott). [Dor6 conglomerate. Unconformity. Archean system: Laurentian series Granites and gneisses intrusive into Keewatin series. Eleanor slate. Helen formation (iron-bearing). W'awa tuff. Gros Cap greenstone. Keewatin series ("Lower Hu- ronian' ' of Coleman and Will- mott). Coleman, ,\. P., Iron ores of northwestern Ontario: Eleventh Rept. Ontario Bur. Mines, 1902, p. 130. !> Tlie Midiiplcoten iron ranges: fniv. Toronto Studies, geol. ser., No. 2, 1902. 47 pp. See also Rept. Ontario Bur. Mines, 1902. pp. 152-llW. c Iron ranges of Michipicolen West; Rept. Ontario Bur. Mines, vol. 14, 1905, pt. 1. pp. 278-3.''», with geologic map. rf Report of the special committee on the Lalte Superior region: Jour. Geology, vol. 13. 1905, pp. 89-104. PRE-ANIMIKIE IRON DISTRICTS OF ONTARIO. 151 The geology of the Michipicoten district is remarkably similar to that of the Vermilion district of Minnesota in regard to hthology, succession, and structure. ABCHEAN SYSTEM. KEEWATIN SERIES. GROS CAP GREENSTONE. DISTRIBUTION. The Gros Cap greenstone is well exhibited just west of Michipicoten Harbor and on the trail to the old fishing station at Gros Cap. The most extensive area of the Gros Cap greenstones is the one extending from Gros Cap eastward to Magpie River and thence north from Michipicoten River to the eastward bend of the Magpie. Other large areas exist northeast of Eleanor Lake, including most of the shore of Loonskin Lake, and along the Josephine branch railway from mile LH to mile 17. Numerous smaller areas have been mapped. There are bands of greenstone and green schist in the Wawa tuff that have the same characteristics. PETROORAPHIC CHARACTER. The Gros Cap greenstone consists of ellipsoidally parted basic igneous rocks formed partly of lava flows, in all respects similar to the Ely greenstone of the Vermilion district of Minnesota. Many parts of the greenstones do not show the ellipsoidal structure and are apparently greatly weathered dia- bases, while still other parts are distinctly schistose; but the three varieties run into one another and can hardly be separated in mapping. The chloritic schists are probably tuffs of the volcanoes which poured out the lavas. The whole series is greatly weathered and saussuritic in thin sections. WAWA TUFF. DISTRIBUTION. The extent of the Wawa tuffs and their boundaries can be given only approximately, partly because of the sand plains covering them and partly on account of the intermixed later eruptive rocks. Beginning at the southwest is a narrow band of quartz porphyry schist and felsite schist along the northern boundary of the Dore conglomerate area, between the latter rock and the Laurentian. Where the Dore conglomerates narrow toward the northeast, the northern fringe of quartz porphyry schist seems to widen correspondingly, though greatly interrupted by later acid and basic eruptives. Still farther northeast the sand plains of the Magpie Valley hide the rocks almost completely, not to reap- pear until near Talbott Lake, where the Wawa schists are extensively developed. From here to the northeast end of the region mapped the Wawa schists are found on each side of the bands of the iron range as the immediately inclosing rocks, except where broken by masses of greenstone or later diabase, and they extend northeast to the end of the region mapped. PETROGRAPHIC CHARACTER. The Wawa tuff generally has the composition of quartz porphyry or felsite, and in some places evidently consists of mashed and rearranged rocks with crystals of quartz and feldspar still to be seen in them. In general, however, the formation apparently consists of tuffs or ash rocks, prob- ably erapted in connection with the quartz porphyry and deposited in water, so that they have a more or less stratified character. A few of them are brecciated, some crashed breccias, others perhaps agglomerates formed of volcanic fragments larger than the ash. Some rare forms have much the appearance of water-formed conglomerates with rounded pebbles, one singular example of the sort occurring on a steep hill slope at the west end of Lake Wawa. In a gen- eral way this resembles a beach deposit with pebbles cemented by a finer-grained greenish or yellowish matrix, but on closer examination the apparent pebbles are found to be really concretions. There is no sharp line between this phase of the rock, which occurs in sipaller amounts at other points also, and varieties like ordinary quartz porphyry schist, so that one may suppose it to be merely a phase of the series of acid schists in which there has been concretionary action. 152 GEOLOGY OF THE LAKE SUPERIOR REGION Since the materials forming the schists were laid down, or else during their deposit, important chemical changes have taken place in them, probably by circulating hot water, bo that sheared and crushed quartz porphyry or porphy- rite has been greatly silicified, at times even transformed into thick bands of pale-gray or green chert or chalcedony, with a small amount of eericite. This phase is similar to parts of the Palmer gneiss of the Marquette district. In other cases a considerable amount of siderite or of a carbonate like aiikerite, dolomite, or calcite has been deposited with cryptocrystalline or microcrystalline silica, suggesting a change to the iron-range rocks which form the uppermost series of the lowor Iluronian. It is probable that this change went on at the time when the original iron-range rocks were deposited and under the same conditions. In a gcneial way it may l>o stated that tlie Wawa tuff is accompanied by lenses or bands of carbonates, inckiding impure siderite.s, dolomites, and limestones. In most places some granular silica also is present, and it may be that these lenses or bands are chemical sediments. In a general way the Wawa tuffs tend to bo more siliceous and to contain more siderite as they approach the iron range, and to be somewhat coarser in grain and gneissoid in look on the sides toward the I.aurentian, as though the proximity of these rocks had influenced their crystalline character and chemical composition. The boundary between them and the Helen iron-range rocks is sometimes quite sharp, a thin sheet of black slate occasionally intervening between the two, but in other cases there are schistose varieties of the siderite of the iron range which form a transition toward the quartz-porphyry schists. STRUCTURE AND THICKNESS. The Wawa tuffs have on the average a strike of 70° east of north, though with considerable local variations, and a dip toward the south of from 50° to verticality. Near the Helen mine they are shown to form a syncline pitching toward the east and inclosing in their trough the iron-range rocks. As the dip is much the same on each side of this synclinal axis, the fold must have been a closed one; and since it was formed erosion has eaten down the Archean sur- face until at various points, such as west of the Helen mine and south of Lake Eleanor, the iron range in the central trough has been completely removed, leaving the lower schists across the whole width. The greatest measured thickness of the schists is to the south of Sayers Lake, where they are known to reach across Lake Wawa, a distance of about two miles and a quarter, which at a dip of 70° would give more than 11,000 feet. HELEN FORMATION. DISTRIUUTION. Beginning at ttie southwest, several bands of the granular sUica variety occur on the Gros Cap Peninsula, the largest being at the Gros Cap mine on the south shore of the peninsula." The materials here are chert and granular silica interbanded with hematite, and the width is in all about 150 feet. To the east another narrower band of rusty siliceous rock is seen, and just around the eastern point, near the beacon, is a third still narrower band, differing from the others in containing magnetite and much pyrite. All of these bands of iron range run about northwest and southeast and have a dip of perhaps 50° to the southwest. A similar band is seen on the west shore somewhat south of the portage across the neck of the peninsula, probably an extension of one of the bands mentioned before. About 150 yards north of the portage are several narrow bands of the rock, usually very pyritous, associated with quartz porphjTy schist and striking about east and west with a dip to the south. This belt probably extends to the east, where an outcrop of brown sandy-looking grained silica occiu's a little inland fi'om the old fishing station. The band just mentioned is nearly parallel to the great area of schist conglomerate to the north and is the nearest part of the iron range to it, so that it may ha\-e furnished part of the pebbles of granular silica in the conglomerate. Two or three small patches of the iron range are found in the greenstone east of Michipicoten Harbor, after which no more is known for about 8 miles, when the Helen iron range begins. All of the outcrops mentioned thus far appear to be inclosed in the greenstones as if swept off eruptively. The prmcipal belt outcrops near the Helen mine. Beginning on the west, the iron range as found at the Helen mine is in two long fingers reaching the shore of Talbott Lake, but not crossing it. The southern finger, long and narrow, possibly reaches a short distance into the water of the lake, but does not appear on the opposite side. It extends eastwardly uji the valley of a small creek until it reaches the main body of the formation near Sayers Lake. Following the boundary northward are several minor folds which are seen to rest on Wawa tuffs. Then crossing the railway track near the outlet of the lake, the range extends westward down to the shore of the lake, where it comes to an end within a few feet of the shore, being bottomed by Wawa tuffs. A comparatively recent development has been the finding of a large band of iron-bearing formation witliin 2 miles to the northwest of the Helen mine in an area Avhich was supposed to have been carefully explored. It is associated with a thick belt of black slate, but most of the inclosing rock is of the Gros Cap and Wawa formations. o Kept. Geol. Survey, Canada 1863-1869, p. 131; also Eighth Kept. Ontario Bur. Mines, pp. 145, 254. PRE-ANIMIKIE lEON DISTRICTS OF ONTARIO. 153 On the north side the range seems to extend quite regularly toward the east, the formation standing almost ver- tically. [From the Helen mine the range] runs for a mile and three-quarters a little north of east, when another inter- ruption occurs, thought by some to be caused by a fault. The evidence for this does not seem conclusive, and more careful exploration may bring to light in the heavily wooded region to the east some links connecting it with the Lake Eleanor band, which commences after a gap of a mile and a half and runs northeast to the Grasett road between Lakes Wawa and Eleanor. The road follows a depression between hills that probably represents a line or zone of fault- ing, for the iron range here jogs three-eighths of a mile to the north and then continues the usual strike of about 60°. Between the two main outcrops and just east of the road are two small ridges of rusty granular silica pointing a little east of north, perhaps remnants left during the dragging of the strata in faulting. The iron range south of Lake Eleanor gives the best exposure of the range between the Helen and Josephine mines. In a general way it suggests that of the Helen mine, though on a smaller scale. The strike of the ii-on-range rocks at the extreme southwest end is not far from north and south, with a dip running from 30° to 90° to the east, pointing toward the two hills of granular silica to the east of the road. Less than 100 paces eastward along the top of the ridge the strike becomes G0° to 80° and keeps this direction until the east end of the little lake is passed, when it changes to 45° for a short distance, and the range ends abruptly in a mass of greenstone. Beyond this it has not been traced, but the country is very mossy and forest covered, so that it is hard to say positively that there may not be exposures of the iron range yet undiscovered. The next point at which the iron-bearing rocks have been found is 2J miles to the northwest of the Lake Eleanor range, where they begin just east of a long unnamed lake and run about 60° east of north past the north side of Brooks Lake almost to Bauldry Lake, a distance of about 2 miles. Here again a fault of great magnitude has been suggested, the plane of faulting running northwest and southeast; and there is much in favor of this view, though it can not be said to have been proved, since very little work has been done on the geology of the country between the two iron ranges. The only rocks known to exist between them are greenstones and green schists. The iron-bearing formation appears again near the south side of Baukhy Lake and extends eastward past the south side of Long Lake. Beginning at Goetz Lake and rimning east through Brooks Lake and Kimball Lake is a considerable belt of iron formation, on vvliich the Josephine mine is located. Ore has been foimd here in small amount by driUing, but has not been mined. STRUCTURE AND THICKNESS. In a general way the rocks of the Helen iron formation, though so narrow, rarely exceeding 1,000 feet in width, are the most distinctive features of the lower Huronian, since they are very easily recognized and nearly always rise as sharp ridges above the sui'rounding region. Except on Gros Cap, where the bands strike about northwest and south- east, the different ridges have a surprising uniformity of strike, about N. 60° to 70° E., the same direction as one finds prevalent in the adjoining schists. Though the general strike is so uniform, it is evident that along with the other rocks of the region the u'on formation has been interrupted frequently by eruptive masses, and apparently also by faults of great magnitude, the effect always being to shift the part east of the fault plane toward the north. It is probable that the bands of iron range are not simple tilted strips of rock but closely folded sheets, only the lower portion of which is still preserved, and it may be that the apparent gaps between the ranges are really due to the erasion of the general rock surface so far down as to cut off the folded upper part of the lower Huronian altogether, leaving only the schists beneath. If this is the case the depth to which the iron-bearing rocks descend may be quite limited, though the amount of mining and diamond drilling done on the range does not give very certain evidence in this respect. The iron-bearing formation at the Helen mine underlying the Boyer Lake basin is peculiar in that the lake bottom is much below the outlet. The origin of tliis is discussed elsewhere (p. 158). PETKOGBAPHIC CHARACTER. Five species of rock may be distinguished in the iron-bearing Helen formation — banded granular silica or ferruginous cherts with more or less iron ore, black slate, sitlerite with varying amounts of sihca, griinerite schist, and pyritic quartz rock. All are found well developed at the Helen mine, and all but the griinerite schist have been found in the Lake Eleanor iron range also ; granular sihca and siderite occur in large quantities in every important part of the range, though small outcrops sometimes show the silica alone. The ferruginous cherts are in many places soft and sandy, like the ferruginous cherts or taconites of the western Mesabi. Jaspery varieties have not been found on this range, but they occur only a few miles to the north. RELATIONS TO OTHER FORMATIONS. The Helen formation is very closely related to the Gros Cap greenstone and Wawa tuff. Its relations to the associated rocks of the Keewatin series are almost identical with the rela- tions of the Soudan formation of the Vermilion district of Minnesota to the adjacent Keewatin 154 GEOLOGY OF THE LAKE SLTEKIOK REGION. rocks. In general the iron-bearing formation from its structure seems to be at upper hori- zons of the Keewatin and to rest on the other rocks, being fol'ded in with (hem; bvit tlie forma- tion lias been also intrudeti by basic rocks wliich have been mapped as Gros Cap greenstone, and some of them may be intei-bedded with the surface flows of the Gros Cap greenstone. For a discussion of the problem the reader is referred to the chapter on the Vermilion district and also to the (Hscussion of the origin of tlie ores. (Sec Chapters V, pp. 118 et seq., and XVII, pp. 460 et seq.) ELEANOR SLATE. In addition to the slates of the Wawa formation, i^lates of a distinctly sedimentary kind occur as thin bands in the northeastern part of the region near Eleanor Lake and elsewhere. .Slate or shale of the kind described is traceable at intervals for a mile along the north shore of Parks Lake, and is found underlying the Dot6 conglomerate north of Eleanor Lake on the Grasett road. They are buff to dark -gray or black rocks with slaty cleavage, sometimes forming an angle of 25° with the well-marked bedding. Some varieties of them are carbonaceous, and at a point east of Wawa Lake such a slate was taken up as a coal mine. \\'hether the black graphitic slate often cormected with the iron ranges belongs with Eleanor slates is not certain, nor has it been determined positively whether the slates are older or yotmger than the adjoining iron-bearing rocks. LAURENTIAN SERIES. The Laurentian series includes various types of granite, quartz porphyry, quartz por- phyrite, felsite, and quartzless porphyry. They are intrusive mto the Keewatin series and in part into the overlymg lower-middle Huronian sediments, but in large part also they he miconformably below the Hm-onian, as is shown by the numerous pebbles of Ijaurentian gneiss and granite included in the basal conglomerate of the Huronian. It is not desirable that all these mtrusives should be classed with the Laurentian, as that term is properly appUed only to the pre-Huronian rocks, but they have not been sufficiently w^eil discriminated and mapped to warrant a separate classification. ALGONKIAN SYSTEM. HTJBONIAN SERIES. LOWER-MIDDLE HUKOXIAX. dor£ COITGLOMEKATE. distribution, topography, and stritcture. The conglomerate occurs fi-om point to point along the shore as far as Dog River, 10 miles to the west, and east- ward to about the third milepost on the railway from Michipiroten Harbor to the Helen mine, a distance of 4 miles; while the greatest width measured during last summer's work is about a mile and a half, on a line due north of the harbor. In general the topography of the conglomerate band is very rugged and hilly, with numerous successive ridges running parallel to the strike, which averages about 70°; and with very steep slopes on each side, but especially toward the north, where the narrow hills often drop off vertically or even overhang. The cause of this is to be found in the unequal resistance of the different layers to weathering and in the fact that the dip is usually very steep, from 60° to 90°, averaging about 75° to the south. Dips to the north have only rarely been noticed. The steep cliffs formed in the way described often have a height of 50 or more feet, and on the north side are frequently unscalable for consider- able distances. Perhaps the most rugged portion of the region is directly north of Michipicoten Harbor, where within 2 miles of the shore there are several of these ridges, with Valleys between, rising finally to over 600 feet above Lake Superior. \\Tiile the general strike is about 70° there are great local \'ariations, especially in the vicinit)^ of eruptive masses. Near the second mile on the railway the strike is nearly north and south for more than 400' yards, but on each side the usual directions of from 70° to 75° are found. There is good reason to believe that in general the strike of the schistosity corresponds to that of the sedimentation, for bands of rock free from pebbles follow the same direction, but in a few cases the schistose structure seems to cross the direction of sedimentation, having a bearing of about 45°, while the general course of the ridges is 70° to 80*". PETROGRAPHIC CHARACTER. Tilie conglomerates are for the most part large and well rounded. They consist of dark- green schist, granite, ferruginous chert, spotted gray-green scliist, porpiijTj-, felsite, and con- glomerate or breccia. All have been more or less flattened during the development of schistosity in the rock. PEE-ANIMIKIE IRON DISTRICTS OF ONTARIO. 155 The conglomerate is in many places penetrated by dikes of quartz porphyry, or sometimes quartzless porphyry, running parallel to the stratification as a rule, and in many cases squeezed or sheared into felsite schist in which the porphyritic structure is almost lost. In addition to the porphyry dikes there are numerous masses and dikes of diabase rising through the conglomer- ate, apparently later in date than the porphyries, since they are seldom squeezed into schists so far as ob.served. The diabase seems to be the most resistant rock of the region with the exception of the iron range of the Helen mine, and accordingly forms in many cases the tops of the highest ridges. THICKNE.SS. The general attitude of the large area of schist conglomerate just described suggests a continuous series of strata, as supposed by Logan, since in most cases the dip and strike are fairly uniform; and any marked variations maybe accounted for by the presence of eruptive rocks. This would give them a thickness of about 7, .500 feet, for the greatest width is 8,000 feet, with an average dip of about 7.5°. However, it is not easy to imagine the mass as tilted bodily, and it is more natural to think of the series as form- ing a close fold, most probably a syncline with the two sides closely squeezed together, and tilted slightly against the Laurentian mass' to the north. In this case we may suppose that the schists were to some extent pulled asunder at the base of the fold, which was in tension, allowing the felsites and diabases to penetrate parallel to the cleavage. There is no doubt, however, that some of the diabase dikes are later in age and cut diagonally across the schistose structure. One feature of the arrangement of the conglomerates supports the view that they form a syncline. Toward the western end of the series of rocks we find bands of well-defined conglomerate along each side with gray and green schists showing few or no pebbles between, as if there was an upper layer of finer sediments nipped in between the two sides of the conglomerate. The absence of pebbles in this central area may, however, be due merely to a greater amount of compression, flattening them beyond recognition. Toward the eastern end there are very few gaps where pebbles have not been seen. RELATIONS TO UNDERLYING ROCKS. The T)or6 conglomerate near Michipicoten Harbor is nowhere found in contact with undoubted Archean rocks, though what look like Wawa tuffs and have been mapped as such occur as a narrow band to the north between the conglomerate and the Laurentian; and schists with some granular silica, certainly lower Huronian, are found near the north end of the peninsula of Gros Cap, though a small sand plain separates them from the conglomerate. The Lauren- tian eruptives have not been seen in actual contact with them on the north, though some belts of green schists in the Laurentian a little way from the hidden contact may be gi'eatly metamorphosed conglomerate swept off at the time of eruption. The relationship to the south is more distinct, and the Gros Cap greenstones appear to be the underlying rock folded into a syncline with them; so that south of the railway half a mile from the harbor the greenstone seems to over- lie the conglomerate, both having a dip of about 70° to the south. The pebbles, however, are clearly derived from the rocks of the adjacent Keewatin and Laurentian. Their variety and large size characterize the conglomerate as a basal conglomerate marking a great unconformity. MICHIPICOTEN EXTENSIONS. Many areas of iron-bearing rocks near the Michipicoten district have been reported and mapped by Coleman, Bell, and others. Their lithology and association are similar to those of the Michipicoten district. No attempt Avill be made here to describe in detail the individual belts. None of them are productive and in few of them have detailed geologic maps been made. J. M. Bell"^ has reported on the iron ranges of Micliipicoten West, covering the northern and western extensions of the producing Micliipicoten iron-range district, adjacent to Micliipi- coten Bay. The northern range lies between Magpie River and the western branch of Pucaswa River, practically continuous with the old Michipicoten range. The western range, separated from the other by granite, lies between Otter Head and Bear River, on the Lake Superior shore, and extends but a short distance north of Lake Michi-Biju. The lithology and succession are essentially the same as in the Micliipicoten district. The Helen formation consists of sideritic and pyritous cherts, jaspers, amplubolitic scliists, siderite, iron ores, cjuartzite phyllites, and biotitic and epidotic sclusts. For the most part the iron-bearing bands are lean ore. Explora- tion has been carried on somewhat extensively at Iron Lake, Frances mine, and Brotherton Hill, a Iron ranges of Micliipicoten West: Rept. Ontario Bur. Mines, vol. 14, 1905, pt. 1, pp. 278-355. 156 GEOLOGY OF THE LAKE SLTERIOR REGION. at the Leach Lake bands in the northern range, and in Laird's claims, the Julia River bands, tlie David Katossin claims, and the Lost Lake claims in the western range, but no important ore deposits have yet been found. THE IRON ORES OP THE MICHIPICOTEN DISTRICT. By the authors and W. J. Mead. GENERAL STATEMENT. The Micliipicoten district has one producing mine, the Helen. The Helen ore bodv Hes in a great anipliitlieater opening westward on Boyer Lake, the east wall of wliicli is formed by iron carbonate, the north by ferruginous cherts, and the south by Wawa tuff. The tuffs and ferruginous cherts stand vertical. Boyer Jjake has been drained, and the basin, a (juarter of a mile long and 130 feet tleep, is apparently cut out of solid rock. A dike of diabase crosses the basin from north to south near its east end, as shown by mining operations, and its outcrop on the edge of the basin can now be seen. Most of the ore mined is east of the dike, but ore^ is known west of it. Alining operations are 300 feet below- the original level of Boyer Lake. The ore body seems to dip eastward as if passing under the siderite hill. A drift under this hill has developed several hundred thousand tons of iron pyrites. Along the south margin of the ore body ocher or paint rock marks the limit against the green schists. To the north the ore runs gradually into lean material, with too much white silica to be worth mining. CHEMICAL COMPOSITION. Following is the average analysis of all ore sliipped from the Micliipicoten district in 1907: Average composition of all ore shipped from the Michipicoten district in 1907. Moisture (loss on drying at 212° F.) 5. 70 Analysis of dried ore : Iron 58. 20 Phosphorus 127 Silica 4. 40 Manganese 165 Alumina 88 Lime 23 Magnesia 14 Sulphur 127 Loss by ignition 10. 40 Chemically the ore most closely resembles some of the more hydrous Mesabi ores. It is low in alumina and high in combined water, which makes almost all of the loss on ignition. MINERAL COMPOSITION. Mineralogically the ores are made up of hydrated iron oxide and silica, with small amounts of clay and other minor constituents. The following approximate mineral composition Vvas calculated from the above average analyses: Mineral composition of Michipicoten ores, calculated from above analyses. Hematite 23. 60 Limonite 69. 60 Quartz ' 3. 36 Kaolin 2. 23 Iron sulphide 24 Apatite 41 Miscellaneovis .56 100.00 PRE-ANIMIKIE IRON DISTRICTS OF ONTARIO. 157 The hydrated ferric oxide is calculated as hematite and limonite in order to afTord com- parisons with other ores similarly calculated. It is known that liydrated iron oxides other than limonite are present, and it is probable that practically all of the ore is more or less hydrated; hence the amount of hematite present is probably less than is indicated above. No phosphorus minerals have been identified, but the presence of calcium sugg:ests calcium phos- phate (apatite). Calculation shows, however, that sufficient calcium is not present to account for all the phorphorus as apatite. Iron sulphide is locally abundant in the ores, occurrmg in pockets of "pyritic sand." PHYSICAL CHARACTERISTICS. Color and texture. — In color the ore ranges from liglit-yellow ocber, suitable for paint, tlu-ough a variety of shades of red and brown to dark brown or nearly black. In texture the ore varies from soft earthy material to rough, slaglike limonitic ore, and locally hard blue hematite is found. Density. — The average mineral density of the ore, calculated from the average mineral composition given above, is approximately 3.85. Porosity. — The porosity of the ore varies to an extreme degree, ranging from a minimum of less than 5 per cent m the dense ore to a maximum of over 50 per cent (locally 60 per cent) in the limonite. The average is rather difficult to estimate, but is probably between .30 and 40 per cent. Cubic feet per ton. — Owing to the extent to wliich it varies in density, jiorosity, and moisture, the cubic content of the ore ranges within wide limits. The average is approximately 1.3.5 cubic feet a ton. SECONDARY CONCENTRATION OF THE MICHIPICOTEN ORES. The Helen mine has impervious walls, but the direction and nature of the concentrating waters are not yet clear. Tlie u'on-bearing formation was originally cherty iron carbonate. The hill east of the ore body exhibits one of tlie largest masses of unaltered carbonate known in the Lake Superior region. The alteration of the iron carbonate can be seen in all its stages, first into ferruginous chert and then into ore, and locally directly into ore. The bottom of the lake basin is partly covered with large masses of yellow ocher dissolved from the carbonate and redeposited. The iron carbonate is thoroughly impregnated with sul])hide minutely disseminated tlirough the carbonate and m veins in it. During the alteration of the iron carbonate to the iron ore this sulphide has remained relatively intact, for it is included with the oxides of iron in large masses. In the deeper levels of the Helen mine the iron sulphide is in such large masses as to constitute a great obstacle to mining. Nevertheless, some of the sulphide has been altered and is rep- resented in the limonite forming the lake bottom. The waters of the lake are liighly charged with sulphuric acid, which has a strong deleterious effect on the pipes. Associated with the limonite in the lake bottom is a peculiar green mud, the composition of which is as follows: Analysis of dark-green mud from lake bottom. SiOj 47. 58 Fe 11. 23 Mn 14 CaO 95 CO, 3.19 158 GEOLOGY OF THE LAKE SUPERIOR REGION. Embedded in tliis imid was found a glacial bowlder consisting largely of serpentine, showing peripheral alteration to a depth of several inches. Analyses made ]>y R. D. Hall of the center and outer ]iortious are as follows: Analyses oj altered bowlder jrom bottom o/ Boyer Lake. SiO;... AljOs.. Fe.Oj. FeO... MgO.. CaO... NasO.. KjO... HaO-. H2O+. TiOs. . SO3... COs... Center of bowlder. 39.36 3.48 0.S4 6.82 31.04 3.22 .n .90 .20 7.44 .13 .18 Trace. Altered portion. 37.80 3.76 9.13 7.76 28.02 3.50 .06 .10 .62 8. 58 .38 .40 .10 Altered portion. assuming FejOa constant. 28.30 2.82 (1.84 5.81 21.00 2.62 .04 .07 .46 6.41 .28 .30 .07 The inner portion of the bowlder was a dense dark-green rock and the altered portion a lighter-green earthy material. The alteration appe.u-s to have been brought about essentially by solution. Oxidation has been practically nil, as has also earbonation. The ferric iron occurs essentially as magnetite. If this mineral is assumed to have been unaltered, it follows from a comparison of the first and third columns in the above table that there has been a loss of all constituents except ferric iron, SO3 and CO,. The nature of the alteration differs essen- tially from typical \yeathering. The abundant evidence of decomposition of the substances of the lake bottom and the presence of sulphuric acid in the lake waters have suggested to Coleman and Willmott" that Boyer Lake represents a solution basin. The bottom of the lake is considerably below its outlet. Though ilecomposition has undoubtedly aided in the erosion of the lake bottom, there is also evidence, summarized by Martin (see pp. 430-431), that the lake basin is a glacial cirque developed largely by mechanical means. 1 The Michipicoten iron ranges: Univ. Toronto Studies, geol. ser., No. 2, 1902, p. 23. CHAPTER VII. THE MESABI IRON DISTRICT OF MINNESOTA.^ GENERAL DESCRIPTION. The Mesabi iron district lies in the jiart of Minnesota northwest of Lake Superior. In shape and trend it is simiLar to the other iron districts of the Lake Superior re(:i;ion. (See Ph VIII, in pocket.) It extends from a point west of Pokegama Lake, in T. 142 N., R. 25 W., east-north- east to Birch Lake, a distance of approximately 110 miles, with a width varj-ing from 2 to 10 miles. Its area is about 400 square miles. To the east from Birch Lake to Gimflint Lake and beyond are small patches of u-on-bearing rocks, constituting remnants of an eastward extension of the ^lesabi district. The main topographic feature of the district is a ridge or "range" parallel to the longer direction of the district, known as the Giants or Mesabi Range.'' Mesabi (spelled also Mesaba and Missabe) is the Chippewa Indian name for "giant." In the west end of the district the Giants Range merges insensibly into the level of the surrounding country, about 1,400 feet above sea level, or 800 feet above Lake Superioi-. Toward the east the elevation with reference both to Lake Superior and to the surrounding country increases; from range 18 to range 12 elevations of 1,800 and 1,900 feet above sea level, or 400 and 500 feet above the level of the surrounding country, are reached. For many miles both north and south of the range there is a comparatively low, flat area, and the Giants Range, particularly its eastern portion, is a conspicuous feature in the landscape. While the general trend of the range is east-northeast, there are many gentle bends in the crest line, and m range 17 a spur known locally as the "Horn" projects in a southwesterly direction for 6 miles. The crest of the range is in places broad and flat, in others comparatively narrow and sharp. The southern slope is very gentle; the northern slope is somewhat less so. At short intervals both crest and slopes are notched by dramage channels. The Giants Range for the most part forms a drainage divide, although it is crossed by drainage channels at several places. The dramage of the district is apportioned among three of the great river systems of the country — tJie Mississippi, St. Lawrence, and Nelson. The succession of formations in the Mesabi district appears in the following statement: Quaternary system: Pleistocene series Deposits of late Wisconsin age. Unconformity. Cretaceous system. Unconformity. Algonkian system: Keweenawan series Great basal gabbro (Duluth gabbro) and granite (Embarrass granite), intrusive in all lower formations. Unconformity. Huronian series: fAcidic and basic intrusive rocks. Upper Huronian (Animi-J Virginia slate. kie group) 1 Biwabik formation (iron-bearing). [Pokegama quartzite. Unconformity. (Giants Range granite, intrusive in lower formations. Lower-middle Huronian. .I^'''''''"'^'''^^^^^'^'^'^"™""'''™®'''^''® formation (equivalent to the Ogishke conglomerate and Rnife Lake slate of the Vermilion district). a For turther detailed description of the geology of this district see Men. V. S. Geol. Survey, vol. 43, and references there given. Mining men and others have cooperated cordially in the preparation of this chapter, hut we would acknowledge jjarticularly our indebtedness to Mr. J. U. Sebenius, who, having been in charge of explorations in the Mesabi district since its discovery and being now chief engineer of the United States Steel Corporation, has perhaps closer knowled^'e of the geology of the iron-tearing rocks here than any other person. i>For the use of the terms "Giants Range" and "Mesabi range" in this report, see footnote on p. 41, also Mon. U. S. Geol. Survey, vol, 43 1903, p. 21. 159 160 GEOLOGY OF THE LAKE SUPERIOR REGION. Unconformity. Archean system: Liiurontian series Granites and porphyries. Keewatin scries Greenstones, green schists, and porphyries. The core of the Giants Ranee is made up principally of j^ranitc of lower-middle Iluronian and Keweenawan a<;e and subordinately of Archean igneous rocks. To the south of the igneous core, for a part of the district, arc lower-middle Huronian sedimentar}'^ rocks, with bedding approximately vertical. Against the southern boundary -f the lower-middle Iluronian, or, where the lower-middle Huronian is lackmg, against the igneous core, he the upper Iluronian sedimentary rocks (Animikie group). They dip gently to the south and underlie the greater portion of the southerly slopes of the range. On the southeast the Huronian rocks are limited by the Keweenawan Duluth gabbro, the north edge of which cuts across the Huronian forma- tions diagonally from southwest to northeast. The Archean, lower-middle Huronian, and upper Huronian are separated from one another by unconformities. Glacial drift cover.s the district so thickly that rock exposures are rare on the lower slopes of the range and only fairly numerous near the crest. ARCHEAN SYSTEM OR "BASEMENT COMPLEX." DISTRIBUTION. The Archean rocks of the Mesabi district are confined to its central portion. They are found north and northwest of Nashwauk; northwest of Hibbing; north and northeast of Motm- tain Iron; in the southerly projection of the Giants Range known as the "Horn," bounded by the cities of Virginia, Eveleth, Sparta, and Mcliinley; north of Biwabik; and eastward nearly to the east Hue of R. 16 W. With the exception of the portion of the Archean area east of Embarrass Lake, exposures are sufficiently common to allow a fairly close determination of the boundaries. East of Embarrass Lake the mapping is based on the presence of abundant Archean fragments in the drift. Included in the areas mapped as Archean north of Mountain Iron are several small patches of lower-middle Huronian rocks. Exposures are so few, they are so mixed in t* e same exposure .with Archean rocks, and they are metamorphosed to such difficultly recognizable forms that their accurate delimitation on the general map is not possible. KINDS OF BOCKS. The Archean is represented, about in order of abundance, by micaceous, chloiitic, and hornblendic scliists, basalts, dolerites, porphyritic rhyolites, granites, and diorites. The basic rocks have commonly a green color and are usually referred to locally as greenstones or green schists. They are given one color on the general map of the ilesabi district and are to be -correlated with the Keewatin series (PI. VIII). The acidic igneous rocks, consisting of the porph}-iitic rhyoHtes and the granites, are mapped under another color and are correlated with the Laurentian. All these rocks have their counterparts in other iron districts of the Lake Superior region. In the .Vermihon and Crystal Falls districts, where especially well developed, Clements has descril>ed each i)hase in great detail. For details of petrography the reader is referred to the description of the Archean rocks in the monographs on the Crystal Falls and the Vermilion districts." Nowhere iu the district have sediments been found which are demonstrably of Archean age, but slate fragments in the basal conglomerate of the lower-midcUe Huronian point to the former existence of Archean sediments. a Mon. U. S. Geol. Survey, vols. 36 and 45. MESABI IRON DISTRICT. 161 STRUCTURE. Most of the Archean rocks show some cleavage, and perhaps about half have enough cleavage to warrant calling them scliists. In general the plane of cleavage is nearly vertical and strikes parallel to the range, about N. 60° E. The hornblendic schists north of Mountain Iron have a cleavage of a linear parallel type, and the lines of the cleavage dip steeply to t-he northeast. In addition to cleavage, there are many joints and faults ^\^th displacements of a few inches or feet, but no regular systems have been determined. ALGONKIAN SYSTEM. HURONIAN SERIES. LOWER-MIDDLE HURONIAN. DISTBIBUTION. Sedimentary rocks of lower-middle Huronian age appear in two considerable areas in the Mesabi district. One vnth an average width of perhaps a mile extends, from Eveleth northeast to Biwabik ; the other, somewhat less than a mile in width, extends from near the Duluth and Iron Range Railroad northeast to near the center of sec. 11, T. 59 N., R. 14 W. In the former belt there are areas of green schist forming the cores of the hills. One of them has been mapped, but others, though their presence is known by isolated exposures, are not sufRcientl}^ exposed to warrant their separation on the map. A number of small patches of lower-middle Huronian sediments are known also in other parts of the district. Granite of lower-middle Huronian age forms most of the core of the Giants Range and, except north of Mountain Iron, where it is interrupted for a short distance by Archean horn- blendic schists, is exposed continuously along the crest to where it is succeeded on the east by the younger Embarrass granite in R. 14 W. This lower-middle Huronian granite, known as the Giants Range granite, thus bounds on the north the other formations for most of the district. Detailed work has not gone farther north than the granite boundary. GRATWACKES AND SLATES. • The interbedded graywackes and slates form the greater part of the lower-middle Huronian sediments. They are dull dark-gray and dark-green rocks which usually weather to a somewhat lighter green or gray or to a dirty hght yellow. The grain is usually fine, although it varies considerably. The bedding, shown by both color and texture, is conspicuous. Parallel to the bedding a secondary cleavage has been developed. As a result of variation in texture, bedding, and secondary cleavage, there appear all gradations between metamorphosed coarse graywackes, banded graywackes, and finely fissile slates. Along the parting plane of some of the graywackes and slates may be seen glistening plates of mica or chlorite, conspicuous because of the fact that they appear in separate spangles on the dark background rather than in con- tinuous layers, although, indeed, some of the more fissile slates show mica and chlorite in the continuous layers characteristic of slates. The graywackes and slates above described have resulted from the alteration of fine mud and feldspathic sand deposits. Some of the mica, especially that in separate clear-cut plates, may have been originall}' deposited in its present position, but most of it, and especiallj^ that in continuous sheets on the parting surfaces, is undoubtedly a secondary development due to dynamic movement in the rock. The intrusion of granite below described has further greatly metamorphosed the graywackes and slates. In approacliing the granite they become more chloritic, hornblendic, and micaceous, and a marked and usually much contorted schistosity obhterates the bedding. Under the microscope may be seen abundant development of secondary chlorite and hornblende and a 47517°— VOL 52—11 11 162 GEOLOGY OF THE LAKE SLTPERIOR REGION. lesser development of secondary biotite and muscovite. Accessories inclmlo tourmaline, stauro- lite, Of which 23.96 is soluble. 1. Specimen 45758. From 250 paces west. 83 paces north, of the west quarter post. sec. 35, T. 59 N., R. 15 W. The finely ground rock was evaporated on the water bath to dryness with 50 cc. of 1-1 UCl. taken up with water slightly acidified with HCl, and filtered. Soluble sdica was then determined in this residue by ijoiling with 5 per cent solution of NasCOa. A determination of soluble SiOs was then made in the rock before treatment with HCl and subtracted from the first soluble SiOs found, which gave the figure for SiO^ in the soluble portion. 2. Specimen 45705. From test pit in Cincinnati mine. The soluble portion was found by evaporating to dryness on the water bath with 50 cc. of 1-1 HCl, and taking up with water slightly acidified with HCl. The residue was then boiled fifteen minutes with a 5 per cent solution of NaaCOa to dissolve any soluble silica, this silica determined and placet! with the soluble portion. The residue was ignited and finally heated for fifteen minutes over the blast lamp, weighed, and then a rough analysis made, which is found in the second column. The small amount of iron shown in the insoluble portion could easily have been carried down mechanically. A determination of soluble silica was then made in the rock before treatment with HCl and found to be 3.3 per cent. Subtracting this from the total soluble silica, 10 per cent of soluble silica remains for the part dissolved in HCl. 3. Specimen 45766. From test pit in Cincinnati mine. The finely ground rock was evaporated on the water bath to dryness with 50 cc. of 1-1 HCl. taken up with water slightly acidified with HCl, and filtered. Soluble sihca was then detennined in this residue by boiling with 5 per cent solution of Na^COa. A determination of soluble Si02 was then made in the rock before treatment with HCl and subtracted from the first soluble SiOo found, which gave the figure for SiOa in the soluble portion. 4. Specimen 45180. From 500 paces west, 100 paces north of the southeast comer of sec. 22, T. 59 N., R. 15 W. Owing to presence of organic matter, tlie determination of ferrous iron is probably high. From the detailed consideration of these results, which is not I'epeated here, it appears that the ferric iron occurs in the rock mainly as sesquioxide, for the soluble silica is accounted for by the ferrous iron and magnesia present, leaving none for the ferric iron; that in tliree slides of the four of the rocks analyzed the ferric oxide may be observed to be present and to be probably secondary, and hence that the iron oxide shown by the analyses is mamly sec- ondary and not to be considered as belonging with the substance of the unaltered granules. It appears further that the alumina and lime are in such small quantity as to be practically negligible. It appears still further that there is far more than enough combined water to com- bine with the ferric iron to form ferric hydrate, and thus that a considerable portion of combined water shown by the analyses may be taken to belong to the green granules. Finally, it appears that the substances which can not be accounted for in any other way and which clearly belong with the green granules are silica, ferrous iron, magnesium oxide in small proportions, and water. It is therefore concluded that the substance of the green granules is essentially a hydrous ferrous silicate with a subordinate amount of magnesium, and that if ferric iron is present at all as an original constituent of the green granules it is in small cjuantity. This conclusion is essentially in accord with that reached by J. E. Spurr in his report on the Mesabi district published in 1894." Having concluded the substance of the green granules to be mainly silica, ferrous iron, magnesium oxide, and water, we may ascertain whether or not there is any uniformity in the proportions of these elements. The ratios of the silica, ferrous iron, and magnesium in the four analyses, calculated on the basis of 100, appear in the table on page 168. The percentage of water is not included for the obvious reason that, while it is certain that much of it belongs with the granules, no quantitative estimate can be made of its amount because of the uncer- tainty as to the portion which belongs with the ferric hydrate. a Bull. Geol. Nat. Hist. Survey Minnesota No. 10. 168 GEOLOGY OF THE LAKE SUPERIOR REGION. 1. 2, 3. i. Average. SiOi : 55.1 42.1 2.8 43.7 47.5 8.8 47.7 44.6 7.8 40.2 50.9 8.9 46.8 FeO 46.3 MgO . . 7.1 The relative proportion of the ferrous iron and silica above shown suggests a combina- tion of the two on the basis of one molecule of each. Theoretically the percentages of the two in such a combination would be — Silica '. 45. 62 Ferrous iron 54. 38 The average of the ferrous iron, 46.3, is about 8 per cent less than tne theoretical percent- age. The magnesium oxide, which has a higher combining power than the iron, more than makes up for this deficiency. On a subsequent page is given an analysis of a rock in which the green granules have been altered to a dark-green and brown amphibole, probably griinerite, apparently through simple recrystallization and dehydration. The alteration has occurred under deep-seated conditions, and it is probable that little if any addition or subtraction of material has taken place other than that involved in dehydration. The composition of the amphibole ought to give a clue to the composition of the original green substance. It is there found that the principal constit- uents of the amphibole are silica and ferrous iron in the following proportions: SiOo. Fed- 47.5 52.5 The correspondence of these percentages with those above given is evident. The above results are not sufficiently accordant to show that the substance under dis- cussion has a definite and uniform composition. On the other hand, the impurities and altera- tions cause such variations that it can not be said that the green granules do not have definite chemical composition. If the granules do have a definite composition, the above results indi- cate the most probable formula to be Fe(Mg)O.Si02.nH20. Dr. Spurr, after his study of the green granules, concluded to call them "glauconite." In view of the fiict that potash is by most mineralogists insisted upon as one of the essential con- stituents of glauconite, the entire absence of potash in the substance under discussion is taken to preclude the application of the term "glauconite." The substance apparently corresponds to no known mineral species. As a convenient term by which to refer to it the name "greenalite" was coined for use in the monograph on the Mesabi district and is used in this report also. The origin of greenalite and the details of the similarities and differences between greenalite granules and granules of glauconite, concretions of iron oxide and chert, and other granule and concretionary structures are discussed in Chapter XVII, on the origin of the iron ores. FERRUGINOUS, AMPHIBOLITIC, SIDERITIC, AND CALCAREOUS CHERTS. The following description applies to the normal types of chert occurring through the central and western portions of the range. The highly metamorphosed chert characteristic of the east end of the range is given a separate description on a subsequent page. The cherts are gray, yellow, red, brown, or green rocks, mth irregular bands and shots and granules of iron oxide varying in quantity from predominance almost to disappearance. A slight brecciation thoroughly recemented may be occasionally observed, and a pitted surface, due to the solution of certain of the constituents, is not uncommon. The iron oxide is mainly intermediate between hematite and limonite, and to a subordinate extent is magnetite, and its color accordingly ranges from red to yellow or to black. The variety of colors of the chert and the iron oxide, their irregular association, and their variation in relative abundance give the cherts most highly varied aspects; yet no phase of the cherts is likely to be mistaken for any MESABI IRON DISTRICT. 169 other rock by anyone reasonably familiar with the iron-bearing rocks of the Lake Superior region. To the casual observer the massive lighter-colored cherts, containing little iron oxide, resemble quartzite, and indeed have been frequently so called. However, the splintery frac- ture of the chert and the absolute lack of rounded clastic grains, aside from the usual content of iron oxide in layers or spots or minute grains, are unfailing criteria for the discrimination of the two. The ferruginous cherts difi'er from the jaspers or jaspilites of the old ranges of Lake Superior ia lacking their even banding and brilliant red color as well as the microscopic features described below. When studied under the microscope it appears that all the rocks hero described as chert are genetically connected. In lookmg over 250 slides but few have been observed which do not show some evidence of the derivuuon of the rock from the greenalite rocks above described. The granule shapes are stUI largely preserved," but the alterations have tended in some places to make the shapes more irregular and partly or wholly to obliterate them. The alteration of the granules has been almost entirely metasomatic, for thero is little evidence of dynamic move- ment resulting in the breaking up of the constituents of the rock. The greenalite has been replaced by cherty quartz, magnetite, hematite, limonite, siderite, calcite, grunerite, cummingtonite, actinolite, epidote-zoisite, or any combination of them. The extent and nature of the alteration replacement vary withm wide limits. The granule may be mainly greenalite, showing incipient crystallization of quartz, griinerite, or actinolite, visible only under crossed nicols. The granules may be represented almost wholly by hematite, limonite, magnetite, intermediate varieties, or any combination of them. The oxides may be arranged irregularly or concentrically. In the iron ores the granules are entirely represented by iron oxide, although their shapes are in part obliterated. The granules may be represented almost wholly by chert, which may be distinguished from that of the matrix by its coarser or finer texture, or, if not by texture, by distribution of pigment. In ordinary light chert granules may be marked by the pigments which in parallel polarized light are completely obscured by the crystallization of the chert, or the granules may not be seen in ordinary light and be conspic- uous under crossed nicols because of the crystallization. Or the crystallization of the chert may have entirely obliterated the granules for much of the slide, both m ordinary and polarized light. The granules may be represented entirely by green, yellow, and brown grunerite, cum- mingtonite, or perhaps actinolite, or aU, which in ordinary light may be scarcely distinguishable from the unaltered greenalite granules but which become apparent under crossed nicols by their double refraction. The granules may be represented by calcite or siderite in rhombs or irregular grains, sometimes showing zonal growth, which for the most part are clearly replace- ments of the granules. Most commonly the granules are represented by a combination of any or all of the minerals above named. Of these combinations, that of chert and iron oxide stands first. The two substances occur in all proportions with a great variety of arrangement. The two may be irregularly intermingled, or the iron oxide may form a rim about a cherty interior, or, though not commonly, the chert and iron oxide may be in concentric layers in the manner of normal concretions, or polygonal areas of fine chert may contain spots of iron oxide in the center of each as well as a rim of iron about the periphery, suggesting an organic structure. The alteration and replacement of the greenalite and the conditions favoring the development of the different minerals are discussed under the origin of the ores (pp. 187 et seq.). In addition to the derivatives of the greenalite granules, there are present a few concentric concretions of iron oxide and chert about quartz, which may have been secondarily developed from some substance other than the greenalite. These are similar to concretions in the iron- bearing formation of the Penokee-Gogebic district, where they have developed from the alteration of an iron carbonate. The secondary concretions in the Mesabi district may also be develop- ments from iron carbonates, which are now associated with unaltered portions of the formation and probably existed formerly in the portions which are at present altered. The secondary concretions are different from the greenalite granules in their beautifully developed concentric 1 Spuir (Bull. Geol. and Nat. Hist. Survey Minnesota No. 10) has applied to this texture the term "spotted granular." t> 170 GEOLOGY OF THE LAKE SUPERIOR REGION. structure. Though a few of the granules themselves have a concentric structure resulting from zonal alteration, this is usually poorly developed and there is ordinarily little difficulty in distin- guishing it from that of the secondary concretion, though in some places it is possible that some of the supposed secondary concretions formed from carbonate may be really secondary alterations of original granules. Spherulites of epidote, rarely to be observed, though in part replacements of the granules, are also clearly secondary developments in the matrix. The matrix of the chert may be a sparse interstitial filling between the granules or it may form most of the rock mass and contain but few isolated granules. The matri.x is similar to that of the unaltered greenahte rocks in that it is mainly chert, but it differs in containing far more actinolite, gri'inerite, cummingtonite, iron oxide, calcite, and siderite, and rarely epidote- zoisite in spherulitic form. Sometimes also green chloritic substances are abundant, either irregularly distributed tlirough the matrix or forming a definite rim about the granule. In the latter case the chlorite is in part in the fibrous form known as delessite and much resembles uralite. The recrystallization of the rock has in some places made the chert in the matrix coarser than that of the granules and in other places the reverse. The leaching out of the car- bonates and greenahte from the matrix has occasionally left cavities which give the pitted char- acter to the weathered surface of the cherts. Accompanying the recrystaUization of the chert has been its frequent adoption of radial or sheaf-hke forms, giving black crosses under crossed nicols. These sheaves, as well as the sheaves of actinolite, griinerite, and cummingtonite, and rarely epidote, frequently lie with their butts against the outhnes of the granules and send their points outward until they interlock with similar projections from adjacent granules. Commonly also one or more of the constitu- ents of the matrix may be observed to lie partly in the matrix and partly in the granule, thus helping to obhterate the granule. Indeed, under crossed nicols the granules may not be observed, while in ordinary light their position may be indicated by the distribution of the fine pigment. All of the constituents in the matrix are secondary except, perhaps, a part of the chert, and even this has been thoroughly recrystaUized. The amphiboles and iron oxide may be observed to have developed by the alteration of the granules and some of the lighter amphiboles by the alteration of carbonate and chert in the matrix. The carbonate is largely though not entirely replacement from without, for it may be observed replacing nearly all the other con- stituents of the rock and occurring in minute veins crossing the rock. The composition and origin of the ferruginous cherts are discussed on pages 186-187. SILICEOUS, FERRUGINOUS, AND AMPHIBOLITIC SLATES. Under this head are grouped a variety of slaty rocks which are interstratified with the other phases of the iron-bearing formation. They include dense black, dark-gray, green, or reddish rocks with a tendency toward conchoidal fracture and the slaty parting poorly devel- oped, if at all; rocks showing banding of dark-green, black, gi-ay, red, or brown layers parallel to the bedding and a well-developed cleavage parallel to the same structure; gradational varieties between these two, between them and the ferruginous cherts, and between them and the iron ores. Any of them may be hard or soft, carbonaceous or noncarbonaceous, fine grained or medium grained. Under the microscope the slates are seen to contain principally cherty quartz, iron oxide, either hematite or magnetite, usually in octahedra, or some hydrated oxide, monoclinic amphi- bole which may be griinerite, cummingtonite, or actinolite, ami possibly even common horn- blende, a small amount of carbonate of calcium or iron, a little zoisite, and possibly also a httle chlorite. From the optical properties and from the analysis of the rock it is thought that the ampliibole is mainly griinerite and cummingtonite. There is much variation in the relative proportion of the principal constituents. Some of the slates consist almost entirely of fine cherty quartz, with subordinate quantities of dark amphibolo in radial aggregates or in irregular masses, and of the iron oxides. Others are composed mainly of in)n oxide, showing but small quantities of the quartz and dark amphibole. Others are composed of a tangled mass of yel- MESABI IRON DISTRICT. 171 lowish, brownish, and greenish amphibole fibers containing minute particles of iron oxide, siUca, and other subordinate constituents. The griinerite is far more abundant than the actinoUte. The banding shown in many specimens is due to the segregation of the above-named elements into layers. ^AHiile it may be convenient in description to refer to tliis or that slaty rock as a ferruginous slate, a siliceous slate, an amphibolitic slate, or an actinolite slate, depending upon the relative abundance of the constituents, usually all tliree constituents are present in one rock, and the rocks are really amphibolitic, siliceous, and ferruginous slates. Perhaps the most char- acteristic feature of the slates as a group is the abundance of the dark amphibole. PAINT ROCK. Tlu-oughout the iron-bearing formation, and particularly adjacent to the ore deposits, are thin seams of paint rock, wliich have resulted from the alteration of the slates above described. The paint rocks are essentially soft red or yellow or white clay. They retain the original bedding of the rocks from wliich they were derived, the structure being marked by alternation of bands of dift'erent color. In place the paint rocks are moist and soft. When taken out and dried they become harder but retain a soft,- greasy feel. The alteration of the paint rocks from slates is proved by the numerous intermediate phases to be observed. For analyses of paint rock see page 191. SIDEEITIC AND CALCAREOUS ROCKS. Associated with the slaty layers in the iron-bearing formation, and particularly with the greenalite rocks, are carbonates of iron and calcium in small quantity. Most of the carbonate reacts readily with cold dilute hydrochloric acid and is certainly limestone, which, from the analysis of rocks containing it, is doubtless magnesian. Some of the carbonate, however, is certainly siderite, as shown by analysis. The carbonates occur in minute clear-cut layers interbedded M-ith the other rocks of the iron-bearing formation, in veins cutting across the bedding, and in irregular aggregates and well-defined rhombohedral crystals in the layers of the iron formation. In the carbonate bands are small quantities of iron oxide, ferrous silicate, and chert, and in the bands of these minerals are small quantities of the carbonate. In some places the carbonates are coarsely crystalhne and fresh and clearly have resulted from the replacement of the other constituents in the rock, particularly the ferrous silicate, or fi-om infiltration along cracks and crevices. In other places, especially where in distinct layers interbedded with unaltered ferrous sihcate phases of the formation, the carbonate layers seem certainly to be original. At the top of the iron-bearing formation and closely associated with the basal horizons of the Virginia slate are several feet of clear calcium carbonate, which is described in connection with the Virginia slate. CONGLOMERATES AND QUARTZITES. At the base of the iron-bearing formation is a thin layer of fairly coarse fragmental material consisting in places of conglomerate alone and in other places of conglomerate and quartzite. THICKNESS. The average thickness of the iron-bearing Biwabik formation is about 800 feet. This figure is based on average dips of the formation, width of outcrop, and drUl records. Local averages are likely to be either larger or smaller. In both the east and west ends of the district the thickness diminishes somewhat, the iron-bearing formation apparently giving way along the strike to slate. ALTERATION BY THE INTRUSION OF KEWEENAWAN (iRANITE AND GABBRO. Through ranges 12 and 13, near Birch Lake, the Biwabik formation is intruded on the north by granite and on the south by the Duluth gabl^ro and has undergone considerable meta- morphism in consequence. This metamorphism has extended even farther west, for, though the gabbro does not come into actual contact with the iron-bearing formation through range 172 GEOLOGY OF THE LAKE SUPERIOR REGION. 14, it abuts against the overlying Virginia slate and has metamoq^hosed both the slate and the iron-bearing formation in this area." In general through the western and central portions of the Mesabi district the iron oxide of the iron-bearing formation is mainly hydrated hematite, and magnetite is in subordinate quantity. Eastward from Mesaba station the iron oxide is mainly magnetite, and hematite is in subordinate quantity. Westward from Mountain Iron the amijliiboles are almost entirely lacking; from Mountain Iron eastward to Mesaba station the amphiboles are present in the iron-bearing formation but are not a])un(lant until Mesaba station is approached ; eastward from Mesaba station they become abundant and make up an important constituent of the formation. In the eastern portion of the range the chert is correspondingly less abundant than in the west- ern and central portions of the district, and in some places is almost entirely absent. The chert becomes also distinctly coarser in this area. In range 12 the grains commonly reach a diameter of 3 or 4 miUmieters, and there are a few smaller particles, and in the central and western por- tions of the district they are seldom greater than 0.1 millimeter and almost invariably are asso- ciated with smaller particles. Toward the east there is a tendency for the texture to become more even, although there are many wide variations from uniformity. The chert grains, instead of being in irregular, roundish, and scalloped cherty forms, as in the central and western por- tions of the district, are in rouglily polygonal shapes and united in a fairly uniform mosaic. Accompanying these changes is a more pronounced segregation of the magnetite and the ampliib- olitic chert into irregular laj-ers and lenses, with the result that the iron-oxide layers, instead of contairdng various other minerals, are comparatively free from them. The characteristic granules of the ferruginous cherts are still conspicuous in the east end of the district, but in the most highly metamorphosed phases of the rocks, as in range 12, they have entirely disappeared, being obscured by magnetite, amphibole, and chert. In the phases not showing the maximum alteration they are marked by magnetite, either as a rim about the granule, as a solid mass filling it, or in evenly disseminated particles through it. Not unconunionly the granules may be observed only in ordmary light and then by distribution of the magnetitic particles; in parallel polarized light they are obscured by the polarization of the amphibolitic and cherty constituents. Finally, in the eastern portion of the district certain minerals have developed which have not been found in the less altered rocks of the central and western portions of the Mesabi district. In the latter areas the amphiboles are entirely grunerite and actinohte, with little or no horn- blende. In the eastern portion of the district the amphiboles include grunerite and actinolite, and in addition green and brown hornblende in considerable quantity. Associated with these minerals are small quantities of biotite, glaucophane, andalusite, zoisite, and garnet. Though hypersthene, augite, and olivine are abundant and characteristic in the true gabbro of range 12 and westward, these minerals are nearly if not quite lacking in the Biwabik formation. Although to the east toward Gunflint Lake the gabbro alone has been able to produce even greater metamorphic effects on the iron-bearing rocks, it is probable that the metamorphism of the iron-bearing rocks in the region untler description has been produced jointly bj' Kewee- nawan gabbro and granite. VIRGINIA SLATE. DISTRIBUTION. The Virginia slate bounds the iron-bearing Biwabik formation on the south from the west end of the district nearly to the east side of sees. 5 and 8, T. 59 N., R. 13 W., where the slate is overlapped by the gabbro. Still farther east, in the SW. i sec. 25, T. 60 N., R. 13 W., drilling has shown altered slate to lie between Keweenawan Duluth gabbro on the south and Kewecnawan diabase on the north, but whether it is an isolated mass at this point in the Keweenawan area or is continuous with the slate to the west explorations or exposures do not yet tell. The slate underlies the lower slopes of the Giants Range and continues south under the low-lying swampy a The metamorphism of the Biwabik formation by the Duluth gabbro in the area adjacent to Birch Lalte and to the east in the ^•ieinit5• of .\lceley and Ciinllint lakes has been described in detail by U. S. Grant, W. S. Bayley, and Carl Zaplle and has been briefly considered or mentioned by N. U. Winchell. 11. V. Winchell. A. U. Elftmann. J. E. Spurr. J. Morgan Clements. C. R. Van Hise.and others. The reader is referred to Chap- ter Vin, on the Cunllint district (pp. 198-204), for a fuller account of the alterations near the gabbro. MESABI IRON DISTRICT. 173 area south of the Giants Range for an unknown distance. The area overlain by slate is so thickly covered with drift that exposures of the slate are almost entirely lacking; its presence and distribution have been determined by drilling and test pitting in the search for iron. Tlirough the central portion of the district enough of such work has been done to show the posi- tion of the slate boundary with a fair degree of accuracy, although even here there are con- siderable stretches where records of the occurrence of slate are wanting. In the western and eastern portions of the district the distribution of the slate is" less well known, particularly in the western end of the district. In drawing the slate line on the map of this portion of the area all that can be done is to connect the separated explorations which reveal slate. Wherever exploration has been detailed it is found that the slate boundary is not straight but in gentle curves, and it is reasonable to expect, therefore, that future work will show numerous additional undulations in the slate boundary for the area at present not completely explored. The normal Virginia slate is usually a gray rock, though in part black, reddish, or brown, with bedding shown by alternating bands of varying color and texture. Some of the beds are almost coarse enough to be called graywackes. Indeed, in the field the rock has been called a banded slate and graywacke. Some of the slate is hard and siliceous; other phases, especially the nonsiliceous and carbonaceous ones, are soft, and can be wliittled with a knife. Near the contact of tlie slate with the iron deposit in the underlying iron-bearing formation, as at Biwabik and in sec. 3, T. 58 N., R. 15 W., the slate becomes iron stained and soft and grades into paint rock. The slate in general has a very poor parting parallel to beddmg planes, and there is little or no development of secondary cleavage. Wliat there is of secondary cleavage has been developed parallel to the bedding planes and is marked by minute particles of mica there found. The rock in general aspect and mineralogical and chemical composition looks like slate, but it differs from true slate in lacking a true cleavage, and as this is one of the essential characteristics of slate it ma}' be doubted whether the term "slate" ought to be applied to the rock. Yet the rock is not a shale, for it is too much metamorphosed and lias too poor a partuag parallel to the bedding. In the Cuyuna district the same formation shows the charactci-istics of a true slate, and the formation both there and in the Mesabi district proper has been known locally and in geologic literature as slate. Hence the term is here retained. Analyses of Virginia slate. 1. 2. SiOs 62.26 10. S9 1.76 4.55 2.95 .42 2.29 3.02 .70 3.8S .60 None. .20 56 Gl AlsOs. 17.76 3.29 5.15 Fe-Oa FeO MgO CaO . • 1.00 NajO K2O 4 04 HjO- H2O + 4 18 TiOj COj PjOs Organic undetemiincd C and c 99.52 99.56 1. Analysis by Oeorge Steiger. of the United States Geological Survey, of a composite sample of the Virginia slate made up bv assembling several specimens from two localities (specimen 45767 from excavation for water tank of Eastern Railway of Minnesota, at Virginia; specimen 45463 from a point south of the Biwabik mine). 2. Analysis of Virginia slate by R. D. Hall, University of Wisconsin, of a sample representing 900 feet of drill core from a drill hole at the south- east comer of sec. S, T. 58, E. 15. CORDIERITE HORNSTONE RESULTING FROM THE ALTERATION OF THE VIRGINIA SLATE BY THE DULUTH GABBRO. In approaching the Duluth gabbro, which overlaps the Virgmia slate in ranges 14 and 13, the slate becomes more crystalline, harder, and characteristically breaks with a conchoidal fracture, and the color becomes darker and in many places is a bluish black. The rock, indeed, 174 GEOLOGY OF THE LAKE SLTPEKIOR REGION. becomes a hornstonc'. Moreover, there appear minute light-colored specks which on tlie weatlicred surface arc likely to have disappeared and to be represented by pits. Under the microscope the wlute specks are found to be cordierite in typical development, standing as numerous phenocrysts in a fine-grained matrix of biotite, feldspar, magnetite, and certain doubtful microlites wliich may be actinolite or sillimanite, or botli.° RELATIONS TO THE BIWABIK FORMATION. Reference has already been made to the fact that tlic relations of tlie Virginia slate to the underlying Biwabik formation are those of gradation, both lateral and vertical. It remains to discuss tliis gradation somewhat fully. The iron-bearing formation contains slate layers tlu-oughout. At upper and middle horizons they are perhaps more numerous than at lower horizons. Just below the solid black Virginia slate there is a zone in which there are many interlaminations of iron-bearing formation and slate, the layers varying in thickness from several feet to a fraction of an inch. Tliis zone is of varying and uncertain thickness. In many places at least the zone of minute interbanding is thin, not more than 15 or 20 feet, but, as already noted, layers of slate are found well down in the iron-bearing formation and layers of the iron- bearing formation are found well up in the slate, so that in a broad way the gradation zone m.'iy be several hundred feet. Drilling shows much irregularity in the alternation of layers. Slate layers are more abun- dant in the eastern end of the district, and westward from Grand Rapids the iron-bearing formation rapidly thins, its place being taken by slate in T. 142 N., R. 25 W. Wliether the iron-bearing formation extends indefinitely southward under the slate or gives place to slate in that direction is not known. All di-ill holes put down near the northern margin of the Vir- ginia slate in the Mesabi district have shown the Biwabik formation below. For reasons cited on pages 517-518, however, it is regarded as not impossible that farther south the iron-bearing formation thins and becomes discontinuous, its place being taken by the black slate. An examination of the map will show the Vii'ginia slate to encroach on the south margin of the iron-bearing formation to greatly varying distances, with the result that the surface outcrop of the iron formation ranges in width from 2 miles or more to less than a cjuarter of a mile. This is due in part to steeper dips at the narrow places than at the wide places in the iron- bearing formation, erosion having thus uncovered less of the iron formation where the dips were steep; it is due in part to faulting, as at Biwabik and eastward; it is due in part to the greater dip of the present plane of surface erosion, either atmospheric or glacial, in places where the formation is wide than where narrow, the greater dip of the surface bringing it more nearly parallel with the dip of the iron-bearing formation, and thus uncovering more of it; but so far as present evidence goes these factors are not adequate to account for the observed variations in width of the iron formation. The known irregular alternation of iron-bearing formation and slate both across and along the beds is therefore regarded as a cause of the varying widths of the iron-bearing formation. STRICTURE. Opportunities for studying the structure of the Virginia slate in place are so few that if the obsei-ver were dependent upon such obsei-vations alone he would be unable to make any statements concerning the structure of the formation beyond the fact that it dips at low angles away from the high land adjacent. THICKNESS. The thickness of the Virginia slate can not be determined in the Mesabi district. The drift covering is thick, mining exploration stops to the south where the slates are encountered, and the southerly extent of the slate belt is thus unknown. o Cordierite in this fonnation was first noted and described by N. II. Winchell, Final Kept. Geol. and Nat. Hist. Surrey Minnesota, vol. S, 1900. MESABI IRON DISTRICT. 175 STRUCTURE OF THE UPPER HURONIAN (ANIMIKIE GROUP). As a whole the upper Huronian (Animikie group) is a well-bedded series of sediments. The bfedding is most pronounced in the mitldle and upper horizons. The beds have gentle dips, averaging between 5° and 20°, though locally greater or less, in southerly and southeasterly tlirections away from the older rocks forming the core of the Giants Range, but locally the dips show much variation both in degree and direction. About the southerly projecting tongue of the Giants Range, in the vicinity of Virginia, Eveleth, antl McKinley, the dips are westerly on the west side of the tongue, southerly at the end of the tongue, and southeasterly on the south- east side — thatis, throughout approximately normal to its periphery. Even more conspicuous than the change of dip at such a place are the minor variations between exposures. Seldom is it possible to get two identical readings in dip at exposures of rock separated by even short intervals, although the direction and amount of the dip come within the above limits. These facts indicate that the upper Huronian beds are tilted away from the core of the Giants Range in directions normal to its trend and that the gently tilted beds are not plane surfaces but are gently flexed. By tabulation and comparison of the dips it becomes further apparent that the greater flexures are not random ones but generally have their axes normal to the trend of the range. On examination of the attitudes of the beds still more in detail it appears that the great flexures themselves are not simple but have many subordinate flexures, some of them transverse to the major ones. The complexity of the structure may be likened to that of water waves. On the great swells and troughs there are smaller waves, on the smaller waves there are stUl smaller ones, and so on down to the tiniest disturbance of the surface. Though perhaps the majority of the minor flexures in tlie upper Hui'onian rocks have attitudes similar to the larger ones, many of them vary greatly in direction. They may be observed at almost any single exposure of the upper Huronian. The great flexures are ver\'- gentle, involving very small changes in degree and direction of dip. Many of the minor flexures superimposed upon the greater ones are sharp and conspicuous. The local dips may vary as much as 50° witliin a few hundred feet and change their direction considerably. Dips as liigh as 45° or even 60° may be seen in the layers of the iron-bearing formation in some of the open pits of the mines, as at the Stevenson, the Sauntry-Alpena, the Kanawha, and the Sparta. At the Hawkins and Agnew mines the iron-bearing formation exliibits steep, sharp fokls. The iron-bearing formation shows more minor contoi'tions than the rest of the upper Huronian rocks, because of the great chemical changes which it has under- gone, but it is not probable that there is any great dift'erence in the major folding. The prevailing gentle southern tflt of the upper Huronian and the manner in which it laps around the salients in the older rocks suggest that the major features of upper Huronian "Structure may be due partly to initial dip as well as to subsequent folding — in other words, that the upper Huronian sediments are essentially in the position in which they were deposited against an old shore and have undergone minor deformation since. Accompanying the tilting and minor folding of the upper Huronian there has been a very considerable amount of fracturing, especially in the comparatively brittle Pokegama and Biwabik formations. Indeed, it seems likely that the folds of the two lower formations of the upper Huronian are mainly the result of lelatively small displacement along fractures, and only to a small degree the result of the actual bending of the strata without breaking. The pondmg of water beneath the Virginia slate would seem to indicate that this formation has been less fractured than the iron-bearing formation because of its less brittle character, and has thus yielded to deformation by actual bending rather than by bi'eaking. On almost every exposure of Pokegama and Biwal^ik formations joints and minute faults are to be obsei'ved cutting almost perpendicularly across the bedding. In each case the joints seem to make up two or more systems crossing each other at various angles, but such sets have little constancy of direction in widely separated exposures, unless we except a set of joints which at a number of places have an average direction of somewhere between N. 60° and 70° E. — that is, approximately parallel to the trend of the range. In the massive rocks the joints are clear cut and continuous for 176 GEOLOCn' OF THE LAKE SUPERIOR REGION. considerable distances. In the well-bedded rocks — as, for instance, in the thin-bedded portions of the iron-bearing formation — the joints are usually more irregular, less continuous, and less conspicuous. In such jjIuccs each individual bed may be more or less jointed witliout reference to the la^'ers above or below. The displacement or faulting along joints has been, in general, small. The displacement is rarel}' 3 or 4 feet, and commonly it is measured by a few uiclies. There is a displacement of about 200 feet along a nearly vertical fault strike running east- ward along the north side of the Biwabik mine parallel to the northern margin of the upper Iluronian past Embarrass Lake. The south side of the fault has droppetl, with the result that the layers of the u'on-bearing formation are somewhat tilted along the contact and the width of the outcrop lessened. The eastward extension of tliis fault carries it tlirough the peculiar point of Pokegama quartzite projecting eastward into the iron-bearing formation cast of Embar- rass Lake. Though the structure has not been worked out in detail east of Embarrass Lake, it seems not unlikely that the peculiar features of the distribution of the quartzite and iron- bearing formation there may be partly explained by faulting, though original configuration of the shore line in upper Huronian time may have something to do with it. Other great faults are almost certainly present in the district, but evidence for them has not been correlated. Certain of the joints and faults have been filled with vein quartz and others have not. It is rather siu'prising that so little vein quartz is to be observed. In the harder rocks, where the joints are clear cut and continuous, the quartz veins also appear so. In the well-bedded por- tions of the iron-bearing formation, where the joints are irregular and discontinuous, the distri- bution of the vein quartz is also irregular and discontinuous, being rather in a confused zone than in a well-defined plane. After the upper Huronian was tilted and folded the upper edges of the beds were eroded awaj', with the result that the rock surface is now in-egular, ^\^th dips corresponding roughly in direction but not in degree with those of the underlying rock strata, being in general less steep. RELATIONS OF THE tTPPER HURONIAN (ANIMIKLE GROUP) TO OTHER SERIES. The upper Huronian lies unconformably upon the Archean and lower -middle Huronian rocks. The proof of unconformity is as follows: 1. The conglomerates at the base of the upper Huronian" contain' fragments derived from the underlying rocks. 2. There is discordance in dip. The underlying formations, where they have any parallel structure at all, are almost vertical. The upper Huronian is well bedded, with a low dip. Moreover, in approaching the contact no change of dip is to be observed either in the upper Huronian or in the underlying rocks. 3. There is a difference in the amount of minor folding, fracturing, secondary cleavage, and further consequent metamorphism of the two boches, the upper Huronian being much less affected than the older rocks. 4. The upper Huronian belt overlies Archean and lower-middle Huronian rocks indiscrim- inately. Near Biwabik, for instance, the northern edge of the upper Huronian lies diagonally across the contact of the Archean and lower-middle Huronian rocks. 5. The lower-middle Huronian sediments are intruded by the Giants Range granite, which composes most of the core of the Giants Range. The u])per Huronian is not intruiletl by the Giants Range granite, and, moreover, in the conglomerate at its base it beare fragments of this granite. The ui)per Iluronian in ranges 12 and 13 is in eruptive contact with the Kewee- nawan granite and gabbro. CONDITIONS OF DEPOSITION OF THE UPPER HURONIAN (ANIMIKIE GROUP). The conditions under which the upper Huronian 0\jiunikic group) was de|)osited are dis- cussed for the Lake Superior region in Chapter XX. It may be noted here that the rocks of this group are believed to be subaqueous deposits grading upward into delta deposits. The Mesabi "Listed in Mon. U. S. Geol. Survey, vol. 43, pp. 94-9S. MESABI IRON DISTRICT. 177 district may represent shore conditions' of deposition as contrasted with the Cuyuna district farther soutli, wliich may represent offshore conditions. The well-assorted sands at the base of the group in the Mesabi district seem to show variation in tliickness and area corresponding to the coiiliguration of the older rock surface. For instance, the point of Pokegama quartzite extending eastward from Embarrass Lake suggests a sand spit,, though distribution may be complicated by faulting. The peculiar conditions determining the deposition of the iron- bearing formation are discussed on pages 499 et seq. KEWEENAWAN SERIES." DULUTH CABBRO A portion of the great mass of Keweenawan gabbro of northern Minnesota comes within the limits of the Mesabi tlistrict. The northern edge of the mass lies diagonally across the east- ern end of the district, extending from near the Duluth and Iron Range track, in range 14, northeastward through ranges 13 and 12 to Birch Lake. Through range 14 the gabbro is in contact with Virginia slate; in ranges 13 and 12 it is in contact with the Biwabik formation, and north of Birch Lake it is in contact with lower-middle Huronian granite. The northern edge of the gabbro forms a conspicuous northward-facing escarpment overlooking the low- lying area of the Virginia slate and of iron-bearing formation immediately to the north. To this the name "Mesabi Range" was first applied. In the neighborhood of Birch Lake the gabbro comes well up on the crest of the Giants Range, and here it does not stand above the adjacent rocks. DIABASE. There are in the Mesabi district certain rocks associated with the Duluth gabbro which are not covered in the above general account. In range 13 exposures of fine-grained diabase appear in the SW. i sec. 25, T. 60 N., R. 13 W., and in the central and northern portions of sec. 35, T. 60 N., R. 13 W. Bowlders of the same material indicate its extension for several miles east and west, and, taken together with the exposures, indicate a belt with a possible width of somewhat less than a mile, a length of at least 3 miles and probably much more, and a trend northeast and southwest — that is, parallel to the general strike of the formation bound- aries in this part of the district. The diabase is a fine-grained dark-gray rock which under the microscope shows a weU-developed ophitic arrangement of plagioclase feldspar crystals and the presence of abundant hornblende and less abundant ilmenite and magnetite. The diabase corre- sponds Uthologically to the diabase sills intruded in the iron-bearing formation in the neighbor- hood of Gunflint Lake, and there supposed to be either offshoots of the gabbro or intrusives both in the gabbro and adjacent rocks. The trend of recent opinion is toward the former conclu- sion. In the SW. { sec. 25, T. 60 N., R. 13 W., south of the diabase, drill holes have recently penetrated altered slate (cordierite hornstone). The relations of the slate to the surrounding rocks are unknown because of lack of exposures and exploration. If the slate is continuous with that to the west, which had not heretofore been known to extend farther east than sees. 5 and 8 of the same range, the diabase must be a sill intruded in the upper Huronian (Animilde group). If the slate is not continuous with the main belt of slate to the west, it must be an isolated mass in the Keweenawan rocks, and the diabase would belong with the main mass of the Keweenawan. From the analogy of its hthologic character with that of the diabase sills to the east, from its distribution, and from the occurrence of slate to the south it is thought that the diabase is probably a siU, but lack of exposures and of sufficient exploration makes it quite impossible at present to show its boundaries on the map. The area south of the diabase, including that in wliich the slate has been found, is therefore mapped as Keweenawan. A httle southeast of the northwest corner of sec. 34, T. 59 N., R. 14 W., E. J. Longyear found diabase at the depth of 984 feet, in a drill hole which had passed through 16 feet of drift, 392 feet of black slate, and 576 feet of iron-bearing formation. Diabase was penetrated for 309 o For a general account of the Keweenawan series of Minnesota see Chapter XV (pp. 366 et seq.) 47517°— VOL 52—11 12 178 GEOT.OGY OF THE LAKE SUPERIOR REGION. feet before the work was stopped. The iron-bearing formation is IioiiihIciI on the north by lower-middle Huronian graj^wackes antl slates, upon the eroded edges of which lies the iron- bearing formation, with perhaps a tliin layer of Pokegama (juartzite between. The fact that the diabase rather than the Pokegama quartzite or lower-middle Iluronian graywacke and slate was reached by the drill below the iron-bearing formation would be in accord with the supposition that the diabase formed a sill intruded into the iron formation at this place. In the NE. i SE. i sec. 13, T. 57 N., R. 22 W., drilUng has penetrated 20 feet of diabase with iron-bearing formation both above and below. EMBARRASS GRANITE. Through ranges 12 and 13 and as far west as sec. 2, T. 59 N., R. 14 "W., a distance of 15 miles, the granite forming the core of the Giants Range is intrusive into the upper Huronian. Wliether it was intruded at the close of the upper Huronian epoch or during the succeeiling Keweenawan is a matter of doubt and indeed is a matter of small consequence. The fact that granite dikes cut the Keweenawan series in other parts of northern Minnesota makes it a plausible assumption that the granite was intruded in Keweenawan time, but no relations of the granite to the Keweenawan have been observed in the Mesabi district. The granite is named the Embarrass granite from its lithologic similarity to granite exposed at Embarrass station on the Duluth and Iron Range Railroad, just north of the Giants Range. The Embarrass granite is a pink hornblende granite. It is usually of coarse grain but shows much variation. In general the grain becomes finer toward the west. The character- istic feature of the granite is its large content of quartz in small and large grains, which are verj^ conspicuous, especially on the weathered surface. The quartzes range in diameter from a few miUimeters to more than a centimeter. The large one>s have a characteristic purplish-blue color. In its content of quartz the Embarrass granite is readily distinguished from the lower- middle Huronian granite (Giants Range granite) in the central and western parts of the range, in which the quartz is exceedingly rare or entirely lacking. Other constituents are pink ortho- clase feldspar, which sometimes occurs as porphyritic crystals almost an inch long, and a rather small amount of hornblende. The relative abundance and coarseness of aU the constituents of the granite of course show the usual variations of a large granitic mass. Cutting the granite are a few dikes of finer-grained, lighter-colored quartzose granite, wluch under the microscope is found to differ from the one just described only in lacking hornblende and the rare elements mentioned. In the Mohawk mine and elsewhere near Aurora granite forms the foot wall of the ore bodies, in one place coming within 16 feet of the rock surface. From tliis vertical dikes cut across the formation. The relations seem to be those of intrusion of granite principally parallel to the bedding but partly across it. These relations may be correlated with those of the Embarrass granite at the east end of the range. CRETACEOUS ROCKS. '>'■■ ^-' ' Distribution and character. — Recent explorations have showTj Cretaceous conglomerates, shales, or iron ores as a tliin mantle over most of the western part of the district and in isolated patches as far east as Embarrass Lake. It is therefore thought inadvisable to attempt to show Cretaceous deposits on the map. Especially noteworthy is the discovery of small con- glomeratic Cretaceous ore bodies overlying the contact of the iron-bearing Biwabik formation and the Virginia slate. From the distribution of the remnants now known it is certain that Cretaceous rocks once overlay all of the district west of range 16, that they may have extended farther east, and that erosion has largely removed thein»from the area they did occupy. It is not unlikely that some of these remnants have been protected because faulted do^Tn in post- Cretaceous time. The rocks consist of conglomerate and shale. The conglomerate in the occurrences known overlies iron-bearing rocks and in some places iron ore. As woukl be expected, therefore, the MESABI IRON DISTRICT. 179 fragments of the conglomerate are (.lerivcd from the iron-bearing formation; in the western part of the range the conglomerate is locally rich enough to mine. The conglomerate fragments consist mamly of heavy ferruginous chert and iron ore, both hematite and limonite. Except locally, and especially where the pebbles are of hard material, they are not well rounded. There are present in the conglomerate also fossils which are described below. The fragments are but loosely cemented. When broken out of the ledge the rock is fairly compact, but on being exposed to weathering it soon disintegrates. The cement is largely ferruginous, but there is present also a considerable amount of white or yellow substance which A. T. Gordon, chemist of the Mountain Iron mine, found to consist of silica and alumina and hence to be essentially a clay. Occasion- ally there may be observed also minute greenish-yellow particles in the cement which may be glauconite grains, so common in the Cretaceous. Analyses disclose abundant phosphorus. The general appearance of tliis Cretaceous iron-ore conglomerate is very like that of "canga" or rubble ores formed subaerially on the surface of iron formations in the Minas Geraes district of Brazil. The shales are soft, thm-bedded rocks of a bluish-gray color when fresh but in many places are of a light color due to bleaching. These, too, contain fossils. Fossils. — Selected specimens of the shale and conglomerate containing fossils were sub- mitted to T. W. Stanton, paleontologist, of the United States Geological Survey, for examina- tion. He pronounced them to be "unquestionably Upper Cretaceous forms, not older than the Benton and probably not younger than the Pierre." In addition to the fossils above noted, the Cretaceous of the Mesabi district has been found to contain small shreds of lignitic material. The presence of this material well up on the Mesabi range suggests the possibility of finding lignite deposits in the low area to the west, north, or south of the Mesabi range. PLEISTOCENE GLACIAL DEPOSITS. The Mesabi district is covered by a mantle of glacial drift, of the late Wisconsin epoch, which effectually conceals the greater part of the imderlymg rocks. On lower slopes the drift is thick, sometimes reaching 150 to 200 feet, and here of course rock exposures are rare: on middle slopes the thickness commonly does not exceed 50 or 60 feet, and 20 to 50 would measure much of it; on the upper slopes of the range the drift is thin or altogether lacking and rock exposures are corre- spondingly abundant. In the eastern portion of the district also, where the Giants Range granite has a higher elevation than to the west, the drift is thm and allows numerous rock masses to project through; toward the west, as the elevation of the Giants Range decreases, the drift becomes thicker, until westward from Grand Rapids it buries even the crest of the Giants Range to a depth of more than 100 feet. The Pleistocene deposits are fully discussed in Chapter XVI (pp. 427-459). THE IRON ORES OF THE MESABI DISTRICT. 1 By the authors and W. J. Mead. . DISTRIBUTION, STBTJCTTJRE, AND RELATIONS. The iron-bearing Biwabik formation rests on the middle south slope of the Giants Range, with a low dip to the south, 4° to 20°, affording an exposure of considerable width at the surface. The elevation of this exposure varies between 1,400 and 1,600 feet. The distribution of the Biwabik formation and the possibilities of westward extension are discussed on pages 164-165. Possibilities of extension southward are mentioned on page 174. The ore bodies are in patches along the erosion surface of the iron-bearing formation, generally less than 200 feet thick, but reaching 500 feet at greatest. The aggregate area of all the iron-ore deposits of present commercial grade known at this writing at the surface is about 15 square miles, constituting a little less than 8 per cent of the exposed surface of the iron-bearing formation in its productive portion between the east line of 180 GEOLOGY OF THE LAKE SUPERIOR REGION. > 4 i\i H X 1^ range 14 on the east and west side of nintje 26 on tlie west. If low-grade ores were counted the area would be approximatelj' douMcd, East of ran^c 14 the nature of the formation is influenced by tlic great Keweenawan Duluth gabbro mass overlying the east end of the district. The ore bodies are few and snuill and are more largely niagnet- itic and amphibolitic than hematitic. Toward the west end of the district also the ores become lower in grade, owinf to increasing content of loosel}' disseminated chert, locally called sand, so abundant in certain of the ores that they require washing to attain the present commercial grade. The rocks immediiitely associated with the ores are mainl}^ ferruginous cherts, locally called "taconite, " form- ing both the walls and basements of the deposits. The ores usually do not rest directly upon the quartzite under- lying the iron-bearing formation. Their lower hmits are locally marked by thin layers of paint rock a few inches thick. A horizontal plan of the Mesabi ore dej>osits is exceedingly irregular both in major outline and in minor features. The deposits are in many places bounded hj intersecting plane surfaces of joint or fault planes. In A^ertical section the ore deposits in general are widest at the top and narrow below, in the form of a shallow basin. The slopes of the basin are rarely symmetrical and few slopes are uniform; a slope is generally a series of steps, some of them overhanging the ore or projecting into it. The bedding of the iron ores is continuous with that of the adjacent ferruginous cherts of the iron-bearing formation except where there has been local slump or faulting at the contact. The shunp is sometimes accompanied by close crumpling of the layers of the iron-bearing formation (PI. IX). It will be shown later that the slump results from the leaching of silica. Obviously the layers have been originally too long for their present positions and have crumpled to ac- commodate themselves to the new conditions. The bed- ding of the ores is thus essentially parallel to that of the upper Huronian of this district — that is, sloping gently south- ward at angles from 4° to 20°, with minor gentle folds whose axes pitch in that direction. A good general conception of the structural relations of the Mesabi ores may be obtained by thinking of the ores as irregular rotted upper portions of the slightly tilted and beveled iron-bearing formation, the rotting having been favored in certain spots, as will be shown later, by the fracturing of the formation or by the minor folds in which the formation rests. (See fig. 16; PI. X, which is a north-south cross section; and PI. XI.) a a a a I CHEMICAL COMPOSITION OF FERRUGINOUS CHERTS AND ORES. AX.VLTSES. of theores and related rocks is « The chemical composit ion g here exhibited by partial and comi)lcte analjses from vari- '^'"'"""°'~'" ous sources. A large number of the analyses employed were kindly furnished ])V the several mining companies. .\.ll the other analyses except those previously published were made hj Lerch Brothers in their laboratories at Ilibbing and U. S. GEOLOGICAL SURVEY MONOGRAPH Lll PL. IX *^. s. >^, 'm^^^ ^■ijf«y',^^.-*;:;f-:.v.s_ A. HAWKINS MINE. B. MONROE MINE. SHARP FOLDING OF BEDS OF IRON-BEARING BIWABIK FORMATION IN MESABI DISTRICT, MINN. See page 180. U. 8. GEOLOGICAL SURVEY MONOGRAPH LM PL. X 5BN..n.20E. , Sec 2e,T, 53N.,R.E0E. Datum 9(J0ft aboy LakeSiipenof ■.^- . _ - -^^ -.Sfe^i ^ Decoinpoaedlaconile Tacoiiile (uaJDt rock) (tl^composedat (with rurruffinous slate and points maihed Ct) 3om^green alats ut h) NORTH-SOUTH CROSS SECTION THROUGH IRON-BEARING BIWABIK FORMATION, MESABI DISTRICT, MINNESOTA. Compiled by 0. B. Warren from drill records. U. S. GEOLOGICAL SURVEY MONOGRAPH LI1 PL. Xr PANORAMIC VIEW OF THE MOUNTAIN IRON OPEN-PIT MINE, MESABI DISTRICT, MINN. Looking east. From photograph presented by J. F. Lindberg, Hibbing, Minn- See pages 180, 497. B. PANORAMIC VIEW OF THE SHENANGO IRON MINE, MESABI DISTRICT, MINN. See pages 180, 497. MESABI IRON DISTRICT. 181 Virginia, Minn. The average cargo analyses for the various grades of ore were obtained from the hst pubHshed by the Lake Superior Iron Ore Association. Nine tj'jjical analj'ses of taconite are given in the followmg table. These analyses include carefully selected samples from several drill holes giving complete sections through. the formation. Partial anali/sts of ferruginous chert {laconilf) from the Memhi range. (Samples dried at 212° F.I La Rue mine, see. 29, T. 57 N., R. 22 W Stevenson mine, sees. 7 and 8, T. 67 N. , R. 21 W Crosby mine, sec. 32, T. 57 N., R. 21 W Do Drill core from three holes in T. 57 N., R. 22 W., in all 800 feet Drill core, 30) feet La Rue mine, sec. 29, T. 67 N., R. 22 W Burt, mine, see. 31 , T. 6S N . , R. 20 W Do Average 32.24 24.99 11.79 19. 5(i 30.24 23.80 32.26 23. 98 32.62 25.71 SiOj. 68.70 P. 0.021 .024 .010 .013 .038 .030 .018 .013 .020 .021 -MiO.. Loss on ignition. 0.37 i .21 .29 I .%i .84 1.20 .30 .91 .42 ..54 0.62 .60 .67 .25 5.16 7.52 .45 1.33 1.07 1.96 The large loss on ignition in tlie drill-core samples is in part due to the presence of CO2 in carbonates. The samples represent the hard phases of the formation, showing little concentration to ore. When all of the iron-bearing formation outside of the available non-ore deposits is aver- aged, including both the hard lean parts shown in the above table and the partly concentrated portions of the formation, the average iron content runs higher. An average of 1,094 analyses, representing 5,400 feet of drilling in the district away from the available ores, gives 38 per cent. This does not include the ores. Because of the great mass of such rocks as compared with the ores, this figure of 38 per cent represents approximately the general average u'on content of the entire formation. The average composition of the Mesabi ore for the years 1906 and 1909 was obtained by combining average cargo anah'ses of all grades mined for each of those j^ears in proportion to the tonnage represented by each grade. In this manner an average analysis was obtained which represents as exactly as possil)le the composition of all of the ore mined in the Mesabi district during the years 1906 and 1909. Average composition of all ore mined in the Mesabi district during the years 1906 and 1909. Moisture (loss on drying at 212° F.). Analysis of dried ore: Iron. Phosphorus. Silica Manganese . . Lime. Alumina.. Magnesia. Sulphur. Loss on ignition . 60.70 .0559 5.58 1.58 '4.' 57" 1909. 12.27 58.83 .062 6.80 .816 .32 2.23 ..32 . 069 4.72 Range in composition of ores mined in the Mesabi district, as shown by average cargo analyses for 1909. Moisture (loss on drying at 212° F.) 7. 15 to 15. 79 Analysis of dried ore: Iron 52. 40 li . 64. 05 Phosphorus 019 to .105 Silica 2.50 to 19. 90 Manganese 20 to 2. 84 Alumina 16 to 5.67 Lime to 1.82 Magnesia to 2. 06 Sulphur 004 to .440 Loss on ignition 1. 71 to 9. 45 182 GEOLOGY OF THE LAKE SUPEKTOI! I!EGTOX. KEPRESENTATION BY MEANS OF TRIANGITLAK DIAGRAM. In figure 17 the triangular method of phitting is employed to show the chemical cf)mposi- tion of the various phases of taconite and ore studied. Here actual percentage weights of the constituents are indicated, and no account is taken of volume or porosity. Each point, by its position in the triangle, indicates an individual analysis. The diagram consists of an equilateral triangle crossed by equally spaced lines, 100 parallel to eacii side. Distances measured i)er- pendicularly from the three sides to any point within the triangle (by means of the divisions in the triangle) represent severally percentages of ferric oxide, silica, and the remaining constit- FERRIC OXIDE MINOR SILICA CONSTITUENTS Principally alumina and water of hydration Figure 17. — Trian;ular diagram showing composition of various phases of Mesabi ores and ferruginous cherts in terms of ferric oxide, silica, and minor constituents (essentially alumina and combined water). The ores and cherts here represented are shown in flguro 21 in terms of percentage volumes of iron minerals, silica, and pore space. uents. Thus any point in the triangle indicates a certain definite combination of these three factors. The grouping of the points in the triangle shows that the principal variation in com- position lies between the iron and the silica. In the process of concentration of ore from the ferruginous chert the percentage of iron increases in proportion to the decretisc in sUica, while the percentage of minor constituents remains practically constant; hence this concentration would be represented by a series of j)()ints in a line parallel to the right-hand side of the triangle. A taconite with a higii content of alumina produces an ore high in kanlin. tmd conversely. MESABI IRON DISTRICT. 183 MINEBALOGICAL COMPOSITION OF FERRUGINOUS CHERTS AND ORES. Mineralogicall}' both the cherts and the ores consist essentially of hydrated oxides of iron, chert, or quartz, aluminum-bearing minerals, usually kaolin, and a small amount of minor constituents. In the calculation of the approximate mineral composition of the various rocks and ores these minor constituents — alkalies, sulphur, phosphorus, etc. — were disregarded, the error thus introduced being small. The iron is present in the ores and cherts as a partly ly'drated ferric oxide. To ascertain in each case the particular hydrated iron-oxide mineral present would be impracticable, but by calculating the iron as hematite and limonite the degree of hydration is expressed by relative amounts of the two minerals. The amount of limonite is found by assigning to the volatile matter or water of hydration available the proper amount of iron, the remainder of the iron being calculated as hematite. The practice of assign- ing to the iron mineral all the water of hydration not in aluminum silicates may introduce minor inaccuracies because of the possible slight hydration of the chert. The mineralogical compositions of the ores and ferruginous cherts of the Mesabi range calculated from the average analysis by the methods describetl above are as follows: Approximate mineral compositions of average ores and ferruginous cherts. Ferrugi- nous cherts. Ores. 1906. 1909. Hematite . . 26.30 12.22 58.07 1.37 2.04 as.oo 27.00 4.10 4.08 1.82 61.81 Limonite 25.95 Quartz . .. .- '4.10 5.30 Miscellaneous . . 2.84 100.00 100.00 100.00 PHYSICAL CHARACTERISTICS OF THE ORES. TEXTURE. The Mesabi iron ores are for the most part soft, somewhat hydrated hematite, though approximately pure limonite ores are present in subordinate quantity. The ores as a whole are of finer texture than those of any other Lake Superior district. Their texture varies from exceedingly fine-grained "flue dust" to a fairly coarse, hard, and granular ore breaking into parallelepiped blocks. Usually the ore needs but little blasting to allow the steam shovel to take it from the bed. The average texture of the Mesabi ores is shown by the following table, repre- senting an average of screening tests on eight grades of typical Mesabi ore totaling 18,313,570 tons in 1909. These screening tests were made by the Carnegie Steel Company and represent the total 3'ear's output of each of the grades tested. The textures of the ores of the several Lake Superior districts are compared in figure 72 (p. 4S1). Textures of Mesabi ores as shown by screening tests. Per cent. Held on J-inch sieve 25. 98 ^-inch sieve 26. 24 No. 20 sieve 11. 54 No. 40 sieve 9. 90 No. 60 sieve 8. 54 No. 80 sieve 2. 16 No. 100 sieve 2. 28 Passed through No. 100 sieve 13. 68 The fineness of many of the ores has required mixture with coarser grades for blast-furnace charges. The average mixture is approximately indicated by the proportions of Mesabi to other Lake Superior ores, which has increased to 69 per cent in 1910. 184 GEOLOGY OF THE LAKE SUPEKlOll KEGIO.X. DENSITY. Several methods were employed in the determination of density — (1) determination of density of finely powdered specimen by means of specific-gravity bottle; (2) determinations of density from hand specimens by the common metliod of weighing in air and in water, the pores of the rock being filled with water by prolonged boiling before weighing under water; (3) calculation of specific gravity of tlie rock or ore from mineral composition by proper combina- tion of tlie thaisities of the several minerals present. The density of the ores or cherts calcu- lated l)y using the density of the iron minerals given by Dana was uniformly liigher than the density iound by gravity methods. The iron minerals in an earthy form have a lower density' than those in the hard ores, and it was found that the two methods could be made to agree by assigning to hematite a density of 4.5 and to limonite one of 3.6. By combining the specific gravities in proportion to the percentages of the minerals the average density of the ferruginous cherts is found to be 3.27. Actual density determinations on eleven specimens of ferruginous cherts gave an average of 3.02. (See table below.) This figure is lower than the average figure computed above, for two reasons: The eleven specimens on which the detenninations were made contained a smaller percentage of iron than the average analysis above. The close texture of the specimens prevented complete saturation by immersion in water and also prevented complete drying; hence both density and porosity determinations are somewhat lower than they should be. For these reasons it is believed that the specific gravity as calculated from the average anah^sis above (3.27) represents most closely the average specific gravity of the taconite. The average specific gravity of the ore, as calculated from the mineralogical composition of the average ore, is fountl to be 4.10. POROSITY. In all rocks and ores of which hand specimens could lie collected the porosity was deter- mined by comparing the weight of the specimen when saturated with water with its weight when dried. This manner of determination is formulated as follows: Weight of water absorbed Weight of rock when saturated Porosity = = moisture of saturation = M. M 1 -M G + M where G equals specific gravity. From this formula it is obvious that a determmsition of density is necessary in connection with each porosity determination. The porosity determinations on eleven specimens of ferruginous chert by tlie method described follow. Porosity detenninations of chert. Specimen No. 44051. 45S88 45309 4S305 40651 4.5021 45596 Specific gravity. 3.25 3.04 2.86 2.88 2.90 3.22 2.92 Porosity (per cent of total volume). 6.5 2.3 9.45 5.1 6.25 6.00 3.75 Specimen No. 45603 45692 4,5672.\ 45.590 Average Specific gravity. 2.80 2.87 2.96 3.07 Poro*:ity (percent of total volume). 3.02 3.50 .3.80 6.45 3.55 4.72 To unconsolidated material, such as a large part of the Mesabi ores, the above method could not be applied. The porosity of such material was found by comparhig its actual density when in place, including jiore space, with the calculated mineral diMisity, which does not include pore space. The actual density of the material in ])lace was detcrmmed by weighing the MESABI IRON DISTRICT. 185 amount removed from an excavation made on a leveled surface of the ore, the volume of the excavated material being determined by measuring the amount of grain necessary to fill the excavation. Another method for the determination of cubic content of the Mesabi ores is one employed by O. B. WaiTcn, of Hibbing, Minn. Mr. Warren used a bottomless box 4 feet long, 3 feet wide, and 1 foot deep. These dimensions were chosen as representing the average volume of a ton of ore. This box is set up on a leveled surface and the ore removed from the inside of the box until the sides are sunk to the level of the surface. In this way exactly 12 cubic feet of ore are removed and weighed, a sample for analysis being taken at the same time. The porosity of the ore may also be determined by saturating a portion in place by an abundant application of water. Placing a sample of the saturated material immediatelv in a closed vessel permits the determination of the moisture of saturation, from which the porosity may be calculated as shown above. Where the ore to be tested is in a vertical wall a small niche should be cut to afford a horizontal surface for the application of the water. It will be seen that this method does not differ essentially from the determination of porosity of hand specimens, except that the material is saturated in place and not after removal from the ground. More than 100 determinations by the various methods show the average porosity of the ore to be approximately 40 per cent of the volume. (See fig. 21, p. 190.) CUBIC CONTENTS. Owing to the wide variation in the three essential factors, densit}', porosity, and moisture, there is a wide variation in the number of cubic feet per ton of the ores. This number ranges from 9 cubic feet per long ton in some of the highest-grade blue granular ores to 17 or 18 in the low-grade limonites. The average for the district is approximately 12 cubic feet per long ton. The method of calculation is discussed on pages 480-484. MAGNETIC PHASES OF THE IBON-BEABING FORMATION. OCCURRENCE. • Eastward from the town of Mesaba, on the Duluth antl Iron Range Railroad, the iron- bearing Biwabik formation becomes progressively more magnetic, more coarsely crystalline, and the red or bro\viiish tones of the ferruginous cherts give way to black and gray colors. Ore deposits are rare. Such as there are consist of mixtures of liematite and magnetite. In the most magnetic and crystalline parts of the formation ore deposits seem to be entirely lacking. In addition to the magnetite and tjuartz, there are present various anhydrous silicates, such as griinerite, actinolite, augite, and others. The parts of the formation rich in magnetite are concentrated into definite layers a few inches to a few feet in thickness and interlayered with layers less rich in magnetite. Mining would require not only hand sorting but presumably also crushing and magnetic concentration. CHEMICAL COMPOSITION. The chemical composition of the amphibole-magnetite rock is about the same as the average of the iron-bearing formation elsewhere in the Mesabi district outside of the iron-ore deposits, as is shown by the following average : Average chemical composition of amphibole-magnetite rock in the Mesabi district- SiO, 60. 51 AUO3 ] . 20 Fe 25. 22 MgO 52 CaO 67 HoO Small. P2O5 05 S 59 MnOo 92 TiOj None. 186 GEOT.OGY OF THE LAKE SUPERIOR REGION. The reasons for the hick of concentration of ore in this part of tlie formation are discussed on page 553. SECONDARY CONCENTRATION OF MESABI ORES. STRUCTURAL CONDITIONS. In the Mesabi district waters faihng on the south slope of the Giants Range have flowed southward, entered the eroded edges of the slightly tilted Huronian series, and flowed through the iron-I)caring formation, following both bedding and joint planes. There are genth' pitching rolls in the formation, but they are so light that their control of the circulation is small as com.- pared with that of the bedding and joints. The result is the extreme irregularity in- the shape and distribution of the Mesabi ore deposits. On the south the iron-bearing formation is overlain by slate. The percolating waters un- doubtedly permeate the iron-bearing formation beneath the slate, but it is altogether likely that there they are ponded and have a relatively slow movement. Drill holes put down through the slate into the iron-bearing formation occasionally meet water under artesian pressure. The principal zone of escape doubtless is the north edge of the slate — that is, the water over- flows to the surface before passing far under the slate (fig. 18). Tliis doubtless explains the comparative lack of alteration of the iron-bearing formation or the existence of ore deposits far under the Virginia slate. The ponding effect of the slate also probably aids in diminishing any possible effect which the southward-pitching synclines in the iron-bearing formation might have on the localization of the ores, for the reason that near the slate flowage of water is controlled by the point of escape at the edge of the slate rather than by the configuration of the basin in which it might othei-wise Iron-bearing forma,tion FlQUBE 18.— Section through iron-bearing Biwabik formation transverse to the range, showing nature of circulation of water and its relations to confining strata. flow, and this point of escape may be higher than the anticlines in the basement, thus ahowing the waters to flow equally well over anticlines and synclines in the basement. The impervious basement in the Mesabi district is usually some laj^er in the iron-bearing formation itself, commonly a shaly layer which has subsequently been altered to paint rock. In no place does the ore rest directly upon the underlying quartzite. The greatest depth of the Mesabi ore deposits must be less than the depth of the iron-bear- ing formation, and as the greatest thickness of the formation is only near the slate margin, where the waters are escaping and are not doing their best work, it follows that the ore deposits are not likely to reach this maximum depth. The greatest depth thus far known in the Mesabi range is 500 feet. The common depths ai-e less than 300 feet. The Giants Range furnishes the head for the percolating waters. Toward the west end of the district the range becomes lower and the grade of the ore becomes correspondingly lower, suggesting that the circulation of the ore-concentrating solutions was less vigorous at the west- ern end because of the lower elevation. The ores have no close relation to the minor hills on the Giants Range slope, though they tend to occur in the depressions, principally because in such places denudation is relatively deep owing to softness. Were it not for the irregular covering of glacial drift, their relations to minor valleys would be more apparent. ORIGINAL CHARACTER OF THE IRON-BEARING FORMATION. The iron-bearing Biwabik formation originally consisted dominantly of greenalite rocks and subordinately of cherty iron carbonate, the characters of wliich are described on pages 165-170. MESABI IRON DISTRICT. 187 The alteration of these rocks to the ore has been accompHshed in two stages, mainly successive but partly overlapping — first, by alteration to ferruginous chert; second, by leaching of silica from the ferruginous chert. ALTERATION OF SIDERITIC OR GREENALITIC CHERT TO FERRUGINOUS CHERT (TACONITE) . Chemical change. — Tiie chemical change consists of oxidation of the iron according to the following reactions : For greenalite — 2FeSi03.nH,0 + O = Fe AnH,0 + 2SiO, ± H^O. For siderite — 2FeC03 + nH,0 + O = Fe^Oj.nH.O + 2C0,. Mineral change. — The greenalitic cherts or greenalite rocks are composed essentially of rounded granules of greenalite in a matrix of chert. The tendency to banding is not as distinc- tive as in the cherty iron carbonates. The greenalite alters to hydrated iron oxide. The silica remains or goes out. Mineralogically the sideritic cherts consists essentially of siderite and chert more or less segregated into alternate layers. The siderite is changed to hydrated iron oxide. Either removal or retention of silica may accompany this change. Secondary siderite, usually differing from original siderite in having coarser grain, is a minor product of alteration of both greenalitic and sideritic cherts. VoluTne cJuinge. — Though the alteration is distinctly of a katamorphic nature, the change is from a light to a denser mineral, and hence involves a reduction in the volume of the iron mineral. Like the oxidation of the siderite, the oxidation of the greenalite involves a change from a lighter to a denser iron mineral and a decrease in the volume. The volume changes involved in the above alterations are as follows: Alteration of siderite to hematite, 49.25 per cent loss. Alteration of siderite to limonite, 18.30 per cent loss. Alteration of greenalite to hematite and quartz, 24.50 per cent loss. Alteration of greenalite to limonite and quartz, 9 per cent loss. As the chert is at first unchanged in the alteration of the greenalite and carbonate to iron oxide, the volume change accompanying these alterations is effective on only a portion of the rock. Chemical analyses of both the sideritic cherts and the greenalitic rocks show that approximately 60 per cent of their volume is chert. Hence the change in volume is effective on only 40 per cent of the total volume of the rock. The loss in volume, then, for the entire rock, taking into account both the iron and the sUica, ranges from 3.6 per cent to 19.7 per cent, according as the original rock bore siderite or greenalite and according to the degree of hydration of the resulting product. Development of porosity. — This volume change, due to oxidation of greenalite or siderite, develops pore space. Determinations of porosity on eight typical .specimens of greenalitic rock and sideritic chert showed the average porosity to be 0.96 per cent of the volume of the rock. An average of twelve determinations on type specimens of ferruginous chert (taconite), from which apparently no silica had been leached, gave a porosity of 4.72 per cent. The porosity resulting from the reduction in volume, due to the oxidation of greenalite, in a rock containing 40 per cent by volume of that mineral should be 9.8 per cent of the volume of the rock when the product is hematite and 3.6 per cent when the product is limonite. The ratio of hematite to limonite in the average taconite is about three parts of hematite by volume to two of limonite; hence the porosity resulting from the alteration of average greenalite rock to average ferruginous chert should be approximately 7.3 per cent of the volume of the chert. This figure does not differ greatly from the observed porosity of the ferruginous chert — 4.72 per cent. It is to be expected that the observed porosity would be less than the porosity as calculated above, for several factors, such as cementation and mechanical agencies, would tend to close openings formed. 188 GEOLOGY OF THE LAKE SUPEKIOll JJEGIOX. ALTERATION OF FEllKUGINOUS CHEKTS (tACONITE) TO ORE. The alteration of ferruginous chert.s (taconite) to ore consists essentially in removal of silica. It has ah-eady been shown that the alteration of ferruginous cherts to ores is essentially later than that of the original greenalite and carbonate rocks to ferruginous cherts. During the change from the ferruginous cherts to ore the iron oxide remains essentially the same in absolute quantity (not in i)ercentage) and in degree of hydration, as will apjjear from some of the following analyses and calculations. VOLUME CHANGES. At many places in the district the actual gradation from ferruginous chert to ore ma}- be observed. In the following table are several series of analyses showing this gradation. Each series represents a series of specimens taken from the same layer of taconite. In no case were the members of one series taken from an area greater than 2 feet in extent, so that approximately uniform original composition was insured throughout each series. The first member of each series represents the least altered phase, each successive member of the same series showing a greater degree of alteration. Alteration of/erniginoiis chert.. Chemical composition. .\pproximate volume composition. Fe. SiO,. P. AljOj. Loss on igniiioa. Pore space. Hematite + limonite. Quartz. Kaolin. f 29.47 J 33.01 1 35.26 I 48.88 f 44.33 1 45. 30 1 48.51 49.18 32.20 38. 84 44.49 52.89 50.08 43.44 23.03 34.24 31.05 20. 42 23.60 50.78 42.69 34.33 0.016 .016 .013 .015 .013 .014 .013 .010 .018 .012 .010 0.62 .33 .40 .21 .37 .23 .33 .32 .30 .24 .22 2.92 1.65 4.48 3.83 1.81 2.73 2.07 2.64 .43 .74 .69 S.OO 16.30 20.30 52.70 38.00 39.70 42.40 43.40 4.00 23.20 24.20 32.33 31.23 33.51 30.81 33.12 34.65 36.05 35.70 33.55 33.73 39.89 57.90 51.40 39.30 •16. 18 28.10 25.20 20.88 18.25 61.10 42.30 35.35 1.74 .93 .92 .34 .77 .48 .70 .05 .90 .61 .57 Series 3 La Rue mine Series 1 was taken near the top and to one side of the ore bod}-; there was apparently no slump, as is shown by the constant volume of the iron mineral. Figure 19 is a graphic repre- Pore space """""^ Pore space ^— ^ Pore space S PoKspzce saica Kaolin Silica KaoJin Silica Kaolin SUica Kaolin Iron minerals Iron minerals Iron minerals Iron minerals Iron 33.01 % Iron 35.26 K Intermediate phases Iron 4S.S^ ft, Ix>w-grade ore Iron 29.J7 yb Ferruginous chert [taconite] Figure 19.— Diagram showing volume changes observed in the alteration of Icmiginons chert to ore. The four specimens represented were collectcil from a single band of ferruginous chert in the .'Jlcvenson mine, Mesabi district, Minnesota. (See analyses, above.) sentation of the series. Both the other series showed slight evidence of slumping, the chert bands being thinner at the most altered end: consequently the increase in volume of the iron mineral was expected. MESABI IRON DISTRICT. 189 Figure 19 shows very well tluit the essential process in the alteration of the taconite is the leaching of silica. This removal of material causes an increase in pore space. The development of porosity beyond certain limits weakens the rock and results in slumping or crushing; lience the volume of silica removed may be greater than the porosity observed. In order properly to compare the various phases of taconite and ore studied, it is necessary to consider them in terms of volume composition rather than of weight. By so doing the factor of porosity is included in each phase studied, the volume composition being given in terms of hydrated iron oxide, silica, pore space, and minor constituents (principally kaolin). The alteration as showTi by tlie average analyses of greenalite, taconite, and ore is expressed diagrammatically in figure 20. . Average greenalite rock Average ore - Figure 20.— Graphic representation of the changes involved in the alteration of greenalite rock to ferruginous chert (taconite) and ore, Mesabi district, Minnesota. The mineral composition of the various phases is represented in terms of volume by vertical distances. The mineral composition of the greenalite rock, ferruginous chert, and ore as represented was obtained by averaging a large number of analyses. METHOD OF EXPRESSING VOLUME CHANGES BY TRIANGULAR DIAGRAM. While the method of representation shown above (figs. 19 and 20) expresses well the average results, it is not a convenient way of handling a large number of detailed figures. In order that many individual comparisons may be made on a single diagram, the volumes of the principal constituents — silica, iron minerals, and pore space — are platted on a triangle (fig. 21) in which all these factors are indicated by position in the diagram. Tlie triangular method of representing percentages of tliree constituents has been described on page 182. In figure 21 the same method is employed to represent the volume composition of the various phases of the iron-bearing formation studied. As is indicated on the triangle, distances measured from the tliree sides represent severally percentage volumes of iron minerals, silica, and pore space. Tlius any point in the triangle represents amounts of pore space, quartz, and iron minerals totaling 100 per cent. In actual analyses, however, it is found tliat these three factors seldom total 100 per cent, a small percentage of minor constituents being present, principally kaolin, which makes it im- possible to represent the volume composition by a single point in the diagram. This difficulty is obviated, however, by representing the percentage volume of each of the three principal constituents by a short line drawn parallel to each side at the proper distance, thus constructing a small equilateral triangle within the larger one. The altitude of this small triangle repre- sents, by the divisions in the large triangle, the percentage volume of the minor constituents. We may then represent by the position and size of a small parallel triangle within the large equilateral triangle the volume composition of any chert or ore in terms of pore space, silica, iron minerals, and mmor constituents. DATA USED IN TRIANGLE. Chemical analyses, together with density and porosity determinations, were procured for 120 taconite and ore specimens, including gradation phases between the taconite and ore, slaty phases of the taconite, and paint rock. These data, when platted on the triangular diagram, 190 GEOLOGY OF THE LAKE SUPERIOR REGION. show the relations of the various ])hases of the iron-bearing formation and cnah'e one to deal with a large number of individual cases as easily as with averages. Each of the small triangles within the large one represents an actual specimen or sample from the iron formation. CONSIDEHATION OF THE TRIANGULAR DIAGRAM. The imaltered taconite is repre.'icnted by the small triangles in the lower left-liand side of the triangle, where porosit}' is low and silica liigh. If the taconite represented by an}' one of these triangles is to be altered to ore, it is necessary that part of the silica be removed to permit an increase in the iron content. If silica is removed, there must be an increase in the percentage IRON MINERALS High-grade soft ore Average ore ( Cargo analysis for 1906 ) Average unaltered ferruginous chert Lowz-grade ore ocherous and clayey SILICA PORE SPACE FiGDRE 21.— Triangular diagram representing in terms of pore space, iron minerals, silica, and minor constituents (clay, etc.) tlie volume compo- sition of the various phases of ferruginous cherts and iron ores of the Mesabi district. For detailed explanation see page 189. of volume either of pore space or of iron. If we suppose the change to lie between silica and pore space, iron being unchanged, the alteration of the chert ^\ill be represented bj^ a succession of triangles reaching across the diagram to the right at a constant distance from the base, silica decreasing and porosity increasing. If the sample selected is high enough in iron and low in silica, sufficient siUca may be removed to produce ore without developing an impossible porosity. On the other hand, if the small triangle selected is near the base of the diagram, representing a taconite that is composed largely of silica and contains only a small amount of iron, it is evident that removal of sufficient silica to brmg the percentage of u'on up to the ore grade without slump would develop a very large porosity. It is probable that the porositj' would increase until the material became too porous to support itself and the weight above, when MESABI IRON DISTRICT. 191 slump would -occur, decreasing the pore space and increasing the percentage volume of iron. This change would be represented on the diagram by an upward movement of the triangle selected. Actual infiltration of iron in solution would also cause decrease in porosity and increase in iron, but field observation shows that infiltration of iron is very sHght in tliis dis- trict, and hence any shortage of pore space must be explained by slump. Calculations sliow that on an average this slump amounts to approximately 45 per cent of the volume of the original taconite, wliich would give a vertical slump of 82 feet for every 100 feet depth of ore. This figure, though apparently large, is well in accord with the observed facts. The degree of slump in an ore body may best be measured by observing the amount of sag in the paint- rock layers which have been bent downward by the slump of the underlying ore. Figure 16 shows a typical cross section of an ore deposit in the Ilibbing district; the amount of slump in the ore beneath the paint rock is seen to be of the same magnitude as the above figures. The diagram (fig. 21) shows that where the original content of iron in ferruginous chert is high the amount of siHca to be leached is small and the resulting pore space is small, but that where tlie iron is low the pore space is proportionally greater. It follows, then, that ferruginous cherts originally low in siUca are much more easily and quickly altered to ore than those high in silica. It is also seen from the diagram that the ores high in alumina or clay (represented by the larger triangles) have a greater porosity in rough proportion to the alumina content. The alumina is very largely in the form of kaolin, a substance characteristically very porous and not so easily affected by slump as the coarser and more granular ores; hence the larger porosity. ALTERATIONS OF ASSOCIATED ROCKS CONTEMPORANEOUS WITH SECONDARY ALTERATION OF THE IRON-BEARING FORMATION. The shaly layers in the original iron-bearing formation become transformed to paint rock or ferruginous slates during the ore concentration. Abinidant phases of the formation inter- mediate between the shales and carbonates or greenalites become, after alteration, either ores or cherts with a pronounced shaly or slaty structure. These are variously called ferruginous slates, slaty ores, or paint rock, according to their kon and clay contents. The nature of the alteration is a leacliing of silica and the more soluble bases, leaving a mixture of clay and iron oxide. Following are typical analyses of the phases mentioned above: Typical analyses of unaltered slaty phase of iron-bearing formation and paint rock. Unaltered slaty ptiase of iron- bearing formations. Paint rock. 1. 2. 3. 4. 5. 6. Si02. 37.11 2,41 17.51 53.86 9.14 23.80 7.95 5.97 9.54 7.00 77.30 20.94 19.01 AI2O3 3 28 FesOa Fe . 15.90 30.88 25 60 FeO 26.13 3.70 .75 .09 .62 .95 2.57 .22 6.16 1.21 .73 32.21 5.89 4.67 .29 .18 } 4.28 .45 MgO CaO NajO K-O H-O - ]■ 13.44 H2O + { {:;;:;:;;: TiOz .61 C02 .14 11. 84 MnO c Volatile 3.35 P2O5 .' .09 .04 .20 1. Specimen 45461 from Moss mine; analysis by George Steiger. 2. Specimen 4.W00 from point near the southeast corner of the NE. J SW. J sec. 21, T. 58 N., R. 20 W.; analysis by H. N. Stolies 3. Specimen 112 (Chem. series No. 240). NE. J SE. J sec. 17, T. 58 N., R. 19 W.; analysis by A. D. Meeds for 1. E. Spurr. (See Geo!, and Nat. Hist. Survey Minnesota, Bull. No. 10, p. 10). 4. Specimen 40661 from Mahoning mine; analysis by George Steiger. 5. Darlc portion of banded red and white paint rock (specimen 4564{;) from Mountain Iron mine; analysis by A. T, Gordon. 6. Paint rock (specimen 4,5594) from Penobscot mine, Ijeneath ore; analysis by H. N. Stokes. Where igneous rocks have been intruded into the formation before its alteration these have suffered similar alterations to the slate. Theh- bases have been leached and they remain essen- tially as clay, retaining the igneous textures. 192 GEOLOGY 01'^ THE LAKE SUPElilOR REGION, PHOSPHORUS IN MESABI ORES. DISTRIBUTION IX THE IKON-BEARING FORMATION. Tho distribution of phosphonis in tlie various phases of the iron-bearing formation is as follows -. J'husphorus in iTon-bmrimj formation. Groenulite rock, average of six typical specimens Ferruginous chert: Average ■ ■ - Iron layers In (erniginous chert (Specimen 440oi ) . . . Chert liivers in ferruginous chert (Specimen 44031).. Iron lavers in ferruginous chert (Specimen 44nriO). . Chert lavers in ferruginous chert (Specunen 440.50).. Iron lavers in ferruginous chert (Specimen 44071 ) . . Chert lavers in ferruginous chert (Specunen 44071 ).. Slate In iron foraiation. typical analysis Paint rock, typical analysis Amphibole-magnetite rock Iron ore, average of 1906 output Ratio of Iron. Phos- phos- phori:s. phorus to iron. 25.05 0.012 0.000479 25.71 .021 .000820 fil.69 .074 .001200 24.50 .019 .000770 51.27 .035 .000680 11.55 .010 .000870 58.39 .052 .00089 38.33 .018 .00047 29.90 .098 .00328 <0.80 .189 .00462 23.56 .0394 .00167 60.70 .0559 .000920 There is a wide variation in the phosphorus content of the several grades of ore. In gen- eral it may be said that the more hydrous ores tend to run high in phosphorus but are not uniformly so. In figure 22 the increase of phosphorus %vith the degree of hydration of the ore is shown, tlie data being average cargo analyses of all grades of ore shipped from the Mesabi FlGUKE 22.— Diagram showing relation of phosphorus lo Uegiee ol hyilralion in .M.saM ores. range in 1906. Percentages of phosphorus and of water of hydration are plattcil respectively as ordinatcs and abscissas. Tiie arrangement of the points on the diagram seems to indicate that high phosphorus is in general associated witli liigh content of combined water. In the Mahoning open-pit mine large round concretions of rather hard yellow ore are fouml embedded in darker ore. The concretions contain in their centers crystalline and botrj^oidal MESABI IRON DISTRICT. 193 quartz and yellow hydrated iron oxide. Analyses of the outer shell and of the core of the con- cretions were made from samples representing a number of individuals. The results of the partial analyses given in the following table show a marked concentration of phosphoras at the center of the concretion. As the concretions are of a distinctly geodal structure, the phos- phorus in the interior was evidently one of the last constituents introduced. Analyses of concretions from Mahoning mine. Outer shell of concretions . Center of concretions Iron. 65.77 53.12 Phos- phorus. 0.058 .143 Ratio of phos- phorus to iron. 0.00088 .00269 In the Oliver open-pit mine, in 1899, a vein of limonite could be seen cutting down from the surface, clearly as a result of an alteration by percolating waters along a fissure, and the per- centage of phosphorus within the vein was much higher than in the ore immediately adjacent. This occurrence of high phosphorus is similar to the high phosphorus in the Mahoning concre- tions, in that it occurs with a more hydrated iron oxide than the surrounding ore, and is evi- dently later than the concretion of the ore. Another instance of the occurrence of high phosphorus with hydrated iron oxide was furnished by Mr. A. T. Gordon, who analyzed hard black hematite and soft yellow limonitic ore in the same hand specimen from the Mountain Iron mine with the following results: Hard ore, iron 61.00, phosphorus 0.077; soft ore, iron 57.98, phosphorus 0.118. To obtain further data on the association of phosphorus with the more hydrated phases of the ore wasliing tests were made on samples of ore from the Sellers and Burt mines at Hibbing, Minn. Each sample was stirred with water in a pail and after the mLxture had been allowed to settle for ten minutes the water was poured off and filtered and a very finely divided reddish-yeUow sediment was obtained. A portion of the remaining ore was then thoroughly washed with water until free of coloring matter. Analyses made in Lerch Brothers' laboratory at Virginia, Minn., of the samples thus obtained gave the following results: Partial analyses from washing tests on Mesabi ores. Iron. Phos- phorus. Alumina. Loss by ignition. Ore from Sellers mine: 46.64 67.92 69.92 49.57 60.65 0. 073 .052 .049 .073 .051 9.06 8.32 3. Dark-colored residue after washing with water . ... 1.34 8.31 2.00 3 54 Ore from Burt mine: 1. Finely rlh-irlpd 'jed'TTipnt hpld in sn riejiosits rcdinck ividt 01i\-ii\e gabbro Slaies and SjK>tLeds!atP6 (luarl z,it es eui.I quartzites Topography from U. S. Lake Survey ANIMIKIE OR LOON LAKE DISTRICT. 205 of an eruptive younger than the gabbro. The coarse-grained rocks between the gabbro and the keratophyxe are intermediate in character between the two and grade into them. They are therefore regarded as a contact product formed by tlie intermingling of tJie gabbro and kera- tophyre magmas. Between the keratophyre and tlie slates and (juartzites of the Animikie group there are tliree zones sliowing different grades of alteration of the sedimentary rocks due to the contact with the igneous rock. ANIMIKIE OR LOON LAKE DISTRICT OF ONTARIO. LOCATION AND GENERAL SUCCESSION. The Animikie district proper includes the area about Animikie or Thunder Bay, on the northwest coast of Lake Superior, but detailed study has been made ])rincipally of the part of the district near Loon Lake, at the east end of the bay, about 25 miles east of Port Arthur (see PI. XIII), and to this part of the district the following description applies. It is taken largely from descriptions by W. N. Smith ° and R. C. Allen.'' The succession of rocks is as follows: Quaternary system: Pleistocene series Glacial drift. Algonkian system: Keweenawan series Conglomerate, sandstone, marl, diabase sills (Logan sills). Unconformity. Huronian series: Upper Huronian (Animikie aroup). .w , '^ . ' llron-bearmg formation. Unconformity. Lower-middle Huronian Graywacke, slate, and conglomerate, with greenstone and granite intrusive rocks. Unconformity. Archean system: Laurentian series Granites and gneisses, intrusive into Keewatin series. Keewatin series Jreen schists, greenstone, mashed porphyries. ARCHEAN SYSTEM. The Keewatin series outcrops along Current River 5 or 6 miles northeast of Port Arthur, along the Canadian Pacific Railway, near milepost 119 and west of it about a mile. It com- prises a variety of green schists and mashed porphyries. Evidence of the extreme deformation to which these rocks have been subjected is found in their folded and schistose structures. The schistosity is nearly vertical with strike N. 70° E. Laurentian rocks are not present in the district itself, but form part of the granitic hills to the north. ALGONKIAN SYSTEM. HURONIAN SERIES. LOWER-MIDDLE HURONIAN. KINDS OF SOCKS. The lower-middle Huronian occupies the central part of the area between the upper Huronian (Animikie group) and the Keweenawan on the east and the Animikie and the Keewatin on the west. The detrital member of the lower-middle Huronian is represented mainly by a great thiclcness of graywacke, which is believed to be correlated with the Knife Lake slate of the Vermilion district of Minnesota. At the base of the graywacke is a considerable thickness of schistose conglomerate carrying fragments of black jasper and of a great variety of green schists. It marks the unconformity between the Keewatin and lower-middle Huronian. a Loon Lake iron-bearing district of Ontario: Rept. Ontario Bur. Mines, vol. 14, 1905, pt. 1, pp. 254-2'>n. 6 Unpublished thesis, University of Wisconsin. See also Silver, L. P., The Animikie iron range: Rept. Ontario Bur. Mines, vol. 15, 1906, pt. 1, pp. 156-172. 206 GEOLOGY OF THE LAKE SUPERIOR REGION. The conglomerate grades up into the graj^wacke, which in its lower horizons is quartzose. The actual contract l)ctwccn the lower-middle Iluronian and Keewatin was not observed, the two usually being separated by a slight topographic depression, which near milepost 119 is but 14 paces broad. The metamorphism of the graywacke has almost obliterated the bedding, but where bedding was ol)scrved it was found to be more or less discordant with the cleavage, which varies in dip from 65° S., a mile or more south of the Canadian Pacific Railway, through the vertical to 6.5° \. on and north of the graywacke ridge which runs |)arallel to and a short distance south of tiic Canadian Pacific Railway. Near the base of the graywacke is found locally a considerable thickness of volcanic tuff and amygdaloid. This formation is best exposed about 1^ miles south of milepost 110 on the Canadian Pacific Railway, in a strip several hundred yards wide and a mile or more long. In places it appears to be conglomeratic, showing a decided banding which looks very much like bedding, and in other places it is vesicular, the vesicules being filled with secondary' minerals. The gradation was but imperfectly observed in a single outcrop, but these tuffs and amj'gdaloids seem to grade both parallel to the strike and across it into the normal phase of the graywacke. INTRUSIVES. 1 The graywacke is intruded by a variety of granites and greenstones. All the granites and some of the greenstones are massive and cut across the strike of the cleavage in the graj-- wacke. Near some of these intrusive masses the graywacke is decidedly more schistose, especially in the area north of the Canadian Pacific Railway, where the intrusion of the granites is more intimate than elsewhere. Here the graywacke locally becomes a hornblende schist. The granite forms the hUls north of the Anunikie district and is correlative in age and topography with the Giants Range granite of the Mesabi district. UPPER HtmONIAN (ANIMIKIE GROUP). GENERAL DESCRIPTION. The iron-bearing Animikie group dips gently to the southeast across the steeply inclined structures of the underlymg series at angles locally varying widely, but averaging from 2° to 7°. It outcrops in two mam areas, the first between Loon Lake and the head of Thunder Bay, the second along the shores of Thunder Bay in the vicinity of Port Arthur. The Animikie sediments comprise two distinct zones, as follows: Thickness. A black slate formation (total thickness not present) 50 to 60 feet. An iron-bearing formation, including: An upper iron-bearing member 250 to 300 feet. An interbedded black slate 25 to 30 feet. A lower iron-bearing member 50 to 60 feet. A thin basal conglomerate 5 to 18 inches. The sediments are intruded by diabase sills varying up to 35 or 40 feet in thickness. IRON-BEARING FORMATION. The iron-bearing formation of this district is believed to be the same as the Gunflint for- mation of the Vermilion district, for it has been seen in almost continuous exposure between this district and the Vernulion district. Conglomerate. — The base of the Animikie group is marked bj- a thin but persistent layer of conglomerate, wiiich, as shown in open pits and by drill cores from bormgs m the vicinitj- of Loon Tjake, varies from 5 to 18 inches in thickness. The pebbles m the conglomerate are small anil predominantly of vein c(uartz. Small patches of it found on the graywacke ridge east of McKenzie antl on the Keewatin schists near Current River, about 5 miles northeast of Port Arthur, attest the original extension of the Animikie group over the entire area. U. S. GEOLOGICAL SURVEY MONOGRAPH LM PL. XIII ;•• •• J "^ ■•'•■"' -la V '. ■■ '. •t'.h-ci '.' '■<■ '■- p. '.9 :'■■'?■ Conglomerate, sandstone,and marl Diatase sills (Log'aa sills) HURONIAN SERIES ' UPPER huronian(animikiegroup) LOWER-MIDDLE HURONIAN 3 Miles + + + + + h + + + -r H + + + + + f + + + -t- ^ vCLl-'--;' IroTi-l:>cririii'^' I'Muiation ajad slate Gra>Tvacke and gre ens tone Granite GEOLOGIC MAP OF THE ANIMIKIE IRON-BEARING DISTRICT, NORTH OF THUNDER BAY, ONTARIO. By W. N. Smith and R, C. Allen- See page 205. - ANIMIKIE OR LOON LAKE DISTRICT. 207 Lower iron-bearing member. — In appearance the lower iron-bearing member resembles the ferrugmous chert or "taconite" of the Mesabi district of Minnesota, but it is peculiar in that it carries a large amount of calcium-magnesium-iron carbonate. The carbonate may be wholly secondary. It occurs in large part as coarsely crystalline siderite. A smgle hand specimen may be found to contain crystalline siderite, iron ore, and typical "taconite," which contains small granules embedded in a cherty matrix, thus closely resembling the altered greenalite rock of the Mesabi district. However, it may be that both the iron silicate and most of the iron carbonate were deposited simultaneously. In the Mesabi district the original iron-bearing rock was predominantly a ferrous silicate, in the Penokee-Gogebic district a ferrous carbonate with very subordinate ferrous silicate. The lower iron-bearing member in the Animikie district may have been originally made up of approximately equal amounts of the sUicate and carbonate. Certain of the layers of this member are sufficiently rich in iron oxide or low in siliceous bands to give thin zones of iron ore. Bands 6 to 8 feet thick contain 30 to 46 per cent of iron. The grade may be easily raised by sorting out the siliceous bands. The possible commercial value of these deposits is in their wide horizontal extent. Ores also appear in small irregular bodies, following the fault plane north of Deception Lake and extending eastward to Silver Lake and south and east of Bittern Lake. Interbedded slate. — Near the top of the "taconite" zone is found a black slate interbedded at more or less irregular intervals with the "taconite" below and the Iron carbonate above. The relations are those of gradation through continuous deposition. Upper iron-bearing member. — The rock making up almost the whole of the upper iron- bearing member is a cherty iron carbonate similar in every way to the iron carbonate of the Penokee district. It exhibits all phases of alteration from iron carbonate to iron ore. Some of it is coarsely crystallized, as though from secondary metamorphism. The iron ores occur principally along the fault zones already mentioned in connection with the lower iron-bearing member. These also cut the upper iron-bearing member. UPPER BLACK SLATE. In its normal phase the upper slate is made of thinly bedded layers, black but weathering to a rusty bro'svn. Locally it bears an abundance of mica. Most of the mica plates lie with their greatest and mean diameters in the plane of bedding, but many of them cut across the bedding at various angles. This phase of the rock has not been studied microscopicaUy, but the mica plates look more like detrital fragments than secondary minerals developed in place, for they occur in separated spangles and not in continuous layers, as commonly sho%vn in rocks having a development of secondary mica. Furthermore, where outcrops of the micaceous slate occur there is no evidence of metamorphic conditions such as commonly develop mica; and where it occurs in contact with intrusive diabase sills the metamorphic eifects of the intru- sion are seen not to extend more than a fraction of an inch from the plane of contact, so the mica is probably not a product of metamorphism attendant upon the intrusion of the diabase. Therefore it is believed to be clastic in origin. KEWEENAWAN SERIES." GENERAL DESCRIPTION. Unconformably above the upper Iluronian (Animikie group) is a succession of conglom- erates, sandstones, and impure marls, to which the term "Nipigon" series has been applied by the Canadian Survey. These rocks, however, are now known to belong to the Keweenawan series, and the name "Nipigon" has been abandoned by the United States Geological Survey. This series is most fully developed east of Loon Lake. The unconformity between it and the underlying rocks is marked in various ways. At the base of the Keweenawan is a coarse con- glomerate containing waterworn pebbles and bowlders of all the underlying rocks, among a See Chapter XV (pp. 366-426) for general discussion of Keweenawan series. 208 GEOLOGY OF THE LAKE SUPERIOR REGION. which, however, granite and the iron-bearing formation are predomiiumt. The Keweenawan series shows comparatively little metamorphism, ev(!n less than the Aniniikie grotij). The strikes and dips of the Keweenawan are always more or less discordant witii the strikes and dips of the underlying formations. The strongest evidence of the great time interval repre- sented by tlie unconformity is, however, the fact that the Keweenawan is found successively overlying both tiie Animikie group and the lower-middle Humnian rocks, thus showing that the entire Animikie group and part of the lower-middle Huronian had been truncated by erosion before the Keweenawan series was deposited. LOGAN SILLS. The Animikie group is intruded, mainly parallel to the bedding, by a series of diabase sills of Keweenawan age, which seem to follow j)referably the slate horizons. By jointing, these sills have been broken up into great columnar blocks, the breaking off of which where the sills are exposed maintains vertical cliffs, a characteristic feature of the topography in this district. These sills are laccolithic in character." At one locality about half a mile south of Deception Lake the diabase outcrojis in the shape of a great flat dome, the overlying slates dipping away from it in all directions. The metamorphic effect of the intrusion on the slates and iron-bearing formation is hardly perceptible more than a fraction of an inch away from the plane of contact. In certain localities the iron-bearing formation in the vicuiity of the diabase is very slightly magnetic, indicating some development of magnetite. The slight metamorphic effect of the diabase intrusions may be ascribed to rapid cooling of the magma. The fmeness of grain of the diabase suggests that the sills were not deep-seated intrusives. Thus, being thin and also near the surface, they cooled rapidly, the heat being conducted away from them by the cooler rocks adjacent. The diabase which forms the laccolithic Logan sills of the Animikie group is also found both overlying and cutting the Keweenawan sediments. STRUCTURAL, FEATURES. The main structural characteristic of the area is the general dip to the southeast ; in this it conforms to its geographic position as a portion of the north side of the Lake Superior syn- clinal basin. The upper surface of the Keewatin series and lower-middle Huronian rocks shares in the general slope to the south, although, as previously noted, this does not apply to the bedding and schistosity of the rocks. The normal strike of the Animikie group is to the northeast, with an average dip of about 7° SE. Locally, however, the rocks have been closely folded and the resulting strikes and dips are widely divergent from the normal. The general strike of the Keweenawan is east of north, with flat dip to the southeast, although it also locally shows the same severe folding and fracturing as the Animikie. Faulting has been an important factor in producing the present structural and topogi-aphic features of the district. The faulting is believed to have been caused by the same general forces that produced the Lake Superior basin. (See pp. 622-623.) The major fracturmg occurred along certain approximately parallel zones, and in the vertical displacements that followed the several fracture blocks acted as independent units, in which the northern units became depressed relative to the southern units, thus producing a system of "block" faults. The greatest vertical displacement defmitely determined is about 300 feet, as shown from diamond-drill records and surface exposures along the east-west fault a short distance south of Loon Lake. GENERAL TOPOGRAPHIC FEATURES IN THEIR RELATIONS TO GEOLOGY. As seen from a point north of Loon Lake on the high range of hills extending from Pearl River station beyond McKenzie, the region as a whole presents a general slope toward Lake Superior. To the north the country rises, the granite hills towermg one above another, and a Lawson, A. C, The laccolitic sills of the northwest coast of Lake Superior: Buli. Geol. and Nat. Hist. Sun-ey Minnesota No. S, 1893, pp. 2-1-48. ANIMIKIE OR LOON LAKE DISTRICT. 209 to the south the hikeward slope is interrupted by the long, narrow McKenzie Valley, beyond the southern rira of which the general slope is continued down to the shores of Thunder Bay. East of Loon Lak§ the range of Keweenawan sandstone hills forming the southern side of the valley swings at a right angle to the southeast, and the valley emerges on a broad flat timbered with spruce and tamarack and sloping gently down to Black Bay. To the southeast the ele- vated and much dissected area of Keweenawan sandstone projects into the lake a distance of 20 or 25 miles, forming a peninsula separating the waters of Black and Thunder bays. This peninsula, crowned at its lakeward end by a great protective cap of diabase, terminates in a bold headland over 1,300 feet high, laiown as Thunder Cape. The great escarpment of sand- stone 600 to 800 feet high forming the northwestern side of this peninsula and extending 2 or 3 miles inland is one of the most striking scenic features of the north shore. West of Thunder Cape, Pie Island, with its great flat protecting top of diabase rising 700 or 800 feet above the water, stands like a sentinel at the entrance to Thunder Bay. North of the island, on the main- land south of Fort William, McKays Mountain, another great flat sheet of diabase, supported on Animikie sediments, rises abruptly from the plain of Kaministikwia River to a height of over 1,000 feet. Thunder Cape, Pie Islantl, and McKays Mountain are magnificent examples of the mesa type of topography, which is a distinct characteristic of the Thunder Bay region. The origin of this mesa-like topography is found in the prevalence of diabase sills underlain at varying altitudes by strata of weaker rocks, the sapping of which maintains a progressive undermining of the great columnar blocks above them, thus producing vertical cUffs with talus slopes beneath. WESTWARD EXTENSION OF THE ANIMIKIE DISTRICT. The Animikie group, containing the iron-bearing formation, extends westward from Animikie Bay to the Gunflint Lake district, with structural and lithologic features like those at its east end, although in the vicinity of Port Ai-thur and thence westward the amount of slate exposed to the south and above the iron-bearing formation is much larger. The slates with their intrusive sills are beautifully exposed in Pie Island and McKays Mountain and many of the hills to be observed along the line of the Port Arthur and Western Railway. The saw- toothed topography characteristic of both the Gunflint and the Loon Lake districts is every- where to be seen, with its gently dipping slopes to the south, usually capped l)y diabase sills, and abrupt slopes to the north. The drainage for the most part follows parallel to the strike. The older rocks on which the Animikie group rests include the same kinds as were observed in both the Animikie and Loon Lake districts, but they have not been mapped in detail for all of this intervening area. THE IRON ORES OF THE ANIMIKIE DISTRICT OF ONTARIO. OCCURRENCE. Iron ores approaching commercial grade are known only in a small area near Loon Lake, 25 miles east of Port Arthur. The ore deposits are thin but extensive layers of hematite in the ferruginous cherts of the lower part of the formation. In one zone, and perhaps in others, ores have developed along fault and joint planes. The thickness of the ore layers which can be mined will depend on the grade wliich can be utilized and on the success with which chert layers may be eliminated by hand sorting. Eight feet is about the greatest thickness of a bed which would run as high as 45 per cent, but with a small amount of hand sorting two or three times this thickness could be used. The commercial importance of the ores obviously depends on their horizontal dimensions. The ores rest upon ferruginous cherts and grade into them lateraUy. One of the beds is capped by a diabase sill intruded parallel to the bedding. 47517°— vol, 52— 11 14 210 GEOLOGY OF THE LAKE SUPERIOR REGION. CHARACTER OF THE ORE. The ore is a lean, banded siliceous lieinatitc, more or less liydrated. Analyses of samples taJien every 3 inches from four exposures representing vertical distances of 6 to 8 feet each are given below. These are from the natural exposures which showed the greatest observed con- centration and include both the hematite and associated siliceous material. Analyses of Animikie ore. Iron Phosphorus Sulphur Silica 45.81 45.22 30.76 .020 .017 .160 .024 .028 .058 31.91 33.13 35. OC 30.21 .256 .036 37.1) SECONDARY CONCENTRATION OF THE ANIMIKIE ORES. Structural conditions. — The movement of waters here has obviously been controlled by the bedding, for the ores constitute merely enriched layers with irregular lateral extent. To some extent also the. waters have been concentrated in the intersecting faults. The formation i? very thin and is subdivided by impervious igneous sills, making such movement of water as i? possible in the formation essential!}" a horizontal one. Original character of the iron-bearing formation. — As described- on page 207, the lower part of the iron-bearing formation of tlie Animikie group was originally a greenalite rock with some carbonate and the upper part was originally an iron carbonate with soine greenalite. Nature of alterations. — The original greenalite and carbonate rocks have altered prin- cipally to ferruginous cherts in the manner described for other ranges. Local and for the most part subsecjuent alteration of the ferruginous cherts by leaching of sihca has devehjped the ore. Coarsely crystalline secondary iron carbonate is abundant. SEQUENCE OF ORE CONCENTRATION. The alteration of the iron-bearing formation has occurred both before and since Kewee- nawan time. Evidence of the pre-Keweenawan alteration lies in the abundant fragments of ferruginous chert and iron ore wliich occur in the Keweenawan conglomerates. Evidence of later alteration is the fact that the deformation which produced fracturing and breccia- tion of the iron-bearing formation, and which in part determined the localization of tlie ore concentration, was later than Keweenawan time, as is shown by the similar phenomena of deformation in superjacent Keweenawan beds. CHAPTER IX. THE CUYUNA IRON DISTRICT OF MINNESOTA AND ITS EXTENSIONS TO CARLTON AND CLOQUET, AND THE MINNESOTA RIVER VALLEY OF SOUTHWESTERN MINNESOTA. CUYUNA IRON DISTRICT AND EXTENSIONS TO CARLTON AND CLOQUET. GEOGRAPHY AND TOPOGRAPHY. The Cuyuna iron district is the most recently discovered range in the Lake Superior region, and as such is receivmg a large share of attention. It trends N. 50° E. along the line of the Northern Pacific Railway, near Mississippi River, in the vicinity of the towns of Brainerd and Deerwood, Crow Wing County; Aitkin, Aitkin County; and Randall, Morrison County, in north-central Minnesota. (See Pis. XIV and XV.) Its boundaries are still being extended and limits can not yet be drawn with certainty in any direction. The area of present greatest activity lies south and east of Mississippi River in Tps. 4.3 to 48 N., Rs. 28 to 32 W. The length is more than 60 miles and the area for exploration amounts approximately to 32,000 acres. The general geologic and geographic relations of the Cuyinia district to the adjacent terri- tory appear on Plate XIV. A larger-scale map of the Cuyuna district itself, showing magnetic belts, is Plate XV. This map is not colored geologically for the reason that the district is heavily drift covered and the distribution of the underlying rocks is known only incompletely from drill holes. A:iy map attemjjting to show geologic l)oundaries would be sadly out of date by the time of publication. However, the magnetic lines follow approximately the distribution of the iron-bearing rocks. The countiy is flat, being not less than 1,150 feet nor more than 1,300 feet above sea level. It is covered with a heavy mantle of glacial drift and dotted with many glacial hills, lakes, and swamps. The rock surface beneath the drift shows slight local variations in elevation, and between widely separated points, because of the general slope of the surface, may show a difl'erence of elevation of as much as 250 feet. Frequently the soft slates are found to be at lower elevations, because of erosion, than the hai-der iron-bearing formation adjacent — as, for instance, near Pick- ands, Mather & Co.'s shaft in sec. 8, T. 45 N., R. 29 W. Notwithstanding these local irregu- larities of the rock surface, it is generally flat. At many jilaces in the district and m adjacent parts of Minnesota Cretaceous deposits are found just above the rock surface and beneath the drift, suggesting that this flat surface may be part of a pre-Cretaceous base-level or peneplain. The Cuyuna district has almost none t>{ the external aspects commonly associated with a Lake Superior iron range. The conspicuous topographic ranges are lacking, as well as the numerous rock exposures. SUCCESSION OF ROCKS. From the information so far available, consisting largely of drill samples, the succession of rocks for the Cuyuna district is as follows: Quaternary system: Pleistocene series Glacial drift of late Wisconsin age, 35 to 400 feet thick. Cretaceous system Sediments, thin and in small areas. 211 212 GEOLOGY OF THE LAKE SLTERIOR REGION. Algonkian eystem: Keweenawan (?) Herios. . .Igneous rocks, extrusive and intrusivo, basic and acidic. • Upper Iluroniiiii (Animikie grou]j t Huronian series: Virginia ("St. Louis") slate: Chloritic and carbonaceous slates, with small amounts of interbedded graywacke, quartzite and limestone. Thickness unknown but great. Where intruded by Keweenawan (?) igneous rocks, this formation consists of gametiferous and slaurolit- ifcrous biotite schists and hornblende schists. Deerwood iron-bearing member of ^'irginia slate, consisting princiiially of iron carbonate where unaltered, but largely altered to amphibole- magiietite rocks, ferruginous slate and chert, and iron ore. Found in lenses in the Virginia slate, presumably near the base. ALGONKIAN SYSTEM. HTJBONIAN SERIES. UPPER HURONIAN (aNIMIKIE GROUP). GENERAL STATEMENT. The upper Huronian rocks of this district, comprising the Virginia ("St. Louis") slate and its Deerwood iron-bearing member, are not separated for much of the district, but are interbedded and have sunilar structure. They are accordingly described together. The slate, hitherto knowai as the "St. Louis" slate, has been correlated with the Virginia slate of the Mesabi district. The name "St. Louis " as apphed to this slate has priority over Virginia slate, but it is preoccupied by the well-known Carboniferous formation of the Mississippi Valley. The formation will therefore be called Virginia slate in this monograph. The iron-bearing rocks in this district have not been satisfactorily correlated with the Biwabik formation of the Mesabi district, and for them the new name Deerwood iron-bearing member is here introduced, from their typical development at and near Deerw^ood, in this district. The iron-bearing beds, being interbedded in the Virginia ("St. Louis") slate, properly constitute a member of the slate and are so treated in this report. DISTRIBTTTION AND STRTJCTTTRE. Sediments of upper Huronian age occupy practically all of the rock surface beneath the drift. They have been bent into repeated folds, as shown by drilling and magnetic work. In the southern part of the district the folding has been so close that the beds generally stand at angles of about 80° with the horizon, though locally varying at the ends of pitchmg folds. Toward the north the folding is less close and flatter dips are common. The folding has been accompanied by the development of cleavage in the softer layers, especially in the softer slates. Wliere the cleavage can be definitely distinguished from the bedding, there is usually a slight angle between them and the cleavage has the steeper dip. The iron-bearing member itself is less aflfected by the cleavage than the slate. The axial lines of folds and cleavage strike east- northeast — that is, about parallel with the axis of the Lake Superior synclihe. The iron-bearing member thus far found seems to be in the fonn of lenses whose longer dimensions are parallel to the higlily tilted bedding of the series. The wall rocks are various phases of the Virginia ("St. Loiiis") slate. Intrusive rocks locally comphcate these relations. Along the strike these lenses pinch out or widen and are locally buckled by the drag type of fold (fig. 12, p. 123). It is dillicult to tell from the present state of exploration just how far the parallel lenses are independent lenses at different horizons in the Virginia slate and Ikuv far the)^ may be the result of duplication by folding. The broader features of ilistribiition are undoubtedly to be explained by folding. There is a narrow zone of iron-bearing rocks known locally as the "south range," extending from a point east of Aitkin southwest past Deerwood and Brainerd and west of Mississi])pi River, as showni by magnetic attractions and by drilling. This is made up of a large nuuiber of short parallel and overlapping belts. \Mietiier these minor belts are repeated by folding or whether they are parallel independent lenses at difTerent NVIMNOSTV i-S £-;i;.; .i.inTTi.jiTi ^rii.ti-.ti) Snrpii[;nii ';»:(>mi ST' #^ 1 I 1 !r. ■Hi? ^ V !>,?- •" MONOGRAPH Lll PLATE XV CUYUNA IRON DISTRICT AND EXTENSIONS. 213 horizons in the slate is not known. Six miles to the north, however, in the vicinity of Rabbit Lake, there is another belt of iron-bearing rocks, known locally as the "north range," which is undoubtedly brought up here by folding, for if it were an independent belt in a monoclinal succession it would imply too great a thickness of intervening strata between the north and south ranges. Still farther to the northwest, between Rabbit Lake and Mississippi River, are at least two more belts of iron-bearing rocks repeated by folding. Whether the folds reappear elsewhere prospectors are now trjang to determine. Inspection of the map (PI. XV) tliscloses a westward divergence of the south range and the north range belts of iron-bearing rocks. The best ores of the district are found in the angle between them. Divergence of strike to the west is also to be noted between certain pairs of the minor belts, though not in all. These facts may indicate either a general anticlinorium with eastward-pitcliing axis or a syiichnorium with westward-pitching axis. The former is regarded as the more probable. LITHOLOGY AND METAMORPHISM. So far as the sedimentary rocks go, the emphasis in description should be placed on the altered phases, for they have all been much metamorphosed. Failure to recognize the scliists as parts of the sedimentary series has caused confusion in the local interpretation of drill records. The changes in the quartzite and slate to scliists are the typical anamorpliic changes of the zone of rock fiowage and igneous contacts. Hall has shown how these slates, toward the south and west, where intrusive rocks are abundant, become garnetiferous and staurolitiferous biotite schists and hornblende schists." When subsequently exposed at the surface, there has been a leaclung out of all the basic constituents, leaving light-colored, soft kaohnic and quartzose schists. This action is most conspicuous in their upper 15 or 20 feet. It is especially confined to the areas near the iron- bearing lenses. Farther south, where anamorpliism was more intense, the rocks were made so hard and resistant that they have been affected but slightly by weathering where exposed at the surface. The iron-bearing member, originally mainly iron carbonate, has also undergone anamor- pliism, resulting in the development of ampliibole-magnetite rocks essentially similar to amphibole-magnetite rocks wherever they are found in other parts of the Lake Superior region. Tliis action, however, was not sufficiently effective to destroy a large part of the iron carbonate constituting the original mass of the member. Where exposed to weathering the amphibole-magnetite rocks have been more resistant than the iron carbonates, but even they have become softer, owing to leacliing of silica, which has resulted practically in the concen- tration of the iron, which remains substantially as magnetite. The iron carbonate has been altered to limonite at the surface. The result is a mixture of hematite, hmonite, and magnetite in the iron-bearing member, soft and granular above and becoming harder and mofe siliceous below and showing more of the unaltered carbonate phases with depth. The gradation phases between the iron-bearing member and the slate have become ferruginous slates. The anamorpliism of the rocks of the Cuyuna district is probably to be explained in large part by the existence of intrusives in the area itself and west and south of it. CORRELATION. The sedimentary rocks of the Cuyuna district probably belong in the same series with the slates and schists of the Carlton, Cloc^uet, and Little Falls areas. They show many similarities in lithology, structure, and metamorphism and are geographically contiguous. Drilling in numerous places in Crow Wing and Aitkin counties shows the same pyritic and carbonaceous phases of slate as have been explored for coal in the vicinity of Mahtowa. Succession and Hthology are in accord mth distribution and general structural relations in pointing to the identity of the rocks of the Cuj^una-Carlton-Little Falls area with the upper Huronian (Animikie group) of the Lake Superior region. The Animikie group as a whole, o Hall, C. W., Keewatin area of eastern and central Minnesota: Bull. Geol. Soc. America, vol. 12, 1901, pp. 343-376. 214 GEOLOGY OF THE LAKE SUPERIOR REGION. where best known in tlie Mesabi aiul Animikie and Gogebic districts, consists of a great slate formation 2 miles or more thick, underlain by and intorbedded in its lower portions with an iron-bearing formation of var^'ing thickness, but averaging perhaps 1,000 feet, and this in turn underlain by quartzite varying from 1 to 200 feet in tliickness. Exploration has not yet gone far enough to warrant a satisfactory estimate of the tliickness of the formations in the Cuyuna district, but the information so far developed is in accord with the figures given for the Animikie group as a whole, except for the iron-bearing member, which thus far has not been found to be as thick as the average for the Lake Superior region. The Cuyuna range is separated from the Mesabi range on the northeast by a flat swamp and lake area about 50 miles wide, which completely lacks rock exposures. The Animikie group in the Mesabi district dips to the south under tliis low, flat area at an angle var\nng from 4° to 20°. It has long been obvious that the group here disappearing under the surface might somewhere be brought up to the south by folding. In tlie Gogebic range, on the south side of Lake Superior, a similar group dips at an average of 60° toward the northwest beneath the Lake Superior basin, and it has long been thought that tills group represents the Animikie group as it comes up again on the south side of the lake. An examination of the general structure of the west end of the Lake Superior liasin, however, shows that the structure of the area between these two districts is not that of a simple syncUne but of a syncline in wliich there are subordinate anticUnes — that is, a synchnorium. One of these subordinate anticlines runs west and southwest from Duluth towanl Little Falls and vicinity on Jklississippi River. If the Animikie group conies to the surface anj-where between the Mesabi range on the north and the Gogebic range on the south, it should therefore appear in tliis subordinate anticlinal fold in the western part of the general synchnorium connecting these two regions, and it was on this hypothesis that the extension of the iron-bearing formation of the Mesabi and Gogebic districts was drawn by geologists, prior to its discovery, through the present Cuyuna district, wliich Ues near the north side of this subordinate anticlinal fold. The existence of a cpiartzite exposure at Dam Lake, near Kimberly, and near Rabbit Lake, as shown by drill records, points to the fact that here erosion has cut down to the lower part of the Animikie group as it would in truncating an antichne. The course of ilississippi River itself suggests the existence of the antichne in the vicinity of the Cuyima range, for after crossing the Mesabi range it flows south until it reaches the Cuyuna district and then turns sud- denly westward as though deflected along the anticline toward a lower point of escape. Where it does break across, as at Little FaUs, rocks are exposed. The slates of the Carlton and Cloquet districts were early assigned by Irving and other geologists to the upper Huronian, but they were later referred by Spurr to the lower Iluronian because of their greater metamorphism and folding than that of the upper Huronian slates in the Mesabi district to the north and because they are intruded by granites supposed to be of lower Huronian age. It is now kno\ra that the upper Huronian (Animikie group) of the Mesabi district is also intruded by granite. The facts developed in the Cuyima chstrict seem to con- &m Irving's view of the correlation. In \new of the probable equivalence of the rocks of the Cu^nma and Carlton areas and the occurrence of small iron carbonate bands and nodules in the slates about Carlton and Cloquet and to the southwest similar to the broader bands in the Cuyuna area, the question naturally arises why erosion should not somewhere in tliis great area of exposed slate between Carlton, Cloquet, and Little Falls uncover the lower part of the Animilde group — in other words, the iron-bearing member. It may be that the crest of the antichne runs parallel with the Cujiina district itself, allowing erosion to cut down here only into the main iron-bearing member, wliile to the south and southeast the tluck capping of slates has not been removed, or it may be that the existence of great masses of intrusive granite and diabase and the intense metamorpliism wliich they have accomplished have prevented erosion of the surface or have made the condi- tions unfavorable for the direct oxidation of the iron-bearing rocks under surface katamorphic conditions. Certainly enough facts are not yet available to warrant the assertion that the iron- bearing member may not yet be found in tliis area. CUYUNA IRON DISTRICT AND EXTENSIONS. 215 KEWEENAW AN SEBIES (?). Igneous rocks are abundant in the area of the upper Iluronian (Animikie jjroup). These inchule granites and basic rocks, many of the latter characterized bj' ophitic structure. Part are sclustose; others are not. The granites outcrop conspicuously (thereby contrasting with the adjacent upper Huronian sediments) in the southern part of the district in a general belt extending from Carlton and Cloquet southwest beyond ^lississippi River. Other exposures are known northwest of the district, in the vicinity of Randall and Motley. Basic igneous rocks of diabase and gabbro types also outcrop, though less abundantly, over the same area. Dikes of the basic rocks, up to 50 feet in width, are conspicuous in the Carlton area. The intrusive character of these igneous rocks as a whole admits of no doubt. Their metamorphic effect on adjacent sediments has already been described. Within and adjacent to the Deerwood iron- bearing member driUing has disclosed much igneous rock, both basic and acidic, of yet unknown extent and with unknown relations. The contacts are sharp, the adjacent members of the upper Huronian have been locally metamorphosed, and no basal conglomerates have been found in the sediments adjacent to the igneous rocks. From these facts it is concluded that the igneous rocks cut in drill holes are probably intrusive into the upper Huronian sediments, just as are the granites to the south. The textures and structural relations of some of the basic igneous rocks suggest the possibility that they may be extrusives contemporaneous with the upper Huronian rather than with later intrusives, but until mining operations disclose more under- ground sections tlus can not be determined. In only three localities are extrusives known. An acidic extrusive rock with amygdaloidal texture, in beds 15 to 25 feet tliick, has been found by drilling to rest across the edges of the Virginia slate and Deerwood iron-bearing member, in sec. 2, T. 44 N., R.31 W.; sec. 6, T. 44 N., R. .30 W.; and sec. 7, T. 45 N., R. 29 W. The igneous rocks intrusive into the upper Huronian and the extrusives resting on the upper Huronian are provisionally classed as Keweenawan, because the Keweenawan is the next period of igneous activity, liecause abundant igneous rocks of Keweenawan age are known elsewhere in the region to cut the upper Iluronian sediments, and because they are especially abundant in that part of the Ci:yuna district which Kes approximately along the central axis of the Lake Superior syncline, largely developed during Keweenawan time. (See pp. 421-422, 622-623.) CRETACEOUS ROCKS. Immediately below the surface, in widely scattered parts of the district in Crow Wing County, remnants of a conglomerate have been found. Some consist of small pebbles of the iron-bearing member in a slaty matrix; others of small pebbles of an extrusive rock. Gen- erally the pebbles are about an eighth of an inch or less in diameter, but on two widely separated properties the oval pebbles measure as much as an inch in their longest dimension. This conglomerate is found resting unconformably, apparently in small depressions, on a rather level erosion surface of the upper Huronian. It does not contain fossil remains to identify it, but it is similar to the Cretaceous of the Mesabi range. An excellent opportunity to examine it was offered when an exploration shaft was sunk in the SW. i SE. i sec. 8, T. 45 N., R. 29 W. More Cretaceous sediments have not been identified, probably because, being poorly cemented, they are chopped and brought to the surface in drilling as churnings. Drillers frequently report imbroken shells in the lower portion of that which is reported as "surface, " and clay immediately above bed rock and below the surface, and frequently the top drill samples are light-colored, unconsolidated, and calcareous material, all of which might well be of Cre- taceous origin. None of this has been very carefully examined. The common occm-rence of large amounts of lignitic material in the glacial drift indicates a once wide distribvition of Cretaceous deposits, possibly with remnants here and there such as are found in the Mesabi range to the north. Cretaceous beds continuously cover the pre-Cambrian rocks of western Minnesota. Those of the Cuyuna district may be regarded as outliers of the main Cretaceous area. 216 GEOLOGY OF THE LAKE SUPERIOR REGION. QUATERNARY SYSTEM. PLEISTOCENE GLACIAL DEPOSITS. The glacial deposits in the eastern part of the district belong, according to Upham," to the eighth moraine and those in the western part of the district belong to the ninth moraine, counted back from the outermost moraine of the late Wisconsin glaciation. The}' vary from 35 to 400 feet in tliidcness. The heavy mantle of weathered material upon the rock surface is a remnant of the product of preglacial weathering, which in the other districts has been removed by glacial erosion. Obviously in the Cuyuna district glacial deposition has predominated over glacial erosion. THE IRON ORES OF THE CUYUNA DISTRICT. By the authors and Carl Zapffe. DISTKIBUTION, STBtJCTTJRE , AND RELATIONS. The Cuyuna ores are scattered through a considerable area beginning a little east of Aitkin, Aitkin County, Minn., and extending southwestward past Brainerd into Morrison County. (See PI. XIV.) The Umits of the ore-bearing district are not yet known. The district lacks the distinct range or ridge characteristic of the other iron-producing districts, though in general it follows a drainage divide. The area is flat, heavily drift covered, and without exposures. The development of the Cuyuna district is still in its exploratory stage. At tliis writing no shipments have been made. In the absence of exposures, information is available from about 2,000 drill holes and two shafts and from magnetic readings. The information is still inadequate to warrant any extended discussion. In the following general outline emphasis is placed on the facts thus far developed. No attempt at proportional treatment is made. This may be possible later. The Deerwood iron-bearing member is magnetic as a whole, and hence its distribution is roughly shown by the magnetic belts outlined on Plate XV and by minor belts which do not appear on this plat. Parts of the member, however, are very weakly magnetic; they are found beneath very weak belts of attraction and extend laterally some distance away from the maxi- mum magnetic line. The ore deposits may be more or less magnetic, usually less magnetic, than the associated iron-bearing member, and hence are not ordinarily situated under the belts of maximum variation, though they are not far from them. Ore deposits of suflicient size and grade to be commercial!}- available Iiave been found in both the north and south ranges, so called. The south-range ores occur at intervals along the magnetic belts from a place a mile east of Deerwood more or less intermittently to the north- eastern part of T. 43 N., R. 32 W., near jMississippi River southwest of Brainerd, a distance of about 30 miles. The north-range ores are in intermittent deposits, in a shorter but wider belt, extending from Rabbit Lake southwestward nearly to Mississippi River. The tonnage of the deposits thus far found is about equal in the two ranges, but on the north range the ores are more largely confined to a few large deposits of good grade, while on the south range the number of deposits is larger and their individual size smaller. The ores are in nearly vertical lenses and layers from a few inches to 125 feet or more wide- on the south range and up to 400 or 500 feet on the north range. The depths on the two ranges are variable as the widths. On the north range the greatest depth known is 850 feet and it is quite likely that this figure may be exceeded, but up to the present time the average depth is about 300 feet. On the south range the greatest depth Icnown is about 250 feet, and it does not seem likely that tins will be greatly. exceeded. The average depth on the south range is about 150 feet, but the higher-grade ores invariably occupy only tlie up])or 100 feet. The strike is east-northeast for distances varying from a few feet to half a mile and to an unlcaown greater distance. a Minnesota Geol. Survey, vols. 2 and 4. CUYUNA IRON DISTRICT AND EXTENSIONS. 217 Whether these lenses pitch in the chrection of strike, following the axes of drag folds, is not yet disclosed by the drilling. (See fig. 49, p. 350.) From analogy with other districts the ore bodies are likely to have a pitch, and this ]ntch is likely to be more or less uniform in direction and degree, affording a guide for exploration. The drilling has not shown the pitch, because where they are vertical the holes are stopped as soon as they run out of ore, and if they go into lean rocks rather than ore they are ordinarily not carried far enough to locate any possible extensions of the pitches. Wliere the holes are put to one side of the ore body and inclined they are stopped as soon as they have penetrated the ore lens. These pitches are, as a matter of fact, extremely difficult to locate by drilling. Closely associated with the ore on one or both walls, or m layers within the ore, is amphibole-magnetite rock. At varying depths, but usually within 125 feet on the south range, the ores tend to grade vertically into cherty iron carbonate rocks, and at these depths also the amphibole-magnetite rocivs contam much more iron ■carbonate than at the surface. It may be found that down the pitch the depth of gradation to iron carbonate is much deeper. The ores, with the associated amphibole-magnetite rocks and cherty iron carbonates, constitute the iron-bearing member of this district. The Deerwood iron-bearing member as a whole constitutes lenses or layers in the great Virgmia ("St. Louis") slate formation, lying parallel, overlapping, or end to end. Each major . lens may be divided into minor lenses by intercalated slate layers. The wall rocks of the ore may therefore be any of the phases of the Deerwood iron-bearing member or any of the phases of the Virginia ("St. Louis") slate. Characteristically one wall may be chloritic or black graphitic slate of the Virginia formation and the other wall amphibole- magnetite rock of the Deerwood iron-bearing member. The association of ore with carbona- ceous slates finds its counterpart in the Iron River, Crystal Falls, and other districts of ^Michigan. Dikes and irregular masses of basic intrusive rocks appear in all parts of this series and are associated with almost every ore deposit yet known. These may constitute one wall of the ore body or may be separated from the ore body on one wall by amphibole-magnetite rock. A characteristic occurrence of the ores is shown i.n plan and cross section m figure 25. It is apparent from this figure that the information furnished from drill holes would depend largely on the angle at wliich the drill penetrates the iron-bearing member. In a vertical lens a vertical hole will tell notlung of the character of the material a few feet away across the strike. An inclmed hole will mdicate the proportions of iron ore, amphibole-magnetite rock, and slate layers, but may not show the greatest depth of the iron-ore lenses, or, on the other hand, it may pass through the carbonate phases of the beds beneath the ore. The ore, where associated with magnetite rocks, is in many places also magnetic. The amphibole-magnetite rocks are somewhat more magnetic than the ores themselves, so that drill- ing on the maximum magnetic attraction is likely to show amphibole-magnetite rocks with the ores a few feet to one side or the other. A not uncommon relation is amphibole-magnetite rock on the maximum attraction, intrusive material on one side of the maximum, and ore on the other. The greatest distance from the maximum attraction at which ore has 3'et been found is one-half mile. It will be shown elsewhere (pp. 552-553) that the magnetic character of the member is not favorable to its richest concentration; this suggests that the best parts of the Cuyuna ore may yet be found farther away from the magnetic belt. The fact that the foot and hanguig walls of the ore deposits of most of the Lake Superior ranges are uniformly different in their lithology has led to the assumption that the foot and hanging walls of the Cuyuna ore deposits are uniformly different. Beginning in slate a few hundred feet either side of the magnetic belt, an inclined drill hole penetrates the iron-bearing member as the magnetic maximum is approached. The slate is ordinarily spoken of as "hang- ing wall." The drill is then likely to penetrate ore more or less mterbedded with slate and amphibole rock. As the magnetic maximum is approached the amphibole-magnetite rock is likely to be more abundant. The drill may go beyond the maxinmm attraction into intrusive, which would be spoken of as "intrusive foot wall." (See fig. 25.) The terms "hanging-waU slate" and "magnetic foot wall" or "intrusive foot wall" therefore signify a certain tendency 218 GEOLOGY OF THE LAKE SUPERIOR REGION. toward uiiil'ormity of reliitions wliich it is well to identil'y by such terms. But the assumption of uniformity imphcd by the use of these terms may lead to misapprehension of the facts. Slate similar to thai of tlio lianjijing wall may bo on citlicr side of the iron-bearing member. If the drills go far enough, they are likely to find slate in both walls. Slate layers witidn the iron-bearuig mem- ber itself, if first penetrated by tlie drill, would be likely to be called "hanging wall." In short, the nature of the foot and hanging w alls will depend on the particular layers in wliich the drill happens to start ami where it stops in tlie interlaminations of slate and iron-bearing member. The re- lations of the intrusive rocks to the ore deposits are still obscure, but it seems not un- likely that these ma}' be found to constitute a definite foot wall for some of the oi^e bodies. The facts just given are disclosed by drilling, but the drilling yet done gives a ver}' incomplete view of the struc- ture, and for the larger struc- tural featiu'es we must rely principally on interpretations of the magnetic field. Tlie existence of five magnetic belts in a zone 7 miles wide north and south suggests that the iron-bearmg member is re- peated by folding. If the dips were monoclinal and the sev- eral magnetic belts represented separate irou-bearuig zones in the slate, the thickness of the series to ho inferred would be greater than is reasonable. On the other hand, the drill cores show variations in the s is inclinccl to the bed- ding, and tiiis relation is itself evidence of fohhng. These folds have a strike east-nortli- east parallel to the Lake Superior axis, to judge from the magnetic belts. Moreover, the discontmuity of these belts, their distribution en echelon, and the varying intensity of the Figure ; -Plan and cross section of the ironKjre deposit in sei.-. 1_', '1'. 4i N. ^^'ing County, Minn. By Carl Zapffe. K. 32 \V., Crow CUYUNA IKON DISTRICT AND EXTENSIONS. 219 magnetic field along a single belt all accortl with the distribution recjuired by pitching folds, which repeat tlie iron-bearmg beds, the number of times diil'ering witii the locality. If tlie ■crests and troughs of the folds were horizontal, the beds would appear as parallel lines upon the horizontal erosion plane, but the actual crest and trough lines of the folds usually have a pitch; in other words, they are cross folded, so that on tlie erosion plane the beds appear to converge in the direction of the pitch. With folding of this type it is apparent that the beds may strike with a considerable variety on the erosion plan(^, according to the section this plane happens to make through the folds. The magnetic belts fail to give all the information desired as to structure, for two reasons: (1) It is not certain that the iron-bearing lenses in all parts of the district are at the same horizons in the slate; indeed, it is known that within a few hundred yards tliere may be several iron-bearing bands, so that the question is raised whether iron-bearing layei"s in other ])arts of the district belong below, with, or above them stratigraphically. (2) It is difficult to tell whether two nearly parallel belts close together represent truncateil iron-bearing layers on the two limbs of a single fold or the axes of two indej)endent folds. The main belts of attraction several miles apart doubtless represent separate folds, but the closely associated minor belts making up each of the main belts may represent either the two limbs of a single fold or two horizons on one limb of a fold. It is concluded, in general, that the iron-bearing member constitutes closely associated lenses and layers along a single general horizon in the slate. The finding of quartzite in a few places near the iron-bearing member suggests that this horizon is near the bottom of the slate formation, but this is not proved. The foldmg of the slates carrying the ii"on-bearing zones, followed by erosion, has developed the present distribution at the surface. CHARACTER OF THE ORES.a j' > GENERAL APPEARANCE. The Cuyuna ores fall into two main groups, hard and soft ores. The soft ores are black, brown, and reddish hydrated hematites, soft and earthy and much like the soft ores of the Penokee-Gogebic district. They have large pore space. These soft ores are of two types — a high-grade ore containing 55 to 63 per cent iron, soft and powdery and of a brown to very dark color, and a lean reddish-purple ore containing 45 to 50 per cent iron. The latter ore is not so soft as the former. It is easily broken do%vn with a juck but retams its ■stratified form and hangs together ia fairly large chunks. In this type cherty layers are scat- tered through the mass at short intervals, the cherty impurity probably accounting for its low grade. This ore also has a large pore sjiace. The hard ores are also of two types. The bulk of the hard ore is a black to very dark brown hydrated hematite. It is closely stratified and has suffered close brecciation as a result of slumpmg caused by the leaching out of silica. This ore varies in iron content, but is mainly high grade, ru;ming fi"om 50 to 60 per cent iron. Although this ore is brecciated it holds together in large masses, owing to the partial cementing of the brecciated pieces by the second- ary introduction of iron. Much of the ore of this type has been classed as soft ore by the drillers because it is fairly easily penetrated by a churn drill and comes to the top broken up in very fine angular pieces. It can be distinguished, however, from the true soft ore, which is washed to the surface of the hole as a fine, even-grained, powdery mass. The Cuyuna hard ore described above must not be compared to Vermilion dense blue hematite of that range. It is much softer and more limonitic. The other type of Cuyuna hard ore, small in amount as compared ^\^th that described above, is a hard blue hematite running about 58 to 63 per cent iron. It is massive and unbrecciated. This is a true hard ore and can only be drilled with diamonds. This ore occurs in layers in the softer ores and is found more frequently close to the intrusives.. ain the description of the ores the writers have drawn on quantitative data assembled by F. S. Adams (Econ. Geology, vol. 5, 1910, pp. 729-740; -vol. 6, 1911, pp. 60-70, 156-180). 220 GEOLOGY OF THE LAKE SUPERIOR REGION. It is impossible to state at lliis incomplete stage of exploration the proportion of hard to soft ore on the Cuyuna nmge. The soft ores prol)al)l3- form flic lar-180 42.10 48.80 4. no 1.72 lSIO-195 40.80 50. 00 2. i;3 3. IB 20J-210 .54. i;o 35. .So 1.. 1.1 1.57 21.5-220 (i8. 20 22. 10 7.63 1.11 230-235 Ii4.10 17.80 14. (;2 1.24 2J5-250 6.1. 10 14.20 28.89 .89 2(;5'270 54. SO Ui. 1:5 25. .35 1.49 295-300 13.85 47.23 30.84 1.87 IRON MINERALS ico% SILICA PORE SPACE Figure 27. — Triangular diagram representing volume ooinposilion o( various phases of iron ores and ferruginous elierts of Ihc Cuyuna district^ Minnesota, .\fler F. S. Adams. For description of metliod of platting and interpretation of diagram, see p. 189. 1 , M:is. analyses are expressed in terms of h(>inatite and limonite. CUYUNA IRON DISTRICT AND EXTENSIONS. 223 Tliis is merely a conventional means of showing the degree of hydration for these ores. The amount of magnetite is so small that its calculation as hematite does not materially affect the result. TEXTURE. The density of the hard ores of standard grade averages 4.09. This includes both types of liard ore. The low figure is due to the hydrated character of the Cuyuna hard ore. Tlic density of the soft ores averages 4.19. The lean soft ore shows an average density of 3.73. Th(> hard blue unbroken type of hematite has an average density of 4.26. The limonitic brccciatcd hard ore shows a density of 3.95. The pore s\rAce. of the hard ores averages 13.13 per cent l)y volume. This includes both types. The soft ore has an average pore space of 36 per cent. The lean soft ore shows 33.3 jier cent pore space. The hartl ores show a range in porosity varying from 9 to 20 per cent by volume. The hard ores of both types average 10 culiic feet per ton. The hard blue hematite varies between 9 and 10.5 cubic feet per ton. The hydrated brecciated hard ore ranges from 10 to 10.8 cubic feet per ton. The soft ores average 11.5 cubic feet per ton. The lean soft ore runs 12.6 cubic feet per ton. An average figure to use in computing tonnage for a large deposit where various ores are represented and a tonnage estimate of each type is out of the question Mould be about 11 cubic feet per ton. Notwithstanding the fineness of much of the ores, the texture is not disadvantageous, for there is probabh' less of it that will act as flue dust in the furnace than there is of the Mesabi ore, for the reason that it is as a whole less crystalline and more earthy and takes on a more coherent texture when comjiressed. SECOND AKY CONCENTRATION OP CUYUNA ORES. Structural covAitions. — The structural relations of the Cuyuna ores are still so imperfectly known ^that any statement concerning them must be made with much qualification. It is nevertheless obvious here that the concentration has been greatest at the surface and less with depth, and that at least in many places it has been very active next to the intrusives rocks which cut the member or along foot-wall slates or am])hibole schist. Also it seems to have followed axes of mmor drag folds. All the rocks have been weathered to a considerable extent. At present glacial drift covers them at depths of 35 to 400 feet, so that water stands much above the rock surface. The present condition is oliviously quite different fi-om that under which the- ores were concentrated. It may be supposed that when the rock surface was exposed waters penetrated into the iron-bearing member as it was exposed on the anticlinal areas betM'een the impervious hanging wall and the impervious foot wall and that where the member M'as cut by impervious igneous rocks they served further to control the circulation. The depth of cir- culation is not yet known, nor is it clear what topographic features may have been present in the past to control the depth of circulation. Original character of the Deeru-ood iron-hearing memier. — The member was originally cherty iron carbonate interbedded with slate. Mineralogical and chemical changes. — The alteration of the original carbonate rocks Avas in different sequence from that in most of the Lake Superior ranges, because before it was exposed to weathermg it underwent folding and intrusion, which partly altered the cherty iron carbonate to amphibole-magnetite rock. Subsequently, when erosion had exposed the mem- ber, the surface agents of alteration therefore had two phases of the mendjer to work upon — unaltered iron carbonate and amphibole-magnetite rocks. The former went through the ordinary cycle of changes to ferruginous cherts and ore. The latter lost some of its silica and amphibole but as a whole was much more resistant than the carbonate. The net result of the alteration is a soft, hydrated ore containing much magnetite along certain bands, both contaming silica as imj)urity and in increased amount with dej)th. 224 GEOLOGY OF THE LAKE SUPERIOR REGION. PHOSPHORUS IN CUYUNA ORES. Phosphorus has been concent rated with the iron during the secondary concentration of the ores. It is probable, for reasons simihxr to those discussed on pages 192-196 for the Mesabi district, that phosphorus, leached from the overlying Cretaceous rocks, has been added to the ore during its secondary concentration. In general there is not sufficient lime in the 'ore to combine with all the phosphorus as apatite, hence some phosphorus is ])robably com- bined with the hydrous aluminum and iron minerals. MINNESOTA RIVER VALLEY OF SOUTHWESTERN MINNESOTA." Pre-Cambrian crystalline rocks of the Minnesota River valley of southwestern Mimiesota appear in numerous exposures along the river, protruding from the drift, from a point south- east of New Ulm to Ortonville on the northwest. The great bulk of the crystalline rocks are granitos and gneisses. These ai)i)ear for the most part in the river bottoms but stand also in a few isolated knobs on the higher ground south and west of the river. There are man}'' varieties of granites and gneisses and all gradations between them. They are taken as a whole to represent the Archean or basement complex. Associated with the granites and gneisses are a much smaller number of exposures of gabbros and gabbro schists. These present many varieties, all of which are believed to have resulted from the alteration of two original forms and their intergradations — a hypersthene- bearing gabbro and a hypersthene-free gabbro. Peridotite is found in one exposure only in this valley, 3 miles southeast of Morton. The relations to the other rocks of the area could not be determined. Cutting the gneisses and gabbro schists throughout the area are numerous dikes of diabase. They vary in width from a fraction of an inch to 175 feet. Their age is probably Kew-eenawan. Southeast of Redstone and near New Ulm are exposures of quartzite associated with coarse c[uartzite conglomerate. Near Redstone the strike of the quartzites is N. 60-70° W. and their dip varies from 5° to 27° N. In New Ulm the strike is N. 15° E. and the dip varies from 10° to 15° SE. The quartzite is beheved to be the same as the quartzite found in a tleep well at Minneopa Falls, near Mankato, Minn., which is covered by a quartzite conglom- erate of Middle Cambrian age. The quartzite of Retlstone and New Ulm is above the Archean granite and gneiss. It is believed to be of Huronian age, but whether upper or lower is unknowTi. The crystalline rocks of the Minnesota River valley are separated from the \iv- ginia slate series of the Cuyuna and St. Louis River areas by a drift-covered area at least partly underlain by granite but partly unknown. Overlying the crystalline rocks are Cretaceous shales and sandstones, whicli appear in rare exposures in the valley, and glacial drift. a For further detailed description see Hall. C. W., The gneisses, gabbro schists, and associated rocks of southwestern Minnesota: Bull. U. S. Geol. Survey No. 157. 1899, 160 pp., with geologic maps. CHAPTER X. THE PENOKEE-GOGEBIC IRON DISTRICT OF MICHIGAN AND WISCONSIN/ LOCATION, SUCCESSION OF ROCKS, AND TOPOGRAPHY. The Pcnokee-Gogebic district lies soutli of the west lialf of Lake Superior, in the States of Michigan and Wisconsin. It extends from Lake Numakagon in Wisconsin about N. 30° E. to Lake Gogebic in Michigan, a distance of about 80 miles. In the accompanying geologic map of the Gogebic range (PI. XVI) the only essential change noted from earlier maps is in the vicinity of Sunday Lake, where faulting and perhaps folds have caused a marked effect in the iron-bearing formation. The succession of formations in the district is as follows: Cambrian system Lake Superior sandstone. Unconformity. Algonkian system: Keweenawan series Gabbros, diabases, conglomerates, etc. Unconformity. Huronian series: Greenstone intrusives and extrusives. Upper Huronian (Animikie group). Tyler slate. Ironwood formation (iron-bearing). Palms formation. Unconformity. Lower Huronian fBad River limestone. ISunday quartzite. Unconformity. Archean system: Laurentian series Granite and granitoid gneiss. Eruptive unconformity. Keewatin series Greenstones and green srhista. This chapter mainly deals with the Huronian series and especially with the upper Huronian (Animikie group). The Huronian series for most of the district has a breadth varying from less than half a mile to 2 or 3 miles. The Huronian series has a simple structure. It consists of water-deposited sediments, the origin of which has been for the most part determined. The rocks have simply been tilted to the north at an angle which is convenient for determination of the succession of belts. They are without foldmg so marked that the belts do not follow in regular order from south to north. The series is terminated on -the east by the unconformably overlying horizontal Cambrian sandstone and on the west by areas in which it has been entirely swept away by erosion, the Keweenawan series coming tlirectly against the southern complex. It is marked off from the underlying granitic ami gneissic rocks on the south and the Keweenawan series on the north by great unconformities. The major features of the topography of the district are dependent upon the relative resistance of the formations. The strike of the harder fonnations largely controls the direction of the ridges. Extending along the southern border of the Huronian rocks is a prominent ridge, the crest of which in the western and eastern parts of the district is fonned by the iron- bearing formation and in the central part of the district by the granitic rocks of the Archean. The Keweenawan igneous rocks north of the Huronian mark a second distinct ridge, the so-called Trap Range. Between these ridges, in the central two-thirds of the district, the soft Tyler <• For further detailed description of the geology of this district see Mon. U. S. Geol. Survey, vol. 19, and references there given. 47517°— VOL 52—11 15 225 226 GEOLOGY OK THE LAlvE SUPERIOR REGION. slate, constitutps level tracts and swampy areas between the more resistant rocks to the south and north. The major lines of drainage are almost directly transverse to the ridges. All the important streams of the district rise in the basement complex, traverse the entire Iluronian series, and break through the Keweenawan Trap Range to the north on their wa\' to Lake Superior. Tiius there are many notches in the east-west ridges. The elevation of the major portion of the dis- trict is between 1,400 and 1,600 feet, but a few points reach an altitude of 1,700 or 1,800 feet. ARCHEAN SYSTEM. GENERAL STATEMENT. The Archean rocks comprise the Keewatin series (greenstones and green schists) and the Laurentian series (granites and gneisses), the latter being intrusive in the fonner. WTien the relations were first appreciated for the Gogebic district the term "Mareniscan" was applied to the greenstones and green scliist series." At that time it was not known that the rocks named "Mareniscan" are equivalent to the Keewatua series of the Lake of the Woods district. Inas- much as the relations between the Keewatin and the Laurentian were worked out by Lawson for the two series of the Lake of the Woods before the tenn "Mareniscan" was proposed, Kee- watin has precedence over "Mareniscan" as a general term. KEEWATIN SERIES. The Keewatin rocks are found in two principal areas, one in the central and the other in the eastern part of the district. They are mainly scliistose basalts, for the most part fme grained and compact. The strikes and dips of the scliistosity vary greatly, in tliis respect contrasting strongly with the strikes and dips of the beds of the Iluronian sediments. The chief mineral constituents of the Keewatin are quartz, a variet}^ of feldspar, hornblende, and biotite, with chlorite, magnetite, sericite, and epidote as subordinate constituents, although locally any one of these latter minerals may be very abundant. In places the schists have a banded appearance and are true gneisses. For the most part the Keewatm scliists are com- pletely crystalline and are allied to igneous rather than sedimentary rocks. Indeed, when the Gogebic district was mapped no material was anywhere found which could be asserted to be sedimentary, although patiently searched for. However, west of Sunday Lake a biotite schist was found which was stated to present in thin section a "strong fragmental appearance." Later work has showTi that south and east of this lake some of the material is banded, weathers white, and appears to be true slate. It seems clear that here there is sedimentary material, but it is difficult to draw a line between the sediments and the greenstones. It is to be noted that the area in which the sediments are foimd is 2 miles from the Laurentian granite. The existence of iron fomiation is reported in the Keewatin area near ^larenisco. This, presumably, is analogous to the iron formation belts so common in the Keewatin in other parts of the Lake Superior region. It has not been examined l)y the authors. LAURENTIAN SERIES. The Laurentian granite occurs m three large areas — in the western, central, and eastern parts of the district. The granites of these areas, like all the other granites of the Laurentian of the Lake Superior region, vary greatly in chemical composition, mineral content, and struc- ture. In general, in the district under discussion the granites are of a somewhat acidic type. However, in the central area, besides the granites there are syenites and even gabbros, and the three rocks seem to grade into one another. Structurally the granites range from rocks which have comparatively little schistosity to those which in general are strongly gneissoid. Aside a Bull. U. S. Geol. Survey No. 80, 1892, p. 490. CAMBRIAN KEWEENAWAN SERIES LEGEN D ALGONKIAN HURONIAN SERIES ARCHEAN LAUftENTIAN SERIES KEEWATm SERIES PENOKEE-GOGEBIO IKON DISTRICT. 227 from the various feldspars and quartz, the most abundant minerals are the micas and horn- blende. There are other subordinate minerals, of which magnetite and chlorite are important. In the dominant, more acidic phases of the rocks the alkaline feldspars, comprising orthoclase, microclme, and acidic plagiociase, are invariably the cliief constituents and in many places com- pose as much as three-fourths of the rock. The gneissoid varieties of the Laurentian may be in part metamorphosed forms of granite. Correlative -with the structural changes are important mincralogical changes. The most mteresting is that by which the feldspars alter into biotite and quartz. Wliere this process has gone far little or no feldspar remains, this mineral being replaced by a fuiely crystallme interlockmg mass of quartz and biotite. This results in a somewhat coarsely crystallme feldspathic rock (normal granite), changing into a finely crystal- line gneissoid biotite-quartz rock. It is interesting to note that identical changes of a feld- spathic fragmental rock in the Tyler slate have formed a mica schist. RELATIONS OF KEEWATIN AND LAURENTIAN SERIES. The fact has already been mentioned that the Laurentian granites intrude the Keewatin schists. It is characteristic for the district that with approach from the Keewatin rocks to the contact of the Keewatm with the Laurentian granite the former rocks become coarser and finally grade into coarse gneisses, not very different from granitoid gneisses. In many jDlaces the granites are foxmd to cut through the schists in dikes and stocks. Indeed, there is between the two series usually a zone of considerable breadth in wliich the two rocks are in approximately equal proportions. In placing the boundary line between the series on the maps the plan has been to mclude m the Keewatin all those rocks the hand specimens of wliich do not have a strong granitic appearance. The relations between the two are plaiiily those which so charac- ' teristically obtaui between the Laurentian and Keewatin. The former rocks are batholithic intrusions in the latter and have cut them intricately. Along the border the granites have pro- foundly metamorphosed the Keewatin, producing marked exomorphic effects, so that the most altered varieties of schists approximate the character of the granite. ALGONKIAN SYSTEM. HURONIAN SERIES. LOWER HURON IAN. The lower Huronian in the Penokee-Gogebic district is represented only by the Smuday quartzite and the Bad Iliver limestone. SUNDAY QUARTZITE. Liihology and distribution. — The Smiday quartzite is so named because of its exposures east of Sunday Lake. It may prove to be the same as the Mesnard quartzite of the Marquette district, but in the absence of defuiite proof that it is the same formation the new name Sunday is here introduced for it. The only known exposures of the formation are those a short distance east of Little Presque Isle Iliver and those near the Newport mine. The former are rather extensive and the latter are small. Probably this quartzite is coextensive with the Bad River limestone, although it is not usually exposetl. Wherever the Bad River limestone occurs there is room between it antl the underlying Archean for the Smiday quartzite to be present. East of Presque Isle River the formation is mainly quartzite, with a thickness of at least 150 feet. Below the quartzite is a basal conglomerate, the fragments of wliich are largely derived fi-om the immediately underlying Keewatin schists. This conglomerate for the most part is but a few inches thick, but in places it has a thickness of 10 feet. The dip of the quartzite is about 30° N. Near the Newport mine the Sunday quartzite is represented by a thm belt of conglomerate clinging to the face of the granite. This conglomerate contains different kinds of granite, 228 GEOLOGY OF THE LAKE SUPERIOR REGION. porphyry, and various basic rocks. From tlio relations of tliis conglonacratc to tlie Palms formatiojT it is believed to be tlic equivalent of the fonfjloiuerate east of Presquc Isle River. Relations to adjacent formations. — The relati(jns of llie Sunday quartzite to the underlj^ing formations, and especially to the Keewatin east of Presque Isle River, show that there is a great unconformity between them. The actual contact between the two is beautifully exposed for some distance. The scliistosity of the Keewatin abuts against the bedding of the quartzite at various angles up to perpendicular. The Keewatin had been formed, metamorphosed, and denuded before the deposition of the conglomerate. The Sunda}' quartzite grades upward into the Bad Kiver hmestone. BAD RIVER LIMESTONE. Distribution. — The Bad River hmestone is so named because of its occurrence at Bad River in the Penokee Gap section. The formation is present at several localities in the western part of the district, at one place iii the central part, and in one area in the eastern part. The eastern area shows the most extensive exposures of the district, the formation here being continuous for several miles. Wlierever the formation is found it strikes approximately par- allel to the formations of the upper Huronian, and the dip is always to the north, being as high as 70° or 80° in the western part of the district and as low as 30° in the eastern part. Lithology. — The formation is called a limestone because that is the predominant rock. The limestone is heavily magnesian and in places approaches a dolomite. It commonly bears sUicates, of wlxich tremolite is the most abimdant, but chlorite and sericite are not rmcommon. The rock is very sihceous. The coarsest varieties of the silica are quartz, but chert is more common. In many places the silica is closely intermingled with the dolomite. In other places it occurs in bands varying from a fraction of an inch to a much greater width, antl in one place a band of siliceous material 45 or 50 feet wide was observed. Thus the chert and limestone are intermingled and iiiterstratified. The cherty limestone is a water-deposited sediment. Whether the original carbonate was of chemical or organic origin we have no definite evidence, but there is no more reason to suppose that life was not concerned in the deposition of tliis cherty hme- stone than of those of later age. MetamorpMsm. — The Bad River limestone has been much metamorphosed since its deposi- tion. During its metamorphism the silica recrystalhzed. It was concentrated into bands. It was rearranged into veinlike forms. During these changes a part of the silica may have been introduced from an extraneous source or at least from parts of the formation now removed by erosion. The abundant tremolite is evidence that the metamorphism took place under deep- seated contlitions when the silica united with the calcium and magnesium to form sihcates, the carbon dioxide being released at the same time. This is an anamorphic change which took place with decrease of volume. Relations to adjacent formations. — The relations of the Bad River hmestone to the Simday quartzite have already been considered. It is probable that everywhere it grades down into this formation, but whether it does so or not the distribution of the limestone at various phices along the southern border of the Huronian, with a strike parallel to the upper Huronian, thus contrasting strongly with the varying strikes and dips of the green scliist and gneisses, leaves no doubt that between the Archean and the Bad River limestone there is a great imconformity. Indeed, as chemical sedimentation at several points for a distance of 60 or 70 miles followed so promptly after the burial of the southern complex below the sea, it appears probable that when the limestone was laid down the Archean was reduced to an approximate plane. The lack of continuity of the limestone formation is due to the erosion which took place after its deposition before the lowest member of the upper Huronian was laid down. Evidences of this erosion are given under the desciiption of the relations of tlie Palms formation to adjacent formations. If formations later than tiie Bad River limestone belonging to the lower Huronian were depos- ited, they were removed by erosion before the deposition of the upper Huronian, as was the larger part of the Bad River hmestone itself. The limestone above the quartzite in the western area has a thickness of at least 200 feet, and to the west the tluckuess is not less than 300 feet. PENOKEE-GOGEBIC IRON DISTRICT. 229 UPPER HURONIAN (aNIMIKIE GROUP). GENEKAL STATEMEITT. The upper Huronian comprises the Pahiis, Ironwood, and Tyler formations. These formations extend continuously from Presque Isle River, east of Sunday Lake, several miles west of Bad River. They constitute a northward-dipping monocUne. Tliis monochne has various minor pUcations which give local variations to the strikes and dips, but they are neither abrupt nor large, the extreme variations in strike usually being between N. 60° W. and N. 60° E. At various places there are cross faults, the most notable of which are those at Penokee Gap, with a throw of at least 900 feet, at Potato River, with a throw of 280 feet, and west of Sunday Lake. Detailed studies of the iron-bearing formation, made in connection with the exploitation of the iron ore, show the presence of very numerous small transverse faults as well as numerous longitudinal faults, with hades parallel to the bedding, or nearly so. The latter were detected by the displaced dikes. Part of the faulting was prior to Keweenawan extrusions because it does not displace the Keweenawan. A notable instance of tlxis appears in the great transverse fault just west of Sunday Lake. Other faults are clearly post-Kewee- nawan, for they affect both Huronian and Keweenawan beds. PALMS FORMATION. Distribution. — The Palms formation is given tliis name because it occurs in typical develop- ment south of the Palms mine. It comprises the lowest of the upper Huronian rocks of the Penokee-Gogebic district. It constitutes a well-marked zone traceable tlirough its entire extent, except in the volcanic area at the east end. It strikes on the average about N. 70° E. Its dip is everywhere north, varjang from about 40° to 75°, the usual dips being between 55° and 65°. For the larger portion of the district the formation is 400 to 500 feet thick, but east of Sunday Lake it is tliicker, the maximum being 800 feet. Lithology. — The Palms formation consists of three members, of which the lowest is a thin layer of conglomerate, the central and dominant mass of the formation is a clayey slate, and the uppermost is a quartzite. The conglomerate is generally less than 10 feet thick and in many places is not more than 1 to 3 feet. The quartzite layer at the top is about 50 feet tliick. The conglomerate varies with the character of the rock with which the Pahns formation is in contact. Where it is next to the Bad River limestone, as would be expected, there are in it very abundant fragments of chert and hmestone, but with these are also granite, gneiss, and schist from the Archean. Where the contacts are with the Keewatin, as at Potato River and the west branch of Montreal River, the dominant fragments of the conglomerate are derived from the schist. W^here the formation is in contact with the granite, as in the central part of the area, the dominant fragments are from tlus formation, but in places — as, for instance, south of the Palms mine — with these fragments there are also pebbles of jasper, chert, and quartz. The central part of the formation is a pelite. It has many facies, varying from fuie- grained clayey slates through novacuUtes to graywackes. For the most part the alterations through wliich the pelite has gone are mainly metasomatic ones, such as quartz enlargement and the alteration of the feldspar to other minerals, especially biotite, chlorite, and quartz. In the western part of the district the feldspathic alteration and recrystallization are sufficiently important so that in places the rocks have become chloritic and biotitic slates. This greater metamorphism is doubtless connected with the intrusions so characteristic of this part of the district. For the most part there seems to be Hthologic correspondence of the main mass of the slate with the immediately underlying Archean rocks, the slate being substantially the same whether north of the Keewatin schists or north of the Laurentian granite. The upper part of the formation is a psammite which has been indurated by the process of cementation to a clean, typical, vitreous quartzite. As this quartzite approaches the over- lying iron-bearing formation it becomes stained with oxide of iron and at the contact it is commonly of a deep brownish-red color. 230 GEOLOGY OF THE LAKE SUPERIOR REGION. Relations to adjacent Jmrnations. — In giviiit; (he relations of tlie Palms to the inferior forma- tions it is necessary to consider separately its relations with the Bad River Jinicstone of the lower Huronian and -with the Archean. The fact that where the belt of conglomerate at the base of the Palms formation lies above the Bad River hmestone it bears much detritus from that hmestone sliows that the limestone after deposition became indurated and was eroded before Palms time. In general, the strikes and the dips of the two formations are approximately parallel, as are tho.sc of corre- lated formations in the Menonunee district, but it is plain that the erosion was sulhcient to remove the major portion of the Bad River limestone and also any later formations that may have been deposited in the lower Huronian. The lack of marked discordance in the bedding of the Bad River limestone and the Pahns formation is no evidence that the time gap between the two was not long enough to have produced a pronounced discordance elsewhere, for the Penokee district at tliis time may have been distant from areas of important folding and thrusting wliich elsewhere may have occurred. Between the Palms formation and the Archean there is a great unconformity. The proofs of tliis unconformity may be summarized as follows: First, the Palms formation and the other sedimentary formations of the upper Huronian strike with considerable uniformity across the country, being here in contact with one variety of the Archean, there with another, everywhere keeping their course, nowhere being penetrated or interfered with by any of the Keewatin or Laurentian rocks, whether scliists, gneisses, or granites. Second, the Archean rocks are either massive ones wloich are presumably igneous or schists and gneisses in which the extreme of foliation and crystalline character is found, whereas the overlying upper Huronian rocks are plainly water-deposited sediments. Third, in a dozen places or more above the Archean are basal conglomerates or recomposed rocks which show the unconformable contacts. The detritus in each place is dominantly the same in character as the rock on w4iich it rests. Where the inferior rock is granite it must be inferred that deep erosion must have exposed it at the surface prior to the deposition of the conglomerate. \^Tiere the basement rocks are Keewatin green scliists their foliation had been tleveloped and has been truncated before Pahns time. This is well illustrated at Potato River, where the conglomerate contains large* flat fragments of green schists which have their scliistosity lying parallel to the bedding of the Pahns, which is at right angles to the scliistosity of the Keewatin below. Fourth, the horizons of the upper Huronian with wliich the Archean is in contact are witliin a zone not more than 300 or 400 feet thick at most. Tliis is the clearest sort of evidence that the underlying rocks were reduced to a peneplain before the beginning of the deposition of the Palms formation. From the foregoing fact it is clear that the break between the Palms formation and the Archean is profound. It included the time represented by the unconformity between the lower Huronian and the Archean, the time retjuired for the (lep<3sition of the lower Huronian, and the time between the lower Huronian and the Palms formation. IRONWOOD FORMATION. Distribution. — The Ironwood formation was given tliis name from the fact that near the town of Ironwood it is well developed, and in this vicinity occur the more important mines. The formation is coextensive in its distribution with the underlying Palms formation. Its strike and dip are conformable with those of the Palms. The belt lor the greater part of the district has a breadth of 800 to 1,000 feet. West of Sunday Lake the surface width of the formation is greater" and north and east of Sunday Lake the belt is narrower. Faults cross and follow the bods. These aflFect the distribution of the ores and the iron-bearing formation, as described on page 237. The average tliickness of the formation is about 850 feet. In the extreme eastern part of the district, where volcanic action prevailed through much or all of upper Huronian time, the Ironwood formation is broken into tliin and impure belts. West of Sunday Lake it is divided into two or more belts by intercalated (luartzito and fjuartz slate beds. In other parts of the district, notably near Upson, the formation is divided b}* slate layers. In the main, in the western part of the district, except for the gaps whore the streams o Recent work seems to show this widening to be due to pre-Keweenawan overthrust folding and faulting from the west. PENOKEE-GOGEBIC IRON DISTRICT. 231 break through it, the Ironvvootl formation is a continuous ridgo, and it was tliis range which first attracted the attention of explorers at Penokee Gap and vicinity. In the central part of the district the formation is softer and the prominent features are made by the Archean rocks to the south. Still farther east, beyond Sunday Lake, the Ironwood formation again consti- tutes prominent bluffs. LitJiology. — The Ironwood is the iron-bearing formation of the district. In the memoir on the Penokee iron-bearing series (Monograph XIX) it w-as simply called the iron-bearing formation, without a geographic name. The greater portion of the formation contains more than 25 per cent metallic iron and there are considerable thicknesses in which the amount of iron averages 37 per cent. (See p. 238.) The ore bodies contain a higher percentage of iron. The Ironwood formation consists of four main varieties of rock — (1) slaty and commonly cherty iron carbonate and ferrodolomite, (2) ferruginous slates and ferruginous cherts, (3) actinolitic and magnetitic slates, and (4) black slates. Tlie iron-bearing carbonates are usually found only near the upper part of the formation, where they have been protected by the Tyler slate. The ferruginous slates and ferruginous cherts are characteristic of the central iron-producing part of the district, and the actinolitic and magnetitic slates are characteristic of the western and eastern parts of the district. The latter also form a belt 20 to 300 feet wide bordering the Keweenawan rocks on the north. In the intermediate areas there are of course gradations between the ferruginous slates and ferru- ginous cherts and the actinohtic and magnetitic slates, as there are also gradations between the cherty iron carbonates and the ferruginous slates and ferruginous cherts. Black slates form thin intercalated layers in the iron-bearing formation. Quartzite is also found in layers up to 100 feet thick well up from the base of the formation near Sunday Lake. The slaty and cherty iron-bearing carbonates are composed largely of iron carbonate and chert, but with these materials are various amounts of calcium carbonate and magnesium car- bonate. Recent reexamination has shown that in these rocks there are also subordinate amounts of greenalite. With these important constituents are other minor constituents, largely second- ary, such as limonite, magnetite, carbonaceous and graphitic matter, iron pyrites, and rarely fragmental quartz. The carbonate is both fine and coarse grained and both origmal and secondary. Coarse-grained recrystallized carbonate is especiallj' abundant near the contact of the Keweenawan in the Sunday Lake area. The cherty iron-bearing carbonate was the original rock of the iron formation. The origin of tliis class of rock is fully discussed in another place (pp. 499 et seq.) and therefore the subject will not be considered here. From it the ferruginous cherts and actinolitic cherts have been pro- duced. The actinolitic and magnetitic cherts were formed under deep-seated conditions largely through the influence of the Keweenawan intrusive rocks, and especially of the great western laccolith. These changes are anamoi'phic ones, which mainly took place in Keweenawan time. Tlie ferruginous slates and ferruginous cherts formed from the cherty iron carbonates by katamorphic changes largely in the belt of weathering and also in part in the belt of cementa- tion. These changes were mainly post-Keweenawan, after erosion brought the iron-bearing formation to the surface, and they have continued to the present day. Previously formed actinolitic and magnetitic rocks were in a much more refractory condition than the unaltered cherty iron carbonates and have been little affected by the alterations of the zone of katamor- phism. The ferruginous slates and ferruginous cherts have silica as their predominant constituent in various forms of crystallization, from amorphous through partly crystalline and chalcedonic material to finely crystalline quartz. With the silica are the various oxides of iron. Hematite and brown hydrated hematite are especially prevalent. Limonite is common and some mag- netite occurs. Where the hematite is in large quantity, to the exclusion of the hydrous oxides, the rocks are genuine jaspers; but this variety is rather unusual in the district. The rocks vary in their stratification from the regular lamination of a slate to irregularity. In many places the laminae have the appearance of having been ilisrupted and recemented. 232 GEOLOGY OF THE LAKE SUPERIOR REGION. The actinolitic, griineritic, and magnetitic cherts and slates, like the rocks of the second variety, have quartz as their (h)ininant constituent. Tliis f|iiartz is crystalline throughout and clearly nonclastic. The actinolite varies in amount from a verj' little to a constituent of great prominence. The iron oxides are mainly in the form of hematite and magnetite. The black slates are carbonaceous fragmental slates in la3'ers in the iron-bearing formation. These exceptionally form the foot wall of the ore deposits. (See p. 242.) Relations to adjacent formations. — The Ironwood formation rests conformably upon the Palms formation. The change from the clastic quartzite to the nonclastic iron-beaiing formation is astonisliingly abrupt. Generally it can not be said that there is any evidence of the transition between them. Locally a thin conglomerate marks the contact. For some reason the clastic deposits of the quartzite ceased and the nonclastic deposits of the Ironwood formation began. Above, the Ironwood formation passes gradually into the Tyler slate. TYLER SLATE. Distribution. — The Tyler slate was given its name from the t3'pical occurrence of the formation along Tylers Fork. It extends from a point about 6 miles west of Bad River nearly to Sunday Lake — that is, it is confined to the central two-thirds of the district. In breadth the formation varies up to 2^ miles at Tylers Fork. The strike of the formation is parallel to that of the iron-bearing formation below. Its dip is also similar to that of the iron formation. At this wider part its dip is from 70° to 75°. It apparently follows, therefore, that for the central part of the district — that is, from Bad River to Montreal River — this forma- tion has a tliickness ranging from 7,000 to 11,000 feet. It is plainly the great formation of the district. It is probable that minor plications partly explain this apparent tliickness. Lithology. — Study of the formation as a whole shows that it is dominantly a pelite but locally it is a psammite, including both arkoses and feldspathic sandstones. There is a general connection between the character of the rocks to the south and those of the slate belt adjacent. The greater part of the belt has received its material in part from the granitic and in part from the schistose areas; the part of the belt west of Penokee Gap has received nearly all its material from the syenitic granite to the south and west. The different varieties of rocks of the Tyler slate may be grouped under three heads — (1) mica schists and mica slates; (2) graywackes and graywacke slates; (3) clay slates or phyllites. Each of these main types has the various phases shown by the follo\\ang tabulation: Mica schist and mica slate: iMuscovitic. Biotitic. Musco\'itic and biotitic. Micaceous and chloritic fChloritic and biotitic. IChloritic and sericitic or muscovitio. Graywacke and graywacke slate: Micaceous jBiotitic. iBiotitic and muscovitic. Micaceous and chloritic Chloritic and biotitic. fChloritic. Chloritic JMagnetitic and chloritic. [Ferruginous and chloritic. Clay slate fChloritic. \Chloritic and magnetitic. It is not necessary to describe in detail the different varieties of these rocks, except as to their alterations. Metamorphism. — In the monograph on the Penokee iron-bearing series the alterations of this slate are discussed. "^ It is there shown that each of the varieties of rocks mentioned above has developed from pclitcs and psammites almost wholly by mctasomatic changes witliin the formation itself, without the addition or subtraction of material from an extraneous source. a Hon. U. S. Geo!. Survey, vol. 19, 1892, pp. 332-345. PENOKEE-GOGEBIC IRON DISTRICT. 233 In general, the eastern part of tlie formation is less altered than the western part. Here the prevailing rocks are clay slates, graywackes, and graywacke slates. From tlie central to the western part of the district the rocks become more crystalline, and at the extreme west end, especially west of Penokee Gap, only mica slates and mica schists are found. Where the rocks are much metamorphosed conlierite is sparingly developed. The parts of the Tyler slate wluch contain large fragmental particles of quartz are those in which the clastic character is easiest to recognize, for the grains of quartz everywhere remain in their entirety. It may be and indeed it is usually true that they have undergone a second growth and have thus become angular; but generally the original cores are easily dis- covered. In the nearly pure feldspar sediments, on the other hand, where the feldspar has changed to other minerals, it is more diilicult and in specimens of the most crystalline mica schist impossible to make out the original fragmental character of tlie rock. On the whole, the major modifications of the formation are those of the zone of anamorphism rather than the zone of katamorphism. This is what would naturally be expected, for at the time these alterations took place the rocks were buried to an unknown deptli below the overlying Keweenawan rocks. As the processes by which a clastic rock alters into a fine-grained crystalline mica schist were first described in detail with regard to the Pcnokee-Gogebic district," the principles involved in the development of this particular rock will be summarized here. As already indicated, the setliments from which the mica scliists were derived were very feldspatliic. Without going into details, the process which has resulted in the development of mica schists has been the alteration of the feldspar into mica, both muscovite and biotite, with the simultaneous separa- tion of quartz. For the change into muscovite the feldspar itself contains all the necessary constituents. For the change into biotite a certain amount of iron and magnesium are neces- ^ry. For the iron it is not so clifRcult to account, as the sediments are ferruginous. In some f^aces also the sediments contain more or less carbonate, and doubtless from tliis source has been derived at least a part of tlie necessary magnesium. At the tune of the recrystallization tlie newly formmg mica flakes developed with a parallel arrangement. At the same time the quartz recrystallized. The total result was to produce from a somewhat coarsely crystalline arkose a finely laminated mica slate or mica schist. The Penokee-Gogebic district is an exceptionally good one in which to work out the changes from the little-altered pelite to a mica schist, because of the very gradiial change in the amount of alteration in passing from the central to the western part of the district. At the time the Penokee-Gogebic monograph was written no reason was assigned for the crystalline character of the rocks at the west end of the district. Later studies on meta- morphism have led us to connect tliis alteration with the great laccoHth of the Keweenawan gabbro, wliich, in the western part of the district, occurs m contact with and cutting the Huronian rocks. The intrusion of tliis rock essentially parallel to the bedding would result in great pressure, as well as in raising the temperature, and it was under these conditions that the recrystallization took place. The absence of similar alterations m the central and eastern parts of the district is explained by the fact that there immediately overlying the slate are the surface Keweenawan lavas, which are locally interstratified with sandstones. It is plain that the alterations of the pelites to mica slates and mica schists took place in Keweenawan time. Relations to adjacent formations. — The Tyler slate rests conformably on the iron-bearing Ironwood formation. It is overlain unconformably by the rocks of the Keweenawan series. UPPER HTJRONIAN (ANIMIKIE GROUP) OF THE EASTERN AREA. In the eastern part of the district — that is, from about 6 miles east of Sunday Lake to Gogebic Lake — the upper Huronian rocks have an exceptional character. In the larger part of the district the conditions were those of quiet sedimentation, but in this eastern area o Van Hise, C. R., Upon the origin of the mica scliists and black mica slates otthe Penokee-Gogebic iron-bearing series: Am. Jour. Sci.,3d ser., TOl. 31, 1886, pp. 453-459. t> 234 GEOLOGY OF THE LAKE SUPERIOR REGION. throiipjliout the greater part of the upper Iluroiiiaii there was eontinuous volcanic action. In conse()uence the rocks are hiva flows, volcanic tull's, conglomerates, agglomerates, and slates, with all sorts of gradations, just such as one would expect if a volcano arose in a sea and volcanic action continued for a great period. Naturally in this area it is not possible to map any continuous sedimentary belts. The dominant rocks are greenstone conglomerates and lavas and massive eruptives. The uppermost formation for the extreme eastern part of the area is a ferruginous slate. This ferruginous slate, though dominantly clastic, contains narrow bands of nonclastic sediments, such as chert, chertj^ ferrodolomite, ferrodolomitic chert. It is believed that the ferruginovis slate is probably at the same horizons as the Ironwood formation to the west and that its dominant fragmental character is due to the presence in tliis area of one or more volcanic mountains which rose above the water and upon wliich the waves were at work after the close of the period of active volcanic outbreaks. KEWEENAW AN SERIES. GENERAL DESCRIPTION. Rocks of the Keweenawan series lie north of and are coextensive with the upper Huronian rocks; indeed, to the west they extend far beyond tlie westernmost kno\\'n outcrop of the Huronian. It is not the purpose here to describe this series more than is sufficient to show its relations to the Huronian. It has already been indicated that for most of the district the appearance of the Keweenawan is marked by a distinct range Ivnown as the Trap Range. For the eastern part the Keweenawan rocks first encountered in traveling north are ordinary basic, amygdaloidal lava flows characteristic of that series. One bed follows upon another and it is easy to ascertam then strike and dip. These bedded lava flows may be very conven- iently seen adjacent to Sunday Lake. Their strike and dip are easily determinable as almost exactly parallel to the beds of the underlying Huronian. In the central part of the district the Keweenawan rocks immediately above the Tvler slate are sandstones and conglomerates. These are seen in Micliigan north of Bessemer and in Wisconsin a few miles west of the State boundary. Above the sedimentary beds of the lower Keweenawan follow lavas similar to those which occur farther east. In the western part of the district the sediments and bedded lavas of the Keweenawan are replaced by the great plutonic basal gabbro laccolitli of Wisconsm, analogous to the laccolith of the north shore of Lake Superior. RELATIONS TO ADJACENT SERIES. The Keweenawan series reposes upon tlie upjier Huronian (^Vnimikie group) imconformably. As the two series are nearly conformable in strike and dip, this fact was only slowlj^ appreciated. The proof of the unconformity rests entirely upon broad field relations. In the central part of the district the Keweenawan is upon a great slate formation (the Tyler slate) wMch has a maximum tliickness of at least several thousand feet. At the east and west ends of the ilistrict the Keweenawan cuts diagonally across these slates and comes into contact with the iron- bearing Ironwood formation. In the west end of the district this relation might be supposed to be explained by the intrusion of the Keweenawan laccolith, but this can not apply to the eastern part of the district, for there the lower beds of the Keweenawan are the surface lava flows. The time gap between the Huronian series and the Keweenawan series must have been sufficient for a widespread orographic movement and tleep denutlation. As the Keweenawan series is largely composed of igneous rocks and rests upon the Huronian series, naturally the latter has been extensively intruded by the former. The intrusives in the Huronian series, so far as known, are mainly doleritcs. ('onsideral)le masses of them in tlu' eastern and western ends of tlic district appear to follow rouglily parallel to the range and PENOKEE-GOGEBIC IRON DISTRICT. 235 seem to be intruded sheets or laecoliths. Some of them may be surface flows contemporaneous in origin with adjacent Iluronian sediments. In addition to those intercalated masses, numer- ous dikes cut the Huronian formations. These dikes are found in all formations, but they have an especial significance and importance in connection with the iron ore. (See pp. 235-238.) In that part of the district which has been the seat of mining operations a large number of these dikes cut the containing formations perpendicularly to the bedding. That these dolerite ilikes are the avenues tlu'ough wliich have passed from deep witliin the earth the vast amount of material which formed the overlying basic volcanic flows of the Keweenawan series of the Trap Range to the north can hardly be doubted, for in chemical composition the lavas of this range are practically identical with the dikes. (See pp. 404-405.) In general, the dolerite dikes are very fresh, except in the lower parts of the Ironwood formation, where they have been subject for a long time to the action of percolating waters. Analyses of the latter rocks show that they have undergone extensive changes, which have been referred to in cormection with the origin of the iron ores. By far the greatest of the intrusive masses is the great gabbro laccolith at the west of Bad River. This was at first supposed to be a great basal flow, but all their later studies lead the writers to believe that it is a plutonic intrusive introduced comparatively late in Keweenawan time, the major dimensions of the mtrusion being nearly parallel to the beds of the Huronian and the lava flows of the Keweenawan, which were separated by the inwelling mass of gabbro. CAMBRIAN SANDSTONE. The Cambrian sandstone is found only in the northeastern part of the district, near Gogebic Lake. It is there found as a flat-lymg reddish sandstone, known as the Lake Superior sand- stone. It rests in horizontal position against the Keweenawan, the Huronian, and the base- ment complex. In one place a basal conglomerate bears detritus from all the lower forma- tions. It is plain that during and after Keweenawan time the Huronian and Keweenawan series were turned up steeply. Lofty ranges, which must have been formed then, were removed by denudation, and the Cambrian sandstone was deposited. Therefore a great unconformity separates the Cambrian sandstone and all the earlier series. THE IRON ORES OF THE PENOKEE-GOGEBIC DISTRICT. By the authors and \V. J. Mead, DISTRIBUTION, STRTJCTUBE, AND RELATIONS. The iron-ore deposits occupy part of the district, extending from a point about 2 miles east of Sunday Lake in Michigan to within 4 miles of Potato River in Wisconsin, a distance of about 26 miles. Ore has recently been developed in sections 15 and 21, T. 47 N., R. 43 W., near Gogebic Lake, far to the east of the previously kno\vn deposits. The iron-ore deposits constitute about 1 per cent of the area of the iron formation. This percentage is less than that in the Mesabi, which is 8 per cent. However, the vertical dimensions are much greater than in the Mesabi district. Ores have already been found to extend to a depth of 2,500 feet, one of the largest ore bodies in the district now being known at that depth. In both the east and west ends of the Gogebic district the character of the formation has been influenced by intrusives, with the result that the iron oxides are largely magnetite dis- seminated tlirough the formation and not concentrated to a commercial extent. The ore deposits come to the surface most largely along the north-middle slopes, locally on the lower slopes, of the topogi'aphic feature known as the Gogebic Range. In the Gogebic district the u'on-bearmg Ironwood formation dips with the other forma- tions of the upper Huronian toward the north at angles averaging about 65°. The under- lying rock is the quartzite, at the top of the Palms formation, and it thus forms the foot wall of the iron-bearing formation; the overlying formation is the Tyler slate. Slate and quartzite 236 GEOLOGY OF THE LAKE SUPERIOR REGION. also form Lnterbedded layers in the iron-bearing formation. Numerous greenstone dikes oi Keweenawan age cut the entire series in such a manner that the intersection of the dikes with the bedding usually pitches to the east. The intersections of the dikes with imjxTvious layers, principally cjuartzite of the foot wall, but also slate layers within the iron-bearing formation, constitute eastward-pitching troughs at angles from 15° to 30°, in wliicli most of the deposits are found (fig. 2S). Westward-pitching dikes intersecting with eastward-pitching dikes or continuous with them, and both intersecting foot-wall quartzite, form canoe-shaped basins for the ore, as illus- trated by the Aurora, Pabst, and Newport deposits. East of the Bessemer the main dike in the Tilden mine has an eastward pitch and the main dike in the Pahnsmine has a westward pitch. On the south the foot wall of the ore is therefore generally quartzite, locally slate, and on the north the ore rests against greenstone dikes. Slate foot walls are seen at the Iron Belt, Mikado, Brotherton, and Sun- day Lake mines, from 250 to 2,000 feet north of the quartzite foot wall. The ore deposits are generally sharply defined along the foot walls and the dike rocks, but in many places vary upward by imperceptible stages into the ferruginous cherts of the iron-bearing Ironwood forma- tion, ^^liere there are a number of parallel dikes, one below the other, there may be several ore bodies one below another — as, for instance, at the Asldanil and Norrie mines. (See fig. 29, a and h.) After many years of mining on upper dikes in the New- port mine one of the largest deposits of the district has been found on a lower dike. The main Norrie dike is over .30 feet thick. The main Aurora-Pabst-Newport dike is from. 20 to 25 feet thick. The mam Colby tlike is over 90 feet tliick. "\Miere a strong dike breaks intomany stringers at a depth, as in the Colby mine, the ore body is also likely to be broken up and become small and perhaps worthless. The dike rocks are altered to soapstone or paint rock along tlieir contacts with the ore l)y tlie leaching of the bases. An ore deposit is likely to have its maximum depth m the apex of a trougli, and from tlus apex a belt of ore may extend to the north along the dike and to the south along the foot wall. In many instances the ore bodies follow the foot walls almost exclusively, as at the Norrie mine. (See fig. 29, c.) Usually where the deposits follow both the quartzite and tUkes the former is larger and more continuous than the latter. Where an ore deposit follows both it may divide before reaching the surface into two parts separated by rock, called the south and ifc 100 JOO 400 Feet FlGUKE 28.— Cross section showing the occurrence of ore In pitching troughs formed by (lilies and quartzite foot wall, in the Gogebic district. Made up from min e plats and slightly generalized. PENOKEE-GOGEBIC IRON DISTRICT. 237 north veins of the mines, but where such deposits are traced below the surface they unite into a sino-le body. The ore grades above or laterally into the ferruginous chert or ferruginous slat«. The conspicuous associa- ^ =• » S g-^ g 5P 2 'S S I I I g & s B' ■■ -"" fi 1 '^ E. ^. -^ ^ lo = » a s- 2 S P) cr n*. ^ fT c - ■ - o ^ -o — . p- o li B s> rV 1 s ffi 2". **j fB O "1 p o a o B fr -•a S & B- E O < 5 S >^ g- " 1=^ I' - ^ ft M ™ 2, S o s "• "> a ° c P £. o "a "• t3 B g. '^ ^ 3 .? Si O ' rr ^ ^ w O 7Q to ^ ^ ■ p cr c -- §■ 1 1 o ^ s- ^ 2, 3- 5 cr ^ ~ ? p Q T E _ ? S a- „ &J I— - f^ D- P o ^ ' •- m :? p O f Pence - Hennepin Montreal Moore and Jo Bourne Ottawa Superior West Gary Gary Windsor Germania Minnewawa Ashland O o P 3 ^ p tion of the ores with pitclung troughs formed by the inter- section of dikes and foot-wall quartzite for a considerable time obscured the importance of fracturing in localizing the ore deposits. From evidence now available it seems likely that tlus factor also is of great consequence. An ex- amination of nune sections taken almost at random in the district shows ore cutting tlu-ough the ferruginous cherts and dikes in a most h-regular way and quite independent of the pitching troughs. Much of it may be directly con- nected with brecciation, fis- suring, or faulting to be seen in the ore and adjacent rocks. It is altogether likely that more fractures exist than are known, because the concen- tration of the ores is of a nature to obscure evidences of them. There are fault planes both parallel with and intersecting the bedding. The displacements have both hori- zontal and vertical compo- nents. The faults intersecting the bedding were first recog- nized because of the ease of detecting the displacements in the bedding. Those parallel to the bedding were for a long time not observed, and proba- bly there are still many to be detected because of the diffi- culty of distinguisliing the evidence. They may be deter- mined only from the displace- ments of the cUkes, and as the dikes are numerous and of varying tliickness a consider- able amount of ground must be opened up before the va- rious fractured dikes may be correlated. The ore in many places lies between the displaced edges of the dike. At the Pabst mine the ore follows down over the broken and displaced ends of the faulted dike toward another dike below, where it again develops into a large body. -I (t -■' o §-•00 3 S"S p P 2. ;rJ G ■a -a ^ S. a£ S P 2. T! o "2. C3 CO O P § S o S- '« I » ^ 52; ■g SB § g w P P ^ C^ s-B P- EO 5* {TO tr o CO OQ 5 &- B- 2 ^ i . a — a: SS 3 ■ o £•? U CO w H „ o- =5- : So . Tyler Fork Annie Shores iron Belt Atlantic Laura Caledonia Imion Emma and Daylight Nome East Norrie Aurora Aurora and Pabst Newport Wisconsin and Geneva Royal Puritan Ironton Winona Jackpot and Yale Colby g Tiiden Palms > 2 EureKa Black River Section 8 Chicago Pike Brotherton Sunday Lake Iron Cnief Castile 238 GEOLOGY OF THE LAKE SUPERIOIl REGION. Another important factor governing the location of tlie ore deposits has only recently been clearly recognized. Certain of the iron formation layers were originally ricJKir in iron than others, and the ores show a distinct tendency to follow these rich original beds. In some d;>posits, like those of tlic Mikado, Brothcrton, and Rvmday Lake mines, this seems to be a con- trolling factor, though the ores of the Brotherton and Sunday Lake mines and less certainly of tJic ^Vlikado min? have suffered more or less secondary concentration along intersections of dike and foot wall and along fissures. CHEMICAL COMPOSITION OF THE FERRUGINOUS CHERTS AND ORES. The following analj'ses represent two completL' sections through tlie iron-bearmg forma- tion. In the Norrie mine a crosscut, extending from foot-wall quartzite to the hanging-wall slate, entirely in ferruginous chert, was sampled in five samples, each sample representing approximately 120 feet of crosscut. In the Atlantic mine a crosscut m feiTuginous chert extend- ing for several hundred feet across the formation was sampled. Anaylses are by Lerch Brothers, Hibbing, Minn. Partial analysis of fcrruijinous chert, Gogebic range. [Samples dried at 212° F.) Fe. SiO". 1'. AljOa. Volatile matter. 35.33 23.39 30.03 27.02 26.81 29.20 43.78 61.22 51.80 54.57 54. 02 52.07 0.143 .034 .037 .040 .074 .037 1.54 .71 1.09 1.78 1.94 .88 1.42 .85 Norrie mine No. 3 . . . . 1.48 1.67 Norrie mine No. 5 . . . . . 1.64 2.89 28.74 53.11 .062 1.32 1.66 It is believed that this average represents closely the true average composition of the unaltered ferruginous cherts. A large part of the ferruginous cherts shows partial alterations to ore. An average of 490 analyses, representing 5,S90 feet of drilling in tliis phase of the formation, wliich is probably nearer to the true average of the formation, is 36.65 per cent. The average composition of the Gogebic ores for the years lOOfi and 1909, calculated from average cargo analyses for each grade, each analysis being weighted in proportion for the tonnage represented, is given in the following table : Average composition of ore mined on the Gogebic range in 1906 and 1909. Moisture (loss on drying at 212° F.). Analysis of dried ore: Iron Phosphorus Silica Alumina Manganese Lime Magnesia Sulphur Loss by ignition PENOKEE-GOGEBIC IRON DISTRICT. 239 Range in percentage of each constituent in Gogebic ores mined in 1909, as shown hij average cargo analyses. Moisture (loss on drying at 212° F.) 4. 51 to 15. 75 Analysis of di'icd ore: Iron 43. 70 to G3. 40 Phosphorus 027 to .206 Silica 4.07 to 23. 52 Alumina 58 to 3. 29 Mangane.se 20 to 7.20 Lime to .87 Magnesia 01 to .79 Sulphur ■. OOG to Loss on ignition . FERRIC OXIDE .022 .56 to 5.80 MINOR SILICA CONSTITUENTS Principally alumina and water of hydration Figure 30.— Triangular diagram showing cliemical composition of various phases of Gogebic ores and ferruginous cherts in terms ot ferric oxide, silica, and minor constituents (essentially alumina and combined water). These analyses include all of the ores and cherts shown in figure 32 and also a number of additional analyses. In figure 30 the triangular method of j)httting is employed to show the chemical composition of the various phases of the chert and ore studied. (See p. 182 for explanation of diagram.) 240 GEOLOGY OF THE LAKE SUPERIOR REGION. MINERALOGICAL COMPOSITION OF THE FEKKUGINOUS CHEKTS AND ORES. The approximate miiu'iiil comijositioii of the ores and clicrts wa.s cak'idatcd from tlie chemical analyses, as follows: Approriinate mineral composition of average ferruginous chert and average ore of the Gogebic range. Average chert. Average ore. 1909 Hematite (i lu-luding a small amount ot magnetite) Limonite (other hydrated iron oxides calcnlated as llnionite) Quartz Kaolin Other minerals 34.00 8.20 51.63 3.3S 2.82 73. SO 14.70 4.31 4.89 2.60 100.00 100.00 77.25 9.30 5.81 4.70 2.94 100.00 PHYSICAL CHARACTERISTICS. GENERAL APPEARANCE. The iron ore of the Gogebic district is a soft red, somewhat hydrated hematite. Much of it is so friable that it can be broken down with a pick, although as taken from the mines a large portion of it is compact enough to hold together in moderately large lumps. These lumps are porous, many of them more or less nodular, and many also roughly stratiform. The strata conform in a general way to the strike and dip of the iron formation. Mingled ^\'ith this soft hematite in a few mines is a small quantity of aphanitic hard steel-blue hematite, which breaks with conchoidal fracture and is of remarkable purity. In general, this exceptionally hard material is found in contact with or close to the diorite dikes of the mines. DENSITY. The specific gravities of the ores and cherts were determined by the two general methods already discussed (see p. 184) — (a) calculated specific gravity obtained by properh* combining the specific gravities of constituent minerals; (6) actual determinations by gravitj" methods. The specific gravities of the minerals as used in determinmg the mmeral specific gra^Tit}' of the ore or chert are as follows: Hematite, 4.5 for earthy ores and chert, 5.1 for crystalline and hard ores; limonite, 3.6; kaolin, 2.62; quartz, 2.65. The average mineral density of all ore mined in 1906, calculated from the above mineral analysis, is 4.33. Following are six analyses of ferruginous cherts and ores with specific gravities determined by both methods, as a check on tlie accuracy of determination: Density of individual cherts and ores determined by calculation and by measurement. Chemical composition. Mineral composition. Specific gravity. Fe. SiOa. P. AljOa. Volatile matter. Mn. Moisture of satu- ration. Hema- tite. Limonite. Quartz. Kaolin. Calcu- lated from analyses. Deter- mined bypyc- nometer. 3.30 30.30 43. 20 40. 80 52.00 63.40 «7.91 48.34 32.88 29.83 12.59 4.00 0.007 .027 .022 .013 .029 .078 6.81 6.30 4.40 1.32 7.40 2.94 0.41 1.S3 1.55 1.08 5.43 1.78 0.15 .50 .45 .85 .25 .70 3.70 9.30 3.35 10.50 7.15 10.50 4.71 43.30 61.70 63.30 57.70 86.20 79.91 17.25 2.73 3.38 3.80 3.83 3.89 4.79 2.68 3.38 3.809 3.89 3.90 4.74 43.24 27.70 28.28 3.89 .54 10.95 11.13 3.34 18.72 7.45 4.21 19.38 5.24 PENOKEE-GOGEBIC IRON DISTRICT. 241 POROSITY. Porosity was determined on hand s])ecimens by the usual method of saturation in water described on page 185. An average of ten determinations on typical specimens of ferruginous cliert gave 4.1 per cent pore space. The average of the porosity of all the ores examined was approximately 34 per cent. CUBIC CONTENTS. The ores vary in cubic content from 7. .5 cubic feet to the ton in the small masses of pure steel ore to 14 cubic feet in the softest yellow ores. The average calculated for the 1906 output is approximately 10.75 cubic feet to the ton. TEXTURE. The average texture of the Gogebic ores is shown by the following table of screening tests. Thes3 were made by the Oliver Iron Mining Company and represent all of the ores mined by that company in the Gogebic district in 1909. Samples of the different ores were taken twice a week, cpiartered down each month according to the tonnage shipped, and at the end of the shipping season quartered to 100 pounds of dry ore, on which the tests were made. The fol- lowing table represents 10 grades of ore, totaling 1,256,557 tons. The texture of the ore is seen to be similar to tliat of the ores of the Marquette district. A comparison of the textures of the ores of the several Lake Superior districts is shown in figure 72, page 481. Textures of Gogebic ores as shown by screening tests. Per cent. Held on J-inch sieve 28. 97 |-inch sieve 32. 30 No. 20 sieve. 16. 08 No. 40 sieve 8. 32 No. GO sieve 4. 03 No. 80 sieve 2. 56 No. 100 sieve 1. 89 Passed through No. 100 sieve 5. 92 MAGNETITIC ORES. At the extreme east and west ends of the Gogebic range the iron-bearing formation consists of dark-gray, green, or black dense crystalline banded rocks, consisting of magnetite, cpiartz, amphiboles, and other silicates in varying proportions in different bands and different localities. Ore deposits are rare or altogether lackmg. For a discussion of i-easons for this condition see pages 552-553. The average chemical composition of these rocks is as follows: Analyses of mngnetitic rocks fi Fe,03 44. 606 FeO 13.811 SiO„ 34. 616 ALO3 588 CaO 1. 802 MgO 2. 166 MnO 1.158 P2O5 018 S 083 HoO 997 99. 845 Metallic iron 41. 95 a Mon. U. S. Geol. Survey, vol. 19, 1892, p. 197. 47517°— VOL 52—11 16 242 GEOLOGY OF THE hAKE SUPERIOR REGION. Wlien this composition is compared witli that of tlic fcrruLciiious cherts of the Gogebic district it is apparent that there is ])ut link' differenco between the two. SECONDARY CONCENTRATION OF GOGEBIC ORES. STRUCTURAL CONDITIONS. The ores of tliis district are probably localized in bands of the iron formation wliich were originality rich in iron, but for most of the district secondary concentrations have so masked the primary distribution in bands that the evidence for it is not clear. Probabh' the clearest case is in the Mikado, Brotherton, and Sunday Lake mines, where the ores seem to follow certain originally rich horizons in the iron formation, the later concentration apparently not having seriously modified their distribution. The secondary alterations of the iron-bearing beds are accomplished (1) by waters follow- ing the pitching trough formed by the intersection of the dikes with impervious quartzite or slate beds below the iron formation layers, and (2) by following fissures or beddhig planes independent of the dikes. The control by the dikes is by far the most conspicuous one for the district as a whole. The movement of the concentrating waters is in general eastward toward lower levels, following the eastward pitch of the trouglis fomied by the intersection of the dikes with foot-waU quartzite or exceptionally foot-wall slate. The waters may thus be brought beneath other dikes. Tliis explains the common occurrence of ores on several dikes one below the other. The movement of the water is controlled to an important degree by bedtling planes, by faults, and by joints, and where so controlled the ores are more or less independent of the dikes and foot wall. The control by faults is especially well shown in one locahty where faulting parallel to the bedding has displaced tlie ends of a dike and the ore follo^vs over the broken end of the dike along the fault plane, obviously a zone followed by percolating waters. Faults and joints may give an eastward pitch of the ore bodies, for many of the fissures along wliich altera- tion takes place pitch in the same direction as the dikes; in fact, the dikes have been intruded along fissures of this kind. That some fissures were there before the intrusion of the dikes is shown further by the fact that the iron formation near Sunday Lake has been displaced by faultmg, whereas the Keweenawan igneous beds to the north, with wliich the dikes are genetically connected, have not been displaced. These early fissures also preceded the Keweenawan folding. If fissures were present in the rocks before the dikes, there is no reason win" some concentration should not have been prior to the intrusion of the dikes in the east and west ends of the district, where the cover of slate was not too great to prevent ingress of water; but evidence of tliis would be extremely difficult to detect because of later alterations since the dikes were intruded. The greatest depth to which the w^aters, and therefore the ore concentration, may be car- ried by the eastward-pitching trouglis or by the fault and joint planes is yet unknown. Large ore bodies have been found to a depth of more than 2,500 feet; one of the largest deposits thus far found in the district was recently developed at tliis depth. Theoretically the de])tli of con- centration is a function of the head detemiuied by the height of the erosion edge of the iron formation and the lowest point of escape; but the difficulty is to determine where the latter point is, for reasons stated above. Even if the head were known, there would be difficulty in calculating the effective depth of the circulation because the medium tlirough which it is flowing is not homogeneous. Further, if the depth of the active circulation could be worked out witliin reasonable limits, this would give us only the maximum depth of the ore deposits, for it might well be that the waters do not carry oxygen abundantly to the maximum dejitli to wliich they penetrate. Theoretically the concentration of the ore should be more effective on the middle slopes of the hills, because these would be places where descending waters are efl'ective, whereas valleys are places where the waters are ascending unless prevented by other structural condi- tions, and not so effective for the purposes of ore concentration. It is unlikel_y that each of the cross valleys should have the same control of the circulation, and it is difficult to tell which of the valleys has been most effective. Also it is to be remembered that the pitch of the dikes to PENOKEE-GOGEBIC IRON DISTRICT. 243 the east is greater than the surface slope and that tlierefore the underground waters, where passing under a valley, would be prevented from escaping b,y the overlying impervioiis dikes, except where faulting would allow the waters to come tlu'ougli. Mining operations actually disclose artesian flows through dikes, as at the Germania niiue. Also, ascending waters are actually observed to follow faults across the dikes, as in the Newport mine. From anything that is now known to the contrary, the faults in the tlikes may be sufficiently numerous to allow upward escape of the water somewhat freely along the cross valleys at the surface. This is especially likely in view of the fact that the cross valleys are observed to have developed along fault planes. These planes must cut the dikes, though some of them are not observed to do so. The cross valley under such conditions is simply the surface expression of the weak faulted zone. It is therefore not to be expected that there is a close relation to be observed between the topography and the distribution of the ores. The ore deposits extend below both eleva- tions and minor valleys, but at some of the principal cross valleys ore deposits are small or lacking. For illilstration, ores extend abundantly under Montreal River at Ironwood, but east of the Newport mine these ores seem to end at a pronounced cross depression northwest of Bessemer, through which Black River flows. It is thought by James R.Thompson, formerly manager of the Newport mine, that the drainage for the Ironwood-Newport group of mines is probably carried eastward and escapes through tliis channel. ORIGINAL CHARACTER OF THE IROX-BEARING FORMATION. Originally the iron-bearing formation consisted largely of cherty iron carbonate inter- layered with sideritic slates and possibly also with banded chert and ferric hydrates. (See p. 2.31.) Some layers were probably richer than others. The alteration of the cherty iron carbonates to ore has been accomplished in the general manner already described as ty|:)ical for the region — (1) oxidation and hydration of the iron minerals in ])lace, (2) leaching of silica, and (3) introduction of secondary iron oxide and iron carbonate from other parts of the forma- tion. These changes may start simultaneously, but change 1 is usually far advanced or com- plete before changes 2 and 3 are conspicuous. The early products of alteration, therefore, are ferruginous cherts — that is, rocks in wliich the iron is oxidized and hydrated and the silica is not removed. The later removal of silica is necessary to produce the ore. ALTERATION OF CHERTY IRON CARBONATE TO FERRUGINOUS CHERT. Chemical change. — The alteration of cherty iron carbonate to ferruginous chert involves the oxidation of iron according to the following reaction: 2FeC03 + nH,0 + O = Fe.Og.nH^O + 200^. Mineral change. — The cherty iron carbonate is practically identical mineralogically with the sideritic cherts of the Mesabi range. The constituent minerals are segregated into alternate layers of siderite and chert. The oxidation of the siderite involves a change to a heavier mineral. Either introduction or removal of silica may accompany this change. Volume change. — The volume involved in the alteration indicated in the above ecpiation is a loss of 49.25 per cent, considering the resulting iron oxide to be anhydrous. If hydration of the iron oxide takes \Aace, the volume reduction is smaller in proportion to the degree of hydration, being only 18.3 per cent when the product is limonite. Approximately 60 per cent of the volume of the cherty iron carbonate is silica; therefore the reduction in volume caused by the oxidation of the iron is efl'ective on approximately only 40 per cent of the volume of the rock. The loss in volume, then, for tiie entire rock, taking into account both iron and silica, ranges from 17.2 per cent to 6.4 per cent, depending on the degree of hydration of the resulting iron oxide. Development of ijorvsity.— The decrease in volume, due to the alteration of the iron minerals, develops pore space in the resulting ferruginous chert. Determinations of porosity on several typical specimens of cherty iron carbonate showed an average of less than 1 per cent pore space. 244 GEOLOGY OF THE LAKE SUPERIOR REGION. A series of ten determinations on typical specimens of ferruginous chert gave a range of 0.9 to 8 per cent pore space, with an average of 4.1 per cent. Evidently the actual porosity is not sufficient to iK'count for the tlieoretical volume change. This may be explained in the following ways: (a) Part of the iron oxide in the ferruginous chert may have been original and not altered from siderite. As the calculated pore space is based on the assumption that all of the iron oxide in the ferruginous chert is the result of the oxidation of siderite, original ferric oxide in the chert would decrease the resulting pore space, (h) Infiltration of iron oxide or silica subserjuent to or accompanying the alteration may have closed part of the openings formed. This is certainly true to at least a small extent, as shown by microscopic examination of thin sections, (c) The difficulty of obtaining saturation and perfect drying in the determination of porosity in the specimens of ferruginous chert may have made the results too low. (J) In the rocks under discussion, both original and secondary, the iron minerals tend to be seg- regated in parallel layers separated by comparatively barren chert. The volume changes in the alteration of the iron minerals would then be largely confined to the ferrugmous la3-ers. If these are assumed to be practically pure iron mineral, the cubical slmnkage should vary between 49.25 per cent and 18.3 per cent (as previously calculated) for the different original and secondary minerals noted above, the linear shrinkage between 6.5 and 20.3 per cent. The shrinkage normal to the layers would probably not result in openings to any large extent, as slumping of the flat layers would close any cavities formed, and as a matter of fact such openings are not observed. On the other hand, slmnkage parallel to the beds is taken to explain the common intersecting sets of cracks confined to the ore layers and breaking them into small parallelepiped blocks when the ore has not suffered general deformation. These by actual measurement give a volume of openings ranging from 12^ to 36 per cent of the volume of the iron layers, wliich would be approximately 5 to 14^ per cent of the volume of the rock. It is believed, then, that the increase in porosity and development of cracks in the ferruginous chert, together with the slump which has obliterated a part of these openings and the infiltration of iron salts, fully accounts for the change in volume which accompanies the production of these cherts from the cherty iron carbonate. ALTERATION OF FERRUGINOUS CHERT TO ORE. The alteration of ferruginous chert to ore is almost identical with the secondary concen- tration of the ores of the Mesabi district. As in the Mesabi concentration, the essential change is the leacliing of silica. The several possibilities resulting in the leaching of silica are dis- cussed on pages 537-538. It is seen that the space left by the removal of silica may remain as pore space and may be partly or entirely closed by slump or may be filled partly or entirely with infiltrated iron oxide. To determine the relative importance of these possibilities, quanti- tative methods similar to those employed in investigation of the Mesabi ore were used. In order to include the factor of porosity in a comparison of ores and cherts, it is necessary to consider their composition in terms of volume rather than of weight. The volume composi- tion of any chert or ore is readily calculated from the mineral composition and the porosity. The volume composition of the average ores anil ferruginous cherts is as follows: Average volume composition of ores and cherts of Gogebic range. Femigi- IIUUS cherts. Hematite.. Limonite... Quartz Kaolin Pore space . 37.30 U.9S 10.43 3.25 3-1.00 99.% 19.60 7.23 ti.OO 4.03 4.10 99.96 PENOKEE-GOGEBIC IRON DISTRICT. 245 The above volume composition is expressed diagrammatically in figure 31. The most important factor in forming ore from the cherts, as sliown by tlie diagram, is tlie removal of silica. I QUARTZ Finely crystalline quartz grading into ' amorphous forms in Iwth the cherty iron carbonat. ' and the ferruginous cherts 1 Reduction of pore SLUMP PORE SPACE -~_^rneehanical packing of ore by weight Porosity is first due to the decrease — ..^matenal above in volume accompanying the oxidation of iron carbonate and later to the removal of silica in solution Partially replacing volume occupied by iron carb onate alterine to clay] Secondary hydrous iron oxide Deposited by iron-bearing solutions from above HYDROUS IRON OXIDE The degrees of hydration of the iron o.vide in the ■ ferruginous cherts and ores may be expressed by ratios of hematite to limonite of -1 .1 and 5 :1 respectively Average ferruginous chert Average ore v^ Average cherty iron carbonate FiGtJEE 31.— Diagrammatic representation of the changes involved in the alteration of cherty iron carbonate to ferruginous chert and ore, Gogebic dfstrict. The mineral compositions of the various phases are indicated in terms of volume by vertical distances. The compositions of the cherty iron carbonate, ferruginous chert, and ore represented are averages of a large number of analyses. IRON MINERALS Average ore (Cargo analysis for 1906) Average ferruginous chert SILICA PORE SPACE Figure 32.— Triangular diagram showing volume composition in terms of pore space, iron minerals, silica, and minor constituents (clay, etc.) of the ferruginous cherts and iron ores of the Gogebic range. See page 189 for discussion of method of nlatting. 246 GEOLOGY OF THE LAKE SUPERIOR REGION. TRIANGULAR DIAGRAM ILLUSTRATING SECONDARY CONCENTRATION Or GOGEBIC ORES. In figure 32 tlie trianfjular method (described on p. IS!)) of i('j)resenting tlie volume relation of ferruginous cherts and ores and intermediate phases is applied to tiie Gogebic ores. As already explained, each small triangle within the large one represents an individual specimen and by its size and position indicates composition in terms of the volume of ])ore space, silica, iron minerals, and minor constituents. The average ferruginous chert, as indicated, is repre- sented by a small triangle in tiie lower left-hand side of the diagram, with low pore space and a large content of silica. The average ore is represented in the upper right-hand j)art of the diagram, and has more pore space, less silica, and more iron than the average ferruginous chert. Scattered about in the area between these two points are intermediate phases between ferrugi- nous chert and ore. In the alteration of ferruginous chert to ore, as represented in the triangle, the following changes have evidently taken place: («) Decrease in silica, (b) increase in pore space, and (c) increase in iron. Obviously the dominant process has been the removal of silica, as this is necessary to an increase of pore space and iron. Removal of silica alone without introducticm of iron or mechanical slump would increase the porosity in proportion to the amount of silica removed. Such a process would be represented on the triangle by a series of small triangles in a line parallel to the base, as the relative volume of iron would remain constant. In the actual case kno\\Ti the relative volume of the iron mineral increases from 26.83 per cent in the cherts to 52. IS per cent in the ores. This could be accomplished in two ways — by mechanical slumping or packing of the material, weakened by too great a porosity, or by infiltration of iron. From the diagram it is impossible to tell wliich of these processes, slumping or infiltra- tion, is more important. Observation shows, however, that slumping has been important, but that introduction of iron has taken place to a much greater extent than it did in the con- centration of the Mesabi ores. ALTERATION OF ROCKS ASSOCIATED WITH ORES DURING THEIR SECONDARY CONCENTRATION. The various conditions and agencies which were effective in the concentration of the ore from the cherty iron carbonates and ferruginous cherts caused alterations of a similar nature in the various rocks associated with the iron-bearing formation — namely, the interbcdded slates, the basic intrusive rocks, and the slates immediatelj- overlying the iron-bearing forma- tion. The alteration of the slates produced paint rock or ferruginous slate similar to that of the Mesabi range. The alteration of the basic dikes by oxidation of the iron, breaking do^\•n of feldspars, and leaching of soluble constituents formed a soft kaolinic product, locally termed soap rock or, if iron stained, paint rock. The following anal3'ses of fresh and altered dike rock are typical of this alteration: . Analysts of fresh and allcnd dikes associated vilh ore. 1 (fresh). ! (altered). Assuming -Vl.Oj constant. SiOj... AI2O3.. FeaOs.. FeO.... MgO... CaO.... NajO... KsQ.... HjO-.. H20+. TiOj... PzO:, . . . CO2.... 47.90 1,1.(0 3.f0 8.41 8.11 9.99 2.C5 .23 .15 2.34 .82 .13 .38 4B.8S 22. f>2 5.12 2.01 1.25 .80 2.cr. 3.12 8.25 1.12 .ir. 1.89 32.20 l.i. CO 3.53 1.39 .86 .55 1.83 2.15 5.(8 .77 .11 1.30 1. Specimen 12880. Unaltered diabase dike rock in iron-bearing formation, from southeast part of sec. 13, T. 47 N., R. 46 W., Uichigan. 2. Specimen 12S78. Altered diabase dike. Same locality aa No. 1. PENOKEE-GOGEBIC IRON DISTRICT. 247 A comparison of the two analyses on the assumption that alumina has remained constant (see third column in the table) shows a loss of silica, iron, magnesia, lime, soda, phosphorus, and titanium and a gain in potassa, water, and carbon dioxide. Except for the behavior of potassn, the alteration is typical of weathering under conditions of oxidation, carbonation, and hydration. Specific gravity and porosity determinations on the specimens analyzed resulted as follows: Sptxiftc (jraritij and porosilij of inialtcrtd and allcrcd phases of diabase. Specific gravity. Porosity. Unaltered diabase Altered phase of diabase. 2.92 2.76 n.so 28.40 On the basis of the specific gravities ami the assumption that alumina is constant, the calculated porosity due to leaching of soluble constituents is 27.1 per cent of the volume / 15.60 2.92 \ ( 1.00 — the bedding, oi- in two directions intei-secting at angles varying from nearly right angles where t-lie pressure has been least to acute angles where it has been strong. In thin section tlie latter rock has an a])i)earanc(' like tjiat of a drawn-out net. The largest areas of mica scliist, representing the most advanced phase of metamorphism of the formation, lie nortli of Michigamme. The greater nictamorphism of this part of the formation is attributed to tlie large masses of intnisive greenstone which have been introduced roughly parallel to the contact of the vSiamo slate and the Negaunee formation. Other consid- erable masses of greenstone are also found within the area of the Si;imo. Evidence of the metamorphic effect of the greenstone is afforded by numerous large secondary crystals of horn- blende in the slate adjacent to the larger masses of greenstone. Relations to adjacent formations. — At the upper and lower horizons the slates tend to becom,e ferruginous. In these phases there is present a considerable cpiantity of iron oxide, generally hematite but in many places magnetite. In the upper part of the formation especially these ferruginous slates have interlaminated layers of material similar to the ferruginous and sider- itic slates and cherts and giiinerite-magnetite schists of the Negaunee formation. The Ajibik quartzite grades up into tlie Siamo slate. It is apparent from the appearance of interlaminated layers of material like the Negaunee formation in the upper parts of the Siamo slate that the transition into the Negaunee is a gradation by interstratification. The fragmental sediments gradually die out and nonfragmental sediments become dominant; this change takes place irregularly, producing interstratification of the two forms of sediments. Thiclcness. — The area perhaps most favorable for detemiining the thickness of the Siamo slate is that adjacent to Teal Lake. If the formation were there assumed to be monoclinal, the tliickness would be from 1,250 feet to 1,300 feet, but as there are an unknown number of sub- ordinate rolls at this locality, and slat}' cleavage has developed, it is probable that the real thickness of the formation is not more than half of this amount. NEGAUNEE FORMATION. Name aiul distribution. — The principal iron-bearing formation of the Marquette district is named Negaunee because in the town of that name and to the south are typical exposures of the formation. The Negaunee formation extends from the northwest end of the district along the north side of the Huronian to the north side of Michigamme Lake. From this place eastward for a distance of 5 miles the formation is cut out by the unconformity at the base of the Upper Huronian. Near Ishpeming it widens out into a broad area and occupies a large portion of the famous T. 47 N., R. 27 W., and also a considerable portion of T. 47 N., R. 26 W. From this broad area a short southern arm, known as the Cascade range, extends to the east and a long arm to the west along the south side of the Algonkian; the formation is found also on both sides of the Republic and southwestern arms. In the western part of tlie main southern belt and in the Republic and southwestern arms the formation is apparently al)scnt for distancps varying from a fraction of a mile to several miles. It is believed tliis lack of continuity is due to the fact that the Negaunee formation was completely removed by erosion liefore the deposition of tlie upper Huronian (Animikie group). Deformation. — The two long arms of the iron-bearing formation of the main belt, as well as the two belts of iron-bearing formation in the Republic and southwestern belts, are the two sides of a synclinorium. The two main arms join in the large area of Negaunee at Ishpeming, showing that it also is in a broad way an east-west synclinorium. This trough pitches to the west. Thus the lower members of the Negaunee formation outcroj) on the east adjacent to the Siamo slate and the higher inembers outcrop on the west adjacent to the Goodrich quartzite. The suiuous contacts between tlie Negaunee and the formations above and below express its folding. ]\IARQUETTE IRON DISTRICT. 263 The salients to the east into the Siamo slate represent synclines and the reentrants anticlines; the salients to the west into the Goodrich quartzite represent anticlines and the reentrants synclines. The Palmer belt of the Negaunee formation, extentling from the main area as a southeastern arm, is also a synclinal fold, which ends to the east in a canoe with a westward pitch. The structure of this syncline is modified by a great fault along the south side of the Ajibik Hills and by faulting at the Volunteer mine. Lithology, including metamorpJiism . — Petrographically the iron-bearing formation com- prises sideritic slates, which may be griineritic, magnetitic, hematitic, or limonitic; griinerite- magnetite schists; ferruginous slates; ferruginous cherts; jaspilite, and iron ores. The ferru- ginous cherts and jaspilite are commonly brecciated, the other kinds less commonly. The sideritic slates are most abuntlant in the valleys between the greenstone masses in the large area south of Ishpeming and Negaunee. These rocks are regularly laminated, are fine grained, and when unaltered are of a dull-gray color. The purest phases of them are approxi- mately cherty iron carbonate, as sliown by two analyses made by George Steiger in the laboratory of the Survey. It is unusual to find exposures of the cherty siderite slates which have not been more or less affected by deep-seated alteration or by weathering processes. The iron car- bonates pass by gradations, on the one hand into griinerite-magnetite schists and on the other into ferruginous slates, ferruginous chert, jasper, or iron ore. The grunerite-magnetite schists consist of alternating bands composed of varying pro- portions of the minerals griinerite and magnetite and (|uartz. Where least modified they have a structure precisely Hke the sideritic slates from which they grade, the grunerite-magnetite belts having taken the place of the carbonate bands. In some places the grunerite-magnetite schists are minutely banded, the alternate bantls consisting of dense green griinerite and white or gray chert, with but a small quantity of magnetite. Certain important kinds appear to be com- posed almost altogether of griinerite, with a little magnetite. In general the griinerite- magnetite schists are found at low horizons, below the ferruginous chert and jaspilite — that is, at or near the same horizon as the sideritic slates. In many places also they are below intrusive masses of greenstone. By oxidation of the iron carbonate the sideritic slates pass into the ferruginous slates, the iron oxide being hematite or limonite, or both. These rocks, in regularity of lamination and in structure, are similar to the sideritic slates, differing fi'om them mainly in the fact that the iron is present as oxide. In the different ledges may be seen every possible stage of change from the sideritic slates to the ferruginous slates. The only necessary change is a loss of carbon dioxide and oxidation of tlie iron. On Meathered surfaces, along veins, and along some of the bedding planes the transfcjrmation may be complete, and between this material and the original rock there are numerous gradations. From the oxidation of the less slaty phases of the sideritic rocks result tlie ferruginous cherts, consisting mainly of alternating layers of chert and iron oxitle, although the iron oxide bands contain chert and the chert bands contain iron oxide (PI. XXXIII, B, p. 466). This iron oxide is mainly hematite, but both limonite and magnetite are locally present. Rarely mag- netite is tlie predominant oxide of iron. In such places the silica is usually coai-sely crystalline. The rocks are folded in a complicated fashion, as a result of which the layers present an extremely contorted appearance. Many of the folded layers show minor faulting. On account of the exceedingly brittle character of these rocks, they are very commonly broken through and through, and some of them pass into friction breccias. In places the shearing of the fragments over one another has been so severe as to produce a conglomeratic aspect. The ferruginous cherts are particularly abundant in the middle and lower parts of the iron-bearing formation, just above or in contact with the greenstone masses. In a number of places they are between the griinerite- magnetite schists or sideritic slates below and the jaspilite above. The rocks here named ferruginous chert are called by the miners "soft-ore jasper" to discriminate them from the "hard-ore jasper," or jaspilite, because within or associated with them are found the soft ores of the district. 264 GEOLOGY OF THE LAKE SUPERIOR REGION. The jaspilites consist of alternate bands composed mainly of finely crystalline, iron-stained quartz iiiid iron oxide (PI. XXX FT, ]). 4CA) . Tiu^ exposures jiresent a brilliant appearance, due to the interlaniinationof llie brij,d it-red jasper and tlie dark-red or black iron oxides. Tiie iron oxide is mainly hematite and includes both red and specular varieties, but magnetite is commonly present. Many of the jasper bands have oval terminations or die out in an irregidar manner. The folding, faulting, and brecciation of the jaspilites are precisely like those of the feniiginous chert, except that in the jaspilite they are more severe. The interstices produced by the dynamic action are largely cemented with crystalline hematite, but magnetite is present in subordinate quantity. In the foldmg of the rock the readjustment has occurred mainly m the iron oxide between the jasper bands. As a result of this the iron oxide has been sheared, and when a specimen is cleaved along a layer it presents a brilliant micaceous appearance; such ore has been called micaceous hematite. This sheared lustrous hematite, present as some form of iron oxide before the dynamic movement, is discriminated with the naked eye or with the lens from the later crystal-outlined hematite and magnetite which fill the cracks in the jasper bands and the spaces between the sheared laminse of liematite. The jaspilite differs mainly from the ferruginous chert, with which it is closely associated, in that the siliceous bands of the jaspilite are stained a bright red by hematite, and the bands of ore between them are mainly specular hematite, whereas in the cherts the iron oxide is earthy hematite. The jaspilite in its typical form, whenever present, usually occupies one horizon — the present stratigrapliic top of the iron-bearing formation, just below the Goodrich quartzite. In different parts of the district it has a varying thickness. With this jasper, or just above it, are the hard iron ores of the district; hence it has been called "hard-ore jasper" by the miners to discriminate it from the ferruginous chert, or "soft-ore jasper." Relations to adjacent formations. — The iron-bearing formation rests conformably upon the Siamo slate or upon the Ajibik quartzite and grades downward into one or the other of these formations through the increase of clastic material and a lessening of the ferruginous constitu- ents. The gradation may occur within a few feet or may require 100 feet or more. The transi- tion is accomplished by interlaminations of material which are alternatively chiefly fragmental and chiefly nonfragmental. The overlymg formation, the Goodrich quartzite, rests unconformably u]:)on the Negaunee formation. The amount of foldmg and erosion of the Negaunee formation accomplished before the Goodrich quartzite was deposited ditt'ers in diflferent parts of the district. In some places the erosion has gone so far as to have removed the iron formation entirely. It therefore follows that the contact between the two formations is here at one horizon of the iron-bearing formation and there at another, ranging from the highest known horizon to the lowest. TliicJcness. — It is evident from these relations that the thiclcness of the formation varies from practically nothing to its maximum. It is, however, difficult to estimate this maximum because of the pervasiveness of the intrusive rocks in the Negaunee. It is rouglily estimated that in the Ijroad area to the east of Ishpeming and Negaunee the thickness may be con- siderably above 1,000 feet, although it is entirely probable that the maximum thickness is less than this amount. Intrusive and eruptive rocJcs. — Within the iron-bearing formation there are numerous intrusive masses of "greenstone," really diabase and its altered equivalents. Tliese occur in the form of both dikes and bosses, and many of the latter are of large size, running up to masses 2 miles or more in extent. These rocks are especiaUy prevalent in the broad area of the iron-bearing formation near Ishpeming, where they occupy between one-third and one-lialf of the area. In many places the greenstones intrude the sedimentary series in a roughly laccolithic fasliion. In consequence of this, where the two have been folded together their relations are roughly similar to those of sedimentary formations, but when examined closely the greenstones are always found to cut the Negaunee formation to a lesser or greater degree. Surface eruptive rocks also appear in the formation in the vicinity of Clarksburg. (See p. 268.) MARQUETTE IRON DISTRICT. 265 UI'PER HURONIAN (aNIMIKIE GKOXJP). The upper Huronian is structurally divisible into a lower belt of conglomerate and quartz- ite, called the Goodrich quartzite, a belt of ferruginous rocks called the Bijiki schist, a belt of slate and schist kno\vii as the Michigamme slate, and, to the south, a mass of volcanic rocks called the Clarksburg formation. The Animikie group as a whole occupies the center of the main Algonkian synclinorium from Ishpeming to the west end of the district. In this part of the region it is the chief surface rock, occupying all the area between the belts of the Negaunee formation. GOODRICH QUARTZITE. Distribution and structure. — The belt of Goodrich quartzite forms a westward-opening U, bordered on the outside principally by the Negaunee formation, with its eastern margin near the city of Ishpeming. The folding is similar to that of the Negaunee formation, though some- what less complex. The sinuous contact of the two formations in the vicinity of Ishpeming expresses the complexity of folding at this end of the synclinorium. Litliology, including metamorphism. — Petrographically the Goodrich is dominantly a quartz- ite, although usiuxlly there is a conglomerate at the base. As the underlying rock is in most places the Negaunee formation this conglomerate is an ore, chert, jasper, and quartz con- glomerate. Wliere the conglomerate is near the Archean this system may furnish material for it — as, for instance, at Palmer, where there are numerous granite, greenstone, and schist bowlders derived from the Archean. Wliere the conglomerate is ore, chert, and jasper conglomerate immediately in contact with the Negaunee formation, the jiarticles have been flattened and schistosity has developed in both the conglomerate and the original basement rock, making it difficult to place the exact line between the two formations. This is illustrated at Ilumljoldt. At several localities the conglomerate resting upon the Negaunee formation has had quartz leached out and hematite and magnetite deposited, developing a material rich enough in iron to be an ore. This is illus- trated at the Goodrich and Volunteer mines. The quartzite is mahily quartz but contains many particles of chert and jasper and usually considerable amounts of feldspar. Cementation by enlargement is an important process in the induration of the rock. In the eastern part of the district dynamic action has not usiuxlly been great enough to give the particles more than undulatory extinction, or at most fracturing. However, these effects are pervasive, not a single clastic particle escaping. The mashing in the central and western parts of the district has been severe and the formation has been transformed to a schist. In the western part of the district, especially in the Republic trough, the alterations have been so great as to transform the fekU spathic quartz rocks into micaceous quartz schists, or locally, where the mica is sufficiently abundant, into muscovite-biotite schists or biotite schists. In this change the feldspar has usually altered into quartz and mica, including both muscovite and biotite, especially muscovite. Relations to adjacent formations. — The Goodrich quartzite rests unconformably upon the Negaunee formation. The evidence of this unconformity consists both in the discordance of strike and dip, varjdng from a few degrees up to perpendicularity, as at the Goodrich mine, and in the existence of conglomerates derived from the Negaunee formation at scores of locali- ties along the contact. At many places, as has already been pointed out, the erosion between Negaunee and Goodrich time cut through the Negaimee formation. In these places the mate- rial of the Goodrich quartzite comes from the underlying formations, the Ajibik quartzite or the rocks of the Ai-chean. There are few Lake Superior formations that have a more complete set of conglomerates at the base or that have clearer proof of unconformity with the rocks upon which they rest. The Goodrich quartzite, by the diminution of coarse fragmental quartz, grades above- into the Michigamme slate, the Bijiki schist, or the Clarksburg formation. The nature of each gradation will be mentioned in connection with these formations. TliicJcness. — The thickness of the Goodrich quartzite varies greatly from place to ])lace. At the Goodrich mine it is calculated to be as great as 1,500 feet, but this is probably much beyond tlie average for the district. 266 GEOLOGY OF THE LAKE SUPERIOR REGION. BIJIEI SCHIST. • Name and distribution. — The Bijiki schist is given this name l)ecaiisc typical exposures occur near the mouth of Bijiki River. It is confined to tliree narrow l)elts in tlie nortli western part of the district. North of tlie northernmost of tliese behs is tlie Goodricii quartzite and between the north and middle belts is the Michigamme slate. These two belts make a synclinal structure. Tlie middle and southern belts unite at (lie east and rej)resent the outcrop of an eroded anticline. Lithology, including metamorphism. — ^Lithologically tlie Bijiki schist comprises two main varieties, one of which is characteristic of the eastern ])art of tlie lielts and tlie otlicr of the western ])art. In the eastern jiart tlie least-altered phases consist of a sideritic chert interbedded \rith the Michigamme slate and jirobably representing a slightly higher horizon than the phase of the Bijiki schist described in the following paragrajiii. Not imcommonly the siderite is the predominating. constituent. This slate has been extensively altered by weathering and meta- soniatic changes into ferruginous slates and ferruginous cherts, with .subordinate amounts of griinerite-magnetite scliist. In a few localities, where the ferruginous material is very abun- dant and the conditions of deposition are favorable, small ore bodies have been found. These are illustrated by the North Phenix, Pascoe, Hortense, Northampton, Marine, PhenLx, and Bessie deposits. These ores differ from the soft ores of the Negaunee formation in that the iron oxide is largely limonite and the associated slates are carbonaceous and graphitic. In the western area, which contains the chief exposures of the formation, the Bijiki is dominantly a banded griinerite-magnetite schist. This rock consists mamly of three miner- als — -rjuartz, griinerite, and magnetite. Here and there a small amount of residual siderite is seen. The rock is discriminated from the griinerite-magnetite schists of the Negaunee foimation chiefly by its exceeding toughness and the dilliculty with which it is broken jiarallel to the stratification. One of the most conspicuous mineralogical features of the iron-bearing Bijiki formation near Michigamme is its content of large garnets, up to 2 inches in diameter, developed late in tlie metamorphism. These have been apparently altered to clilorite and amphibole, early described by Pumpelly as chlorite pseudomorphs after garnet." Microscopic examination .shows that although much of the matrix material is chlorite, the garnet is largely replaced bv green amphibole and magnetite. Porphyritic biotite in a chloritic matrix is also a very conspicuous mineralogical feature of these rocks, giving them in the hand specimen a brilliantly spangled appearance. The garnet may be really a poikilitic development later than clilorite. The two chief phases of the Bijiki schist may be in part at separate horizons, but there seem also to be gradations between the ferruginous slates and cherts and tlie griinerite-magnetite schist. As the schists are largely confined to the western ])arts of the belts, where there are important masses of intrusive igneous rocks, and occur in the part of the district where the Negaunee formation is also changed to a griinerite-magnetite schist, it is believed that the schist represents the original sideritic formation altered under the influence of igneous rocks while deeply buried and largely by the process of sihcation, whereas the eastern part of the formation, consisting tif ferruginous slates and cherts and containing ore bodies, was altered after the formation was exposed at the surface, later than upper Iluronian time, by the proc- esses of weathering. Relations to adjacent I'ocks. — ,Uong the northern belt where tlie base of the Bijiki schist is exposed, roundetl fragmental quartz appears near the bottom of the formation, and with an increase of this material the member grades downward into the Goodrich quartzite. The Bijiki schist grades above into the Michigamme slate. In the central and eastern parts of the ilarquette district the Bijiki has not lieen detected. Apparently in the greater portion of the district between the time of the Goodrich quartzite a I'liiiipelly, Riiplmel, On pseudomorphs of chlorite after garnet at the Spurr Mountain iron mine, Lake Superior: .\m. Jour. Sii., lid ser., vol. 10, July, isrs, pp. 17-21. MARQUETTE IRON DISTRICT. 267 and tlie Michigamme slate the conditions were not favorable for the deposition of the iron- bearing formation. The u-on-bearing Bijiki schist, though not tliick or economically of as great consequence as the Negaunee, is of considerable significance in the matter of correlation, for it occurs at tlie same horizon as an important u'on-])oaring formation in other districts — notably the Menominee, Gogebic, and Mesabi. Thickness. — The Bijiki schist apparently has a maximum thickness of about 520 feet and from this it ranges down to the disappearing point. MICHIGAMME SLATE. Name, disfrihution, and correlation. — The name Michigamme is given to the upper slate and mica schist formation because extensive exposures of it occur on the islands of Lake ilichi- gamme and on the mainland adjacent to the shore. The ^lichigamme slate is mainly in a single great area, which extends from a ]>oint about a mile west of Ishpeming along the axis of the Marquette synclinorium to the west end of the district. To Lake Michigamme the breadth of this belt is for the most part less than 2 miles, but at Lake Michigamme it broadens out into an area 5 miles or more in width, from which extend the Republic and southwestern arms. Beyond the limits of the Marquette district proper the formation continues to widen and covers a great expanse of country, extending to the Crystal Falls district on the south and well toward the Gogebic district on the west. It is the ecpiivalent of and is contmuous with the slate to which the name "Hanbury" has been given in previous reports. It is also probably the equivalent of the Tyler slate of the Penokee- Gogebic district, to judge from its relations with associated formations and from the probability (indicated by known outcrops) of direct areal connection, though outcrops are not sufficient h' numerous to establish this connection absolutely. Deformation. — The Michigamme slate in most of the district forms a great synclinorium, the secondary folds of which are, however, not sufficiently large to bring up the lower rocks to the erosion surface except in a central anticline at the east end of Lake Michigamme, where the Bijiki schist and Goodrich quartzite appear at the surface. Litlwlogii. — The formation is a pelite, which now comprises two main ^•arieties — slates and graywackes and mica schists and mica gneisses — each of which includes both ferruginous and nonferruginous kinds. The slates and graywackes occur east of Lake Michigamme and the mica schists and mica gneisses at Lake Michigamme and to the west, including the Repubhc and southwestern arms. The slates and graywackes differ from each other chiefly in coarse- ness of grain, the two being interlaminated in many exposures. There are all gradations from aphanitic black shales or slates to a graywacke so coarse as to approach a cpiartzite or even a conglomerate. In color the rocks vary from gray to black. Where fine grained they have a well-developed slaty cleavage. In places they are graphitic, pyritic, and ferruginous. Two specimens showing the maximum amount of graphite analyzed 15.69 and 18.92 per cent of carbon. The slates and graywackes differ in no essential respect fi-om the similar rocks of the Siamo slate (see pp. 261-262) or from the Tyler slate of the Gogebic district (see pp. 232-233), there- fore they will not again be described. MetamorpMsm. — The slates and graywackes by increase in metamorphism pass into chlo- rite schists, mica schists, and even mto mica gneisses. The process of alteration for the mica schists is identical with that already described in connection with the development of similar rocks for the Siamo slate and the Tyler slate. (See pp. 232-233, 261-262.) In many places where the rocks are completely crystalline garnet, staurolite, chloritoid, and andalusite are plenti- frdly present. In the more coarsely crystalline rocks much feldspar has developed, and the rock thus becomes a gneiss. This material appears in bands which seem to be altered beds of the formation but which resemble granitic material. The appearance is that of a rock pegmatized throughout. These bands grade into ordmary mica schists. No independent granites have 268 GEOLOCxY OF THE LAKE SUPERIOR REGION. boon discovered in oonnoction with tliis extremely metamorphosed variety of rock, but it can not be asserted that such rocks are not somewhere present. Where the rocks have become schists the ferruginous constituents have been largely transformed to magnetite. Relations to adjacent formations. — The Michigamme slate grades downward into the Bijiki schist or the Goodrich quartzite. TliicTcness. — Tlie thickness of the Michigamme slate is considerable, as is shown by the wide area which it covers. There are, however, so many subordinate folds and the mota- morphism is so extreme that it is ini|)ossible to make even an approximate estimate of its thickness. Within the area described the thickness of the formation may not be more than 1,000 or 2,000 feet, or may be greatly in excess of this. CLARESBXTIIG FORMATION. Distribution. — ^The Clarksburg formation differs from the other Algonkian formations of the Marquette district in that it is dominantly a volcanic formation. It is confined to the south side of the Iluronian area, extending from the region north of Stoneville to a point some- what west of Champion, the largest and most typical areas being east of Clarksl)urg. It is clearly a local formation, not only in its eastern and western extent but in being confined tip one side of the district. This is explained by its volcanic character, the vents being on the south border of the Algonkian area. Lithology. — Petrographically the formation comprises massive greenstones of the general character of diorites; lavas that are interbedded with sediments and tuffs; tuffs that grade off imperceptibly into sediments, the material of which is mainly of volcanic origin; and, finally, greenstone conglomerates and fine-grained sediments, the material of which is mainly volcanic but has evidently been arranged by water. All these rocks are extremely altered and in places so much so that they are now schistose. The pyroclastic material may have been partly sub- aerial, but doubtless a large part of it fell upon the water. The volcanoes of Clarksburg time were very plainly of explosive type. The center of volcanic activit}^ was east of Clarksburg, and m this vicinity are found the largest amounts of massive and coarse material, lavas, breccias, and conglomerates. Toward the east and west the formation becomes thinner and its material finer, imtil it dies ovit in both directions into the Michigamme slate. It is not the purpose here to describe in detail the many different varieties of rocks of this volcanic formation. These are discussed in Monograph XXVIII of the United States Geological Survey." This volcanic formation is similar to that of the volcanic formation at tlie east end of the Gogebic district, the chief difference being that the latter is much less meta- morphosed. It is notable that both occur in the upper Iluronian and mainl}- take the place of the great upper slate formation (Michigamme slate), although the beginning of the volcanic outbreak was early in upper Huronian time or earlier. In the eastern part of the district a small amount of volcanic material appears also to be associated with some of the earlier formations, especially with the Siamo slate. Eelations to adjacent formations. — The volcanic outbreaks of the Clarksburg began early in Goodrich time^ or perhaps even in late Negaunee time, but the main volcanic deposits were in Michigamme time. Later in Michigamme time, by the dying out of volcanic activity, the sediments became more largety ordmary material, and thus the Clarksburg grades above into the Michigamme. Thicl-ncss. — There is no way to ascertain the maximum thickness of tiie formation, but east of Clarksburg it must be several thousand feet thick. From this maximum it ranges down to a knife-edge. INTRUSIVE IGNEOUS UOCKS. Into all the formations of the Huronian series igneous rocks are intruded. These are of at least two ages; the older probably belong to the Huronian and the later to the Keweenawan period. Much the larger number of intrusive masses are distinctly of post-Hin*onian and 1 Van Hise, C. R., and Bayley, W. S., The Marquette Iron-bearing district ot Michigan: Mon. U. S. Geol. Survey, vol. 28, 1897, pp. 4(»-186. MARQUETTE IRON DISTRICT. 269 probably Keweenawan age. Many of them are distinctly bo.sses, laccoliths, and sills which in their upward movement have been stopped by the massive competent layers of the Negaunee or Goodrich qoartzite, and therefore on the present erosion surface are likely to show close areal relations with the Negaunee formation. This is especially conspicuous in the vicinity of Ishpeming, Negaunee, and Spurr. The intrusive rocks have been described by various authors under the terms diorite, diorite schist, chlorite schist, magnesian schist, soapstone, and paint rock. Part of them have been regarded by some geologists as metamorphosed sediments, but microscopical study of all the varieties shows that they were originally basic rocks of the composition of diabases. The great bosses of greenstone, commonly known as diorite, are a prominent feature of the topography in the general area covered by the iron-bearing Negaunee formation, and the relations of these greenstones to the genesis of the ores has already been described. During the folding there was much differential movement between the greenstone masses and the surrounding forma- tions, and also the contact plane is one favorable to the action of percolating waters. As a result of this it is a common thing for the periphery of the greenstone knobs to be schistose. In the area around Ishpeming and Negaunee the schistosity has obviously been the result of differential movement between the greenstones and the overlying Goodrich quartzite. The Goodrich quartzite has moved in the usual direction upward along the limbs of the folds, developing cleavage dipping more steeplj- than the contact of the greenstone and the quartzite. Wliere not heavily stained by iron these rocks are commonly called chloritic schists. Adjacent to the iron-bearing formation the rocks, besides having a schistosity, have been much leached and modified m composition and are commonly known as soapstones because of their greasy feel. The much-altered greenstones that have a strongly developed schistosity and have been stained by iron oxide are called paint rock by the miners. Even in the massive varieties of dikes, laccoliths, and bosses the original augite has extensively changed to hornblende and consequently the rock in the district has generally been called diorite. In the western part of the district, both within the intrusive greenstone masses and in the adjacent formations, there have been important contact effects. This is shown by the extensive development of garnet m both the intrusive and intruded formations, by the less common development of biotite, and by the metamorphism of the iron-bearing formation mto griinerite-magnetite schist and of the Michigamme slate into a mica schist. Griinerite has formed to some extent within the intrusive rocks also. The intrusive character of these igneous rocks of Huronian age is shown not only by con- tact effects but by the manner in wliich they cut across the bedding of adjacent rocks and project dikes into them. However, evidence of this kind is not available for all the igneous masses, especially those of laccolithic and sheet form, and it is regarded as not at all unlikely that some of them may be really extrusive rocks put down contemporaneously with the adjacent sediments. The latest intrusive rocks are fresh diabase dikes which are probably of Keweenawan age. They cut all the other formations of the district, including the older greenstones wliich have just been described. These rocks include diabase, quartz diabase, olivine diabase, porphyrites, and basalts. CAMBRIAN SANDSTONE. Upper Cambrian or Potsdam sandstone is exposed in an east-west belt along Carp River to the south of the city of Marquette and Mount Mesnard, where it rests unconformably upon the Kona dolomite. QUATERNARY DEPOSITS. The district is more or less covered by Pleistocene deposits. On the southeast it is so thoroughly covered that the bed-rock geology is not well known. The Pleistocene is discussed in Chapter XVI (pp. 427-459). 270 GEOLOGY OF TPIE LAKE SUPERIOR REGION. THE IRON ORES OF THE MARQUETTE DISTRICT. By the authnr-i and W. J. Mead. DISTRIBUTION, STRUCTURE, AND RELATIONS OF ORE DEPOSITS. The cliief iron-bearing formation ul the Marquette district is the Negaunee. It bears ore at various horizons. Ores also occur at tJio basal horizon of the Goodricli quartzitc, where it rests upon and h:is derived debris from tlie Negaunoe formation. Snitdl cjuantities of oi-e are found in tlio iron beds of the Bijiki schist, associated witli the Micliigiumue slate. Workable iron-oie deposits have been found at many places from a point east of Negaimee to Michi- gamnie and Spurr. The Marquette district differs from the ilesabi and Gogebic districts in not having long stretches of nonproducing iron-bearing rocks. The maximum depth of concentration of ores in the Marquette district is still imknown. On the Teal Lake range the depth is not more than 700 feet; in the Ishpeming and Nogaunee areas depths as great as 1,500 feet are known. In the Champion area ore has been foUowed 'ii ore/i Figure 36,— Ore deposits of the Marqiiettfi district. (Both ore exploited and ore now in mine are represented as ore, as the purpose of this figure is to show themanner of the development of the ore rather than the present stage of exploitation.) a, Generalized section in Marquette district, showing relations of all classes of ore deposits to associated formations. On the right is soft ore resl ing in a V-shaped trough between the Siamo slate and a dike of soapstone. In the lower central part of the figure the more common relations of soft ore to vertical and inclined dikes cutting the ja,sper are shown. The ore may rest upon an inclined dike, between two inclined dikes, and upon the upper of the two, or be on bothsidesof a nearly vertical dike. In the upper central part of the figure are seen the relations of the hard ore to the Negaunee formation and the Goodricjiquartzite. .\t the left is soft ore resting in a trough of soapstone which grades downward into greenstone. (From Mon. U. S. Geol. Survey, vol. 28, 1S97, PI. XXVIII, fig. 1.) 6, Cross section of Section lij mine. Lake Superior mines, in the Marquette district. On the right is a V-shaped trough made by the junction of a greenstone mass and a dike. The hard ore is between theseand below the Goodrich quartzite. On the left the hardoreagain rests upon a soapstone which is upon and contains bands of ore-bearing formation. The ore is overlain by the Goodrich quartzite. Scale: 1 inch=220 feet. (From Mod. U. S, Geol. Survey, vol. 28, 1897, PI. XXIX, fig. 1.) down 2,000 feet and is laiown to extend fartlier. The Negaunee formation constitutes a part of the westward-pitchmg Marquette trough, and west of Ishpeming and Xegaunee the central part of the trough goes beneath a considerable thickness of upper Huronian sediments. Bectiuse of this dee]) burial ))ut little drilling has been done to ascertain whether or not the ores go down here, but the discovery of a large ore deposit at the very bottom of the Xegaunee formation near Negaunee has led to deep drilling west of Ishpeming and Negaunee with such results as to indi- cate that the ores extend to unlooked-for deptlis in this direction. In general the ores come to the rock surface along the middle slopes of the hill.s, l)ut they In general tlie ores come to also go under the lowest ground. MARQUETTE IRON DISTRICT. 271 The ore deposits of the Xegaunee formation and the associated ores may be divided, accord- ing to stratigraphic position, into three chisses — (1) ore deposits at the bottom of the iron- bearing formation; (2) ore deposits within the iron-bearing formation (these ores in many places reach the surface but are not at the uppermost horizon of the formation); (3) ore deposits m the top layers of the Negaunee formation and the bottom layers of the Goodrich quartzite. (See fig. 36.) This last class of deposits runs past an unconformity. Some of these ore bodies are almost wholly in the Goodrich quartzite. Stratigraphically these deposits ought to be separately considered, but they are so closely connected genetically and in position with the Negaunee ore deposits that they are treated with the deposits of that formation. The first two classes of ore are generally soft, and the adjacent rock is ferruginous chert or "soft- ore jasper;" the deposits at the top of the iron-bearing formation are hard, specular ores and magnetite and the adjacent rock is jaspilite, also called " sjiecular jasper" and "hard-ore jasper." Although the larger number of ore bodies can be referred to one or another of the three classes above given, it not infrequently happens that the same ore deposit belongs partly in one and partly in another. Also the upper part of an ore deposit may be at the topmost horizon of the iron-bearing formation and be a specular ore, whereas the lower part may lie wholly within the iron-bearing formation and may be soft ore. In some places there is a grada- tion between the two phases of such a deposit, but more commonly the two bodies are sepa- I'ated by dikes, now changed to soapstone or paint rock. 1. The ore deposits at the bottom horizon of the Negaimee formation have been mined principally where the lowest horizon of the formation outcrops — that is, they are confined to that part of the formation resting upon the Siamo slate or the Ajibik quartzite, along the outer borders of the Negaimee formation. The best examples of these deposits are those occurring at the Teal Lake range and east of Negaunee. East of Negaunee the ore bodies occur at places where the slate is folded into synclinal troughs which pitch sharply to the west. Here the iron- bearing formation is in places cut by a set of steep vertical dikes, and the conjunction of these dikes with the foot-wall slate forms sharp V-shaped troughs, as in the Cleveland Hematite mine, where the ore bodies are found between a series of vertical dikes and the Siamo slate. By com- paring this occurrence with the ore deposits of the Penokee-Gogebic district, it will be seen that they are almost identical, there being on one side of each of the ore bodies an impervious dike, the two uniting to form a pitching trough. The ore deposits of this horizon are being found by deep drilling to be extensive. The opening of the Maas mine at the east end of Teal Lake and the discovery of ore by deep drillmg at this horizon in the western part of the Ishpeming area suggest that the beds of this horizon at gi'eat depth may ultimately be foimd to carry a larger tonnage of ore than those of any of the other horizons. 2. The typical area for the soft-ore bodies within the Negaunee formation is that of Ish- peming and Negaunee. Here are the Cleveland Lake, the Lake Angeline, the Lake Superior Hematite, the Salisbury, and many others. The large deposits rest upon a pitching trough composed wholly of a single mass of greenstone or on a pitching trough one side of which is a mass of greenstone and the other side a dike joining the greenstone mass. The underlying rock is called greenstone where unaltered; that immediately in contact with the ore is known by the miners as paint rock or soap rock or soapstone. The greenstone changes by minute gradations into the schistose soapstone, and this into the paint rock. Many of the thinner dikes are wholly changed to paint rock or to soapstone, or to the two combinetl. The larger number of these troughs are found along the western third of the Ishpeming-Negaunee area. Plate XVII (in pocket) shows several westward-opening bays occupied by the iron-bearing formation in the masses of greenstone. Conspicuous among these are the Ishpeming basin, the northern Lake Angeline basin, the southern Lake Angeline basin, and the Salisbury basin. The iron-formation embayments open out and pitch to the west. At Lake Angehne an eastward dike cuts across the basin south of the center, and this combined with the gi-eenstone bluffs to the north and to the south forms two westward-pitching troughs, the northern of which has the greatest ore deposits of the Marquette district, containing many millions of tons of ore. 272 GEOLOGY OF THE LAKE SUPEKIOK REGION. 3. The hard-ore bodies, mainly specular hematite but in some deposits including nuuli magnetite, are at the top horizons of the iron-bearing formation, immediately below and in the basal members of the Goodrich quartzite. Examples of this class are the Jackson mine, the Lake Superior Specular, the ^'ohulteer, the Michigamme, the Kiverside, the Champion, the Republic, and the Barnum. Also, as interesting deposits, giving the history of the ore, may be mentioned the Klomau and the Goodrich. In all these deposits the associated rocks of the iron-bearing formation are jaspilite or griinerite-magnetite schist, usually the former. Many of these ore deposits weld together the Goodrich quartzite and the Negaunee formation and can not be separated in description. As in classes 1 and 2, all the large ore deposits belonging to this third class have at their bases soapstone or paint rock. Wliere the soapstone is within the Negaunee formation it is a modified greenstone mass or this in conjunction wdth a dike or dikes. Where the ore deposits are largely or mainly in the Goodrich cjuartzite the basement rock may likewise be a greenstone or it may be a layer of sedimentary slate belonging to tlie Goodrich quartzite. These different classes of rocks are, however, not discriminated by the miners, but are lumped together as soapstone and paint rock. Wlierever the deposits are of any considerable size the basement rock is folded into a pitching trough, or else an impervious pitching trough is formed by the union of a mass of greenstone with a dike, or by the union of either one of these with a sedimentary slate. Perhaps the most conspicuous example of this is at the Repubhc mine, but it is scarcely less evident in the other large deposits. A few small deposits of ore (chimneys and shoots) occur at the contact of the Negaunee and Goodrich formations, where no basement soapstone has been found. As examples of ore deposits which are largely or wholly witliin the Goodrich quartzite may be mentioned the Volunteer, Michigamme, Champion, and Riverside. These are partly recomposed ores and differ in appearance from the specular hematite or magnetite of the Negaunee formation in having a peculiar gray color anil in containing small fragmental particles of quartz and complex pieces of jasper; in many of them also sericite and chlorite are discovered with the microscope. Ore deposits in the Bijiki schist, associated with the Michigamme slate, have slate as foot and hanging walls. They are illustrated by the Beaufort, Bessie, Ohio, and Imperial mines. Although these different classes of ore bodies have the distinctive features indicated above, they have important features in couunon. They are confined to the iron-bearing formations. They occur upon impervious basements in pitching troughs. The impervious basement may be a sedimentary or an igneous rock, or a combination of the two. Wliere the ore deposits are of considerable size the plication and brecciation of the chert and jasper are usual phe- nomena. In many places this shattering was concomitant wth the folding into troughs or with the intrusion of the igneous rocks. In any of these classes the deposits may be cut into a number of bodies by a combination of greenstone dikes and masses. A deposit which in one part of the mine is continuous may in another part of the mine be cut into two deposits by a gradually projecting mass of green- stone which passes into a dike, and each of these may be again dissevered, so that the deposit may be cut up into a numl)(>r of ore bodies separated by soapstone and paint rock. Some of the ore deposits have a somewhat regular form from level to level, but the shape of the deposits at the next lower level can never be certainly predicted from that of the level above. Horses of "jasper" may appear along the dikes or within an ore body at almost any place. The ore bodies grade al)()ve and at the sides into the jasper in a variable manner. As a result of the comljination of these uncertain factors, most of the ore bodies have extraordinarily irregular and curious forms when examined in detail, although in general shape they conform to the above descriptions. MARQUETTE IRON DISTRICT. 273 CHEMICAL COMPOSITION OF MARQUETTE ORES. The following average partial analyses were calculated from cargo analyses in shipments for 1906 and 1909: Arcrar/c partial analyses of Marquette orcsfi calculated from cargo analyses for 1906 and lS09.b Composition of ore dried at 212° F. Loss on of total pro- Fe. P. SiOj. -\IjO3. ignition. duction. 100.0 59.55 0. 107 8.21 2.28 1.66 100.0 57.05 . 105 10.16 2.18 2.31 21. S 59.60 .078 8.47 2.13 .57 37.0 61.40 .094 6.40 2.34 2.61 39.5 59.20 .082 S.U 2.54 2.20 2.0 53.70 .290 11.30 1.17 7.05 Moisture (loss on drvinsat 212° F.). Average of entire district: 1906 1909 Upper horizon. 1906 Middle horizon, 1906 Lower horizon, 1906 Bijiki formation, 1906- . . . 9.04 9. .52 1.24 11.75 11.32 8.30 a Including ores of Swanzy district. ^ Calculated from analyses from I^ake Superior Iron Ore Association booklet. In addition to the constituents listed above the ores contain small amounts of manganese, lime, magnesia, sulphur, soda, and jjotassa. The range for the various constituents of the ores as shown by average cargo analyses for 1906 and 1909 is as follows: Range of percentage of each constituent in the Marquette ores for 1906 and 1909.'^ Moisture (loss on drying at 212° F.) . Analysis of dried ore : Iron Phosphorus Silica Alumina Manganese Lime Magnesia Sulphur Loss by ignition 0.51 to 14. 33 0.50 to 15. 75 90 to 64. 61 029 to .402 21 to 34. 20 .09 .04 .18 to .09 to .004 to .18 to 6.26 2.72 2.00 1.18 .062 7.07 40. 20 to 05. 69 .018 to .387 3.25 to 40. 77 to to to to .42 .00 .00 .00 .003 to 4.32 2.78 2.09 .039 .10 to 11.40 a Calculated from analyses from Late Superior Iron Ore Association booklet. The magnetites do not differ essentially m composition from the dommant hematites and limonites except m having less water. CHEMICAL COMPOSITION OF IRON-BEARING NEGAUNEE FORMATION. An average of 1,727 analyses representing 11,025 feet of drilling from the district away from the available ores gives 35.12 per cent of iron. This includes both the lean jaspers and the partly altered jaspers, but not the ores. Because of their great mass compared with the ores, this figure represents nearly the general average composition of the entire formation. If the unal- tered jaspers alone are taken, the average is somewhat lower. The composition of a typical amphibole-magnetite-quartz rock is as follows: Average analysis of griinerite-magnetite schist." Loss 1. 03 SiO, 50.02 AUOj 97 Fe^Oj 10. 05 Feb 28. 29 MnO 74 CaO 2.63 MgO 4.13 CuO Trace. Na,0 P265 CO2 H.^6 (above 110°) 0,S 09 1 55 42 100. 00 Total Fe 29. 20 It mil be noted that this differs but little from the average composition of the jaspers. TrMT" 4751 a Calculated from analyses given in Mon. U. S. Geol. Survey, vol. 28, 1897, p. 338. -VOL 52—11 IS 274 GEOLOGY OF THE LAKE SUPERIOR REGION. MINERAL COMPOSITION OF MAKQUETTE ORES. The ores of the Marquette district are doiiiinantly hydrous lieniatites and sulxirdinately anhydrous specular hematites and magnetites. Owing to the presence of magnetite, tlie mineral conijjosition can not be calculated fi-om analyses in which ferrous and ferric iion are not separated. The coarse specular hematites are made up mainly of large, closely fitting flakes of hematite, most of which take an imperfect polish and have, therefore, a gray, sheeny, sj)otted appearance. The flakes, which are parted along the cleavage, reflect the light hke a mirror. The large number of individuals of this kind is appreciated only by rotating the sections under the micro- scojje. This In'ings successively different flakes of hematite into favoi-able positions to reflect the light into the microscope tube. In some sections cut transverse to the cleavage the scliistose character of the rock is apparent in reflected light, innumerable laminae of hematite giving fine, narrow, parallel dark and light bands, which are comparable in appearance to the ])oIysynthetic twummg bands of feldspar. As both the magnetite and the hematite are usuafly opaque, the two minerals in general can not be discriminated, although in some sections the crystal forms of magnetite are seen and a small part of the hematite, much of it in little crystals, shows the characteristic blood-red color. The important accessory minerals are quartz, griinerite, feldspar, and muscovite. Some of the small, detached areas of cjuartz and feldspar appear to be frag- mental. The muscovite occurs mainly in small, independent flakes, but some of it is apparently secondary to the feldspar. The fine-grained specular hematites dift'er from the so-caUed micaceous hematites chiefly in that much more of the hematite is translucent and hence at the edges and in sjiots in the slides is of a l)rilliant red color. The "slate ores" m reflected light show the laminated character of the rock, while the massive ores give the peculiar spotty reflections, exactly the same as magnetite. The mottled red and black specular ores in reflected hght present a pecuhar appearance, the true specular material giving the usual briUiant, spotty reflections, whereas the soft hematite has a brownish-red color. The soft hematites in transmitted light show in many slides the characteristic blood-red color of hematite, although for the most part the sections are so thick as to give a brownish appearance or are ojjaque. In the softest ores m reflected light a dark brownish-red color is every^vhere seen, which is much less lirilliant than that presented bj' the same mmeral in trans- mitted light. In some of the soft hematites, however, within the mass of red material are many small areas which reflect the light m the same maimer as the specular ores. The limonitic hematites differ from the pure hematites onlj^ in that, m both transmitted antl reflected hght, in many places the reddish colors are not so bright. Lender the microscope the magnetites are opaque m transmitted light ; in reflected Ught they give the characteristic spotty appearance of that mineral. Where not pure the usual mmerals contamed in the iron formation appear with their ordmary relations. Those most plentifully seen are quartz, griinerite, muscovite, and biotite. Here and there garnet and chlorite as an alteration pi'oduct are alnmdant. On the borders of the mcluded material the magnetite invariably shows crystal outlmes. As a result each area of included mmerals has a serrated form. With the magnetite there is always more or less of hematite, a large part of wiiicii in many places results from the alteration of the magnetite. The liematite ranges from a subordinate to an important amount. Also at many places with the magnetite are varying cjuantities of pyrite and garnet and alteration jiroducts of the latter, chlorite and amphibole. The magnetites range in color from ])lack to gray. PHYSICAL CHARACTERISTICS OF MARQUETTE ORES. The magnetites and specular hematites are called hard ores b}' the miners, and tiie iivtirous red hematites are called soft ores. The magnetites range from very coarsely granular to finely granular magnetite. MAKQUETTE IRON DISTRICT. 275 As the ores are made up essentially of ii'on minerals and quartz, the mineral density varies directly with the iron content, ranging;; from as high as 5.1 in some of the dense hard ores to as low as 3.5 m some of the low-grade limonitic ores. Owing to the witle variation iji the mineral composition of the ores, an average figure for the district would have no significance. The avei'age density of fhe soft hematites, calculated from the 1906 cargo analyses, is 4.14. The porosity varies from less than 1 per cent m the hard specular ores to over 40 per cent in the limonitic ores. The average moisture content of the ores of the middle horizon indicates a porosity of approximately 35 per cent, assuming the mmeral density to be 4.14. This is probal)ly n(jt far from the true figure. The number of cubic feet per ton varies from 7 in the pure hard hematites to as high as 14.5 m the limonitic ores. The average for the soft red hematites is approximately 11.9 cubic feet per ton, calculated from a mineral density of 4.14, a porosity of 35 per cent, and a moisture content of 11.75 per cent. The following table, showing an average of a number of screening tests on the soft ores of the Marquette district, gives a good idea of the average texture of these ores. A comparison of the textures of the ores of the several Lake Superior districts is shown in figure 72, page 481. The screening tests, of which the following is an average, were made by the Oliver Iron Mining Company on 11 typical grades of ore mined in the Marc|uette district in 1909 and aggregating a total of 746,779 tons. For each grade of ore tested a sample was taken biweekly, quartered down monthly in proportion to the number of tons mined, and at the end of the year quartered down to 100 pounds, dried, and tested. The average was obtained by combining the results of the 11 screening tests in proportion to the number of tons represented by each of the 11 grades. Composite of screening tests on typical soft ores of the Marquette district. Per cent. Held on J-inch sieve 28. 15 ^-inch sieve 42. 22 No. 20 sieve 10. 98 No. 40 sieve 4. 90 No. 60 sieve 2. 90 No. SO sieve 1. 23 No. 100 sieve 1. 15 Passed through No. 100 sieve 7. 19 SECONDARY CONCENTRATION OF MARQUETTE ORES. Structural conditions. — The structural conditions controlling the circulation of water in the Marquette district are various. At the lower horizons of the Negaunee formation the impervious basement is formed by the pitching folds of the Sianio slate, as on the Teal I^ake range. At the middle and upper horizons of the Negaunee formation the irregular bosses and intnisive masses of greenstone constitute impervious basements in the reentrants of which the ores are found. The greenstone and its altered form, soapstone, accommodated themselves to folding without extensive fractures and, while probabh" allowing more or less water to pass through, acted as practically impervious masses along which water was deflected when it came into contact with them. It is a common opinion among miners that a few inches of soap rock is more effective in keeping out water than many feet of the iron-bearing formation. On the other hand the brittle siliceous ore-bearing formation was fractured by the folding to which it was subjected, so that where this process was extreme water passes through it as through a sieve. It is evident that the tilted bodies of greenstone, or soap rock, especially those that occur in pitching S3'nclines or that form pitching troughs bj^ the union of dikes and masses of green- stone, must have converged downward-flowing waters. It is also clear that the weak contact plane between the Goodrich quartzite and the Negaunee formation was one of accommodation and shattering, favorable for the free movement of waters. Finally, the ores in the Bijiki schist of the upper Iluronian have been developed by the percolation of waters along impervious slate basements with which the Bijiki schist has been folded. 276 GEOLOGY OF THE LAKE SUPERIOR REGION. Chemical and mineralogical changes in secondary concentration of Marquette ores. — The soft ores and the associated ferruginous cherts of the middle and lower horizons of tlie Negaunee formation arc similar physically, chemically, and mineralogically to the ores of the Penokec- Gogebic district. They are derived by the same processes, under similar conditions, from cherty iron carbonate rocks which arc practically identical with those of that district. The hard ores have undergone not only this change but the additional anamorphic changes of deep bu-rial antl igneous intrusion, the result being that the hard ores differ from the soft ores chemically only in that they have less water and a little less oxygen, mincralogicallv in that they have developed in them certain anhydrous silicates and some magnetite, and tex- turally in that they are coarsely crystalline and in places schistose. To some slight extent also similar hard ores may have been developed directly from the original cherty iron carbon- ates by deep burial or igneous contact action, but it is shown elsewhere that such action usually results in lean silicated iron-bearing rock rather than in rich ore bodies. The associated ferru- Quartz (chert) V N S . s V N S N \ Pore space, slump, secondary iron oxide and silica {relative proportions not known ) I>or.^. [>,-.<■. Ro^uit,I,^■fr..n, SoIUtl'jn -.'f ;lllL.i and reduction in volume of iron mineral \ \ \ s Quartz (chert) \ N Aluminum silicate Quartz JS Fore space QuarU Iron carbonate Kaolin Secondary iron oxide replacing iron carbonate- Amphibole Hematite dehydrated by pressure Aluminum silicate Hydrated iron oxide from oxidation of iron carbonate Hydra ted iron oxide from oxidation of iron carbonate Sideritic chert Ferruginous chert Soft ore Hard ore FiGUKE 37.— Graphic representation of the volume composition of the principal phases of the iron-bearing Negaimee formation, showing the changes in volume and mineral composition involved in the concentration of the ores from the cherty siderite and the production of hard ore from soft ore by dynamic agencies. ginous cherts or soft-ore jaspers umlergo similar changes so far as the iron oxide laj-ers are concerned. The chert beds are recrystallized, but not othermse changed. The result is a hard-ore "jasper or jaspilite differing from the ferruginous cherts in being more crystalline, having less pore space, and being less hj'drated, and accorduigly having red rather than yellow or brown colors. Volume changes in secondary concentration of Marquette ores. — The volume changes in the concentration of the ores and the development of the hard ores are shown in figure 37. The volume composition of the four phases of the iron-bearing formation is represented, thus permitting a consideration of porosity as well as mineral composition. The mineral compo- sition of the sideritic chert is calculated from a typical analysis." The mineral composition of the ferruginous chert is calculated from the sideritic chert analysis, allowing for oxidation of the iron mineral. The result is about an average for ferruginous cherts, as shown by analyses. The indicated volume compositions of the soft and hard ores represent actual average partial analyses of all ore as mined and averages of porosity determinations. \^^len subjected to oxidizing solutions, the siderite of the chert}'' siderite is oxidized to a more or less hydrated iron oxide, involving a considerable reduction in volume (see Gogebic discussion, pp. 242 et seq.) ranging from 49.25 per cent when the product is hematite to 18.3 per cent when limonite is produced. If no iron were introduced, the actual amount of oxide resulting would be intermediate between these two figures and j)robably would not tlitl'er greatly o Mon. U. S. Geol. Survey, vol. 28, 1897, p. 337, second analysis. MARQUETTE IRON DISTRICT. 277 from the hyclrated oxide of the soft ores, which is represented by a ratio of hematite to Umonite of 7 to 1. Even if a considei'able amount of iron were introduced, the resulting rock would be banded ferruginous chert having a larger pore space than the original cherty siderite. The reduction in volume of iron mineral accompanying the alteration of the carbonate is partly compensated by several factors, the relative importance of which is not known — by mechanical slump and by the introduction of secondary iron oxide and quai'tz. IRON MINERALS High-grade hard oie Low-grade hard ore Soft ore SILICA PORE SPACE Figure 38. —Triangular diagram showing the volume composition of the several grades of ore mined in the Marquette district in 1900, in terms of pore space, iron minerals, and silica. The altitudes of the small triangles show in each case the amount of minor constituents (amphibole, clay, etc.) The development of ore from the sideritic chert involves, in atldition to the oxidation of the iron in place, the removal in solution of a considerable amount of quartz. This gives a still larger pore space, which agam is partly compensated by slump and by infiltration of iron. Observation shows that the oxidation of the iron carbonate in place, producing ferruginous chert, mainly precedes the removal of the larger amount of silica. The oxidation of the iron is chemically more readily accomplished than. the solution of silica; and, further, the conse- quent development of pore space affords opportunity for more abundant flow of solution to accomplish the solution of silica. When the passage of the ore bodies into the chert or jasper is examined in detail it is found that a siliceous band, if followed toward the ore, instead of remaining solid becomes porous and may contain considerable cavities. These places in the 278 GEOLOGY OF THE LAKE SUPERIOR REGION. transition zone are lined with iron oxide. In passing toward the ore deposit more and more of the silica is found to liavc been removed, and iron oxide has partly replaced it. An exami- nation at many of the localities shows this transition from the l)andcd ore and jasper to take place as a consequence of the removal of the silica and the partial substitution of iron oxide. In many such instances the fine-grained part of the ore is that of the original rock, and the coarser crystalline material is a secondary infiltration. It is not uncommon, however, for the ore deposits to terminate abruptly along joint cracks or fractures. The solution of quartz and the introduction of iron oxide ultimately produce the soft ores from the ferruginous ciierts. These soft ores, as the diagram shows, have an average porosity of about 36 per cent and are made up essentially of hydrated iron oxide, quartz, and cla\-. The iron oxide largely represents siderite oxidized in place, but partly represents iron secondarily introduced. The development of the hard ores is accomplished by pressure or igneous contact action on the soft ores, causing a reduction in volume of approximately 40 per cent or less, by decreasing the porosit}^, dehydrating the iron oxide, and developing some magnetite and certain meta- morphic ferromagnesian, aluminum-bearing minerals, such as amphil)ole and garnet. Representation of ores and jaspers on triangular diagram. — The volume compositions of the various phases of the iron-bearing formation are represented in the triangular diagram, figure 38. (For explanation see p. 189.) The lines of demarcation between the hard and soft ores and between the low and high grade hard ores are not as sharp as the grouping of the small triangles would indicate. Tyj)ical specimens of each grade were selected and intermediate phases were neglected. If all phases were represented the entire upper corner of the large triangle would be covered with small ones, indicating complete gradation between the various classes of ore. SEQUENCE OF ORE CONCENTRATION IN THE MARQUETTE DISTRICT. 1. The alteration of the Negaunee formation began before upper Huronian time, when the formation had been slightly folded, eroded, and intruded by igneous rocks. Prior to upper Huronian time all the phases of the iron-bearing formation now known, except the specular hematites, had been developed, for all of them appear as pebbles in the basal conglomerate of the upper Huronian, and it is unhkely that such closely intermingled diversity of pebbles could have been developed from a single type of iron-bearing material after it had been deposited as pebbles in the conglomerate at the base of the upper Huronian. Erosion was not deep, and ores seem to have been developed' only near the erosion surface which bevels at a low angle the upper beds of the Negaunee formation and now constitutes the horizon exposed nearest to the overlying upper Huronian conglomerate. That ores M'ere formed at this time and place is indicated by the fact that at this horizon occur specular hematites having a secondary cleavage developed during the folding which followed the deposition of the upper Huronian and which preceded the second great period of ore concentration. 2. Inter-Huronian alteration of the formation was inten-upted b}' the deposition of the upper Huronian (Animikie group), the base of which was made up of conglomerate carrving fragments of ferruginous chert and iron ore derived from the Negaunee formation. A higher formation (Bijiki schist) contained iron carbonate. 3. The deposition of the upper Huronian was followed by severe folding and both intrusion and extrusion of the basic igneous rocks. Much of the intrusion preceded the folding, for the cleavage in the sedimentary beds developed during the fokling, and, having an attitude deter- mined by the differential movement between the folds, affects also the intrusive rocks. Many of these post-upper Huronian (Keweenawan) intrusive rocks are now found in the area of the Negaunee formation. It is certain that some of them — as, for instance, those in the vicinity of Michigamme — represent laccolithic masses which were unable to penetrate above the massive Goodrich quartzite and spread out in the upper portion of the Negaunee formation. The intrusion and folding, with varying relative efi'ectiveness in ilifferent parts of the range, anamor- MARQUETTE IRON DISTRICT. 279 pliosed the iron-bearing formations, but witli widely differing results, depending on the condi- tions of the iron formation before the anamorphism. The ferruginous cherts and ores of the upper horizons of the Negaunee formation were changed to hard hematites and jaspers, becom- ing specular when folded. The iron-bearing conglomerate at the base of the Goodrich quartzite was similarly affected. The iron carbonate of the Bijiki schist of the upper Huronian was changed into a coarsely crystalline amphibole-magnetite rock. Portions of the formations farther removed from the intrusive rocks were less anamorphosed. These would include the part of the Bijiki schist near the Bessie mine and the lower part of the Negaunee formation, both of which up to this time still remained as iron carbonate. Post-Keweenawan erosion exposed all phases of the iron-bearing Negaunee formation, together with the ferruginous detrital base of the upper Huronian and the still unaltered car- bonates higher in the upper Huronian. The iron carbonates, both of the lower parts of the Negaunee formation and of the Bijiki schist, now for the first time exposed, became altered in the ordinary manner, producing soft ores associated with soft ferruginous cherts, now found typically along the Teal Lake range and in the Bessie mine of the western Marquette district. The other phases of the Negaunee formation, which had been previously altered to chert, jasper or iron ore, or amphibole-magnetite rocks, were also attacked to some extent, principally by the leaching of siHca, which can be conspicuously observed in the loss of chert pebbles from the conglomerate at the base of the upper Huronian, and by alteration of garnets and amphibole to chlorite. The total effect of the alteration at this time on these harder phases, however, was probably not so essential in the concentration of the ore deposits as that which had gone on before. The great varieties of phases of the iron-bearing rocks of the Marquette district are therefore the results of katamorphic and anamorphic processes described in earlier pages, acting alone or successively on different parts of the iron-bearing formations. OCCURRENCE OF PHOSPHORUS IN THE MARQUETTE ORES. DISTRIBUTION OF PHOSPHOKUS. The ores of the Marquette range are as a whole higher in phosphorus than those of the Vermilion, Mesabi, or Gogebic districts. They also show a greater range in phosphorus content than the ores of any of these three districts. Of the total shipments of ore from the Marquette range in 1906 approximately 18 per cent was of Bessemer grade. The lowest phosphorus grade was Sheffield (Fe = 64.61, P = 0.029, P/Fe = 0.000448), and the highest phosphorus grade was Cambridge (Fe = 59.60, P = 0.570, P/Fe = 0.00957). The phosphorus and iron contents of the ores of the Marquette range are shown in the following table : . Phosphorus and iron content of Marquette ores. Iron. Phospho- rus. Ratio of phospho- rus to Average total sliipraents for 1906 ._ Average ore from Ijottom horizon of the Negaunee formation — Average ore from the middle horizon of the Negaunee formation Average of ore from upper horizon of the Negaunee formation. . . Average ores from upper Huronian Bijiki schist 69.55 58.38 57.22 59.00 65.91 0. 1072 .103 .063 .369 0.00180 .00176 .OOlliS . 00107 .00642 Six hundred partial analy.ses of jasper carrying between 20 and 50 per cent of iron, repre- senting 10,450 feet of drill holes in the area south of Negaunee, showed an average of 35 per cent of iron and 0.050 per cent of phosphonis. 280 GEOLOGY OF THE LAKE SUPERIOR REGION. The local distribution of phosphorus in the ores is extremely irregular. In many ore bodies fho pliospliorus coiitont is found to increase as the greenstone or soap roek (altered greenstone) walls are approached. This is shown by the foUowmg analyses of ore and greenstone collected from the Cliicago shaft of the Lake Superior Iron Company: Partial analyses of ore and greenstone from Chicago shaftM P. AljO^ CaO. Ore 2 feet from foot wall Paint rock (altered greenstone) 2 feet from contact (Jreen.stune foot wall, soft, S feet from contact Greenstone foot wall, hard. S feet from contact Greenstone foot wail, hard, 3;i feet from contact Greenstone foot wall, hard, 70 feet from contact. . . Altered greenstone (soap reel;) at contact" Fresh greenstone 80 feet from contact " 0.112 .192 .132 .134 .064 .106 .181 .090 1.12 4.67 6.41 15.30 0.19 .15 .22 .14 a The last two samples were from another part of the deposit. ^ s M 3S T m s m+ rpc m s ™ ffi m TTTp m I iiiiiiiii ;;; : ;: \ : : : : ;;;; ::: 1 :;; M :::: : ::: ::: HI lli:: V-' s i £ S ■■ j T- S ; ii i 4 s izJ S ^ ^ S - — - tttt ! 1 1 x::t: t — -^ CC O I a. in O X Q. U- O \- Z ' \ j i 1 , ii o cc 111 a. 1 i I'. 1 ._ __. — 1 ■ ■ ' mumm \m\\ ^ ftii^ itt i :^ m M h 1 M NT m LO as 3N GN Tl m £& i a m^ liiiiiiiFi-i i ;'i J 1 1 1 1 ;i ffmff a a ittiiMfTfrtl -f iiiiri^H Figure 39. — Diagram showing relation of phosphorus to degree of hydration in Marquette ores. Local variations in ])hosiihorus also occur apparently intle])endent of relations to green- stone walls or channels of flow, being due, perhaps, to origmal difl'erences in the iron-bearing formation. Typical of this is an occurrence in the Volunteer mine, where high-phosphorus ore is found against hangmg-wall jasper and low-])hosphorus ore against tlie jasper foot wall. The increase of phosphorus with degree of h3'dration is showTi in figure 39. Two wasliing tests similar to those made on Mesabi ores (see p. 193) were made on sam])les of soft red hematite ore from the Lake Angehne mine and the Hartford mine. The results of these tests are shown in the following table : MARQUETTE IRON DISTRICT. Partial analyses from washing tests on Marquette ores. 281 Fe. P. AI2O3. HjO. Lake AnReline mine: Heavy residue 65.00 02.53 61.20 02. 23 00. :i9 60.07 0.078 .100 .100 .120 .100 .080 0.89 1.56 2.20 2.30 2.27 2.64 2 12 2.43 Finest material 3 40 Hartford mine: 1 82 Medium 2 i6 Finest material 2 32 The test on the Lake Angeline ore gave results similar to those obtained from the tests on Mesabi ore, showing the association of phosphonis with the more hydrated parts of the ore. The washing test on the Hartford ore, however, does not show this relation. MINERALOGICAL OCCURRENCE OF PHOSPHORUS. Phosphoi-us is known to occur as apatite, dufrenite, and as aluminum ])hosphate. It probably occurs in a variety of combinations with iron, magnesium, calcium, and aluminum and in forms too minute to be identified. Apatite has been identified by Prof. Seaman " and others at a number of localities in the Negaunec formation and in the upper Huronian iron-ore deposits. In the chemical determination of phosjihoiiis it is found that only a part of it is soluble in hydrocliloric acid, the insoluble portion remaining with the sihceous residue. Tliis seems to indicate that phosphonis is present in at least two combinations. The soluble phosphorus may be present in a variety of combmations, as iron phosphate, calcium phosphate, alid some aluminum phosphates are soluble in hydrochloric acid. Charles T. Mixer and II. W. Dubois * analyzed the insoluble residue remaining after treating ore with hydrocldoric acid (1.10 specific gravity) and found its composition in percentages of original residue to be AljOj 9;55, CaO 0.92, P2O5 4.10, from wliich they concluded that the insoluble phosphorus is to a large extent combined with alumina. What this aluminum phosphate is it is impossible to say. It is of interest to note that the relative amounts of soluble and insoluble phosphonis are not uniform in the various ores; in some the insoluble form is entirely absent, but in others it makes up the greater part of the phosphorus present. It is believed by some of the chemists of the iron range that the insoluble phosphorus is highest in ores liigh in alumina. In order to ascertain the possibility of the phosphorus being present as apatite, the percentages of calcium oxide and phos- phoras, in the difl'ercnt grades of ore produced in 1906, were platted as ordinates and abscissas in figure 40. The diagonal line indicates the relative amounts of the two constituents in apatite. It may be seen that most of the points fall below the liuje, indicating an excess of iime over the amount requii'ed to combine with the phosphorus present as apatite. It is of interest to note that the high phosphorus ores are correspondingly high in lime, indicating rather strongly the possibility of at least a large part of the phosphorus being present in apatite. PHOSPHORUS IN RELATION TO SECONDARY CONCENTRATION. As shown in the table on page 279, there is apparently a gradation in the phosphorus content of the ores of the Negaunee formation, from comparatively low pliosphorus in those of the upper horizon to high phosphorus in those of the bottom horizon. The difference is most marked between the hard ores of the upper horizon and the soft ores of the middle horizon. The difference between the ores of the two lower horizons is very small and may be apparent rather than real. In explanation of the difference in phosphorus content between the hard and soft ores may be cited the opportunity for leacliing of pliosphorus from the upper strata during the erosion interval previous to the deposition of the Goodrich quartzite. Another possibility may be an original difference in the phosphorus contents of the ores at the two horizons. a Personal communication. i> Jour. Am. Chem. Soc, vol. 19, No. 8, p. 619. 282 GEOLOGY OF THE LAKE SUPERIOR REGION. ■ti- +!■■■ fl- : T,' ; J. f i: 1^ jiiip =^ • — 1 m 1 - i - U- " : O uJ -M — r ii..-. ' ^ 1 — — 1 "^ 'x. - ■n - \s^^^ 1 • iiir m i 1 ...-^ ' ■■ ,_ i ■ ^^\ SWANZY DISTRICT. 283 The abundant slaty phases of the Michigamme may have some bearing on the high plios- phorus of the ores, as in all the iron districts the slates are higher in phosphorus than the iron- bearing formation proper.. The local occurrence of high-phosphorus ore near greenstone contacts is believed to be due to dh-ect transfer of tliat constituent, leaclied from the greenstone during its alteration to soap rock or paint rock and deposited in the neighboring oi-es. The analyses on page 280 show tJiat there is actually a loss of pliosphorus in the alteration of the greenstone if alumina is assumed to have remained constant, although the actual percentage of phosphorus increases. Local variations, apparently not related to greenstone contacts, are probably due to origi- nal differences in the phosphorus content of the formation and not to secondary transfer or infiltration. SWANZY DISTRICT. GEOGRAPHY AND TOPOGRAPHY. The Swanzy iron district lies about 16 miles south of the city of Marquette, in T. 45 N., R. 25 W. (fig. 41). In 1908 the productive area was less than 2 miles long and about half a mile wide and contained five producing mines. Future exploration and development will undoubtedly extend the district to the south and east, but northward and westward extensions are apparently cut off by the granite area that bounds the district on these sides. The towns within the producing area are Gwinn and Princeton, both reached by the Munising Railway. The district occupies a range of hills typical of the granite area, and slopes on the south and east to a flat sand-covered phxin above which stand a few monadnocks of pre-Cambrian rocks. GENERAL SUCCESSION AND STRUCTURE. The succession is as follows: Quaternary system: Pleistocene Glacial deposits. Cambrian sandstone. ■ Ordovician limestone. Unconformity. Algonkian system: Huronian series: "Michigamme slate. Upper Huronian (Animikie group) . . Bijiki iron-bearing member. In lenses and layers near ba^^ of Michigamme slate. Goodrich quartzite. Quartz slate and quartzite, grad- ing down into arkose or recomposed granite. Unconformity. Archean system; Laurentian series Granite. The Swanzy district consists of a southeastward-pitching synclinorium of upper Huronian rocks, bounded on all but the southeast side by Archean granite. It is about 2 miles long; its width is for the most part not more than three-quarters of a mile, and at the narrowest point, near the Stevenson mine, is only half a mile. To the southeast it widens, but in this du-ection the structure is not known because of the deep overburden. The pitches of the minor folds at the Stegmiller, Princeton, and Swanzy mines are toward the northwest. The slates have developed a good cleavage, usually crossing the bedding. This structure does not affect the quartzite and the iron-bearing member. ARCHEAN SYSTEM. The Archean forms the basement upon which the Huronian sediments lie. It is repre- sented by granites similar to tlie basal granites of the neighboring iron ranges. The Archean bounds the district on the north, west, and southwest sides. Isolated expos- ures stand as monadnocks above the flat sand plains of the district. 284 GEOLOGY OF THE LAKE SUPERIOR REGION. NVIWNOOIV NV3HDyV 1 ■s Si" • * 4 SWANZY DISTRICT. 285 ALGONKIAN SYSTEM. HXJRONIAN SERIES. UPPER HURONIAN (aNIMIKIE GROUP). GOODRICH QTJARTZITE. The Goodrich sediments lie unconformably upon the Archean rocks. They consist of a coarse arkose or recomposed granite at the base, which grades upward tlirough quartzite and quartz slate to the Brjiki iron-bearing member of the Michigamme slate. The arkose horizon represents a shore phase of sedimentation where disintegration was very active and rapid transportation of the disintegrated material prevented decomposition. In places the arkose is distinguished from the granite with difficulty. The quartzite is petrographically very similar to the Goodrich quartzite of the Alarquette range and exhibits all phases of gradation between the arkose below and a thin-bedded quartz slate above. Both the quartzite and the quartz slate are locally iron stained, and in places the impregnation is so strong as to have attracted prospecting operations. The arkose phase is best exhibited in drill cores. The quartzite and quartz slate phases are well exposed in abundant outcrops on the north slope of the range of hills wliich crosses sec. 19, T. 45 N., R. 24 W. The quartzite also outcrops in a small hill near the northeast corner of sec. 18, T. 45 N., R. 24 W. The thickness of the quartzite and the quartz slate varies and locally the slate and jasper lie directly oh the recomposed granite or on the granite itself. MICHIGAMME SLATE. The Michigamme slate is best exposed at the old Swanzy open pit, near the center of sec. 18, T. 45 N., R. 25 W., where it is found in contact with the Bijiki iron-bearing member. It both underlies and overlies the iron-bearing beds, which are therefore treated as a member of the slate. The Michigamme forms much the larger part of the upper Huronian. The iron-bearing member is a banded ferruginous chert or "soft-ore jasper" similar in appearance to part of the Bijiki schist of the Marquette range. Locally it grades into a ferru- ginous slate. It apparently occurs in lens-shaped beds in and near the base of the Michigamme slate, and therefore it is treated as a member of that formation. Drilling and mining operations have shown jasper with slate above and below, or slate above and quartzite below, or in places the iron-bearing member is found directly above the arkose and overlain by slate, the quartzite and quartz slate being absent. The iron-bearing member is exposed at several places in the vicinity of the Princeton, Stegmiller, and Austin mines and also in the old Swanzy open pit. An exposure near the center of the SE. J sec. 18, T. 45 N., R. 25 W., shows typical banded soft-ore jasper with a nearly vertical dip. Near the southeast corner of the same section, just west of tlie Stegmiller mine, is a similar exposure. An exposure about 600 feet west of Princeton station shows the member folded and contorted. PALEOZOIC SEDIMENTS. On the east side of the district flat-lying sandstones antl limestones belonging to the Cam- brian and Ordovician overlap the pre-Cambrian formations unconformabh". The nearest exposure of limestone is in the northeast corner of sec. 18, T. 45 N., R. 24 W., where a small hill of quartzite has a few remnants of a limestone capping. QUATERNARY DEPOSITS. Pleistocene sand flats of glacial origin cover most of the district. (See Chapter XVI, pp. 427-459.) 286 GEOLOGY OF THE LAKE SUPERIOR REGION. CORRELATION. Tlie upper Iliironian (Animikie f^roup) is very similar, both in stratigraphy and in litliolojjy, to the upper Huronian of the Marquette district on tlie north and tlie Crystal Falls and Menom- inee districts on the south. THE IRON ORES OF THE SWANZY DISTRICT. By the authors and W. J. Mead. GENERAL DESCRIPTION. The ores of the Swanzy tlistrict are in the Bijiki iron-bearing memljer, which is interbcdded with the lower part of the Michigamme slate of tlie upper Huronian and rests upcm the Arcliean o-ranite with only a comparatively thin intervening zone of quartzites, quartz slate, or recom- posed granite, constituting the Goodrich quartzite. The upper Huronian constitutes 'a south- eastward-pitching synclinorium, ])ut some of the minor folds on its limbs pitch to tlie northwest. They are of the drag type so common to the Lake Superior region. (See fig. 12, p. 123.) The iron-bearing member takes part in this general structure. The ores therefore appear as much-folded deposits with foot wall of slate, quartzite, recomposed granite, or granite and with hanging wall of black slate. All the ore deposits reach the erosion surface either at the border of the s3'nclinorium or on the eroded minor anticlines in the main synclinoriiun. Five mines are in operation and several additional ore deposits are known. (See map, fig. 41.) The ore is a soft hydrated non-Bessemer hematite containing a rather high percentage of moisture. The following is the average composition of ore sluppetl in 1906: Average composition of ore shipped from Swanzy district in 1906. Moisture (loss on drying at 212°) 13. 50 Analysis of dried ore: Iron 58.60 Phosphorus 211 Silica 10. 20 Manganese 71 Alumina 1. 05 Lime 1. 15 Magnesia 46 Sulphur 012 Loss by ignition 1. 25 SECONDARY CONCENTRATION OF SWANZY ORE. The structural conditions governing the concentration of the ores in the Swanzy district are a foot wall of granite, quartzite, or slate and a hanging wall of slate, conforming to the structure of a sj'nclinorium that has a gentle southeastward pitch with many minor variations. Erosion has exposed the iron-bearing member near the borders of the synclinorium and along the arches of the minor anticlines. The circidation of the iron-bearing solutions has obviously been controlled n(jt only by the impervious basement but by the overlapping impervious forma- tions which determined tlieir points of escape. The ores and ferruginous cherts have been derived from the alteration of sidcritic cherts and slates, accompanied by the removal of silica and the development of pore space. MONOGRAPH Lii Plate «» GEOLOGIC MAP OF DEAD RIVER AREA, MICHIGAN H^ A.K.SKAMAN ScalR aiA>(i s//,r// I I u. ^k^lii^ LEGEND ALGONKIAN (Huronlan sei-ies) UPPER HURONIAN TANIMIKtE GROUP* OLE HUFtONIA Aus N«Ratiiii»« fbniinlii.in Siamo tilatr- Q witk Httfltc "H;. ,3S ~*. 33*4. * -^-^.^* '-^. ^ v_ ^-■ -'1' . '■V-.J.~:- -:^: L s^-^^;;>^r GEOLOGY OF THE LAKE SUPERIOR REGION. 287 DEAD RIVER AREA." The Dead River area lies north of the Marquette district along the Dead River. Its {greatest extent is 18 miles west-northwest and east-southeast. Its maximum width is 6 miles. (See PI. XX.) The basin is largely a low, flat sand-covered plain with an amphitheater of rock-exposed hills about it. GENERAL SUCCESSION. The general succession is as follows : Quaternary system: Pleistocene deposits. Unconformity. Algonkian system: Huronian series: Upper Huronian (Animikie group) Slates and conglomerate. Unconformity. {Negaunee formation (iron bearing). Siamo slate. .\iibik quartzite. Unconformity. Archean system: Laurentian series Graiiite intrusive into Keewatin series. Keewatia series, including Kitchi and Mona schifits. The Laurentian and Keewatin rocks occupy the liills siu-rounding the basin; the middle Huronian rocks outcrop along the margin of the basin, anci the upper Huronian (xVnimikie group) occupies nearly all of the basin itself. ARCHEAN SYSTEM. KEEWATIN SERIES. The Keewatin series forms hills along the northeast and southeast sides of the basin. The series includes on the south side the Kitchi and Mona schists, already described for the Mar- quette district, and on the north side schists entirely similar in aspect, even to their content of iron-bearing sediments consisting of jasper, cherty siderite, and cherty slate. Slate and con- glomerate are weU exposed at the German exploration in sec. 35, T. 49 N.-, R. 27 W., and in the Holyoke mine on the south side of the hiU. LATJBENTIAN SERIES. Laurentian granites and gneisses bound the Dead River district on the southwest, west, and northwest and also for a short distance along the southeast end. They are not different from the rocks of the northern complex of the Marquette district. ALGONKIAN SYSTEM. HURONIAN SERIES. Middle Huronian. — The middle Huronian is exposed along the south and southeast sides of the district and also at the extreme west end bordering the amphitheater of Keewatin and Laurentian rocks. The best exposed of these rocks is the Ajibik quartzite, forming the base of the middle Huronij^n, and showing unconformity between Laurentian and middle Huronian by discordance in structure and by conglomerates. The Siamo slate outcrops in a narrow belt along the north side of the Ajibik quartzite where it follows the south boundary of the district. The Negaunee formation (iron bearing) is exposed in only one area, in sec. 15, T. 48 N., R. 26 W., along the railway track and in pits. Here the iron-bearing formation, mth the under- o Mapped by A. E. Seaman, Michigan College of Mines. 288 GEOLOGY OF THE LAKE SLIPERIOR REGION. lying Siamo slate and Ajil)ik quartzite is much folded. Oveilyint; tlie iron-bearin<; formation (in direct contact in pits) is the basal conglomerate of the ui)per lluronian, containing fragments both of middle lluronian and Keewatin. Upper lluronian (Animikie group). — The upper lluronian consists princijjall}- of slates, similar in all respects to the Michigamme slate of the Marquette district. They outcrop in isolated exposures over the area and their presence is further indicated by the prevailing low relief of the basin. The base of these rocks is probabh' marked l)y the conglomerate resting unconformably on the Keewatin series at the Holyoke mine and eastward at intervals to the east end of the basin; also by the conglomerate covering the Negaunee formation, alreadj' referred to. The slates have not been connected directly with tlie conglomerate, l)ut tlie fact that the conglomerate contains fragments not only of Keewatin but of middle Hurunian rocks seems to require its correlation' with the upper Huronian. Greenstone dikes cut the slates. One of them constitutes the falls of Dead River where it cuts through the slates in sec. 9, T. 48 N., R. 26 W. PERCH LAKE DISTRICT (IXCLUDIXG WESTERN IMARQl'ETTE). GEOGRAPHY AND TOPOGRAPHY. The Perch Lake district includes territory extentlLng west from the Marquette district and north from the Ciystal Falls and Iron River districts to a line extending from L'^Vnse Bay on the northeast to the south end of Lake Gogebic on the southwest. The area thus defined includes roughly 1,200 square miles. (See fig. 42; PI. XXI, in pocket.) A topographic map has been prepared of the area around Perch Lake, extending from SS° 30' to 88° 45' west and 46° 15' to 46° 30' north. The remainder of the country has not been surveyed topographically. As a whole the country is characterized by morainal topography with much local irregularity, but has no consj^icuous ranges characteristic of the principal ore-pi'oducing districts. GENERAL, SUCCESSION. The succession is as follows, from the top downward: Quaternary system: Pleistocene or glacial deposits. Cambrian sandstone. Algonkian system: Huronian series: Upper Huronian (Animikie group). . Middle Huronian . Michigamme slate (slates and graywackes with pos- sible iron-bearing lenses). Equivalent and areally continuous with the Michigamme slate of the Crys- tal Falls, Iron River, and Menominee districts. Goodrich quartzite (quartzites and conglomerates). Intrusive diorite. Xegaunee formation (iron bearing). Siamo slate. .\jibik quartzite. Unconformity. Archean system: Laurentian series Granite and syenite. ARCHEAN SYSTEM. LAURENTIAN SEKIES. The Laurentian granite and syenite bound the district on the northeast. They show no features different from the Laurentian of the contiguous Marquette district. The rocks are abimdantly exposed. The topograph}' of the Archean area is as a whole rougiier and more irregular than that of the Algonkian on its southwestern margin, affording a very satisfactoiy guide for discrimination in the field mapping. The Archean underlies the Huronian uncon- formably. PERCH LAKE DISTRICT. 289 ALGONKIAN SYSTEM. HUBONIAN SERIES. Middle Huronian. — Between the upper Huronian slates and graywackes (Michigamme slate) and the Archean granite on the northeast there appears a belt about 5 miles long extend- ing from the Marquette district northwest, in which are exposed middle Huronian sediments and upper Huronian Goodrich quartzite. (See fig. 42.) The middle Huronian Ajibik quartz- ite and Siamo slate show no features different from those of the Marquette district. They rest unconformably against the Archean. On the northwest and along their trend they become covered by glacial materials until they can no longer be followed. Presumably they extend R.31 W. Eruptive diaba-se aiid diorite ///a Tvliclii^amme slate and Bijiki scliist (iron tjf-'oririg) Goodrich quartziie Negaunee formation {iron hearing) Siamo slate Ajibik qiiartzite 1# Granite 3 Miles Figure 42.— Geologic map of west end of Marquette district, Michigan. By W. N. Merriara and M. H. Newman. considerably farther than the map indicates. The Negaunee formation also is similar to the Negaunee formation of the Marquette district. It is followed, however, principally by mag- netic observations to the point indicated on the map, where it is lost beneath the covering of later drift. Wliether it extends farther or whether this represents the end of the originally deposited iron-bearing lens is not known. Upinr Huronian {AnimiJcie gi'oup). — The district is underlain principally by upper Hui-o- nian slates and graywackes, known as the Michigamme slate. On the northeast they rest unconformably against the Archean granite and middle Huronian rocks. On the northwest they are overlain by Cambrian sandstone, the relations of the two locally being obscured by faulting. The Goodrich quartzite of the upper Huronian is exposed only in the northeastern part of the area bordering the middle Huronian and at the northwest end, presumably over- 47517°— VOL 52—11 19 290 GEOLOGY OF THE LAKE SUPERIOR REGION. lapping on the Archean. Its characteristics are similar to those of the Goodrich quartzite of the Marquette district. The Michigamme slate covers much the larger part of the Perch Lake area. Exposures are fairly abundant, especially in the Perch Lake district. Presum- ably contemporaneous basic volcanic rocks are associated with these slates, to judge from the facts observed to the south, but their detailed (hstribution is not known. There Ls difli- culty in identil'ymg horizons in the slate and graywacke, and thei'efore in working out the structure of this area. From the abundance of exposures, however, it is probable that this may be accomplished in the future. The locations of most of the exposures have been noted in commercial surveys, but the Geological Survey has not examined this area in detail to work out the structure. From the promising development in similar series in the adjacent Iron River district, it would seem that this area would warrant careful examination for iron-bearing lenses. QUATERNARY DEPOSITS. Pleistocene glacial deposits cover all of this area. (See Chapter XVI, pp. 427-459.) CHAPTER XII. THE CRYSTAL FALLS, STURGEON, FELCH MOUN- TAIN, CALUMET, AND IRON RIVER IRON DISTRICTS OF MICHI- GAN AND THE FLORENCE IRON DISTRICT OF WISCONSIN. The Crystal Falls, Sturgeon, Felch Mountain, Calumet, and Iron River iron districts of Michigan and tlie Florence iron district of Wisconsin together form the ore-producing area between the Marquette district on the north and the Menominee district on the south. (See fig. 43.) The ores of all these districts occur in the upper Huronian (Animikie group) and have many similarities in kind and relations, and the limits of the several districts are poorly defined. They are accordingly grouped together in one cliapter. CRYSTAL FALLS IRON DISTRICT."^ LOCATION AND AREA. The Cr\'stal Falls district is centered in the town of that name in the Northern Peninsula of Michigan. (See PL XXII, in pocket.) As the term is here used it includes an area of about 540 square miles, covering all the territory between the Marquette and Menominee districts as • these have been limited on the maps of the United States Geological Survey. In commercial parlance the Menominee district includes the Crystal Falls and southwestward extensions, and reports of shipments for the Menominee district include these districts. However, they are geologically and structurally more or less independent and have been treated in two reports,' hence here the Crystal Falls district will be treated independently of the Menominee district. The Felch, Sturgeon, and Calumet troughs bordering the Crystal Falls district on the southeast are also discussed in this chapter, as well as the Iron River and Florence districts, which lie to the south and southwest. GENERAL, SUCCESSION AND STRUCTURE. The succession is as follows: Quaternary system: Pleistocene drift. Cambrian sandstone (in southern and eastern parts of district). Algonkian system: Huronian series: Upper Huronian (Animikie group). Unconformity (?). Middle Huronian (?) Volcanic rocks interbedded with slates. Michigamme slate. Thickness unknown, but proba- bly several thousand feet. Vulcan iron-bearing member, 300 feet. Negaunee (?) formation (iron bearing). Ajibik quartzite. Hemlock formation (volcanic), 1,000 to 10,000 feet. Includes at top iron-bearing slate member, 1 to 1,900 feet thick, formerly called "Mansfield slate." Unconformity (?). T Tr • (Randville dolomite, 500 to 1,. 500 feet. Lower Huronian i . ' , [Sturgeon quartzite, 100 to 1,000 feet. Unconformity. Archean system : Laurentian series Granites and gneisses. o For further detailed description of the geology of this district see Mon. U. S. Geol. Survey, vol. 36, 1899, and references there given. 6 Clements, I. M., and Smyth, H. L., The Crystal Falls iron-bearing district of Michigan: Mon. U. S. Geol. Survey, vol. 30, 1S99. Bayley, W. 8., The Menominee iron-bearing district of Michigan: Mon. U. S. Geol. Survey, vol. 46, 1904. 291 292 GEOLOGY OF THE LAKE SUPERIOR REGION. The northeastern part of the area is underlain by Archean granites. Bordering tliis main Archean area on the southwest, with longer axes parallel and striking north-northwest an,d c a .5 3 C P ^'^fl < . ^ Vf "o.!P ■o ^ '^ C 0) D Q. O Q. CD O •1 S o 1 S>E L oE 5S 39 N. N 9VX N SI' J. M 6C X N 9e i 2 ^ 1 south-southeast, are two minor oval areas of Archean granite. Iluronian sediments and basic igneous rocks, exposed ]irinci])ally in the western part of the district, lap around the Archean ovals and against; the main Archean area to the northeast, and their general structure is deter- CRYSTAL FALLS IRON DISTRICT. 293 mined by their relations to the Archean ovals. The Crystal Falls antl Amasa districts are on the southwest side of one of tlicse Archean ovals. Therefore both tlie di]) and (he pitch of the minor folds of the upper Iluronian occui)ying these areas are in southwesterly directions. ARCHEAN SYSTEM. LAUBENTIAN SERIES. The Archean or basement rocks occupy the northeastern part of the district, filling the angle between the Crystal Falls antl Marquette districts. To the west of this they also appear in two elliptical cores with longer axes north-northwest and south-southeast, approximately parallel to the axes of the major folds of the district. The Archean rocks consist mainly of massive and schistose granites and of gneisses. No- where in them have any rocks of sedimentary origin been discovered. They have been cut by igneous rocks, both basic and acidic, at diflerent epochs. These occur in the form both of bosses and of dikes, the latter in places cutting but more ordinarily showing a parallelism to the foliation of the schistose granites. The Ai-chean granites and gneisses and the earlier intrusive rocks alike have been profoundly metamorphosed, and at several places have been completely recrystallized. In the westernmost oval there is to be observed a distinct arrangement of feldspar crystals with their longer dimensions parallel to the contact with the Huronian rocks. ALGONKIAN SYSTEM. HURONIAN SERIES. LOWER HXJEONIAN. STURGEON QtTARTZITE. In the central part of the district the Sturgeon quartzite is represented only by thin frag- mental layers at the base of the overlying Randville dolomite. These are too thin to be mapped. Its principal outcrops are to the southeast in the Felcli Mountain and Sturgeon districts, de- scribed later in this chapter. RANDVILLE DOLOMITE. The Randville dolomite completely surroimds the Ai-chean oval northeast of the town of Crystal Falls. Here it constitutes the base of the sedimentary series and rests directly upon the Archean with only thin intervening layers of fragmental quartzose dolomite and recom- posed granite, all more or less altered to cpiartz schist and in many places difficult to distin- guish from schistose phases of the granite itself. On the west side of the western Archean oval the dolomite is poorly exposed and its thick- ness is not estimated. On the east side the belt is about half a mile wide and the thickness about 1,500 feet. The formation constitutes here an eastward-dipping monocline with mmor plications. In the scattered outcrops of the Michigamme Mountain area the dolomite strikes and dips toward all points of the compass as a result of the gentle arching from the general northwest-southeast axis, combined with sharp local folds which run nearly east and west. Petrographically the formation ranges from coarse saccharoidal marbles, in places very pure but usually filled with secondary silicates, to fine-grained, little-altered limestones, which are here and there so impure as to be calcareous or dolomitic sandstones and shales. The prevalent colors are white, but various shades of pink, liglit and deep blue, anil pale green occur. Some of the varieties are oolitic. This structure does not seem to have been previously noted in limestones of pre-Cambrian age in the Lake Superior region. 294 GEOLOGY OF THE LAKE SUPERIOR REGION. MIDDLE HURONIAN (?). HEMLOCK FORMATION. Distribution and general character. — The volcanic Hemlock formation occupies a large area in the Crystal Falls district. It is believed almost to surround the westernmost Archean oval and also to occur in a great area northwest of Crystal Falls and in one isolated area near the Mastodon mine, south of Crystal Falls. Its general stratigraphic position is conformably above the Randville dolomite and beneath the upper Huronian slates, but like most volcanic formations its relations differ in different parts of the district in ways wliich will appear below. Well-bedded cherty slates, iron-bearing lenses, and limestone are mterbedded with tlie Hem- lock foi'iiiation and also both underlie and overlie it. The volcanic extrusions may be regarded as interruptions of otherwise continuous deposition of sediments. The lack of con- tinuity of the volcanic flows antl of the interbedded sediments, and the difhculty of correlating the beds of either in different parts of the district, make it practically impossible to use geologic names for these sediments which will have anything more than very local significance. One of the prmcipal local sedimentary units within the Hemlock formation has been described and mapped in the United States Geological Survey monograph on the Crystal Falls district" as the "Mansfield slate." Limestone and slate layers appear abimdantly in the Hemlock formation near Hemlock River immediately northeast of the town of Amasa and in several other localities. Area south and west of the westemrnost Archean oval. — Exposures of the Hemlock formation are numerous west and south of the western Archean oval, and where erosion has removed the di'ift the formation has a marked influence on the topography. The thickness is estimated from the dip to reach 23,000 feet, but this is probably illusory because of reduplication due to fold- ing. The formation here consists mainly of bedded surface basic extrusive rocks and crystalline scliists derived fi'om them. Sedimentary rocks play a subordinate part. The Hemlock rocks are similar in all respects to the Keewatin volcanic rocks and to the volcanic Clarksburg formation of the Marcjuette district. The formation is cut by a few acicUc chkes and by numerous dikes and enormous bosses of basic rock. On the former Survey map of the district '' certain of these were discriminated, but they are not discriminated on the accompanying map (PI. XXII, in pocket), because more study has sho^\^l a most intimate association of extrusive and intru- sive phases of the formation tliroughout the area. The acidic intrusive rocks mclude rhyohte porphyry and aporhyolite porphyry. The rhyolite porphyry shows interesting micropoikihtic textural characters. Acithc pyroclastic rocks are scarce and were derived from the aporh3'olite. The basic lavas correspond to the modem basalts. They are much altered and are called " metabasalts." The basic lavas include nonporphyritic, porphyritic, and variolitic and eUip- soidal types. Clements " has described the ellipsoidal textures and concludes that basalts possessing this structure were origmally very viscous and correspond to the modern aa lavas, probably of submarine origin. The pyroclastic rocks comprise eruptive breccia, includiug friction breccias and flow breccias, and volcanic sedimentary rocks. The colian deposits, which are described as tuffs, grade from fine dust up to those in which the fragments are bowl- ders. The water-deposited volcanic fragmental rocks are known as volcanic conglomerates, and likewise range from those of which the particles are minute to those of which the fragments are very large. At many places occur clastic rocks which are now schistose and whose exact mode of origin — that is, whether eolian or water-deposited — could not be determined. The crystalline schists of Bone Lake include rocks of completely crystalline character, which by field and microscopic study have been connected Math the volcanic rocks and are considered to have been derived from rocks similar in nature to them. In general some of the volcanic rocks are submarine. The greater proportion, however, were derived from volcanic vents, which could not be located, but were probably situated near the Huronian shore line. Clements suggested that volcanic activity began in the north and a Mon. U. S. Geol. Survey, vol. 36, 1899, pp. 54-73. b Idem, PI. III. « Idem, pp. 112-124. CRYSTAL FALLS lEON DISTRICT. 295 moved to the south, and that some of the volcanic deposits to the north are contemporaneous with the so-called "Mansfield slate." Fence River area. — In the Fence River area the Hemlock formation occupies a belt between 2,000 and 3,000 feet in width, between the Randville dolomite on the west and the Negaunee formation on the east. The best exposures occur on the sections made by Fence River. No folds have been observed within the formation. The tliickness probably ranges up to 2,300 feet as a maximum. The rocks of the formation in tliis area are cliiefly chlorite and ophitic schists, with which are associated schists bearing biotite, ilmenite, and ottrelite, greenstone, conglomerates or agglomerates, and amygdaloids. As evidence of the origm of these schists several facts may be cited. First, they include no rocks possessmg any sedimentary characters; next, lavas and also greenstone conglomerates or agglomerates are undoubtedly present in the series; furtherinore, the minerals wliich compose the schist are those wliich would result from the alteration in connection with dynamic metamorphism of igneous rocks of basic or inter- mediate chemical composition; and finally, the grain and character of the groundmass and in some sUdes the presence of plagioclase microlites disposed in oval lines point dii-ectly to an igneous origin and to consohdation at the surface. The conclusion is reached that the Hemlock formation of the Fence River area is composed of a series of old lava flows varying in compo- sition from acichc to basic. Other areas of the Hemloclc formation. — Other areas of volcanic rocks similar to those of the Hemlock formation appear to the north and west of the town of Crystal Falls, near the Mastodon mine, and elsewhere, as shown on the accompanymg general map of the Crystal Falls district (PL XXII). Wliether these are of the same age as the main mass of the Hemlock formation and owe their distribution to folding, or whether they are later extrusions, is not yet known. Ironr-hearing slate member {"Mansfield slate") of the Hemlock formation. — The so-called "Mansfield slate," wliich is interbedded near the top of the Hemlock formation, is best exposed in the vicmity of the town of Mansfield. It here occupies a valley through wliich flows ^lichi- gamme River. Petrograplucally the member includes graywackes, clay slate, phyllite, siderite slate, chert, ferruginous chert, and iron ores, with several metamorphic products derived from them. The strike is north and south and the dip on an average 80° W. The maximum thick- ness of the belt is 1,900 feet. Southward from the point of maximum tliickness it rapidly thins out and disappears. The iron-bearing beds form a belt 32 feet wide or less between black slate walls. The strike and dip are the same as those of the slate. A single ore body of commercial importance has been mined. (See p. 324.) The Hemlock formation both east and west of the main belt of this slate carries thm bands of slate with similar strike and dip. In general, in this vicuiity, there is a monoclinical west- ward-dipping succession of volcanic rocks extending 2 miles or more east of Mansfield and about the same distance west, containing interbedded layers of slate, which in the vicinity of Mansfield are in considerable abundance and include also iron-bearing beds. These rocks may be best seen on the hill just east of Michigamme River, southeast of the Mansfield mine, where eight or ten layers of cherty slate from a few inches to 10 feet or more in width are interlay ered with westward-dipping ellipsoidal basalt flows. The centei-s of the flows are usually homogeneous and coarse grained, and the ellipsoidal structures appear only within a few feet of the top or bottom of the flow immediately next to the slate. As a whole the contact between the basalt and the slate is a plane surface, making it possible to follow a bed of slate even 2 feet thick for hundreds of feet. In detafl, however, the contact may be very irregular, following interstices in the ellipsoidal surface as if deposited upon an initially irregular surface. Slates mapi)ed as "Mansfield" by Smyth also outcrop on Michigamme Mountain and thence at intervals for 6 miles to the northwest. The area northwest of Michigamme Mountain, mapped as Pleistocene on Plate XXII, is believed to be largely underlain by slate from its appearance in a few pits and exposures. The information is so meager, however, that it is not thought desirable to map this area as slate. On Michigamme Mountain the geologic position of the 296 GEOLOGY OF THE LAKE SUPEIUOR REGION. so-called "Mansfield" rocks is free from doubt. In the principal synclinc of sec. 32, T. 44 N., R. 31 W., they overhe tlio dolomites and j)ass downward hitcj them hy a relatively slow gradation ; on the borders of the MicJugamme Mountain syncline they underlie the iron-bearing Negaunee ("Groveland") formation. The j)assagc to the higher formation likewise is graded, though rapidly, and is marked m certaui bands by an increase m clastic grains and by changes in the character of the matrix in whicii these are set. The average thickness of the formation in tliis mountain is not less tliau 400 feet. NEGAUNEE 0) FORMATION. Magnetic helts northeast of Fence River. — By reference to tlie map (PI. XXII, in pocket) it will be noted that there is a magnetic line marked "A" along the west side of the mam north- eastern area of Ai'chean rocks. That tliis magnetic line is caused bj' and marks the position of the ii-on-bearing Negaunee formation there can not be much doubt, according to Snij-th," for that rock outcrops in a few scattered localities, occurs abundantly in the drift, and has been found in test pits and drill holes here and there along this Une. The underlying quartzite outcrops beneath tlie non-bearing formation near the north end of the Une, but farther south it is entirely covered by the drift, so far as the territory has been examined. The overlying upper Huronian rocks ai"e also known to be present just west of the Negaunee formation as far south as sec. 19, T. 46 N., R. 30 W. The dip along the "A" line is probably therefore, on the whole, toward the west, although the observed dips at the few locahties where deter- muiations have been made are either vertical or slighth' mchned fi-om the vertical toward the east. In an east-west section of driU lioles in sees. IS, 19, 29, and 30, T. 46 N., R. 30 W., cutting the magnetic belt "A," the iron-bearing formation is found to be amphibole-magnetite rock cut by intrusives. Ai-ound the immecUately adjacent Archean oval on the west the magnetic line "B" has been traced for 25 miles without a single exposure. The known facts with reference to the "B" line, according to Smyth,'' are these: (1) It represents a magnetic rock; (2) this magnetic rock completely eucncles an Archean core. It may further be inferred with practical cer- tainty that this formation, winch carries such constant magnetic properties for 25 miles, must be sedimentary. With regard to its structure the foregomg considerations would necessarily' involve the conclusion that it dips away from the Archean core on all sides, ami this conclusion is fortified by the unsymmetrical separation of the horizontal maxima on tlie magnetic cross sections. East of the "B" line, between it and the "A" line, is found the basal member of the upper Huronian. The rock which is manifest m the "B" Ihie must, therefore, be older than any member of the upper Huronian. The Negaunee formation, represented in the "A" fine, dips west, but the rock of the "B" hne dips east. They are both older tlian the basal member of the upper Huroruan and are both younger than the Archean. They are l)oth strongly and persistently magnetic. For 8 or 10 miles the}' run parallel to each other less than half a mile apart. Their broad structural relations to the Archean basement of the region are precisely similar. Therefore, although the rock that gives rise to the "B" line has never yet been seen, it may be concluded with confitience that it is the Negaunee formation, and that the "A" and "B" lines represent this rock brought up in the two limbs of a narrow and probably deep synclmal fold. Negaunee (?) formation at MicMgamme Mountain and in tlie Fence Jiiver area. — The known outcrops of u'on-l)earmg formation (previously mapped as "Groveland" formation) in tlus belt are limited to three localities — the vicinity of Michigamme Mountain, in sec. 33, T. 44 N., R. 31 W., and sec. 3, T. 43 N., R. 31 W.; the exposures and test pits at the Sholdice explora- tion, in sec. 21, T. 45 N., R. 31 W.; and the test pits at the Doane exploration, in sec. 16, T. 45 N., R. 31 W. The last two localities are 1 mile apart, and the more southern is 8 miles north of Michigamme Mountain. <■ Van Hise, C. U., Clements, J. M., and Smyth, H. L., The Crystal Falls iron-bearing district of Michigan: Jlon. U. S. Geol. Survey, vol. 3fi, 1899, p. 453. b Idem, p. 454. CRYSTAL FALLS IRON DISTRICT. 297 Magnetic lines connect the outcrops on Micliigamme Mountain witi: tliose to the north. The magnetic line also extends beyond the outcrops around the north side of the western Ai-chean oval. The eastern belt was not traced farther than a mile southeast of Micliigamme Mountain. In the central and southeastern portions of T. 43 N., R. 31 W., however, in the direct prolongation of the anticlinal axis, is a broad belt of slight magnetic disturbance, along the western margin of wliich he volcanic rocks, dipping west. In sec. 26, T. 43 N., R. 31 W., this magnetic belt splits into two branches, one of which runs directly east for a mile and then southeast mdefinitely, while the other maintains a general southerly coiu-se to the south luie of the townsliip. In sec. 26 large angular bowlders, evidently tierived fi"om the iron-bearing formation, are found in the zone of magnetic disturbance, but no outcrops have been discovered. There can be little doubt that these disturbances roughly outline the position of the Vulcan formation m the axial region. Except in Micliigamme Mountain, the most elevated pomt of the district, the n-on-bearing formation is not topographically proimnent. In the Fence River area it produces a more subdued and somewhat lower-lying surface than the underlj-mg formation, but the difference is slight and is of Uttle moment in comparison with the confusing effects of glaciation. At Micliigamme Mountam the iron-bearuig formation caps the hill hi a well-marked syncluie, the axis of which runs northwest and southeast. The structure is distmctly shown by the attitude both of the ferruginous rocks and of the underlyuig phjdUtes ("Mansfield slate"). At the Interrange exploration, half a mile to the south, is found a secondary but more open embayment of the same syncluie. These are the only folds of the Micliigamme Mountain area sufficiently deep to include the iron-bearmg rocks. The thickness of the formation can only be guessed at, as no complete section is exposed, and the data for determmmg its upper limit are decidedly shadowy. The magnetic observations mdicate a breadth of 400 to 600 feet, aind as in the Fence River area it is certainly much thinner than the two lower formations its thickness may be approximately 500 feet. The rocks are interbanded ferruginous quartzite and actinolite and griinerite scliists, which still contain evidence of detrital origin. The formation contains less iron than the Vulcan formation of the Felch district, and consecjuently the Ughter-colored varieties are niore abundant, it contains more detrital material, and in the Michigamme Mountain area the texture is generally closer and less granular. Moreover, in passing north from the Micliigamme Mountain area to the Fence River area we find at the Sholdice and Doane explorations that the lower portion of the formation is composed of ferruginous quartzite, which is succeeded higher up by actinolite schists and griinerite scliists similar in all respects to the characteristic rocks of the Negaunee formation in the western Marquette district. The stratigrapliic position of the iron-bearing formation is above the Hemlock formation on Micliigamme Mountain ; to the west of the mountain the formation is apparently below the Hemlock formation; to the north of the mountain, in the Fence River area, it is above the Hemlock formation. In the last-named area nothing is known of the nature of the overljang rocks. Tliis iron-beaiing formation is doubtfully called Negaunee because of its lithologic character and because it comes ^\•ithin 2 miles of the ''B" line of attraction, regarded by Smyth as Negau- nee, suggesting that it is the same belt brought up again on the west side of this intervening gap of 2 miles by synchnal structure. On the other hand, it is nearly connected by a magnetic belt around the north side of the oval with the Vulcan formation and for this reason its correla- tion has been regarded as doubtful. However, by reference to the map (PI. XXII, in pocket), it \\"ill be noted that this belt of supposed Negaunee, extending around the north side of the oval and south as far as the north line of T. 45 N., R. 33 W., fails to connect b}^ nearly 2 miles with the known Vulcan formation, which is represented by a magnetic line running as far north as sec. 16, T. 45 N., R. 33 W. Moreover, at the north end, near the Red Rock mine, the Vulcan is associated with conglomerate carrying fragments of an earlier iron-bearing formation very suggestive of unconformity. Still further, the iron-bearing Vulcan formation where last seen. 298 GEOLOGY OF THE LAKE SUPERIOR REGIOTs^. is associated with red slates and apparently unaltered, while the rooks associated with the magnetic line to the north, supposed to represent the Negaunee, are micaceous and amphibolitic slates and scliists showing a mucii higher degree of metamorphism. Ferruginous quartzite associated with irorir-b earing formution north of Michigamme Mountain. — Ferruginous quartzite is found in isolated exposures in sees. 27 and 34, T. 44 N., R. 31 W., Michigan, lying inuneOGY OF THE T.AKE SirPERIOR REGION. in pocket.) Here tlie rocks are well exposed in numerous outcrops. The dip is about vertical iind the strike sH<;iitly west of nortli, wliicii is the direction of cloiij^ation of the field. Under the microscope tlie rocks are seen to contain innumcraljie small f^rains of maf^nctite associated with iilmndaiit chlorite and finely crystalline quartz and considerable siderite. A ina<;netic field of about the same size and shape occurs in the SW. \ sec. .33, T. 43 N., R. 34 W. (see PI. XXIV), but here the field is elongated in a northwest-southeast direction, which is likewise believed to indicate the strike of the rocks at this place, although no exposures occur. Local magnetism occurs also in separated patches in sees. 35 and 30, T. 43 X., R. 34 W. Here the magnetic rock is mainly a graywacke carrying abundant magnetite associated with chlorite, biotitc, and siderite. To the west of the Iron River district projjcr a belt of magnetic attraction has been traced in an area of heavy drift from a point near the center of T. 43 N., R. 37 W., westward to the Michigan boundary and thence probably into Wisconsin. Slate and i ■' graywacke INTRUSIVE AND EXTRUSIVE ROCKS IN THE UPPER HURONIAN (ANIMIKIE GROUP,. Igneous rocks of basaltic type are abundant in the upper Hu- ronian. The distribution of those now known is indicated on the accompanying map of the Iron River district. (See PI. XXIV, in pocket.) There is much difficulty in determining the general distribu- tion of tliese rocks, because the relations to the slates are so intricate that it is never safe to conclude that adjacent exposures are or are not separated by slate. The rocks are principally of extrusive type and have surface textures, especially the ellipsoidal and agglomj-ratic textures, that are characteristic of the Hemlock formation and of the volcanic rocks associated with the upper Iluronian of tlie C'lystal Falls dis- trict. Some of these extrusive rocks arc distinctly contemjioraneous with the slates. Southwest of Atkinson agglomeratic and tuffaceous phases of the greenstone are interbedded with upper Iluronian slate SE and iron-bearing member (fig. 44). In the southern part of the dis- FiGURE44.-sectionshomng roughly ^j.^^^^ j,^ gg^. 93 T. 42 X., R. 34 W., elHpsoidal and tuffaceous green- the succession of beds in the \ i:!- 1 c i tt • 1 • 11 can iron-bearing member near Ai- stoue occurs north of the Upper Huroniau slates m a uorthward- kinson, in the Iron River district, clipping series. From the lack of contact metamorphism and the abundance of tuffaceous phases and effusive rocks they were prob- ably nearly all deposited contemporaneously with the sediments. The deposition was prob- ably submarine. (See pp. 510-.512.) Definite evitlence of relations is lacking for many of the greenstones, especially those not adjacent to slates or some of those which have been developed by mining operations and explorations. Iron formation Slate Iron formation Tuff Iron formation Black slate RELATIONS OF UPPER HURONIAN (ANIMIKIE GROUPS TO UNDERLYING ROCKS. No direct evidence of the relations of the upper Iluronian with the undcrlnng Saunders formation is yet available. Certain slates conformable with the Saunders formation in Sheriilan Hill may be upper Huronian slates and may therefore indicate the conformable relations between the upper Huronian slates and the Saunders formation. The fact that rocks of the Saunders type form a continuous belt between the upper Huronian slates and the supposed Archean shore to the south is evidence of nearly conformable relations. It is noteil in the sections on the Crystal Falls, Menominee, Felch Mountain, and Calumet districts tluit the succession from underlying quartzite and dolomite to the upj)er Huronian shows similar relations. (For dis- cussion of correlation and nomenclature, see pp. 597 et seq.) IRON RIVER DISTRICT. 319 ORDOVICIAN ROCKS. Remnants of flat-lying Paleozoic rocks occur in the southern part of the district, on Sheri- dan Hill and vicinity and farther southwest in the SW. i sec. 27, T. 42 N., R. 35 W., also in the SE. } sec. 24, T." 44 N., R. 35 W. The base of these rocks on Sheridan Hill is a conglomerate made up almost entirely of material from the underlying Saunders formation. Angular fragments of chert and vitreous quartzite up to 2 inches in diameter lie in a matrix of materials of the same general composi- tion, but finer grained. The rock is cemented mainly with iron oxide and calcium carbonate. The tliickness of the conglomerate is unknown but is not great. The rock has not been found in natural exposure, but is abundant on the dumps of pits wliich have been sunk through it into the Saunders formation. The conglomerate is overlain by a coarse quartz santlstone of buff and red color and gen- erally very friable texture. The cement is mainly iron oxide. Under a slight tap of the hammer the rock falls apart into its constituent sand grains. The thickness of this sandstone is not known, but it probably ranges from a knife-edge up to perhaps .S5 or 40 feet. In the southeast corner of sec. 24, T. 44 N., R. 35 W., a film of red sandstone is found mantling black slate. Here the rock carries considerable iron oxide, doubtless derived from the Vulcan member occurring about a quarter of a mile north of it. The conglomerate and sandstone of tliese areas have the lithologic characters of the lower- most Cambrian beds in the Menominee district and were formerly correlated w^th the Cambrian. Also Seaman has suggested that they perhaps represent the base of the upper Iluronian. Recent fossil discoveries, however, in flaggy limestone beds in the S. 5 SW. | sec. 27, T. 42 N., R. 85 W., have fixed witliin narrow limits the age of these rocks. In this area there is one natural exposure on the east side of Brule River and several pits, all showing nonmagnesian dove-colored to buff flaggy hmestone of the same general characters. The rock seems to be flat-l3ang, although the beds iii the outcrop on the Brule, where observations were made and where most of the fossils were found, have been disturbed by slump, following undercutting by the river. From the position of this outcrop in reference to an exposure of the Saunders formation on the west side of the river about 500 paces south, it woul^l seem that these rocks are not far above the eroded surface of the Saunders formation. 'VMiether they are underlain by the conglomerate and sandstone of Sheridan Hill is not known. The beds are practically undisturbed in both areas, but the lowermost kno\vn occurrence of the conglomerate on Sheri- dan Hill is about 150 feet higher and the uppermost known beds of sandstone are about 300 feet liigher than tjie hmestone outcrops on Brule River in sec. 27. It would seem from this that the conglomerate and sandstone on Sheridan Hill are stratigraphically higher than the limestone of sec. 27. Doubtless the conglomerate originally formed a continuous mantle at the base of the Paleozoic rocks, but owing to the rugged character of the surface over wliich the sea advanced there was probably a considerable time interval between the submergence of the lower areas and that of the tops of the liills. Consequently the relative age of the basal mendaer formed at any point is a function of its altitude at that place. The occurrence of sandstone on Sheridan Hill at an altitude of about 1 ,760 feet makes it certain that the entire district was almost if not entirely covered by a Paleozoic sea. The lowest exposure of the Paleozoic beds is the limestone member in sec. 27, T. 42 N., R. 35 W. Tliis hmestone is correlated by E. O. Ulrich on paleontologic grounds with the Lowville of New York and the Platte\'ille hmestone of Wisconsin — that is, with the Middle Ordovician. The following is Mr. Ulrich's report to T. W. Stanton: I beg leave to report as follows on the fossils collected in the Iron River district, Michigan, by R. C. Allen and forwarded to the Survey for examination and report by C. K. Leith November 18, 1909: This discovery of fossils in northern Michigan is of great interest, as it adds an important link in proving the former connection of the early Mohawkian limestone of Minnesota and western Ontario across northern Wisconsin. In discussing the Lowville limestone in my paper on revision of Paleozoic systems I state my conviction that this ind perhaps other Mohawkian formations must have originally extended from New York through Ontario, northern 320 GEOLOGY OF THE LAIvE SUPERIOR REGION. Michigan, and northern Wisconsin to Minnesota and Iowa. Tliis direct westerly connection wa.s indicated by the great similarity in fauna and lithology noted in comparing the Lowville limestone in New York and the more typical part of tlie Platteville limestone of southern Minnesota, Iowa, southern Wisconsin, and northwestern Illinois. I objerted to comraiitiication via southeastern Wisconsin because there the beds supposed to correspond in age to the Lowville are dolomites instead of pure limestone, with no indication of transition in lithic characters northward. Hitherto the northern connection could not be established farther west from New York than Escanaba, Mich. This Iron River occiurence, which is of the .same fine-grained noinnagnesian dove-colored limestone everywhere charac- terizing the Lowville and lying well up on the old "Wisconsin Peninsula," may therefore ju.stly be regarded as tending to establish a \iew hitherto based only on inference. The following 20 species are more or less confidently identified. All are older than the Trenton limestone and younger than the latest Stones River. ?Coreniatocladus densus. Tetr.uiium cellulosum (fragment of tube only). Rhinidictya cf. nicholsoni and mutabilis-minor (fragment). R. cf. major (fragment). Escharopora angularis. ?Homotrypa arbuscula. Raftne?quina minnesotensis. Strophomena incurvata (Lowville var.). Zygospira recurvirostris (Lowville var.). Ctenodonta sp. undet. (near C. levata). Leperditia fabnlites. Lcperditella tumida. L. germana. Bythocj'pris granti var. Eurychilinia reticulata. E. subradiata. E. n. sp. Isotelus cf . obtusus. Thak'ops ct. ovatus. Pterj'gometopus sp. undet. (pygidium). The fossils of the above list indicate a horizon at the extreme top of the Platteville limestone in the Lead district. Compared with the New York section the bed corresponds in age to the uppermost beds of the Lowville, asdescribed by Gushing, or to the cherty bed at the base of the Black Ri^'er limestone, as defined by the same author. FLORENCE (COMJklONWEALTII) IRON DISTRICT OF WISCONSIN. LOCATION AND GENERAL SUCCESSION. The Florence district is the westward geographic extension mto Wisconsin of the Menomi- nee district bej'ond Menomuree River. It is essential]}" included between the two tributaries of the Menominee, the Brule on the north and the Pine on the south (PI. XXV, in pocket). On the east it is separated from the Menominee district, as this is limited on the geologic map, by Menominee River, ^he area is one of low relief, hke the Iron River district to the north- west. Exposures are relatively few except along the rivers and lakes. Part of the Florence district has been studied by members of the United States Geological Survey, and a complete outcrop map of the district has been prepared by ^Mi-. W. N. Merriam and assistants for the OUver Iron Mmmg Company. As yet, however, the district has not been studied with sufficient exhaust iveness to definite^ estabhsh the succession and structure. Such a study is now being conducted by W. O. Hotchkiss, State geologist of Wisconsm. So far as the facts are now known, including those developed in recent work of Hotchkiss, the succession in the Florence district seems t(5 be as follows : Quaternary system: Pleistocene deposits. Paleozoic rocks Patches of sandstone, probably Cambrian. Algonkian system: Keweenawan(?) series Granite and gneiss. (Juinncsec schist, intrusive and extrusive green- Huronian series: stones and green schists. Upper Huronian (Animikie group) . . . P> "•'■gamme slate, includmg the \ u can u-on-beanng member (inlerbedded with base of the slates), and also quartzites and conglomerates of doubtful age but believed to be phases of the slate. FLORENCE IRON DISTRICT. 321 ALGONKIAN SYSTEM. HtTBONIAN SERIES. UPPER HUEONIAN (aNIMIKIE GROUp). MICHIGAMME SLATE. General character and distribution. — The Animikie group seems to occupy nearly all the area of the Florence district north of the Qumnesec schist belt, except where small patches of intrusive or extrusive greenstone appear at the surface. Tlie rocks are cliiefly slate. In less quantity occur conglomerate, quartzite, tuffs, and iron-bearing rocks. It has not been proved that all these rocks belong to one group, but as yet they have not been certainly separated. The Michigamme slate is poorly exposed in the district as a whole, except along Brule River, in the vicinity of Keyes Lake, and northwest and southeast of Florence. It is ahnost identical in petrographic characters with the u]iper Iluronian slates of the Menominee and Crystal Falls districts, and has been regarded as belonging to the same formation. Quartzites, associated with more or less conglomerate, appear m three main areas — (1) at Island Rapids, on Menominee River, m sees. 1.3 and 14, T. 40 N., R. 18 E.; (2) in a belt running north of Keyes Lake; and (3) in a belt running through sec. 28, T. 39 N., R. 18 E., north of Pine River. The quartzite at Island Rapids stands vertical or dips steeply to the south, and the top is to the south. In tlie Keyes Lake belt the rock is vertical or dipping steeply to the southwest. The relations of these two belts witli the slates are not known defi- nitely, but are probably conformable. The southern belt of quartzite just north of Pine River dips southwestward at a lower angle. It is thought by Hotchkiss to rest unconformably upon the slates to the north of it. If this is true the so-called upper Iluronian of this district con- sists really of two groups, the correlation of which is doubtful. The southern quartzite is overlain conformably by slates which upward become uiterbedded with tuffs and eruptives belonging to the Quinnesec schist. Vulcan iron-hearing member. — The Vulcan iron-bearing formation is somewhat widely dis- tributed through the upper Huronian area, but here it is so interbedded with the slates that it is difficult to map independently. In this district, therefore, as in some other districts, it is treated as a member of the Michigamme slate. In the Florence district tliere are only five areas in which the ferruginous phases of the upper Huronian are now known sufficiently well to warrant a separate color on the map — one is immediately northwest of Florence in sees. 20 and 21, T. 40 N., R. 18 E., and in a belt extending northwestward to Brule River; two south- east of Commonwealth, in sees. 33 and 34, T. 40 N., R. 18 E. ; one extending east and west south of the greenstone belt in sees. 8 and 9, T. 39 N., R. 19 E. These three exposed areas are connected by a belt of magnetic attraction, indicating that the ii'on formation is probably continuous from Brule River on the northwest nearly to Menominee River on the southeast. Another area is in the vicinity of the Buckeye mine, just to the southwest of Commonwealth. This connects with a magnetic belt running southeastward to Menominee River, in sec. 22, T. 39 N., R. 19 E. To the east, across the river, this magnetic line connects with the principal iron-formation l)elt of the Menominee district. Another belt of iron-bearing formation out- crops west of Keyes Lake, whence it is followed by magnetic Imes to the southeast to about the east side of T. 39 N., R. 18 E., and northwestward toward the northwest corner of T. 40 N., R. 17 E. Belts of attraction not connected with any well-exposed areas of iron formation are known elsewhere m the tlistrict. Particularly to be mentioned are the belts extendmg north- westward from Pine River from sec. 28, T. 39 N., R. 18 E. The iron-bearmg member is magnetic in places, especially along the contacts with the intrusive greenstones. The map shows a number of disconnected magnetic lines which have been traced in this area. Some of these may represent altered iron-bearing rock. The Vulcan iron-bearing member consists of (1) ferruginous chert, siderite, and hydrated hematite; (2) various phases intermediate between these and the slates, called sideritic slates 47517°- VOL 52— 11 2] 322 GEOLOGY OF THE LAKE SUPERIOR REGION. and ferruginous slates; and (3) griineritic and magnetic slates. They are similar, except for type 3, to the rocks of the Vulcan iron-beuring member in tlio Iron River and Crystal Falls districts. Iron ores are exploited at the Florence mine, immeiliately northwest of the town of Florence; at the Commonwealth and Badger mines, southeast of the town of Commonwealth; and at the Buckeye mme, south of Commonwealth. (.See p. 323.) The ores seem to be in minor drag folds, pitching steeply northwestward m the Florence and Commonwealth mines. The major trend of the iron-bearing exposures of magnetic belts and of exposures of other rocks is north of west in this district, a trend which would tend to connect the iron-bearing belts with those of the Menominee district on the southeast and with those of the Mastodon area in the southern part of the Crystal Falls district on the northwest. (See p. 292.) Ex])loration has been very slight, as there has been little to guide it. However, there is a large territory along the trend here noted which inust soon receive attention. The horizon in the upper Huronian slates at which the iron-bearing member of this dis- trict occurs has not been determined. The proximity to the upper Huronian iron formation of the Menominee district suggests its occurrence near the base of the upjier Huronian. INTRUSIVE AND EXTRTTSIVE GREENSTONES AND GREEN SCHISTS. Quinnesec schist. — The Quinnesec schist outcrops in an east-west belt 1 to 3 miles wide along the south side of the district, probably constituting the northwestern extension of the southern Quinnesec schist belt of the Menominee district. The best exposures are along Pine River, especially in sees. 29 and 30, T. 39 N., R. 18 E. The schists are cliiefly hornblendic gneiss, locally micaceous. They are cut by basic and acidic intrusive rocks, the former being the more abundant. The detailed petrographic description of these schists given in the Menom- inee chapter will suffice for this district. The continuation of these schists along the south side of the Menominee district has been assigned to the Keewatin series of the Archean in previous reports of the LTnited States Geo- logical Survey. ° Later work showed this assignment to be a very doubtful one, and the question of the correlation of the schists has been largely left open for the Menominee district. The work of Hotchkiss along the south side of the Florence district shows clearly an interbedding of upper Huronian slate with tuffs and cruptives of the Quinnesec schist in a manner showing the main body of schist to be later in origin than the upper Huronian to the north of it. Intrusive and extrusive greenstones and green schists other than Quinnesec. — Massive and schistose intrusive and extrusive greenstones appear in several small areas in the upper Huronian. Two of them cross Menominee River on the east, where they join the northern Quinnesec schist area of the Menominee district. Another group is exposed along Brule River and others between the Brule and Florence. Isolated outcrops of green schistose and tuffaceous rocks of doubtful structural relations are somewhat widely distributed through the district. They are in places associated with amphibole-magnetite schists, some of which represent phase s of the intrusive rocks, but some of which doubtless also are metamorj)hosed phases of the Huronian ferruginous slates. Petrographically these rocks are very similar both to the Hendock formation and to the Quinnesec schist, and the description of the northern Quinnesec schist area of the Menominee district will apply to them. The areas of intrusive rocks are longer from east to west than from north to south. Evi- dence of the intrusive character of the greenstones is found along Brule and Menonunee rivers in T. 40 N., R. 18 E. Especially good evidence is the area just west of Keyes Lake. In sec. 9, at several points along the Brule, are to be found outcrops of the massive greenstones in contact with the slates. Invariably the slates are more micaceous near the contact than elsewhere. In fact, they become mica schists, and here and there is seen a slight development of some secondary niineral, probably garnet. In every outcrop along the Brule the contacts of the greenstones and sediments are not sharply defuied, the greenstones being schistose and chloritic at the contacts. In sec. 13, T. 40 N., R. 18 E., greenstone is found in contact with a micaceous oMon. U. S. Geol. Survey, vol. !f.. 190^; Monominoo sppclM folio iNo. 02), Geol. Atlas V. S., C. S. r.eol. Survey, 1900. IRON ORES OF CRYST.-VL FALLS, IRON RIVER, AND FLORENCE DISTRICTS. 32S quartzitc. The actual well-defined contact may be seen here, and the intrusive character of the greenstone is clearly shown. A wedge of the greenstone cuts the quartzite at 1 ,650 paces north antl 200 paces west of the southeast corner of sec. 13, T. 40 N., R. 19 E. The quartzite at this place is much fissured and shattered. Brule River, where it crosses the E. i sec. 9, T. 40 N., R. 18 E., is a favorable place to see the way in which the intrusive greenstones stand out prominently as hills in the slate area. The river here cuts through the slates and greenstones, giving a well-exposed cross section. The conclusion is here forced on the observer that the outcrops of the greenstones of this area represent with a very fair degree of accuracy the actual distribution of the greenstones. The greenstone outcrops are many times longer east and west than north and south, as has been noted. This, however, does not justify the correlation of greenstone knobs because they happen to align in the direction of their long dimensions. The areas mapped as intrusive and extrusive greenstones and green schists on the Florence map (PI. XXV, in pocket) may there- fore be regarded as containing much slate in lower, covered ground. GRANITE AND GNEISS INTB.USIVES. Bordering the Quinnesec schist on the south is an area supposed to be underlain by granites and gneisses. Exposures are few, but to the east, south of the Menominee district, they are more abundant. The relations are those of intrusion into the Quinnesec schist, and the rocks are doubtfully correlated with the Keweenawan. PALEOZOIC SANDSTONE. A few patches of Paleozoic sandstone he unconformably upon the pre-Cambrian rocks. These are well shown just west of the Buckeye mine and north of Keyes Lake. QUATERNARY DEPOSITS. This district is covered by Pleistocene glacial drift. (See Chapter XVI, pp. 427-459.) THE IRON ORES OF THE CRYSTAL FALLS. IRON RIVER, AND FLORENCE, DISTRICTS. By the authors and W. J. Mead. DISTRIBUTION, STRUCTURE, AND RELATIONS. The principal ores of this region are found in iron-bearing layers infolded \vith upper Huronian slate in the vicinity of Florence, Commonwealth, Crj^stal Falls, Amasa, and Iron River, and in the middle Huronian slate near Mansfield. These districts are usually considered as a part of the Menominee district in returns of ore sliipments, and their ores are similar, geologically and structurally, to those of the Menominee district. Though not chrectly continu- ous with the iron formation of the Menominee district, so far as explorations yet show, they mainly belong in a formation which is closely correlated with that iron-bearing formation (the Vulcan), and is given the same name. Also the upper Huronian slate with wliich this iron- bearing formation is associated is similar to and continuous with the Michigamme ("Hanbury") slate of the jNIenominee district, and is therefore called by the same name. The Micliigamme slate over this great area is remarkably uniform in character, anil it is difficult to tell at what horizon in the slate formation the ores occur in any particular locality. In tlie vicinity of Crj'stal Falls and Amasa the upi)er Huronian slate rests upon greenstones of the Hemlock formation, so that in tliis part of the district it is easy to determine the base of the upper Huronian, and the occurrence of the ore at a short though varying distance from the volcanic Hemlock formation shows that for this locality at least the iron-bearing rocks occur at^ a fairly persistent horizon near the base of the upper Huronian slate. Most of the ore deposits of these districts are accompanied by black and pyritiferous slate walls, in places associated with greenstone, or they maj^ be separated from such walls, especially the hanging wall, by a small amount of lean cherty iron-bearing rock. Along the trend of the 324 GEOLOGY OF THE LAKE SUPERIOR REGION. iron-bearing member and in (l("i)tli the iron-ore layers jiass info lean clierty layers. The ore bodies throughout show a strong tendency to follow the steeply inclined and uniformly trending bechiing of the iron-bearing member, liaving tlius distinct linear shape and distribution at the surface and tabular or lens shape in three dimensions. In certain of the Crj'stal Falls deposits these characteristics are much more apparent than in others. For instance, the ores at the Hemlock mine at Amasa constitute a lens in a narrow band of iron-bearing rock, with consid- erable extent vertically and horizontally, parallel to the strike of the upper Iluronian. The same is true of the ore deposits in the so-called "Mansfield slate." Though minor folds are present in both of these deposits, they are subordinate to the general tabular sha{)e of the deposits. Other ore bodies follow the axial Hues of drag folds, thus jiitching at various angles beneath the surface. Their shape, considered in three dimensions, tends to be linear rather than tabular. As few of these axial lines are uniform for long distances, offsets of the ore body are common. The ores of the Florence district seem to be in drag folds, with ])itches to the northwest. Their distribution suggests sharp offsets by drag folding. The iron-bearing rocks, and therefore the ore bodies, are usually not more than 300 feet tliick, though locally the thickness may be much increased l)y buckling. It will be noted by figure 12 (p. 123) that folding of that type multiplies the thickness by 3. The depth to which mining has thus far extended is 1 ,000 feet, but exploration has shown ore to a greater depth. It can not yet be said what the maximum depth of the ores may be. At the Florence mine the formation becomes pyritiferous below this depth, although it is not demonstrated that the pyritiferous portion continues indefinitely. The iron formations near the main area of the Hemlock formation in the Crystal Falls district and part of those in the Florence district are distinctly magnetic. Elsewhere in the Crystal Falls district and in the Iron River district the formations are weakly or not at all magnetic. The structural relations of the ores of this group are less satisfactorily known than those of almost any other district in the Lake Superior region, partly because of the lack of sufficient development and partly because of the uniformity of the slate, making it difficult to find recog- nizable horizons as a basis for working out the structure. Because of the lack of continuity of the iron formation in tliis great slate area and the covering of a large part of the area by glacial drift, it seems altogether likely that there are still many deposits to be found through the slate. Magnetic work sometimes indicates places to begin exploration, but much of the exploration must begin blindly. CHEMICAL COMPOSITION. The ores of these cfistricts, with the exception of the Mansfield deposit and the Amasa- Porter, south of Amasa, are non-Bessemer hydrated hematites of medium to low grade. The average composition and range for each constituent of the ores minetl in these districts in 1907 and 1909 are as follows: Arcniye rhcmicd} composition of ores from carrjo anahjscsfor 1907 arid 1909. Crystal Falls dLs- trict. Iron Kiver dis- Iriut. Florence district. 1907. 1909. 1907. 1909. 1907. 1909. 8.46 8.42 8.23 8.34 10.86 9.76 -Analysis of ore dried at 21'2° F.: 54.10 .437 0. 27 1.27 2.94 2. 62 2.15 .050 5. 89 54.79 .495 7.71 .799 2.50 2.63 2.16 .071 4.11 55.70 .390 8.62 .20 2.54 .92 .76 .057 5.25 54.35 .404 8.77 ..30 3.07 1.34 1.49 .056 5.74 54.50 .32 0.72 .26 3.35 1.51 2.40 .1.12 5.20 54.70 .319 6.89 .08 4.17 I.ime 1.80 2.86 .173 S.20 IRON ORES OF CRYSTAL FALLS, IRON RIVER, AND FLORENCE DISTRICTS. 325 Range in pcrrentage of each constituent in ores mined in 1909. Crystal Falls districl. Iron River district. Florence dis- trict. Moisture (loss on drying at 212°) Analysis of ore dried at 212° F.: Iron Phosphorus Silica Man^nnese . . . .- Alumina Lime Macnesia Sulphur Loss on ignition 2. S3 to 13. 75 35.74 to 57. 20 .04010 1.28 5..';i to 30. .13 .15 to 2.93 1.20 to 3.41 1.20 to 4.96 .71 to 2. NO .007 10 .100 l..-)S to 7. CO 49.87 to 50. 07 .70910 3.13 5.35 to 14. IB .18 to 2,10 .99 to .40 to .20 to . 009 to 2.45 to 4. 23 2.74 2.40 8.46 to 9.1 53. .30 to.M.OO . 297 to . 410 0. .50 lo .00 to 2.S2 to 1.01 to 2.74 to .11 to 5.05 to S.05 .20 4.47 2.03 2.88 1.87 .5.80 MINERAL COMPOSITION. The ore of these districts is cliiefly soft red hematite, though in places it is hydrated and graded as brown hematite (limonite). Goethite has been identified at Iron River. In achlition, there are quartz and some kaoUn, with small amounts of magnetite, calcium, and magnesium carbonates, and minute amounts of sulphides. The average mineral composition of the ores of these districts, calculated from average analyses for 1909 given in the above table, is as follows: Approximate mineralogical composition of ores, calculated from the average analyses for 1909. Crystal Falls district. Iron River district. Florence district. 71.90 7. 50 4. 311 4.70 3.50 4.00 2. 00 1.44 54.00 27.80 4.82 5.80 3.80 .45 2 12 \.2\ 62.42 18.10 1.26 7.70 6.20 2 85 Apatite (all phosphorus calculated as apatite) 1.65 100. 00 100.00 100.18 The above mineral compositions are necessarily only approximate, as ferrous and ferric iron are not separated, and the combined water, COj, ami a possible small amount of organic material are included together under loss on ignition. All the phosphorus with proper amounts of limestone was calculated as apatite; the remaining lime with proper amounts of magnesia and water was calculated as dolomite. The remaining magnesia with alumina, silica, and water was calculated as chlorite. The alumina not used in the chlorite, together with sufficient silica and combined water, was taken as kaolin. Sufficient iron was combined with the remain- ing water to form limonite and the remaining iron figured as hematite. Hematite and limonite probably do not exist in the ores, but as a means of comparison and to show the degree of hydration the hydrated iron oxide is calculated in terms of these two minerals. PHYSICAL CHARACTERISTICS. The ore is very porous and shows many crystal-lined cavities. At places a hard steel hematite ore is fomid, which rims high in metallic iron. It breaks into a mixture of small blocks and soft ore similar to the ores of the Menominee district. The average mineral density of the ores, calculated from the above analyses, is 4.38 for the Crystal Falls ores and 4.30 for the Iron River ores. The porosity of the ores ranges from less than 5 per cent to over 40 per cent of their volume. The cubic contents of the ores vary from 8.5 to 15 cubic feet to the ton, with an average of about 1 1 cubic feet. The volume composition of these ores, in comparison with those of the Menominee district, is represented in figure 50 (p. 352). 326 GEOLOGY OF THE LAKE SUPERIOR REGION. SECONDARY CONCENTRATION OF THE ORES OF THE CRYSTAL FALLS, IRON RIVER, AND FLORENCE DISTRICTS. Structural conditions. — The ores of the Crystal Falls, Iron River, and Florence districts are enrichments of narrow bods and lenses of iron-bearing roc'ks, as a rule not more than 300 feet wide, usually between steeply inclined walls of slate, generally graphitic and pjTitiferous near the contact, and commonly associated with greenstone. The iron-lx'aring member Uiay trend in the same direction for considerable distances and yet be closely corrugated by minor folds of the drag type illustrated in figure 12 (p. 123). These steeply pitching drag folds furnish an impervious basement of slate along which the waters have followed the o])enings in the iron- bearing member in especial abundance and have effected the concentration of the ore. The iron-bearing rock is brittle, but the slate is not, the result being that breccias are common in such troughs, greatly favoring the flow of water. The folds are of various magnitudes and the concentration may follow either the minor or the major folds. The circulation has been controlled by the fracture openings in the iron-bearing member and the bedding in it, and the confining strata hav(> been foot-wall slates, hanging-wall slates, and iron-bearing member. The essential parallelism of the ores to the trend of the iron-bearing member shows the obvious tendency of the waters to follow that trend but to be deflected by the minor bends in it. This is especially well seen along the main belt of iron-bearing rocks along Iron River. The depths to which the waters have acted is yet largely unknown. The deepest mines 0])erate to a depth of 1,000 feet in the Gystal Falls district, 500 feet in the Iron River district, and 950 feet in the Florence district. In certain deposits the ore has apparently given out with depth. It is possible that in some mines it has been lost because of considerable offset by the folding. Deeper exploration is warranted. The topographic relief of the region is so great that different ])arts of the iron-bearing member may differ as much as 300 feet in elevation. The ores are as a rule closely associated with the hills but seem to follow, indifferently, crests, slopes, and adjacent valleys. In the Iron River district the ores favor especially the valleys. These are discernible with difficulty tlirough the thick drift, but are being found by drilling. The depth to which a head given by the observed topography would carry a vigorous circulation through the iron-bearing member can not be woi'ked out theoretically because of the imcertamty of the factors mvolved. Certainly nothing is now known which would prevent exploration as deep as in other districts of the Lake Superior region, although here, as in other districts, many of the deposits have certainly been found to be only a few hundred feet deep. Ohemical and mineralogical changes. — The iron-bearing member was originally pyritiferous iron carbonate interbedded with more or less slate. The alteration to ore has occurred in two phases — first, the oxidation of the iron without removal of silica, producing ferruginous cherts; second, partly simultaneous and more local, the leaching of the silica, leaving the iron oxide concentrated as ore. The phj^sical and chemical features of these alterations have not been worked out tjuantitatively as they have for other districts, but qualitativel}' they are knowTi to be similar to those of other districts in all respects. Time of concentration. — The ores were concentrated after the upper ITuronian folding and before the Cambrian deposition, and since their concentration they have been little affected by further folding. THE IRON ORES OF THE FELCH MOUNTAIN AND CAI.LTMET DISTRICTS. By the authors and \V. J. Mead. The Felch Mountam and Cahmiet districts are eastward branches of the Crystal Falls district. Except for low grade and low ])hosphorus, their ores are the same in horizon, relations, and mineralogical and j)liysical character as the ores of the CVystal Falls and Menominee districts. The shipment from these districts has been small. IRON ORES OF FELCH MOUNTAIN AND CALUMET DISTRICTS. 327 FELCH MOUNTAIN DISTRICT. Iron ores have been mined at two localities in the Felch Mountain district near Groveland and near Felch. In both these localities the iron-bearing Vulcan formation lies in a closely compressed s\'ncline with basement of impervious slate or schist, called "Mansfield" schist by Smyth, but called Felch schist in this report. The lenses at the east end of the Felch Moimtain trough are now largely worked out. At the Groveland mine dikes of granite cut the ore body. The average composition of the ores mined in the Felch Mountain district in 1907 is as follows : Average analysis of ore mined in the Felch Mountain district in 190/. Mointure (loss on drying at 212°) 4. 05 Analysisof ore dried at 212° F.: == Iron 52. 50 Phosphorus 040 Silica 11. 22 Manganese 1. 10 Alumina. . . .- 2. 49 Lime 3. 51 Magnesia 4. 62 Sulphur 008 Loss by ignition 5. 29 The volume composition of these ores, in comparison with the Crystal Falls, Menominee, Iron River, and Florence ores, is given in figure 50 (p. 352). CALUMET DISTRICT. Ore is mined in the Calumet district only at the Calumet mine, a comparatively recent development, where there is a steeply southward-dipping succession beginning with Archean granite on the north, followed successively by Sturgeon quartzite, Randville dolomite, Felch schist, Vulcan formation (iron bearing), and Michigamme slate. The strike of the ore body is parallel to the bedding. The bedding trends east and west, but has minor folds with steep pitches parallel to the strike. The ore body with its associated iron-bearing formation is divided longitudinally into three parts by layers of slate, from north to south 60, 15, and 60 feet thick. The foot wall is slate, quartzite, and dolomite. The hanging wall is slate or iron- bearing formation. Along the strike the ore abuts irregularly against unaltered iron formation. The depth of mining operations to the date of writing is 200 feet. The possibilities of the extension of the deposits are discussed on page 324. The iron ore is banded cherty hematite and limonite and some magnetite. It is nonmagnetic in individual pieces, but collectively it exerts a powerful magnetic pull. The ore runs from 40 to 45 per cent of iron and is sold on the basis of 0.028 per cent of phosphorus. Its density is about 4 and its porosity IS per cent; it averages about 10.5 cubic feet to the ton. The average composition of the ores mined in the Calumet district in 1907 is as follows: Averaije analysis of ore mined in the Calumet district in 1907. Moisture (loss on drying at 212°) 5. 00 Analysis of ore dried at 212° F.: =^ Iron 42. 82 Phosphorus 028 Silica 32. 27 Manganese 20 Alumina 2. 53 Lime .- .74 Magnesia 1. 06 Sulphur Oil Loss by ignition 1. 86 The volume composition of these ores, in comparison with Crystal Falls, Iron River, Florence, and Menominee ores, is given in figure 50 (p. 352). 328 GEOLOGY OF THE LAKE SUPERIOR REGION. SECONDARY CONCENTRATION OF THE FELCH MOUNTAIN AND CALUMET ORES. Structural conditions. — Tlie iron-bearing Vulcan formation of the Fclch Mountain district is in closely compressed synclinal folds in the upper lluronian Felch schist. It stands out as erosion remnants forming the crests of the hills. The concentration has evidently been con- trolled bj' the impervious basements of slate, and also to some extent by the o[)enings along fracture planes, especially north-south fracture planes crossing the axis of the trough. The granite dikes at the Groveland mine may also have been influential in controlling circulation. In the Calumet district there is no essential difference in the structural relations governing the How from those in the Crystal Falls and Iron River districts. The dip is steep and the forma- tion has the usual drag type of corrugation. Che^nical and mincralogical changes. — The iron-bearing member was originally iron car- bonate mterbedded with more or less slate. The alteration to ore has occurred m two phases — first, the oxidation of the iron without removal of silica, producing ferruginous cherts; second, partly simultaneous and more local, the leachmg of the silica, leaving the iron oxide concentrated as ore. The physical and chemical features of these alterations have not been worketl out quantitatively as they have for other districts, but quahtatively they are known to be similar to those of other districts in all respects. CHAPTER XIII. THE MENOMINEE IRON DISTRICT OF MICHIGAN." LOCATION AND EXTENT. The portion of the Menominee district covered by tiie accompanying map (PI. XXVI, in pocket) is bounded on tlie west by Menominee River, on the soutli by tlie same river and the south hne of T. 39 N., on the north by the north Una of T. 40 N., and on the east by the east hue of sees. 10, 15, 22, 27, and 34, T. 39 N., R. 28 W. The area thus outlined constitutes a tongue of sedimentary deposits lying between a granite area to the north and a greenstone schist area to the south. At about tlie line between Rs. 27 and 28 W. the characteristic rocks of the Menominee trough become so deeply buried under later sediments that they can be traced no farther by outcrop. Lines of magnetic attraction, however, have been obtained still farther east, and these are taken to mean that the Huronian deposits continue for a considerable distance beyond the places where they are last seen on the surface. The area of sedimentary rocks belonging in the Menominee trough is about 125 square miles, entirely within the State of Michigan. This area is narrowest in the vicinity of Vulcan, where it measures about 4 miles in width from the contact witli the granite on the north to the contact with the greenstone schist on Menominee River to the south. To the east the area widens gradually, until m the eastern portion of R. 28 W. its width measures about 7 miles. To the west it also widens gradually and finally loses its identity as a distinct trough at al)out the center line of R. 30 W., where it merges, with the Calumet trough, into the wide area of Huronian sediments on the west. TOPOGRAPHY. There are thi'ee important ridges in the district with axes parallel to its length, a northern one and two others, nearly parallel, near the central part of the district. The northern ridge is composed of Archean granite and the Sturgeon cpiartzite. The central ridges are composed of the Randville dolomite and the ii-on-bearing Vulcan formation, capped in much of the dis- trict by Cambrian sandstone. The higher points of these ridges range in altitude from about 1,000 feet to nearly 1,600 feet. The valleys between the ridges, as well as the valley to the south of the main central ridge sloping to Menoininee River, are composed mainly of the Michi- gamme ("Hanbury") slate. The southern lowland area of the Michigamme slate continues into the area of the Quinnesec schist. The lower areas have altitudes varying for the most part from 800 to 1,000 feet. The minor streams follow to a considerable extent the valleys of the Michigamme slate, and the same is true of the chief stream of the district, the Menominee, for a considerable part of the area, but this and a number of the other more important streams, such as Sturgeon River and Pine Creek and some of its branches, flow transverse to the ridges. Several of even the smaller branches break through either one or both of the iron ranges and the cpiartzite and granite range to the north. Sturgeon River crosses all the formations of the district. SUCCESSION OF FORMATIONS. The rocks of the Menominee district belong to the Archean, Algonkian, Cambrian, and Ordovician systems. The oldest rocks bordering the Menominee tongue are greenstone schists and granite. These are regarded as Archean. Resting unconformably upon the Archean rocks o For further detailed description of the geology of this district see Men. U. S. Geol. Survey, vol. 46, 1904, and references there given. 329 Huronian scries: Upper Huronian (Animikie group) 330 GEOLOGY OF THE LAKE SUPERIOR REGION. are Algonkian sediments, which belong to the Huronian series. These are (hvisible into lower Huronian, middle Huronian, and upper Huronian, and an^ separated by unconfornjities. The Paleozoic rocks comprise horizontal Cambrian sandstones and Cambro-Ordovician Umestones. These occur in patches on the tops of the hills, capping the closely folded and truncated Huronian rocks. The Huronian series is divisible into a number of formations, each representing a time during which the conditions of deposition were apjjroximately uniform. The following table gives the list of the formations arranged in descending order according to age. The members of the Vulcan formation are distinguished in the descriplion but not on the map. Cambro-Orclo\dcian Uermansvillc limestone. Cambrian system Lake Superior sandstone. Unconformity. Algonkian system: Keweenawan series Granite (?). Quinnesec schist and other green schists representing surface eruptions overlying and interbedded with Michi<;amme slate. Michigamme ("Hanbury") slate, including iron-bearing beds. Vulcan formation, subdivided into Curry iron-bearing member, Brier slate member, and Traders iron-bearing member. Unconformity. Middle Huronian Quartzite, not separated from Rand\'ille dolo- mite in mapping for most of the district. Unconformity. T „ ■ [Randville dolomite. Lower Huronian <„^ ^ .^ ISturgeon quartzite. Unconformity. Archean system : Laurentian series Granites and gneisses, cut by granite and diabase dikes. Keewatin series (not separated in mapping from Laurentian) Green schists. The Quinnesec schist is so named because the formation is typically developed at the Quinnesec Falls, on Menominee River. The Sturgeon cjuartzite is so called because this forma- tion in the Menominee district has been traced almost continuously to a like formation in the Crystal Falls district which has been called the Sturgeon quartzite. The dolomite in the Menominee district is called the Randville dolomite because it has been practically connected with the Randville dolomite of the Crj^stal Falls district. In the upper Huronian the Vulcan formation is so named because it occurs in typical development with full succession and fine exposures near the town of Vulcan. The "Hanbury" slate was thus named because in the vicinity of Lake Hanbury this formation is better exposed than anywhere else in the district. This slate, however, has been proved to be equivalent to and continuous with the Michigamme slate of the Marquette district, and the older name, Michigamme, is therefore used in tliis report. ARCHEAN SYSTEM. LAURENTIAN SERIES AND UNSEPARATED KEEWATIN. The complex north of the Menominee district is composed largely of Laurentian rocks. They are principally gneissoid granites and finer-grained banded gneisses. In addition to these there are also present in subordinate qiuintity hornblende schists and certain feklspathic green- stone schists identical lithologically with some of the mashed eruptive rocks among the Quin- nesec schist. These are intruded by Laurentian granites and are believed to represent the MENOMINEE IRON DISTRICT. 331 Keewatin series. They have not been sejjarated in map])ing. Mica scMsts arc founfl only in a few exposures in the interior of the Ai'chean area north of the region shown on the map (PI. XXVI, in pocket). The granites, gneisses, and schists are cut by small dikes and veins of granite, pegmatite, and aplite, by numerous quartz veins, and by coarse granite, massive basalt, diabase, and gabbro. Some of the hornblende schists (Keewatin) and some of the gneisses appear to be older than most of the granites. Others of the scliists are unquestionably mashed intrusive rocks that are younger than some of the granites. The aplites, pegmatites, and some of the basic intrusives are the youngest rocks belonging exclusively in the complex, but even these, as they are not known to cut through the Huronian deposits, are thought to have taken their present position before the sediments were deposited. The latest of all the intrusive rocks are certain coarse-grained massive diabases and gabbros. These rocks not only occur as members of the complex but are found also in the lower division of the Huronian series, overlying the Archean complex. There is no reason to believe that any of these rocks are metamorphosed sediments. Most of them are clearly of igneous origin. The massive granites and the gneissoid granites differ from each other in no essential respect. The latter are merely schistose phases of the former. They both embrace medium- grained to fine-grained gray and pink rocks with a granitic texture that locally approaches in appearance the texture of some quartzites. The banded gneisses consist of alternate bands of pink and gray material, each band having the look of granite. These bands, though appearing to be approximately parallel in the ledges, are found on close inspection to run parallel to one another for short distances only and then to anastomose or interlace. The red layers cut across the gray gneiss as if they were veins of granitic material. The only difference that can be discerned between the banded gneisses and the fine-grained gray gneisses cut by red granite veins is that the latter are irregularly injected by the granitic material, while in the former the injections are largely parallel. The hornblende scliists (Keewatin) are usually lustrous greenish-black schists with the normal characteristics of such rocks. They are cut by the granites in some places. In other places large blocks are found included in granite. Plainly they are older than the granites, and probably they are the oldest rocks in the northern complex. A second kind of hornblende sclust exists in which the rocks are so related to the granites and gneisses that they must be regarded as dikes. In some places they a]^pear as bands cutting across the banding of the gneisses, and in others as bands conforming in strike and dip with the lighter-colored bauds of these rocks. These schists are therefore looked upon as mashed intrusive rocks. ALGONKIAN SYSTEM. GENERAL CHARACTER AND LIMITS. The Algonlvian rocks constituting the Menominee trough, though strongly metamorphosed, are recognized as mainly sediments. The greater mass of these sediments is mechanical, clastic textures being still plainly apparent. The iron-bearing formation is largely mechanical, but with the mechanical material an important amount of chemical and organic material was deposited, and some of the jaspers of the formation may be wholly chemical or organic. The limestones are chemical or organic sediments. The sedimentary rocks have been intruded by a few coarse-grained and some fine-grained igneous rocks. The latter are now usually scliistose. The lowest formation of the Algonkian system has at its bottom basal conglomerates, which rest unconformably upon the Ai'chean rocks of the northern complex. These conglomerates may be seen at a number of places along the border of the trough, and notably at the falls of Sturgeon River. 332 GEOLOGY OF THE LAKE SUPERIOR REGION. The formations of the Algonkian system are likewise separated from the overlying Cambrian sandstone by a profound unconformity. The Algonkian rocks are folded ; the Cam- brian sandstone is horizontal and thus lies across the truncated ends of the eroded folds. Its lowt-r layer is formed largely of the debris of the more ancient rocks. Hence the Algonkian rocks formed a land surface for a vast period of time before the deposition of the Cambrian sandstone. HTJKONIAN SEKIES. LOWER HURONIAN. SUCCESSION AND DISTRIBUTION. The lower Huronian is divided into two formations — the Sturgeon quart zite and the Randville dolomite, the former bemg the older. These formations are observed only in the center and on the north side of the Menominee trough. On the south side of the trough no evidence of their existence is obtainable. This may possibly be due to the thick covering of drift that blankets the rocks north of the southern area of Quinnesec schist ; but it is thought to be more probable that these formations are not present at the rock surface in this portion of the district. STURGEON QUARTZITE. Distribution. — The Sturgeon quartzite forms a continuous border of bare hiUs on the south side of the northern complex. The formation lies between the Archean complex and the northern belt of dolomite. Prominent bluffs of the typical quartzite may be conveniently studied northeast of the Loretto mine. Lithology. — At many places at the base of the Sturgeon quartzite there is a conglomerate made up of bowlders and fragments of granites, gneisses, and hornblende schists identical with the corresponding rocks in the adjacent Archean complex to the north. The matrix in which these are embedded is in some places a quartzite, in others an arkose composed of the fine-grained debris of granitic rocks. In many places this matrix is schistose and a large quan- tity of a micaceous mineral has been produced by alteration of the feldspar of the original sediment, so that the matrix is now lithologicaUy a sericite schist. The major portion of the formation consists of massive beds of a very compact, vitreous quartzite, usually white, but here and there tinted with some shade of pink or green. In its upper portion the cement between the quartz grains is locally calcareous. This calcareous con- stituent increases in quantity as the overlying dolomite is approached, until the rock becomes a calcareous quartzite and finally a quartzose dolomite. The change from the cjuartzite to the dolomite is thus a transition. This indicates a gradual deepening of the waters during the later part of the Sturgeon epoch. Deformation. — The main belt of the Sturgeon quartzite is a nearly vertical southward- dipping monochne. The outcrop of this monocline varies in strike, thus indicating that cross folding has taken place to some extent. At the west end of the district the quartzite turns northward, ^Tapping around the Archean complex and then passing eastward into the area of the Calumet trough. On the turn to the north several small folds are developed, the synclines of which are now represented by embayments extending eastward into the Archean. The dips of the quartzite beds may vary a few degrees — 25° in one place — from perpendicularity. There arc almost as many northern dips toward the granite and gneiss complex as there are southern dips toward the center of the trough. Relations to adjacent formations. — The Sturgeon quartzite rests unconformably upon the Archean rocks of the northern complex. This is shown by the character of the lower bed of the quartzite, which, as already said, is a basal conglomerate. This basal conglomerate con- tains almost every variety of fragment derivable from the rocks of the northern complex. Some of this material in its original position must have been formed at great dei)th in the earth. Therefore there was deep-seated denudation of the Archean before the deposition of the MENOMINEE IRON DISTRICT. 333 quartzite. Upward the Sturgeon quartzite grades into the Randville dolomite. The nature of ^,he gradation is discussed in tlie section on tliat formation. Thickness. — Two difficulties stand in the way of determining the thickness of the Sturgeon quartzite. The first is the inqiossibihty of deciding how much of the apparent thickness of the many rock layers in a closely folded district, like the Menominee, is due to the duphcation of beds in consequence of close folds. The other difficulty is the impossibihty of fixing the upper limit of the formation. There is everywhere between the c[uartzites and the nearest ledges of the overlying dolomite a belt of country without exposures of any kind. If we assume that the southward-facing cliffs, which in so many places mark the southern limit of the quartzites, are cliffs of differential degradation, that the low ground at the base of the cliffs is underlain by the dolomite formation, and that the exposures are monochnal, the maximum thickness of the formation is between 1,000 and 1,250 feet. RANDVILLE DOLOMITE. Distribution. — The Randville dolomite occupies three separate belts, whose positions and shapes are determined by the folding to which the formation has been subjected. These will be referred to as the northern, central, and southern belts of dolomite. The northern belt is south of the belt of Sturgeon quartzite. Only a few exposures are found in this area, but they are so uniformly distributed that on the map (PI. XXVI, in pocket) the whole belt has been colored for the formation. It is quite possible, however, that in some places erosion has carried away the dolomite and that the upper Huronian rests immediately upon the quartzite. The central belt of dolomite borders the north side of Lake Antoine for a portion of its length, passes eastward between the Cuff and Indiana mines, and ends at the bluff known as Iron Hill in the E. i sec. 32, T. 40 N., R. 29 W. This belt is well marked by numerous and large exposures. The southern belt of dolomite extends all the way from the west side of the sandstone bluff west of Iron Mountain to the village of Waucedah, at the east end of the district. Where not exposed the rock has been found in mines, test shafts, and pits, so that there is a reason- able certainty that it exists throughout this distance of 16 miles. Where there is any doubt of its existence at the surface this is due to a considerable thickness of overlying Cambrian sand- stone. From Iron Mountain as far east as Sturgeon River tlie country underlain by the dolo- mite is a range of high hills, broken only at a few points by north-south gaps. On the southern slope of this ridge are the principal producing iron mines of the district. LitTiolo'jy. — The Randville dolomite is composed of a heterogeneous set of beds Lq which dolomite is dominant. With the pure dolomites are siliceous dolomites, calcareous quartzites, argillaceous rocks, and cherty quartz rocks. The Randville dolomite, lying upon the Sturgeon quartzite, grades dowmward into it. The intermediate rock is a calcareous quartzite. The predominant rock of the Randville dolomite is an almost massive, apparently homo- geneous, fine-grained white, pink, blue, or bufl' dolomite, occurring m beds from a few inches to many feet in thickness. This is interstrattfied with beds of siliceous dolomite m which are observable numerous grams of quartz. In many places on the weathered surfaces of the dolo- mites are thin projecting Ijands of vein cpiartz parallel to the bedding, which the microscope shows to be calcareous quartzite. In other places projecting bands anastomose or run irregu- larly over the weathered surfaces, here and there intersecting the bedding planes of the rock at acute angles. Their abundance proves clearly that the dolomites, in spite of their homo- geneous appearance, have been extensively fractured and crushed. In many places the crush- ing has produced a breccia of dolomite fragments in a siliceous matrix. In a few localities the fragments are rounded, so that the rock is a pseudoconglomerate. The greater part of the argillaceous rocks interstratified with the dolomite is soft light- gray or dark-gray slate. Another part is typical black slate, still plainlj^ marked by bedding lines. Still other parts are purplish-pink schistose argillaceous dolomite. Many of the thin 334 GEOLOGY OF THE LAKE SUPERIOR REGION. layers of the pui'plisli-piiik shitelike material between massive dolomite beds appear to be- largely the selvajje of the softer lajers of dolomite, rendered schistose by the movement of accommodation between the stronger beds. Dcfommiion. — Structurally the northern belt of dolomite is a southward-dipping mono- cline. The central and southern belts are anticlines. The three belts are separated by two sjTiclines. In the anticlinal belts the beds at first sight appear to bo isoclinal, but a close examination of the southern belt reveals the existence of a number of minor folds having almost vertical pitches. In the western part of the district the folds are overturned to the south, the axial planes dipping northward at high angles. In the central and eastern parts of the district, east of Quiimescc, the minor and major folds have their axial planes steeply inclined to the south. Although the minor folds are rather easilj' recognizable, it is only on the south side of the southern belt that they become prominent. Here the synclines open out, forming basins in which the ore bodies lie. The small folds, with a few exceptions, pitch west in the western portion of the range and east in the eastern portion. The attitude of minor folds is, as is well known, an indication of the attitude of the major folds on which they are superimposed. By using this principle, it is concluded that the major anticlines in this district disappear to the east and to the west by plunging beneath the upper Huronian sediments. From the above statements it is clear that, in addition to the major east-west anticlines and synclines that are so prominent in the district, the dolomite formation is also affected by a^ gentle but large cross anticlinorium whose axis runs approximately north and south. It is- remarkable that eiosion has nowhere exposed the Sturgeon quartzite in association with the central belts of dolomite. Relatione to adjacent formations. — The dolomite formation is nowhere seen in actual con- tact witli the Sturgeon quartzite, nor are ledges of the two formations seen in close proximity. It is knowii, however, that the upper layers of the quartzite are calcareous and that the lower beds of the dolomite are quartzose. The inference seems to be safe that the two formations are conformable, and that they grade into each other through calcareous quartzites. The rela- tions of the dolomite to the overlying formations are discussed in connection with the upper Huronian. Thicl'ness. — At no place i^dthin the area mapped is the dolomite known to be exposed from the bottom to the top of the formation. On the north side of the trough the formation is bordered by the Sturgeon quartzite on the north and the Vulcan formation on the south, but exposures between these limits are so few that it is not ceitain that the dolomite occupies the entire breadth, and on this account and because of the minor folds it is impossible to give anything like an exact estimate of the thickness of the formation. By making calculations so as to obtain a minimum figure, 1,000 feet or less could be obtained. If, on the other hand, calculations were made on the supposition that all of the isoclinal beds are different layers, the estimate might be as great as 5,000 feet. Probably the truth is much nearer the lower figure than the higher. The original thickness of the dolomite is probably somewhere between 1,000 and 1,500 feet. MIDDLE HURONIAN. The identified middle Huronian of tJic Menominee district consists entirely of chert}' quartzite resting in a thin film, from a few feet to 70 feet thick, on the Randville dolomite near its contact with the upper Huronian (Animikie group), and it is included with tlie Rand- ville dolomite on the general map of the district (PI. XXYI, in pocket). These rocks were formerly regarded as a part of the dolomite formation, but recent work has shown them to be sci>aral)lc from the dolomite. The quartzite has been separated from the dolomite in the mapping for several small areas near Norway and the east end of Iron Hill. (See fig. 4.').) The chcrty quartz rocks are fine grained, drusy in places, and white, light red, or dark purple. The darker colorcil kinds look very much like some varieties of jaspilite. Under th& MENOMINEE IRON DISTRICT. 335 LEGEND Cambrian sandstone microscope the cherty quartz rocks seem to be composed almost exclusively of a fine-grained crystalline aggregate of quartz which incloses a few grains of hematite, magnetite, and other iron compounds. Here and there a fragmental quartz grain may be seen, but usually no trace of fragmental constituents can be discerned. Pebbles in the conglomerate at the base of the upper Huronian are partly jasper and iron ore, obviously derived from some preexisting formation not now appearing. A reasonable inference is that these pebbles represent fragments of the Negaunee formation, which would normally lie above the middle Huronian quartzite. In the previous report on this district" several masses of iron-bearing rocks were doubtfully referred to the Negaunee. Subsequent work has demonstrated these to be upper Huronian. The middle Hiu-onian quartzite rests unconformabl}- on the Randville dolomite, with discordance in bedding. The quartzite may be observed to fill fissures and depressions in the dolomite. At Norway Hill erosion cut ofl^ 100 feet or more of the dolo- mite before the quartzite was de- posited. On the south side of Iron HiU there is a thin film of conglomerate, taken to represent the base of the micklle Huron- ian quartzite, plastered against the dolomite escarpment. The quartzite is not shown directly above the conglomerate but ap- pears a few hundred feet to the east, resting against the dolomite escarpment. (See PI. XVII, ^4, of Monograph XL VI.) In fact, much of the mitldle Huronian quartzite itself on Iron Hill is conglomeratic and brecciated, and a considerable part of it may possibly represent a coarsely fragmental basal phase of the middle Huronian. The intri- cacy of the relations of the middle Huronian quartzite with the Rand- ville dolomite on Iron HiU is rep- resented in figure 45. The hill is a normal anticlinorium, of the type to be expected in com- petent formations of this type. It contrasts in every essential feature with the abnormal anticlinorium in the weak, incompetent beds of the Michigamme slate on Han bury Hill. MichigammeC'Hanbury"^ slata with iron formation (upper Huronian) Quart-zite (^middle Huronian) Randville dolomite (lower Huronian) Direction and pitch of ' axial lines of minor folds a5| Strike and dip Outcrop Outcrop with strike ^S5S3ffiS3.ff Figure 45.- Axis of folds Cross section A-B, looking east Geologic map and cross section of Iron HiU, Menominee district, showing relations of lower and middle Huronian. UPPER HURONIAN (aNIMIKIE GROUP). All the formations between the Randville dolomite and the unconformity at the base of the Cambrian sandstone are placed in the upper Huronian. For the purpose of the present monograph the group may be divided into two formations; the lower, the Vidcan formation, includes all the known iron-bearing rocks of the district except the conglomerate beds at the base of the Cambrian; the upper, the Micliigamme ("Hanbxiry") slate, comprises the great upper slate formation of the upper Hiu-onian. VULCAN FORMATION. Subdivision into members. — The u'on-bearing Vulcan formation embraces tlu-ee members; these are, from the base up, the Traders iron-bearing member, the Brier slate member, and the Curiy iron-bearing member. In this monograph the three members are mapped as a single formation because they are not so well exposed that they can everywhere be separately out- u Mon. U. S. Qeol. Survey, vol. 40, 1904, pp. 273-279. 336 GEOLOGY OF THE LAKE SUPERIOR REGION. lined. However, at several places the three members are known to exist, and ran be separately mapped. The Traders iron-bearinlainly detrital there are all grada- tions. It is difficult to ascertain whether the fragmental or the nonfragmental material is the more abundant in the Curry iron-bearing member as a whole, for it is poorly exposed. The ferruginous quartzose slates are beheved to have been derived largely from the erosion of the lower Iluronian. But mingled with tliis detrital material m many places was apparently- a considerable amount of nonfragmental material. There are, therefore, in the Curry iron- bearing member all gradations between clastic and nonclastic sediment. Deformation. — The Vulcan formation occupies a position on the upper sides of the dolomite anticlines. Its major folds, or folds of the first order, correspond exactly to the major folds of the RandviUe dolomite. The folds of the second order correspond also with those of the dolomite. The troughs on the south side of the southern dolomite area are occupied by the members of the iron-bearing formation. Moreover, within the Vulcan formation are numerous still smaller folds of the thirtl order, which, because of the hardness of the rocks and the perfec- tion of the bandmg, are well exliibited. These small folds may be observed at nearh' every place where minmg has progressed to any considerable extent and at many other places wliere only lean ores have been developed. The folds of the third order pitch in the same direction as those of the second order, upon which they are superimposed, but the strikes of their axes may diverge slightly. Still smaller folds are superimposed on the folds of the third order m the same \vay in which the latter are superimposed on the folds of the second order. On exposed surfaces the folds of the higher orders appear as a series of crinkhngs or flutings, with heights ranging from one-quarter inch to 5 or 6 inches from trough to crest. Even in the troughs of these minute folds, under favorable circumstances, iron ore was deposited, especially where crusMng and brecciating took place in connection \\\i\\ the folding. Wherever folding is observed within the iron-bearing formation it is noticeable that the bedding is best preserved in the siliceous bands. The iron-ore layere between the siliceous laj^ers, while yielding to the stresses that produced the folding, were mashed and sheared and became schistose. Where the compressing forces were very powerful a slat}' cleavage developed in both the iron-ore and the siliceous layers. In the western part of the south belt of the iron-bearing formation cariying the piincipal ore bodies the minor folds show considerable regularity of pitch to the west at angles ap])roaching 30°. The ore bodies follow these axial lines. Not uncommonly these nainor folds pass into overthrust faults. The distribution of the formation suggests that more overthnist faidts are really present than have been found. In this area the rocks to the south have moved westward and upward with reference to the rocks to the north, developing drag or buckle folds and thrust faults in the relatively incompetent upper Huronian beds near the contact with the relatively competent lower Huronian. The eastern part of the south belt shows some eastward pitches. Relations hetween the members of the Vulcan formation and the Michigamme slatt. — Where the relations between the Traders iron-bearing member and the Brier slate member are normal o Mon. U. S. Geol. Survey, vol. '.3, 1903. ' Idem, vol. 3C. 1899. MENOMINEE IRON DISTRICT. 339 the Traders grades into the Brier by diminution of tlie amount of ferruginous material and by increase in the number and tiiickness of the quartzose beds. At the same time there is an increase in tlie proportion of shity material. Where the ferruginous material is much reduced in quantity the Traders iron-bearing bed becomes the Brier slate member. This gradation occupies only a veiy short vertical range, so that the line between the iron-bearing member and the slate member is usually determinable within a few feet. Where marked disturbances have occurred, as in the vicinity of Norway and for several miles to the east, the relations between the two members are vevy different. Wherever it can be seen the contact between the Traders and Brier members is shaqj. In many places the contact seems to be slickensided and locally to be a plane of differential movement. At the o])en pits of the Norway mine and those north of the Curry mine and between this mine and the West Vulcan the Traders rocks are in ])laces ])seudoconglomeratic. The Brier slate member also may be brecciated. Moreover, the brecciation is not confined to these two members, but the underlying dolomite is at some places likewise brecciated for a short distance beneath its upper surface. The phenomena wherever studied appear to indicate that at the time of folding fault slipping occurred along the contact between the upper Huronian and the lower Huronian and between the Traders and Brier members. The dolomite was brecciated to some extent, the Traders detrital ores were crushed and brecciated, and in several places the lower portions of the Brier slate member were likewise included wiihin the zone of movement and were frac- tured and brecciated. Later the breccias were enriched by the deposition of hematite and other iron compounds, and both the Traders member and the lower part of the brecciated Brier slate member became sufficiently ferruginous to warrant mining. The Brier slate member passes upward into the Curry iron-bearing member by the diminu- tion of argillaceous material and the introduction of ferruginous material, especially bands of jaspilite, the somewhat ferruginous quartzose Brier slate meml)er thus becoming heavily ferru- ginous. At one place this transition is seen to occur laterally as well as vertically. No strati- graphic break has been discovered anywhere within the Vulcan formation. The relations between the Vulcan formation and the overlying Michigamme slate are those of conformity. The contact is usually very sharp. No difficuUy^ is experienced in defining the upper limit of the iron-bearing formation. The slates, however, are in places so very schistose on the upper side of the contact that their bedding planes can not be recognized, suggesting fault slipping. The bedding of tlie iron-bearing formation, on the other hand, is still almost perfectly preserved and is parallel to the contact. The relations of the Vulcan formation with the lower and middle Huronian are discussed on pages 342-343. ThicTcness. — A number of sections offer opportunities for determining the thickness of the separate members of the Vulcan formation, but in only a few can its total thickness be deter- mined. All along the south side of the southern dolomite belt, from the Aragon mine eastward to Sturgeon River, the iron-bearing formation stretches as a narrow belt, which for much of the distance appears to be without important folds. At several places mining operations have afforded excellent sections from the base of the productive portion of the Traders iron-bearing member to the top of the Curry iron-bearing member, and at a few places the sections extend downward to the top of the Randville dolomite. At Brier Hill, where practically the whole formation can be seen on the surface, its thickness is about 600 feet. At the Curry shaft No. 2 it is 700 feet thick and at the Aragon mine about 675 feet. At a number of sections the thickness of the individual members comprising the formation is easily measured. The Brier slate member has been measured at seven places, yielding results between 100 and 360 feet. Five of these measurements fall between 320 and 360 feet. Eight measurements of the Curry member have given results vaiying between 100 and 225 feet. Six of these fall between 160 and 225 feet. Measurements of the Traders iron-bearing member have yielded no such concordant results. In the first place, its thiclvness probably varies widely, as should be expected of a formation composed largely of detrital deposits. 340 GEOLOGY OF THE LAKE SUPERIOR REGION. Moreover, only a few sections reach as low as the dolomite; hence the exact position of the contact between tliis rock mikI the iron-bearing formation must be guessed at. Only three ineasurement.s have been made from the known top of the dolomite to the known top of the Traders member. These give 170 feet, 85 feet, and 155 feet. An interesting feature of these figures apjiears when the estimated thickness of the Brier and Currj' members is compared with the total thickness of the two. In almost every section where the estimated tliickness of either of these members falls below tlie average of all the measurements for that member the tliickness of the other member exceeds the average, and the total of the two is fairly constant. Thus, whereas seven estimates of the tluckcess of the Brier slate member vary between 240 and 360 feet, and eight estimates for the Curry iron- bearing member vary between 112 and 225 feet, measurements of the total thickness of the two vary only between 400 and 5.30 feet. The apparent greater variation in tliickness of the two members separately than in that of the two combined may be partly explained as due to the gradation between the two and the consequent difficulty of fixing the exact place at which one ends and the other begins. From a careful consideration of the figures given above and a few others that are not here recorded, it is estimated that the average thickness of the Vulcan formation is approxi- mately 650 feet, divided as follows: Traders u-on-bearing member, 150 feet; Brier slate mem- ber, 330 feet; Curry iron-bearing member, 170 feet — that is, the two ore-bearing members combined about equal in thickness the intervening slates. It is conceded, however, that the Traders member departs considerably from this average and that the total thickness of the ormation varies accordingly. MICHIGAMME ("HANBURT") SLATE. Di-ttrihution. — The Micliigamme slate occurs mainly in three large belts constituting valleys which correspond wdth synchnes between the older rocks. It occupies nearl}- all the low ground in the Menominee trough, forming a plain broken only by heaps of glacial material deposited upon it, by the protrusions of a few liillocks composed of the harder slates, or by equally resistant greenstones. The slate areas are narrowest at the east and gradually' widen toward the west. The northern belt is divided into two portions by the western area of Quinnesec scliist. The northern part turns northwest and leaves the Menominee district at the northern limit of the mapped area; the southern portion coalesces with the middle belt and crosses Menominee River into Wisconsin. East of Iron Hill the two northern belts again coalesce and extend as a single belt to Sturgeon River. Near the longitude of Waucedah all the slates disappear to the east beneath the Paleozoic beds. Name. — In previous reports on this district this slate has been called the Hanbury slate, but the formation has been proved to be equivalent to and continuous witli tlie Micliigamme slate of the Marquette district, and the older name, Micliigamme, is therefore used m this report. Lithology. — The formation is dominantly a pelite. It comprises black and gray clay slates, gray calcareous slates, graphite slates, graywackes, tldn beds of cjuartzite, local beds of ferru- ginous dolomite and siderite, and rarer bodies of ferruginous chert and iron oxide. Tilie formation is cut by dikes of schistose greenstones, and in one or two places sheets of the same rock have been intruded between the sedimentary beds. The predominant rocks of the forma- tion are gray clay slates and calcareous slates. The latter are more abumlant m the lower portions of the formation and the former in the upper portions. The exact vertical relations of the two rocks have not been made out, because of the scarcity of exposures and the very intricate folding to which they have been subjected. The clay slates are normal argillaceous slates, in wliich there is always more or less ferruginous matter. Those exposed to the weather are light in color and have a slialy character, iluscovite then becomes prominent. Their iron components are decomposed to red ocherous compounds. \Yliere most altered the rocks are light-red sericite slates or shales. The weathering of the slates that contain small quantities of calcareous components is somewhat different. They tend to bleach to a very ])ale-green or white color and to become porous tlirough the loss of their calcareous cement. The ferru- MENOMINEE IRON DISTRICT. 341 ginous components oxidize, forming red ocher, and this lies in an irregular pattern on the light- colored background. The result of tliese changes is a red and wliitc or pale-green mottled friable slate, known locally as "calico slate." By the Sedition of carbonates the argillaceous slates pass into the carbonate slates. These in places contain as much as .50 per cent of carbonate as a cement. With an increase in the carbona;te the slates lose their slaty character, become more massive, and finally pass into beds of f#rodolomite and siderite measuring from a few inches to 20 feet in thickness. On many of the weathered surfaces both the dolomite and the calcareous slates are coated with a skin of brown ocherous limonite, which on some of the massive dolomites reaches a tliickness of an inch or more. Much of the limonite is pseudomorphous after the carbonate siderite. The ferruginous cherts and u"on oxides are not known to be present in the Michigamme slate in large quantity. Indeed, they are as a rule only locally developed in association with the sideritic dolomites and calcareous slates where these have been severely crushed or folded. The source of the iron oxides is clearly iron-bearmg carbonate in the calcareous slates and the dolomites. The cherts are wliite or yellow massive rocks with finely granular texture. They occur as tliin seams and veuis traversing the slates and dolomites, and as tliin beds inter- laminated with the tlucker beds of the last-named rocks. Wlierever the cherts occur there is usually found also a greater or less quantity of some iron oxide. Tliis occurs as small veins of pure hematite cutting through the cherts, as coatings of hematite on the walls of cracks traversmg the slates, as small vugs inclosed in shattered cherts, as druses covering the walls of the cavities in an extremely porous chert, m distinct bands interlaminated with bands of graywacke or cpiartzite, and in the form of a mixture of oxides anil hydroxides impregnating slaty material. In short, the iron oxitles occur in all forms characteristic of deposits precipitated from percolating waters. The slates impregnated with ferruguious matter are naturally dark red. Those that are but slightly ferruginous still plamly exliibit their true character. In those containing a large proportion of the iron oxides, how- ever, but few traces of the original slate remain and the rock resembles a slaty ocher. The grapliite slates are black, very fissile, tliinly laminated rocks. They appear to be limited to the lower portions of the Micliigamme slate. At any rate, they have been seen only in association with the underlying Curry iron-bearing member and at horizons a few hundred feet above the base of the slate formation, but they do not everywhere occur at the base of the formation. Their association with iron-beaiing beds at many places in tliis and other districts probably has some significance as to the origin of the ore. (See p. .502.) The graphite slates appear to grade laterally mto the normal gray slates, of wliich they seem to be local modifications. The graywackes and quartzites of the Micliigamme slate are normal rocks of their kind, requiring no special description. They both occur in comparatively tliin beds, more commonly in the lower part of the formation than in the upper part. The quartzites are more abundant than the graywackes, but neither are common. Deformation. — The major folding of the Michigamme slate seems to correspond with that of the underlying formations, and the slate therefore lies in three major synchnes. This structure is inferred from areal relations to older rocks rather than from structures seen in the slates themselves, which are poorly exposed, lack easily identifiable horizons, and have their bedding much obscured by cleavage. Many of the folds are of the abnormal type characteristic of incompetent strata. The hmbs are thinned and the crests thickened, as would be expected in folds of this type, con- trasting in every essential detail with folds in the competent quartzites and dolomites, as, for instance, in Iron Hill. (See fig. 45, p. 33.5.) The strong north-south compression of the slate beds, producing the close east-west folds, also produced in all the weaker members of the formation a perfect slaty cleavage with a nearly east-west strike and a dip that varies but a few degrees on either side of the vertical. There is also a set of fracture planes or joints at right angles to the cleavage. These joints intersect the rocks at approximately equal intervals of several inches. In some places they 342 GEOLOGY OF THE LAKE SUPERIOR REGIOX. are bordered by narrow shear zones in which the total displacement of tlio slate beds is an inch or more. On some flat horizontal surfaces two sets of these joints are seen cutting each other at acute angles, and about each slight faulting has occurred. Extensive thrust faults are suggested by the close folding of cleavage, but these have not been identified. Tliickness. — No even approximately correct estimate of the thickness of the Michigamme slate can at present be made. The similarity of the beds and their redujjhcation in consecpience of the close folding render it impossible to determine what proportion of the apparent thickness of the formation is due to folding and what proportion is due to successive deposits. There is no (l(iul)t that the Michigamme slate is little thicker than any of the other formations in the district. RELATIONS OF TIPPER HrRONIAN TO TINDERLYING ROCKS. Relations between Vulcan formation and the lovxr Ibtronian. — The ir<)n-l)eariIlg^'ul(■an forma- tion, except in very small areas, rests upon the Randville dolomite or middle Huronian (juartzite. If the Vulcan formation exists in the do-.btful areas adjacent to the Quinnesec schist, it there rests against that schist. Where the Vuican formation rests upon the middle Huronian quartzite or Randville dolomite the lower layers of the younger formation appear to he conformably upon the older one, with an extremely sharp hne of definition between them. In j)laces the contact rock is a talc schist derived from the dolomite or cherty cpiartzite. The basal member of the Vulcan formation is either a quartzite which in i)laces contains ore and jaspilite fragments, or an ore and jasper conglomerate containing large and small pebbles of ore, or a breccia con- taining fragments of all the adjacent rocks. The relative abundance of autoclastic rocks and true water-deposited conglomerates is uncertain. The Traders iron-bearing member appears to be nearly conformable in attitude with the underlying dolomite, but detailed work discloses distinct though slight discordance. Contacts between the Randville dolomite or middle Huronian quartzite and the overlying formation are found in many of the mines, but they are nowhere discoverable on the surface. In the little ravine just east of the old Brier Hill mine the dolomite and the lower members of the iron-bearing formation are very close together, but their actual contact is covered. The space between the ledges of the two forrnations is filled with loose fragments, and among these fragments are large pieces of quartzite holding pebbles of jaspilite, quartzite, granite, and other members of the Archean. In many of the mines and the open pits a similar conglomerate or a coarse quartzite is found lying upon the dolomite or quartzite. The dolomite near the contact is usually schistose, so much so that in most places it is a pure talc schist. The calcium of the dolomite has been removed and much of it has been deposited in the ore bodies as calcite, while the magnesium has remained in the talc. A surprisingl}^ similar schist has been formed from the middle Huronian quartzite, though on the whole it is more siliceous and less talcose. This talc schist serves as an impervious lining to many of the folds in which the ore deposits lie, and afforded better conditions for the concen- tration of the ore material than were afforded by the massive and shattered dolomite under- lying the ore formation at many places. The schist was probably formed in connection with movement along the contact plane after the upper Huronian deposits were laid down, contem- poraneously with the folding and mctamorphism that affected both the lower Huronian and upper Huronian. The contact between the schist and the superjacent quartzite is extremel}' sharp, and in many places the plane of contact is slickensided. In those places where the basal member of the iron-bearing formation is not a coarse quartzite, it is usually a bedded red slate, or more nearly a schist composed of small grains of quartz, considerable dolomite, and locally talc. Alternate bands are composed of layers in which dolomite and talc are predominant and those in wliich siliceous material predominates. Tlie contacts between the schist and the rocks on both sides of it are usually covered. There is in some localities an apparent gradation between these underh'ing rocks and the rocks Ij'ing above them, but in others the line of division between them is well defined. MENOMINEE IRON DISTRICT. 343 In earlier reports certain dense jaspilites were diseriniinated from the fragmental and micaceous jaspilites of the Vulcan formation above them and were regarded as belonging to the middle Huronian, unconformably below the Vulcan formation. The principal evidence of the existence of such a formation is the presence of fragments of jaspilite in the conglomerate at the base of the Vulcan formation. In general, then, there is a slight structural discordance between the beds of the Vulcan formation and the middle and lower Huronian, and schistosity and autoclastic rocks seem to inilicate that this has been a plane of considerable faulting. Also the fragmental phases of the Vulcan formation point to a preceding erosion interval, though evidence of great differential erosion is lacking, and so far as these fragmental j)hases are autoclastic this evidence is weakened. The general significance of the unconformity will be discussed in the chapter on general correlation (pp. 597 et seq.). Relations between AficMgamme {"Ilanbury") slate and the middle or lower Ilurnnian. — The Michigamme slate rests upon the Vulcan formation conformably. Where the Vulcan formation is absent the slat.e rests directly upon the Randville dolomite or the middle Huronian quartzite. This relation is seen for a short distance in the central part of the southern belt of the slate, and it is the relation which prevails generally in the two northern belts, for in this part of the district the Vulcan formation occurs only locally. Contacts are not exposed. At Iron Hill, in sec. 32, T. 40 N., R. 29 W., there is at the top of the middle Huronian quartzite a conglomerate the debris of which is derived largely from that formation and which may be a basal conglomerate of the Michigamme slate. At other locahties also there is a breccia which appears to be a brecciated conglomerate. The absence of the Vulcan formation east of Quinnesec could be explamed by the hypothesis that the Michigamme slate had been thrust over the lower formation of the upper Huronian so as to rest upon the Randville dolomite. The absence of the Vulcan formation between the Michigamme slate and the middle Huronian quartzite at Iron Hill might be similarly explained only here it would be necessary to believe that folding accompanied the faulting, else the manner in which the slate wraps around the east end of the central belt of dolomite would l)e inexplicable. There are undoubted mmor faults in the Menominee district, but most of them are extremelv small, that in the Pewabic mine being the only one of sufficient magnitude to be mapped on the mine plats. It is clear that certain crushed zones of the Traders and Brier members near Vulcan are due to faulting. Further, there have been marked movements of accommodation between the different formations at their contacts, which might be called faidting. AJl these faults are local, and in none of them is the displacement of the faidted beds known to be great. On the other hand the existence of overthrust folds grading into faults, so clearly indicated in the distribution of the southern belt of the iron-bearing formation, is the best of evidence of the extensive relative displacement of the upper and knver Huronian beds, a displacement brought about largely by the close deformation of the lower beds of the upj)er Huronian as they are crowded against the competent beds of the lower Huronian. It is entirely likel}' that more faults are present than ha^-e been found, and there is little difficulty in believing that overthrust faulting may have been a large factor in this deformation and may have thrust the slate locally over the iron-bearing formation against the dolomite, or, on the farther side of the fold, may have carried the dolomite up and over the slate. Although faultmg is doubtless a factor m determming the distribution of the Vulcan formation, from present evidence faulting is inadequate to explain the uniform absence of the formation through such long belts of country where it might be supposed to exist. The presence of doubtfid conglomerates at the base of the Michigamme slate where it rests upon the middle or lower Huronian suggests unconformable overlap of the Michigamme. It is possible also that the iron-bearing formation was originally deposited in discontinuous lenses, with intervening slate, resting directly upon the lower or middle Huronian. 344 GEOLOGY OF THE LAKE SUPERIOR REGION. IGNEOUS BOCKS IN THE ALGONKIAN. QtrlNNESEC SCHIST. The Quinnesec schist lies along and adjacent to Menominee River, from the sharp north- ward bend in the river due west of Iron ih)untiiin to the eastern limit of the area mapped. The river is bordered by scliistose greenstones and various rocks that cut them, except at a few places where rock ledges are absent. The Quinnesec Falls and Sturgeon Falls are on some of the harder ledges of these rocks. South of the river, in Wisconsin, at a distance ranging from half a mile to - miles, is the north side of a large area of granite. This granite sends apoi)hyses into the greenstone schists, and consequently is of later age. For the most part the schists are arranged in belts striking a little north of west at Sturgeon Falls, but trending more toward the north as they pass up the river, until at the Upper Quimiesec Falls they strike about north- west. Their schistosity is, as a rule, nearly vertical. The Quinnesec schist is composed of schists of two classes, basic and acidic. The ba.sic scliists comprise greenstone schists, chlorite schists, and amphibolites. Elhpsoidal and other extrusive structures are common. The acidic schists comprise gneissoid granites, porphyritic cneisses, felsite schists, and sericite schists. Associated with the schists are both basic and acidic massive rocks. The basic rocks include gabbro, diorites, diabases, and basalts. The acidic rocks include granite and granite porphyry. The greenstones and the basic scliists are closely allied, as are also the granites and granite porphyries and the acidic schists. A microscopic study " of the basic scliists shows that they comprise schistose gabbros, diorites, diabases, basalts, and basalt tuffs. Where the schistosity is not strongly developed the original structures of the massive eruptive rocks may be recognized, so that there is no doubt that the greenstone schists, chlorite schists, and amphibohtes are merely altered phases of the greenstones. The amphibolites are limited in their distribution to the neighborhood of the sreat granite mass of Wisconsin, and nearlv all of them occur directlv in contact with this granite. It is clear that the schistosity in these rocks has developed in connection with the folding of the district and that the extreme phase of metamorphism represented by the amphib- olites has taken place in connection with the intrusion of the great batholithic granite of Wisconsin. The acidic schists are limited principally to the neighborhood of Horserace Rapids and Big Quinnesec. The sericite schists in many places grade into the felsite schists. They occur mainly in bands parallel to the trend of the bands of basic schists. The coarser-grained gneissoid granites and porphyritic rocks clearly represent metamorphosed phases of the great granite mass to the south in Wisconsin, but some of the felsite schists and the sericite schists may represent acidic lavas contemporaneous with the basic igneous rocks. From the field relations and microscopic study of the Quinnesec schist and associated rocks it must be concluded that all are of igneous origin. Many of tlieni were lava flows; some were beds of volcanic ashes, or tuffs; others were dikes cutting through the bedded deposits. A few small dikes cutting the scliists are normal diabases and basalts, identical in com- position with some of the rocks cutting through the iron-bearing beds. Within the Menominee district itself there are no contacts between the Quinnesec schist and the Huronian sediments. A sand plane covers the area of contact. Exposures and explora- tions indicate that upper Huronian slates are the rocks nearest to the Quinnesec scliist, and these have not been found nearer than 200 3'ards. In earlier reports on the Menominee district * the Quimiesec schist was provisionally cor- related witli the Koowatin scries of the Archean because of its relatively high degree of meta- niori)hisni and similarity to certain schists in the kiio\\Ti ^Vrchean on the north siile of the district. The apparent absence of the Vulcan formation at the base of the upper Huronian was exjilained by overlap, and later it was suggested that faulting might play a part. During the simimer o Williams, G. U.. The treenstone scliist areas of the Menominee and Marquette regions of Michigan ; Bull. V. S. Geol. Survey No. 02. 1S90. i> Mon. r. S. Geol. Survey, vol. -Id, 1904; Menominee special folio (No. 02), Geol. .Vtlas U. S., U. S. Geol. Survey, 1900. MENOMINEE IKON DISTRICT. 345 of 1910 the Wisconsin Geological and Natural History Survey, under direction of W. O. Hotch- kiss, mapped what is prol^al^ly the continuation of the Quinnesec schist to the northwest along the south side of the Florence district of Wisconsin, and determined the green schists there clearly to overlie the upper Huronian sediments to the north of them and to be locally inter- l)cdded with upper Huronian sediments. However, it is yet possible that the Quinnesec schist in the Menommee district may be really pre-Huronian, for continuous exposures do not connect the two areas, and green schists of this type are known in at least tlu'ee different horizons in the pre-Cambrian of Michigan. GREEN SCHISTS AT FOTJRFOOT FALLS. Another area of igneous rocks of Algonkian age occupies about .5 square miles, extending from about the center of sec. 15, T. 40 N., R. 30 W., to Menominee River. The Fourfoot Falls are on the south side of the area, and the old village of Badwater at its northern edge. The rocks of this area are mainly schists, but they are cut by altered diabase dikes. The schists are gi-ayish-grecn fine-grained greenstones, in which schistosity is nearly every- where noticeable. In some places the rocks are well-defined schists, with a cleavage almost as perfect as that in slates; in other places they are nearly massive. On many of the exposures a typical ellipsoidal structure is discernible. The ellipsoids vary in diameter from a few inches to 3 or 4 feet. There is no striking contrast between the material of the ellipsoids and that of the matrix between them. In both the rock is a dense grayish greenstone without any distinct textural features. The matrix is usually slightly more schistose than the ellipsoids, but otherwise it is like them. At the Fourfoot Falls the exposures consist of alternating beds of massive, schistose, and slaty rocks, striking about N. 80° W., almost at right angles to the course of the river, and yet these rocks are mostly schistose on the Wisconsin side of the river and mostly massive on the Michigan side. The microscopic examination of tliin sections shows that some of the rocks in the western area are altered dolerites still preserving their characteristic textures. Others are so much changed that their original nature can only be inferred fi-om the character of their alteration products. Some of these appear to have been fine-grained dolerites and others perhaps glassy basalts. A few others were originally basic tuffs. All are now aggregates of actinolite, uralite, zoisite, epidote, quartz, and other well-known decomposition products of basic igneous rocks. TMs area of schists, at the time the Menominee monograph was written, was supposed to be equivalent in age with the Quinnesec schist of Menominee River, then regarded as Archean. Later work by G. W. Corey and C. F. Bowen " has shown that they are really intru- sive and extrusive in the upper Huronian or in part contemporaneous flows. That these igneous rocks antedate the chief folding of the district is shown by the fact that they are so extensively transformed to schists. The only other large masses of igneous rocks which have been found m the Huronian series are in the Micliigamme slate and the Sturgeon quartzite. In each of these formations in a number of places are found greenstones, locally in the form of dikes, in other places as siUs, and in others as interbedded eruptives. In the Michigamme slate the form of the igneous bodies is known in but few places. In then- present condition they are much-altered diabases or basalts comj^osed of uralitized augite or hornblende, decomposed plagioclase, and a consider- able quantity of quartz that is probably entirely secondary. PALEOZOIC ROCKS. Small areas of Paleozoic sediments in horizontal sheets lie on the eroded edges of the Huro- nian and Archean rocks. The Paleozoic rocks are represented by two formations, one of Cambrian age and the other of Cambro-Ordovician age. The lower formation consists mainly of red sandstone, and is known as the Lake Superior sandstone. The upper formation is a porous arenaceous limestone, identified by Rominger as corresponding to the Chazy and "Cal- ciferous" of the Eastern States, and designated the Hermansville limestone. The sandstones a Unpublished field notes, 1905. 346 GEOLOGY OF THE LAKE SLTPERIOR REGION. and limestones were at on(> time spread continuously over the entire Menominee district. To the east of the district tiicy still cover all the older rocks. West of Waucedah, however, they have been generally eroded from the valleys, leaving remnants as isolated patches on the tops of the higher hills. CAMBRIAN SYSTEM. LAKE SUPERIOR SANDSTOITE. Lithohr/i/. — The Lake Sii])erior sandstone consists of a lower portion partly cemented by an iron oxide and conscciucntly red in color and an upper portion in wliich the cement is partly calcareous and the color white. The total thickness is estimated by Rominger" at 300 feet. Several ])ieces of the sandstone have been obtained, wliich according to reliable authority came from the ledge tlu-ougli wliich one of the Pewabic mine shafts, near Iron Mountain, was driven. These contain numerous fragments of fossils, some of which were determined by Wal- cott as " the heads of small trilobites, probably Dicelhcephalus misa; also fragments of a large species of DiceUocephalus." According to Walcott, "These indicate tlie Upper Cambrian horizon of the Mississijipi Valley section." Relations to adjacent formations. — The relations of the sandstone to the underlying forma- tions are everywhere practically tiie same. Whetlier on the tops of hills or in tlie depressions between the hills, the horizontal beds of the younger rock rest unconformably upon tlie up- turned and truncated layers of the older series. Moreover, the basal layers of the sandstone contain a great deal of material derived from the immediately subjacent formations. Where the underlying rocks belong to the Vulcan formation the basal member of the sandstone is an ore and jasper conglomerate, composed of huge rounded bowlders of ore and large sharp-edged fragments of ferrughious quartzose slate and jasper in a matrix consisting of quartzose sand, numerous small pebbles and fragments of ore-formation materials, quartzite, and a few pebbles of white ((uartz, of granite, or of other Archean rocks. In a few places their proportion of ferruginous material is so great that they have been utilized as sources of iron ore. CAMBRO-ORDOVICIAN. HERMANSVILLE LIMESTONE. The general character of the Hermansville limestone "is that of a coarse-; rained sandstone, with abundant calcareous cement, in alternation with pure dolomite or sometimes oolitic beds." The limestone may be seen near the top of the hill east of Iron Mountain, on the bluff northeast of Norway, and at several places on the hills north of Waucedah. Its maximum thickness, according to Rominger,'' is about 100 feet, but this maximum is rarely reached in the Menominee district. Only a few fossils have been reported from it. Romin^-er states tliat it has yielded a few fragments of molluscan shells. To these may now be adiled a broken OrtJioceras. a frag- ment resembling a piece of a fh/rtoceras, a gastropod, am.! several otlier fra.Lrmeiitary forms found in the top layer on the ])\iiiY northeast of Norway. THE IRON ORES OF THE MENOMINEE DISTRICT. By the authors and W. J. Mead. DISTBIBUTION, STRUCTURE, AMU RELATIONS. The ore deposits of the Menominee district occur in the two iron-bearing members of the Vidcan formation known as the Traders iron-bearing member and the Curry iron-bearing member. These are separated by the Brier slate member. Much the larger tonnage of ore mined lias come from the Traders member, lying south of the southern dolomite belt. The ores may occur at any horizon within these members, but otlier comlitions being equal they are more likely to occur at low and high horizons than at middle horizons. A number of the o liomlDger, Carl, Paleozoic rocks: Geol. Survey Michigan, vol. 1, pt. 3, 187», p. '<1. » Idem, p. 71. U. S. GEOLOGICAL SURVEY MONOQRAPH Lll PL. XXVIl 6 -5 i&im No 3 SHIFT Si Ih leittl / Un i V, s / \*^ 'Ta/c-schist , ^ VERTICAL NORTH-SOUTH CROSS SECTIONS THROUGH THE NORWAY-ARAGON AREA, MENOMINEE DISTRICT, MICH,, ILLUSTRATING GEOLOGIC STRUCTURE. After Bayley. See page 347 MENOMINEE IRON DISTRICT. 347 large ore bodies extend entirely across the members in which they occur. The deposits of large size rest upon relatively impervious formations, which are in such positions as to constitute pitching troughs. A pitching trough may be made (a) by the Randville dolomite or middle Huronian, underlying the Traders iron-bearing member of the Vulcan formation; (b) by a slate constituting the lower part of the Traders member; or (c) by the Brier slate member, between the Traders and Curry iron-bearing members. (See PI. XXVII; figs. 40, 47, 48.) The dolomite or quartzite formation is especially likely to furnish an impervious basement where its upper portion has been transformed into a talc schist, as a consequence of folding and shearing between the formations. These pitching troughs are minor folds of the drag type. In tJie western and central parts of the south belt of Vulcan formation carrying the principal ore bodies there is consider- ► No3 SHAFT ■ No 2 SHAFT Figure 40.— Horizontal spotion of the Aragon mine at the first level, Menominee tlistrict, Michifran. Scale, 1 inch = 250 feet. After Bayley. able uniformity of pitch to the west, resulting from the westward and upward shearing of the southerly beds with reference to the northerly beds. At the east end of tiie district some of the pitches are eastward. Any portion of the iron-bearing member may have yielded to the shear- ing by a series of tliese drag folds. The ore bodies following the axial lines may thus be in a series of parallel shoots, one pitcliing below the other along the strike. This is well illustrated in the Chapin and Millie mines. In these folds the strike of the shoots at the surface is at a slight angle from the strike of the bedding, as shown in 'figure 49 (p. 3.50). The wall rocks of the ore l)0tlies may be unaltered phases of the iron-bearing member, especially the ferruginous cherts, or any of the rocks forming the impervious basement. The ber the most part on the middle slopes of tlie ridges formed by the Kandville dolomite and middle Huronian quartzite, but they also go beneath the lower ground. .... .... V Ol ii "% WMM&MmifimMif^:-iM'Mi0!yMi "-%. :'■;.' •.W-.'.';.;.;.V;.'-:.;'.V' 'y\yy:'\:'-'-'^]{-yrr\:\'-:y^-yy^\y\^y'-y'.\-'^-^ fk fm^i^!Miy^^M$M0-^M^^^ ,-—~—~ — ^^ . -___^ — ^^— ^_— ^_ Figure 49. — Sketch to show pitch of a drag fold in a monodinal succession. The ores in some places follow the axis of the fold. It will be noted that: the strike of the ore body, measured at th? surface, is at a slight angle to the strike of the bedding, notwithstanding the fact that the ore body follows throughout a single bed or set of beds. CHEMICAL COMPOSITION OF THE ORES. The averages of the cargo analyses of ore shipped from the district in 1907 and 1909, with the range for each constituent, are as foUows: Average chemical composition of ores from cargo analyses for 1907 and 1909, with range for each constituent. Average. Range. 1907. 1909. 1907. 1909. Moisture (loss on drying at 212° F.) 5.92 6.67 2.16 to 7.92 1.07 to 8.77 Analysis of ore dried at 212° F: 50.70 .0.38 1.54 21.15 .17 .28 2.12 .013 1.92 52.23 .074 L41 Hi. 77 .19 1.31 2.70 .012 2.52 39.00 to 59. 90 .010 to .084 .80 to 2.73 5.70 to41.M .03 to .47 .35 to 1.70 .14 to 3.71 .005 to .022 .60 to 3.50 3S. 46 to 01. 20 Phosphorus 008 to 620 .86 to 2.28 Silica 4 ''t to "iQ 14 .07 to 1.27 Lime 63 to 2 90 .70 to 3.9S Sulphur . 006 to . 041 Loss on ignition , 90 to 4 30 Comparison of analyses of all the ores of the district shows tlie rich ores to consist principally of slightly hydrated hematite, witli additional varymg amoimts of magnetite, silica, alumina, lime, magnesia, carbon dioxide, phosphorus pentoxide, and water. Most of the ores contairt also manganese, potash, and soda, and a few of them titanium and carbon. MENOMINEE IRON DISTRICT. Following are three complete analyses of high-grade Menominee ore: Complete analyses of Menominee ores." 351 1. ■ 2. 3. Fe ' 60.64 65.63 o7 03 FP.O3 ■ 85.44 .47 L.'iS .76 1.26 3.02 .004 .060 4.M .15 .002 91.51 1.97 1.53 80 15 FeO 1.10 AI^Gi 3 88 MiijOj CaO ..■i6 .21 .57 .03 3.03 .021 .099 .38 .27 .17 MgO .48 K.O 2 29 Na.O .30 SiC). 10.72 .074 P^Oi s 146 CO, .08 H-OCabove 100°) 2.75 56 99. 842 99.980 99.950 a Mon. U. S. Geol. Survey, vol. 40, 1904, p. 383. 1. Chapin ore: analysis furnished by E. E. Brewster. 2. "Soft specular" Quinnesec ore. 3. "Soft specular blue ore" from Cornell ir.ine. AVERAGE IRON CONTENT OF THE IRON-BEARING FORMATION. An average of 1,681 analyses, representing 5,287 feet of drilling from the district away from the available ores, gives 37.93 per cent of iron. Ores of this class are so much more abundant than the "available" ores that the average of the entire formation, including ores, is not much higher than this figure. The composition of the lean, unaltered jaspers where not altered to ore' has not been averaged, but presumably the iron content is about 25 per cent, as in other districts. MINERAL COMPOSITION OF THE ORES. The approximate mineral composition of the average ores of the Menominee district, calculated from the preceding average chemical analysis, follows: Approximate mineral composition of average Menominee ore, calculattrl from average cnemical analysis. Hematite (including a small amount of magnetite) . Limonite Quartz Kaolin Serpentine and talc Dolomite Miscellaneous 100.00 The richer ores are usually bluish black, porous, fine-grained aggregates of crystallized hematite. These rich ores grade into leaner phases containing more or less hydrated hematite, with varying amounts of quartz, soijientine, talc, clay, and carbonates of calcium and magne- sium, rangmg in color from the bluish black of the richer ores through various shades of red and brown to yellow. All the minerals occurring as constituents of the ores are found also as visible masses either hi veins cuttmg the ore bodies or in ^'ngs or pores within them. Dolomite, calcite, and pyi-ite occur locally in excellent crystals, and serpentine as large, white, almost pure masses. Talc also occurs in thick scams of almost ideal jiurity, and chalcopyrite in small crystals associated with pyrite. The carbonates and sulphiiles are found near watercourses and the silicates mainly in the lower portions of the ore bodies. 352 GEOLOGY OF TPIE LAKE SUPERIOR REGION. Tlie ores when exposed to the action of the atmosphere become coated with a white elllorescence, consistiiifi of a mixture of the sulphates of sodium, magnesium, and calcium, in wliicli tire sodium sulpluito is greatly in excess. PHYSICAL CHARACTERISTICS OF THE ORES. The lean ores differ vcmt little in appearance from the jaspilites, of which they are essentially a part. They are banded, brecciated, and m places specular. The brecciated ores may consist of jas})er fragments in a mass of hematite, or of hematite fragments in a mass of dolomite, or fragments of ore, jasper, and slate m a mass consisting largely of slate debris that has been strongly ferruginized. IRON MINERALS SILICA PORE SPACE Figure 50.— Triangular diagram representing the volume composition of the various grades ot ore mined in the Menominee. Crystal Falls, and neighboring districts in 1907. M, Menominee; CF, Crj'stal Falls; IR, Iron River; F, Florence; FM, Felch Mountain: C. Calumet. The pore space fur each grade was calculated from the average moisture content, and hence represents the true pore space only when the moisture in a particular grade was at a maximum. The true porosity of the various grades of ore would therefore be slightly greater than is shown. For description of the method of platting on triangular diagram, see page 189. The average texture of the Menominee ores is shown by the following table of screening tests, made by the Oliver Iron Mming Companj' on six typical grades of ore representing a total of 1,033,491 tons. Each test was made on a sample of 100 pounds, representative of the entire 3-ear's output of that grade. A comparison of the textures of the ores of the several ranges is shown in figure 72 (p. 481). MENOMINEE IRON DISTRICT. 353 Textures of Menominee ores as shown by screening tests. Per cent. Held on ^-inch sieve 39. 44 ^-inch sieve 30. 63 No. 20 sieve 11.56 No. 40 sieve 4.73 No. 60 sieve - 1-31 No. 80 sieve !■ I'J No. 100 sieve 1-35 Passed through No. 100 sieve 9-67 The mineral density of the ores varies with the iron content. The average mineral density of the ores calculated from the average of the cargo analyses for 1907, by computing the mineral composition and properly combining the densities of the component minerals, is 4.28. To test the accuracy of this method of computing mineral density, pycnometer determina- tions were made on the average pulp samples '^ of Ajax antl Cluipin grade for 1907, with the following results: Mineral density of Menominee ores. Determined by means of pycnometer. Calculated from chemical analysis .... Ajax grade. 4.21 4.34 Chapin grade. 4.601 4.607 The porosity of the ores ranges from 1 per cent or less in some of the lean jaspilite ores to as much as 45 per cent in some of the richer hematites and especially in the limonitic ores. The cubic contents of the ores vary greatly. The bulk of the ores, however, lies between 9 and 14 cubic feet to the ton. Volume comparisons of the Menominee ores with each other and with ores of the Crystal Falls, Iron River, Florence, Felch Mountain, and Calumet districts are made in figure 50. IRON ORE AT BASE OF CAMBRIAN SANDSTONE. The basal conglomerate of the Cambrian sandstone where it rests upon tlie iron-bearing . formation contains abundant fragments of that formation. In a few places the proportion of ferruginous material is so great that the conglomerates have been utilized as sources of iron ore. A deposit of this kind was formerly worked by the operators of the Quinnesec inine, and another has recently been worked by the Pewabic company. The latter was reached by the open pit in the SE. J sec. 32, T. 40 N., R. 30 W., known as the Pewabic pit. Although at this place the rock immediately underlying is dolomite, the amount of iron ore in the conglomerate is so great that the company operating the pit felt warranted in erecting concentrating works on the property for the separation of the ore from the sandstone. SECONDARY CONCENTRATION OF THE MENOMINEE ORES. Structural conditions. — The ore deposits in the Menominee district rest upon steeply dipping impervious basements of sheared dolomite or slate. The hanging wall may be of slate or iron- bearing formation. The greater dimensions of the deposits are parallel to the bedding. Fold- ing, of the type illustrated in figure 12 (p. 123), develops minor corrugations in the foot wall and other rocks, with pitches parallel to the main strike of the formation. In tliese pitching folds the ore deposits are hkely'to be larger and better concentrated than elsewhere. It is obvious that the flow of water concentrating the ore has been principally parallel to the bedding, that it has been especially strong where the bedding has been folded into pitching troughs, and that the fracturing of the brittle iron-bearing rocks during tliis folding has aided greatly in the circu- lation of waters in pitcliing troughs and elsewhere in the formation. The ores are associated with marked topographic relief, affording abundant head for the waters. The larger number o Kindly furnished by Mr. J. H. Hitchens, chief chemist for the Oliver Iron Mining Company at Iron Mountain. 47517°— VOL 52—11 ^23 354 GEOLOGY OF THE LAKE SUPERIOR REGION. of them are on the upper or middle slopes of the rock elevations, though some of them extend })eneath tiio depressions. Mincralofjical and chemical changes.— The iron-bearing formation was originallj' iron car- bonate and greenalite interbedded with more or less slate and containing much detrital ferric oxide at the base of the formation. The alteration of the chcrtj' iron carbonate and greenalite to ore has been acconij)lisiic(l in the general manner already described as typical for the region — (1) oxidation and hydration of the iron minerals in place, (2) leadiing of silica, and (3) intro- duction of secondary iron oxide and iron carbonate from other parts of the formation. These changes may start simultaneously, but 1 is usually far advanced or complete before 2 and 3 are conspicuous. The early products of alteration therefore are ferruginous cherts — that is, rocks in which the iron is oxidized and hydrated and the silica not removed. The later removal of silica is necessary to produce the ore. SEaUENCE OF ORE CONCENTKATION IN THE MENOMINEE DISTRICT. The first considerable concentration of ore in the district which is now minetl did not take place until the erosion period following upper Huronian time. As indicated in the general discussion, the process was well advanced before Cambrian time and has practically continued to the present. CHAPTER XIV. NORTH-CENTRAL WISCONSIN AND OUTLYING PRE- CAMBRIAN AREAS OF CENTRAL WISCONSIN. NORTHERN WISCONSIN IN GENERAL. The only work done by the United States Geological Survey in northern Wisconsin is in the Florence district; the southern extension of the Menominee district, in the northeastern part of the State ; the Penokee range, in the northern part of the State ; and the Keweenawan belt crossing the northwest corner of the State. These districts are described on other pages. Other areas in northern Wisconsin have been examined in reconnaissance work by members of the Survey, but no detailed mapping has been done. Outside of the areas named, the chs- tribution of the rocks of northern Wisconsin shown on the general map (PI. I) is taken from the Wisconsin Geological Survey reports, particularly that of Weidman '^ for north-central Wis- consin. The recent map of Douglas County made by Grant '' for the Wisconsin Geological Survey is used in place of the earlier map by Irving. Granites and gneisses, with subordinate amounts of sedimentary rocks and basic igneous rocks, constitute a highland in the northern part of the State, roughly oval in its outline, extend- ing from the vicinity of Grand Rapids and Stevens Point, on the south, to the State boundary, on the north, and from Barron County eastward to the Micliigan boundary. The area is bounded on the northwest by the Keweenawan rocks described in Chapter XV, and on the north and northeast by the Huronian formations of Michigan; on the southeast, south, and southwest it is overlapped on the lower ground by Paleozoic sediments which outcrop in wide belts sur- rounding the pre-Cambrian core. The predominating granites and gneisses were called Lauren- tian and the sedimentary rocks Huronian by the geologists of the first Wisconsin Geological Survey (1882). The highlands as a whole have been often referred to as a "Laurentian liigh- land." The drift cover is heavy, exposures are few, except in certain localities, antl much of it has been difficult of access even to the present time. The only published detailed work is that of Weidman, ° which is summarized below. WAUSAU DISTRICT. LOCATION, AREA, AND GENERAL GEOLOGIC SUCCESSION. The pre-Cambrian area in north-central Wisconsin mapped and described by Weidman includes the counties of Marathon, Portage, Wood, Clark, Tajdor, Lincoln, and adjacent parts of Ru.sk, Price, and Langlade, containing in all about 7,200 square miles. From 90 to 95 per cent of the pre-Cambrian rocks of this area are of igneous origin. The following table is compiled from the succession worked out by Weidman. The rocks he classes doubtfully as lower and middle Huronian we classify doubtfully as middle and upper Huronian, respectively. The names of the formations are those used by Weidman. Quaternary system: "Wisconsin drift. Pleistocene series. Third drift. Second drift. First drift. Alluvial deposits (contemporaneous with drift). a Weidman, Samuel, The geology of north-central Wisconsin; Bull. Wisconsin Geol. and Nat. Hist. Survey No. 16, 1907. t Grant, U. S., Preliminary report on the copper-bearing rocks of Douglas County, Wisconsin; Bull. Wisconsin Geol. and Nat. Hist. Survey No. 6, 2d ed., 1901. 355 356 GEOLOGY OF THE LAKE SUPERIOR REGION. Unconformity. ■ Cambrian system Upper Cambrian or Potsdam sandstone. Unconformity. Algonkian system : Huronian series: fNorth Mound oonglomerate and quartzite. Upper Huronian? ("Middle Huronian?" or "Upper sedimentary group," of Weid- man). (Stratit;raphic relations unknown; formations presumably contemporaneous.) Arpiu conslomorate and quartzite. Mosince confjlomerate. Marshall Hill conglomerate. Marathon conglomerate. Unconformity. , . f3. Granite and nepheline syenite series. Intrusive igneous rocks. (In order of in-L, ^ , , i ,• •, ^ " ^ ^2. Gabbro and diorite senes. trusion) Middle Huronian? ("Lower Huronian?" or "Lower sedimentary group," of Weid- manV (Stratigrapliic relation.^ unknown.) 1. Rhyolite series. Rib Hill quartzite. Powers Bluff quartzite. Hamburg .slate. Wausau graywacke. Unconformity. Archean system (?) Gneiss and schists. ARCHEAN (?) SYSTEM. The basal rocks, believed to be the oldest aud to belong to the Archean system, consist of a complex mixture of rocks, such as contorted and crumpled granite gneiss, diorite gneiss, granite scliist, syenite scliist, and diorite schist. The gneisses and scliists form a belt which can be fairly well outlined, extending from the vicinity of Stevens Point and Grand Rapids in a northwesterly chrection through Neillsville. The rocks are closely intermingled with one another, and have been subjected to extensive folding and metamorphism. The zone in which they are largely comprised lies between areas of later igneous and sedimentary rock to the north and to the south, and hence appears to have the position of the arch of an anticline. These basal rocks are intruded by later formations of rhyolite, diorite, and granite. Sedimentary rocks have not been found in contact with the basal rocks. ALGONKIAN SYSTEM. HURONIAN SERIES. MIDDLE HUKONL\N (?). The rocks next succeeding are of sedimentary origin, and consist of quartzite, slate, and graywacke. They include the quartzite of Rib Hill and ^^cinity, the quartzite of Powers Bluff and in the vicinity of Junction and Rudolph, a wide belt of slate in northwestern Mara- thon County, and graywackes in the vicinity of Wausau. These rocks are almost entirely of fragmental origin, and only rarely contain phases of carbonaceous, calcareous, and ferruginous deposits. The basement upon which these sediments were deposited can not be defuiitely determined, for all the observed contacts with associated rocks are those either of later intru- sive igneous rocks or of later overlving conglomerate. The quartzites are throughout extremely metamorphosed and to all appearances completely recrystallized. The slates and gra3'wackes do not reveal as much metamorpliism as the quartzite, although in places rocks presumably belonging with the slate have been changed to schists bearing staurolite, cordierite, and garnet. These sedimentary rocks appear to bear the relation of great fragmentary masses intersected and surrounded by later igneous intrusive rocks. They constitute the lowest and oldest sedi- mentary rocks of this area. NORTH-CENTRAL WISCONSIN AND OUTLYING PRE-CAMBRIAN AREAS. 357 ROCKS INTRUSIVE IN MIDDLE IIURONIAN (?) AND ARCIIEAN ( ?). The next younger rocks are of igneous origin. They form about 75 per cent of the rocks of the area, and in the order of their intrusion arc (1) rhyoHte; (2) a basic scries of diorite, gabbro, and peridotite; (3) a series consisting of granite, quartz syenite, nephehne syenite, and related rocks. Of these the last-named series is the most abundant, the granite alone forming about 50 per cent of the surface rocks of the area. The three series are intrusive in the Archean(?) of the area and also in the middle Huronian (?). They are in turn overlain by later Algonkian sediments. The period involved in the intrusion of the igneous formations must have been a very long one, and evidently constituted an important portion of the pre- Cambrian era, for the granite and syenite series itself represents a complex magma of varymg though related rocks, intruded at different dates. In the stratigraphy of this area, therefore, these igneous intrusives play an important part and occupy a well-defined position between the upper Huronian ( ?) and the middle Huronian ( ?) sediments. UPPER HURONIAN (?). The latest Algonkian rocks of the area consist mainly of conglomerate and quartzite over- l3dng all the other rocks above referred to. North of Wausau, at Arpin, and at North Mound they are represented by conglomerate and quartzite, and at Marathon City and Mosinee by conglomerate. CAMBRIAN SYSTEM. In the north-central area the pre-Cambrian was worn down to base-level by subaerial erosion before the much later Upper Cambrian or Potsdam sandstone '^ was dejjosited upon it.* BARRON, RUSK, AND SAWYER COUNTIES. In Barron, Rusk, and Sawyer counties the pre-Cambrian rocks are largely of igneous origin. The most prominent sedimentary areas are the prominent ridge of quartzite at the junction of Flambeau and Chippewa rivers and the numerous quartzite I'idges along the divide of Cliippewa and Red Cedar rivers. In general, these quartzites dip westward, away from the crystalline and schistose area, with strongly marked eastward escarpments overlooking the nearly flat plain of older rocks. Although no final conclusion has been reached concerning the relative age of these quartzites, Weidman is of the opinion that there are here represented at least two and probably three series. The quartzite in the small outcrops along the railroad about .3 miles east of Weyerhauser is greatly metamorphosed and is correlated with the Rib Hill quartzite at Wausau. The prominent ridge of quartzite at the junction of the Flambeau and the Chippewa is correlated with the upper sedimentary series in north-central Wisconsin and the Baraboo quartzite. The prominent ridges of quartzite in eastern Barron County and in the adjacent parts of Rusk and Sawyer counties are but slightly metamorphosed, the bedding is in general nearly flat-lying, and the formation has a much younger aspect than the other two quartzite formations in the region and may be Keweenawan. o The term Potsdam sandstone is here used in a quotational sense from the Wisconsin Geological Survey. <> Weidman, Samuel, The pre-Potsdam peneplain of the pre-Cambrian of northKjentral Wisconsin: Jour. Geologj', vol. 11, 1903, pp. 289-313. 358 GEOLOGY OF THE LAKE SUPERIOR REGION. VICINITY OF LAKE WOOD. Quartzites arc known in the, vicinity of Lakewood, indicating the presence of Huronian rocks in tliis district. Practically all that is known concerning the distrii)iition and structure of these quartzites is shown on the accompanying sketch (fig. 51). They stand up as monad- R. 16 E. R.I7 E. 20 21 22 ?3 24 19 20 P P 21 22 23 24 19 20 21 22 29 28" 27 26 , 25 30 29 26 27 26 25 29 b\°*'S'?b ?27 32 Gr 33 34 3S 36 31 32 % ^V 35 * Or 32 33 34 /^ Gr*Gr 4 .8 ^B 2 1 6 5 4 3 Z 1 6 5 4 3 8P* 9 10 M 12 7 8 9 10 II 12 7 6 9 10 17 16 15 14 13 18 17 16 15 14 13 IS 17 16 15 FlGtTRE 51.— Sketch map showing occurrence of quartzites of Huronian age in Tps. 33 and 34 N., Rs. 15, 16, and 17 E., Wisconsin. B. Quartzite and quartzite breccia; C, conglomerate; D, diabase; Gr, granite; P, porphyry. nocks above the surrounding drift-covered surface. Associated with them are granite, por- phyry, and diabase in isolated exposures. NECEDAH, NORTH BLUFF, AND BLACK RIVER AREAS. At Necedah, in Juneau County (see figs. 52 and 53), and at North Bluflf, in Wood Count}'-, are quartzite exposures projecting tlirough the Cambrian. R. 3 E. R. 4- E. DRILL HOLE O Quartz dionte at 229' 14 23 D/f/LL HOLE o Granite and diorite^ at 202' 13 Necedah, NIV ft 24 " DR/LL HOLE O Quartz dionte at 192' 1/2 zMiles FiatniE 52.— Sketch map shownng occurrence of Huronian quartzite near Necedah, Wis. Drilling at Necedah has cUsclosed the presence of granite, probably intrusive into quartzite. The quartzite is highly metamorphosed and is lithologically similar to the Huronian rocks. BAEABOO IRON DISTRICT. 359 111 the Black. River vallej', north of Black River Falls, are exposures of gneiss, granite, hornblende schist, magnesian schist, and ferruginous quartz schist, mapjied l)_y Irving ° in 1873 and by Hancock * in 1901. The relation of these rocks to one another is not defmitely Ivnown. All are pre-Cambrian. BARABOO IRON DISTRICT.'^ LOCATION AND GENERAL GEOLOGIC SUCCESSION. The Baraboo district constitutes an outher in the Paleozoic rocks and centers in tlie town of Baraboo, in the south-central part of Wisconsin. (See fig. 53.) It is south of the area R.4-E;. R.SE. PALEOZOIC R.6E. R.7E. R,8E. R9E.. RIOE. HURONIAN SERIES fMIDDLE ?) RUE, R.IZ.E. R.I3E. R.I4-E. LAURENTIAN? SERIES (This also covers part of areas mapped as Huronian) <%' Seeiey slate, Freedom dolomite, including iron-bearing member Granite and metarhyolite ZO MILES FiGUEE 63.— Sketch map showing Baraljoo, Fox lliver valley, Necedah, Waushara, and Waterloo pre-Cambrian areas o( south-central Wisconsin. shown on the general Lake Superior map (PI. I), but a brief description is here given because the district is producing iron ore and is similar lithologically and structurally to the iron- producing area of the Lake Superior region. a Irving, R. D., The Necedah quartzite: Geology ol Wisconsin, vol. 2, 1873-1877, pp. 523-524. 6 Hancoclv, E. T., The geology of the area at Blacli River Falls, Wisconsin: Unpul)lished thesis, Geol. Dept. Univ. Wisconsin, 1901. cSee Weidman, Samuel, The Baraboo iron-bearing district of Wisconsin: Bull. Wisconsin Geol. and Nat. Hist. Survey No. 13, 1904. 360 GEOLOGY OF THE LAKE SUPERIOR REGION. The area is elliptical in outline, extending 20 miles east and west and ranging in width from 2 to 12 miles. It is essentially a quartzite syncline. The succession is as follows: Quaternary system Pleistocene deposits. {Trenton limestone. <" St. Peter sandstone. " Lower Magnesian " limestone. Cambrian system Potsdam sandstone." Unconformity. Algonkian system: Hurouian series: Upper Huronian (?).. .Quartzite. Unconformity. fGranite, intrusive into lower formations. Freedom dolomite, mainly dolomite, including an iron-bearing member in its lower part. Seeley slate, 500 to 800 feet. Baraboo quartzite, 3,000 to 5,000 feet. Middle Huronian (?). Unconformity. Archean system : I.aurentian (?) series Granites, rhyolites, tuffs, etc. The principal structural feature is an east-west synclinorium of middle Huronian ( ?) rocks resting on the -Archean basement, carrying in the trough unconformable upper Huronian ( ?) c[uartzite and Paleozoic sediments, and surrounded by Paleozoic sediments. The edges of the basin, composed of hard, resisting middle Huronian (?) c[uartzite, form ridges known as the north and south Baraboo ranges, standing 700 to 800 feet above the surrounding country and the intervening valley. (See fig. 54.) 10000 500O 4.00O -3000 -2000 4-1000 SealeveZ 1000 HOOO -3000 ..+000 Lsooo 4 Miles Figure 54. — Generalized cross section extending north and south at-ross the Baraboo district. 1, Potsdam sandstone; 2-4, Htironian (2, Freedom dolomite; 3, Seeley slate; 4, Baraboo quartzite); 5, rhyolite and granite (Laurentian?). Alter Weidman. ARCHEAN SYSTEM. LATJBENTIAN SEKIES. The Laurentian roclis outcrop in isolated areas bordering the outside of the Baraboo syncline. The surface volcanic phases are best exposed west of the Lower Narrows of Baraboo River on tlie northeast .side and near the town of Alloa on the southeast side. Thov are similar to tlie surface volcanic rocks of the Fox River valley. Granitic rocks appear in isolated areas on the south side of the belt. Some of these rocks, previously considered as Archean, have recently been found to be intrusive into the rhiddle Huronian ( ?) . o Used la a quotational sense from the Wisconsin Geological Survey. BAEABOO IRON DISTRICT. 361 ALGONKIAN SYSTEM. HTJKONIAN SERIES. MIDDLE HURONIAN (?). BAEABOO QUARTZITE. The Baraboo quartzite is a massive though well-bedded formation, considerably jointed, faulted, and brecciated, but showing no cleavage as e\ddence of rock flowage except along certain thin slate beds in which readjustment has been concentrated during folding. Cross-bedding, ripple marks, and thin conglomeratic layers are numerous. In the north range the beds stand nearly vertical; in the south range they dip gently toward the south. Isolated exposures in the north-central side of the trough are thought to be brought up by minor folds. There is, however, a possibihty that faulting has been a factor. SEELEY SLATE. The Baraboo c{uartzite passes up into the Seeley slate, a soft, gray, finely banded chlorite slate, known principally by drilling along the south hmb of the syncline. The cleavage is some- what steeper than the bedding, corresponding to the normal development of cleavage in such relation to a syncline. FREEDOM DOLOUITE. The Freedom formation consists principally of dolomite but contains near its base slate, chert, and iron ore and all gradational phases between these kinds of rocks. The lowest mem- ber is a ferruginous kaolinitic slate, well exposed in the Illinois mine, representing a fernigi- nated gradation phase of the Seeley slate into the Freedom dolomite. The next overlymg member of the Freedom dolomite is banded ferruginous chert and iron ore, known principally along the south hmb of the syncline, but occurring also in the east end of the basin and in several explored areas on the north side. Interbedded mth tlie chert, especially near its upper parts, are calcite, siderite, kaolin, and dolomitic slates. Minor folding and brecciation are conspicuous in this member, part of it at least resulting from secondary chemical changes, causing slump in the formation. The cherty dolomite making up the upper member and by far the greatest part of the Freedom formation is a fine-grained banded rock similar in some of its phases to the ferruginous cherts but usually softer. It grades locally into ferrodolomite. UPPER HURONIAN (?). Upper Huronian (?) cjuartzite has been found only by drilhng in the deeper parts of the east end of the trough. Only recently has it been discriminated from the Cambrian sandstone above it or the middle Huronian ( ?) quartzite below. When the drill penetrated tlie Cambrian sandstone and conglomerate and reached quartzite below it was usually assumed that this was -the middle Huronian ( ?) quartzite and the drilling was stopped. When tliis quartzite was pen- etrated by the drill, however, it was found to overlap the edges of all the middle Huronian (?) rocks and to have conglomerate at its base. The thickness of tliis quartzite, as shown by the drilling, is not more than 50 feet. Its attitude is not definitely known, but from the way it lies over all the earlier formations it is beUeved to be not much tilted. No exposures of the formation are recognized as sucli. It seems to remain simply as a residual patch in the deeper part of the trough where protected from erosion. However, some of the quartzite on the so-called Baraboo ridges may be upper Huronian ( ?) rather than middle Huronian ( ?) . Still more recently red slate has been found above tliis upper Huronian (?) quartzite. . PALEOZOIC SEDIMENTS. The Paleozoic rocks consist, from the base upward, of the Potsdam sandstone, the "Lower Magnesian" Umestone, the St. Peter sandstone, and the Trenton hmestone. The Potsdam sandstone occurs on the lower flanks of the quartzite ranges and in the valley bottom; the 362 GEOLOGY OF THE LAKE SUPERIOR REGION. succeeding formations lie at higher elevations. The Paleozoic beds rest horizontally uj)on the more or less folded Huronian beds, a conspicuous basal conglomerate marking the great uncon- formit}'. QUATERNARY DEPOSITS. Pleistocene deposits cover about the northeast half uf tlie district. (See Chapter ^Yl, pp. 427-459.) THE IRON ORES OF THE BARABOO DISTRICT. l»y the authors and W. J. Mead. OCCURRENCE. The iron-bearing beds, which are a part of tiic Freedom dolomite, have been productive thus far on the south limb of the basin. They dip northward at angles ranging from .50° to 70°. The foot wall is Seeley slate; the hanging wall is cherty dolomite, with small amounts of slate and iron carbonate. The iron-bearing member itself consists of ferruginous chert, iron carbonate, ferruginous slate, and iron ore. There is a gradation from this member into both hanging and foot walls. It is thin, for the most part not more than 200 feet thick, and the productive ore bodies are still thinner, 20 to 30 feet. The ores stand as lenses arranged end for end or overlapping paraUel to the layers of chert. These have been found by drilling at a maximum depth of 1,500 feet, but mining operations do not yet go beyond 500 feet. Their lateral extent has been found to be at least 2,000 feet. Deep drilhng down the dip discloses minor folds. Also to the south of the main outcrop the ore may be repeated by an additional minor fold. The iron-bearing member has been found also on the north side of the basin, where it stands almost vertical or dips south, but so far it is nonproductive here. Only one mine has operated to the present time, the Illinois mine (see fig. 55), although three other shafts are now being sunk. CHEMICAL COMPOSITION. The following is a complete analysis of the Baraboo ore : " Chemical composition of the Baraboo ore. Ferric oxide (FejOs) 88. 62 Ferrous oxide (FeO) 92 Alumina (AUOj) 68 Manganese monoxide (MnO) 26.5 Silica (SiO,) K06 Lime (CaO) 12 Magnesia (MgO) None. Titanium oxide (TiOo) None. Sulphur (S) Trace. Chromium oxide (CrjOj) None. Water at 110° 21 Water at red heat 55 Carbon in carbonaceous matter 04 Carbon dioxide (CO,) 51 Phosphoric oxide (P2O5) 004 99. 979 Total iron 62. 75 o Weldman, Samuel, The Baraboo iron-bearing district of Wisconsin: Bull. Wisconsin Geol. and Nat. Hist. Survey No. 13, 1904. p. 12S. BARABOO IRON DISTRICT. 363 A commercial analysis showing the average grade shijDped for 1007 is as follows: Partial analysis showing average grade of ore shipped for 1907. [Sample dried at 212° F.] Fe 53.47 P 043 StO, .' 18. 51 Mn 22 Moisture U. 36 An average of 1,517 analyses, representing 4,814 feet of driUing in the iron-bearing member away from the available ores, gives 36.40 per cent of iron. This includes both the lean jaspers and the partly altered jaspers but not the ores. Because of their great mass compared with the ores, they represent nearly the general average composition of the entire iron-bearing member. MINERALOGICAL CHARACTEB.o The Baraboo iron ore is mainly red hematite with a small amount of hydrated hematite. There are also small amounts of iron carbonate in isomoiphous combination with varying amounts of manganese, calcium, and magnesium carbonate. Next to hematite in abundance is quartz or chert, which occurs either in bands in the ore or somewhat uniformly distributed throughout the ore. Chlorite, mica, and kaolin also occur in the ore in vaiying but usually small quantities. The vein material in the ore is to a very large extent quartz, to a small extent calcite or ferrodolomite, and to a very small extent iron sulphide and iron oxide. The quartz of the veins has the usual interlocking granitic texture of vein quartz and is generally very much coarser than the finelj^ granular cherty quartz in the ore and in the banded ferruginous chert. The carbonate of the veins is also much coarser than the carbonate of the beds. PHYSICAL CHARACTER. 6 The common phases of the Baraboo ore are soft granular ore, hard banded ore, and hard Uue ore. The soft granular phases generally carry the highest percentage of iron, the banded and hard blue ore containing usually a larger amount of silica. The ore in its prevailing aspects is more like the hard varieties of ore of the old ranges of the Lake Superior district than the soft, hydrated hematite ore of the Mesabi district. SECONDARY CONCENTRATION. Structural conditions. — The circulation of waters in this district is controlled by the imper- vious foot-wall slate on the one hand and the impervious dolomite on the other. The zone between is a narrow one. The shaft of the Illinois mine (see fig. 55) goes down in the foot-wall slate. In walking from the shaft in the drift toward the ore body one notes the conspicuous dryness of the slate as contrasted with the extreme wetness of the drift where it passes through the iron-bearing member. Water is circulating at the present time through the iron-bearing member with great rapidity. The point of escape of the waters is not known; neither is it possible to tell what the depth of circulation has been. Ores have been found to a depth of 1,500 feet, but the deep ores were not so rich as those at the surface. The Baraboo quartzite ridges control the major circulation. The ores, however, are a considerable distance from the foot wall of these ridges on a comparatively flat area, although the hanging walls are usually in still lower ground. Original character of the iron-hearing member. — The iron-bearing member was at -least in larger part iron carbonate, as shown by the residual iron carbonate into which the member grades below, but it may have consisted also in part of banded ferric hydrate and chert. o Weidman, Samuel, op. cit., pp. 127-128. l> Idem, p. 127. 364 GEOLOGY OF THE LAKE SUPERIOR REGION. Mineralogical and chemical changes. — The alterations of tlie iron carbonate have been accomplished through the usual processes as described on earlier pages. All stages of altera- tion arc to be observed and all criteria for determining these alterations are known to be present. Weidman believes that the iron ore of the Baraboo district was originally a deposit of ferric hydrate, or limonite, formed in comparatively stagnant shallow water under condi- tions similar to those existing where bog or lake ores are being formed to-day, and that J^hrough subsequent changes long after the iron was deposited as limonite, while tiic member was deeply buried below the surface and subjected to heat and pressure, the original limonite became to a large extent dehydrated and changed to hematite, and that therefore its structural relations are not j)rimarily con- trolled by the necessity of later water circulation. Though this district is widely separated from the principal Lake Superior ranges and may have the different origiji outlined by Weidman, its close similarity in lithology and stnicture to the Lake Superior ranges is believed to be a priori evidence of simi- larity in origin. The theorj* of origin of the Lake Sujierior ores adequately explains the origin of the Baraboo ores and is combated by no facts yet shown in the Baraboo district. Moreover, recent deep drilling has shown an abundance of original iron carbonate. Certainly development work has not been nearly sufficient in the Baraboo dis- trict to warrant any conclusions at variance with those for the older Lake Superior ranges at the present time. Shale bed -V^ ^V//////////Ai -— ^ -— •— /'//Second /eve/ FlOUBE 55.— Vertical section of liilnoismine. (After Weidman, Bull. Wisconsin Geol. and Nat. Hist. Survey No. 13, 1904, fig. 1, pi. 15.) WATERLOO QUARTZITE AREA. The mapping of the Waterloo quartzite at Portland, Hubbleton, Mudlake, and Lake Mills (see fig. 53) by Buell ° and subsequently by J. H. Warner * shows that the outcrops of this quartzite have a distribution and stiiicture such as to suggest that they represent part of a great eastward-pitching syncline of quartzite. The quartzite is lithologically almost identical with the Baraboo quartzite and its synclinal axis has the same direction as the axis of the Baraboo syncUne. There is little reason to doubt that the Baraboo and Waterloo quartzites are of the same age. If this is the case, one would expect to find slate and ferru- ginous dolomite formations within the Waterloo quartzite syncline, as in the Baraboo syn- cline, but drilling has thus far failed to locate them. Like the Baraboo quartzite, the Waterloo quartzite is referred to the Huronian, and its similarity with the middle Huronian is emphasized. Well drilling outside of the Waterloo syncline shows the presence of a granite basement. a Buell, I. M., Geology of tlie Waterloo quartzite area: Trans. Wisconsin Acad. Sci., vol. 9, 1893. pp. 255-274. i Warner, J. n., The Waterloo quartzite area of Wisconsin: Unpublished bachelor's thesis, Dept. Geology Univ. Wisconsin, 1904. GEOLOGY OF THE LAKE SUPERIOR REGION. 365 FOX RIVER VALLEY." Several small isolated outcrops of pre-Cambrian crystalline rocks project through the PAleozoic sediments in the Fox River valley at Berhn, Utley, Waushara, Marquette, Montello, Observatory Hill, Marcellon, and Endeavor. (See fig. 53, p. 359.) The rocks are mainly acidic extrusives; metarhyolites, showing gradation mto rocks of more deep-seated origin; rhyolite gneiss; quartz rhyolite; and granite, all of them cut by basic dikes. The characteristic fea- ture in the metarhyolites is the presence of abundant and well-jircserved surface volcanic textures, such as fluxion, perlitic, spherulitic, and brecciated textures. The hthologic simi- larities of the rocks, the presence of the surface textures, and their composition, as shown by analysis, indicate clearly their consangumity with one another and with certain of the igneous rocks on the north and south sides of the Baraboo range. In the Baraboo district these rocks have been found by Weidman '' to lie unconformably below the sedimentary rocks, and hence the volcanic rocks of Fox River may be supposed to be pre-Huronian. a Hobbs, W. H., and Leith, C. K.. The pre-Cambrian volcanic rocks of the Fox Eiver valley, Wisconsin: Bull. Univ. Wisconsin No. 158 (Sci. ser., vol. 3, No. G), 1907. pp- 247-27S. b Weidman, Samuel, The Baraboo iron-bearing district of Wisconsin: Bull. Wisconsin Geol. and Nat. Hist. Survey No. 13, 1904, p. 21. CHAPTER XV. THE KEWEENAW AN SERIES." GENERAL CHARACTERISTICS. The Keweenawan is tlie upper series of the Aljjonkian sj^'stem in the Lake Superior region. Its most cliaraeteristic feature is that its abunchint effusive rocks are as widespread as the series itself. Indeed, they probal)Iy compose from a third to a lialf of the series. The Keweenawan contrasts with the Huronlan in that in tlie latter scries tiie efTusive rocks are largely concen- trated m a number of localities, although in these areas they may he of veiy great thickness. In short, the Keweenawan was a period of regional volcanic activity and the Huronian was a period of local volcanism. It results from these facts that in the earliest studies of the Kewee- nawan the igneous rocks were noted and described. In the Huronian, on the other hand, the sediments were more conspicuous and were especially studied in the early years, and it is only recently that the extent and magnitude of the igneous rocks of that period have been appreciatctl. In the following discussion of tlie Keweenawan no attempt will ho made to give detailed petrographic descriptions. The most salient petrographic features will be mentioned, and a review of the petrography and chemistry, with reference to nomenclature, \\'ill be presented by A. N. Wmchell. In order to give a somewliat more definite impression of the series, the more important districts will be briefly described. DISTRIBUTION. The Keweenawan rocks border the major part of the shore of the western half of Lake Superior, occupy islands in the eastern half, and are found on the mainland at the extreme east end of tlie lake. They extend to a maximum distance of 120 miles northwest of Lake Superior. To the southwest Keweenawan rocks have been penetrated by drills at Stillwater, and still farther southwest, at St. Paul and vicinity, certain red sandstones have been drilled which may be Keweenawan. On the south side of the lake they occur mainly witliin 12 miles of the shore. Sandstones, Keweenawan or Cambrian, are known also at the east end of the Felch Mountain trough. This distribution shows that tliis series once occupied the greater portion of the Lake Superior basin and from it extended for varying distances. In much of the basin at present the Keweenawan rocks are overlain by Cambrian sandstone. The total present exposed area of the Keweenawan rocks is approximateh" 15,000 square miles. To obtain tlie original, area there must be added a very large but unknown portion of the Lake Superior basin. Further, there must be added the numerous masses, large and small, of the rocks of Keweenawan age intrusive into the Huronian and Archean of the Lake Superior region. Irving '' estimated the area of the Keweenawan, aside from the rocks intrusive in older series, at 41,000 square miles. It is thus evident that Lake Superior in Keweenawan time was an aiea of regional activity extending east and west for more than 400 miles and north and south for scarcely a less distance. SUCCESSION. A broad study of the several Keweenawan districts leads to the conclusion that a threefold division of the seiies as a whole may be made, beginning at tlie bottom, as ft)lIows: (1) Lower Keweenawan, comprising conglomerates, sandstones, dolomitic sandstones, shales, and marls; a For further detailed description of the Keweenawan rocks of the Lake Superior region see Mon. C S. Geol. Survey, vol. 5, and references there ^iven. In the descriptions of the. several districts accounts of local features of the Keweenawan are given. ^ Irving, K. D., The copper-bearing rocks of Lake Superior: Mon. U. S. Geol. Survey, voL 5, 1SS3, p. 27. 366 THE KEWEENAWAN SERIES. 367 (2) middle Keweenawan, comprising extrusive and intrusive igneous rocks with Important amounts of interstratified sandstones and conglomerates and subordinate amounts of shale; and (3) upper Keweenawan, comprising conglomerates, sandstones, and shales, represented in northern Wisconsin and Michigan. In only one district, nortliern Wisconsin and Michigan, is the full succession found. In the area of Black and Nipigon bays and Lake Nipigon, in Minnesota, and at the east end of Lake Superior the lower and middle Keweenawan appear. On Isle Koyal the upper and middle Keweenawan occur, and on Michipicoten Island only the middle Keweenawan is found. BLACK AND NIPIGON BAYS AND LAKE NIPIGON. LOWER KEWEENAWAN. The rocks belonging to the lower Keweenawan occupy the peninsida between Thunder and Black bays and the neck between Nipigon and Black bays from the northwest corner of Nipigon Bay to a point 20 miles west of Black Sturgeon River. They consist of quartzose sandstones, dolomitic sandstones, and red marls. According to Logan ** their thickness is from 800 to 900 feet. Bell, however, estimated it from 1,.300 to 1 ,400 feet. Bell's section' is as follows : Section of lower Keweenawan rocks near Black and Nipigon bays. Feet. Alternating red and white dolomitic sandstone, with a red conglomerate layer at the bottom, occurring on Wood's location, Thunder Cape"^ 40 Light-gray dolomitic sandstone, with occasional red layers and spots and patches of the same color. These sandstones occur along the southwest side of Thunder Bay and on Wood's loca- tion d 200 Red sandstones and shales, interstratified with white or light-gray sandstone beds, frequently exhibiting ripple-marked surfaces, and also with conglomerate layers composed of pebbles and ' * bowlders of coarse red jasper in a matrix of white, red, or greenish sand 500 Compact light-reddish limestones (some of them fit for burning into quicklime), interstratified with shales and sandstones of the same color , 80 Indvu'ated red and yellowish-gray marl, usually containing a large proportion of the carbonates of lime and magnesia. « This di\dsion runs through the center of the peninsula between Thunder Bay and Black Bay, and may, in this region, have a thickness of 350 feet or more 350 Red and white sandstones, with conglomerate layers, the red sandstones being often very argil- laceous and variegated with green spots and streaks, and having many of their surfaces ripple- marked. These rocks are found all along the northwest side of Black Bay as far up as the township of McTa\-ish 200 There are no lavas interstratified with the Black Bay and Nipigon Bay rocks, but at numerous places they are cut by diabase dikes similar to those which cut the upper Huronian (Animikie group) . The lower Keweenawan occurs on the shore of the southwestern part of Lake Nipigon in relatively small areas and irdand from Lake Nipigon in a large area of which Black Sturgeon Lake is the center. This division is called the Nipigon formation by Wilson./^ It comprises basal conglomerates which rest unconformably upon the Archean, sandstones, shales, and dolomites — green, ferruginous, and white. For this area Wilson gives the succession, in descending order, as follows: Section of lower Keweenaioan rods in Nipigon basin. Feet. Dolomites and dolomitic shales 400 Grits and sandstones 150 Basal conglomerate 4-6 o Logan, W. E., Report of progress to lS(i3, Geol. Survey Canada, 18fj3, p. 70. t Bell, Robert. Report of progress from 1800 to 18C9, Geol. Survey Canada, 1870, p. 319. c Macfarlane finds the red sandstone to contain 12.5 per cent of carbonate of lime and 11 percent of carbonate of magnesia. d Macfarlane found them to contain 13 percent ofcarl)onate of lime and 12 percent of carbonate of magnesia. c The amount varying, in the specimens analyzed by Macfarlane, from 21 to 34.5 per cent of the carbonate of lime, and from 7.5 to 13.5 per cent of the carbonate of magnesia. /Wilson, A. W. G., Geology of the Nipigon basin, Ontario: Canada Dcpt. Mines, Geol. Survey Branch, Memoir No. 1, 1910, pp. 69-70. 368 GEOLOGY OF THE LAKE SUPERIOR REGION. Nothing is said by Wilson as to the dips of the lower Keweenawan rocks, but it is apparent from liis descriptions that they are relatively flat. The district al)ovo described is of interest as being the only district in wliich the accu- mulation of detrital material before the outbreak of the Keweenawan lavas covers any considerable area. It is believed that these rocks really represent the first deposits of the transcressin"' Keweenawan sea and antedate the igneous epoch of the Keweenawan altogether. The absence of material derived from the Keweenawan lavas led some of the earlier geologists — for instance, Macfarlane" and Hunt'' — to question whether these rocks really belong with 'the Keweenawan. These lower Keweenawan rocks pass under the middle Keweenawan diabases and amygdaloids, which form the southern half of the peninsula southwest of Black Bay. On the north they are overlain, according to Bell,'^ by columnar trap. MIDDLE KEWEENAWAN. Aside from the area occupied by the lower Keweenawan sediments the remainder of the Black and Nipigon bays and Lake Nipigon district is occupied by the middle Keweenawan, consistmg of basic igneous rocks with subordinate amounts of interstratified clastic material. These igneous rocks are partly flows and partty intrusions. BLACK AND NIPIGON BAYS AND ADJACENT ISLANDS. Black and Nipigon bays are noted for their conspicuous and interesting topography, wliicli has originated in essentially the same way as the topography of Thunder Bay. In both locahties the sediments are interleaved with great sills of diabase, sedimentary and igneous rocks ahke being in nearly horizontal attitude. The rocks of the middle Keweenawan constitute the shores of the outer parts of Black and Nipigon bays and of the adjacent islands, including those from the size of St. Ignace to small rocks, and from the shore they extend considerable but varying distances inland. Over large areas these rocks present f acies which are similar to those of the Beaver Bay area of the Mimiesota coast, described on pages 371-374. Locally they show spheroidal weathering, as at Fluor Island. They are cut by red rock, which metamorphoses the diabase to an orthoclase gabbro, just as on the Minnesota coast. The sediments are subordinate. In places the diabase clearly intrudes the sediments and locally the latter are somewhat modified at the contact, the color changmg toward the intrusive rock from red to gray or white. For the most part the dip of the rocks of the areas of Black and Nipigon bays is very gentle, here m one direction and there in another, but near the shore of Lake Superior there is the usual gentle and persistent lakeward slant of 8° to 10°. Locally, however, the dips go up to 20° or 30°, to 60° or 70°, or even to the vertical. These steep dips occur at places where the diabases intrude the sediments or the amygdaloids, and thus disturb their normal attitudes. LAKE NIPIGON. The middle Keweenawan igneous rocks extend tliroughout the Lake Nipigon district, except in the areas of the lower Keweenawan already mentioned. They occupy about half of the shore line on the east and north sides of Lake Nipigon, where they mainly constitute the pen- insulas and headlands. North of the lake they extend 40 miles or more to the Hudson Bay divide. They occupy all the hundreds of islands of the lake, varying in size from those which are several miles long and wide to those which are mere rocks. The midtlle Keweenawan of Lake Nipigon consists mainly of great masses of diabase, which Wilson says are in sheets and dikes, and with these are later acidic dikes. English Bay is an area of granite porphyry, which Wilson places with the Archean, but which, it may be suggested from the association, may. belong with the Keweenawan. Macfarlane, Thomas, Canadian Naturalist, new ser., vol. 3, 180S, p. 2S2; vol. 4, 1809, p. 38. b Hunt, T. S., Special report on the trap dikes anil Azaie rocks of southern Pennsylvania, pt. 1: Kept. E, Second Geol. Survey rennsyl\-ania, 1878, p. 241. c Bell, Robert, Report of progress from 180C to 1809, Geol. Survey Canada, 1870, p. 338. THE KEWEENAW AN SERIES. 369 There has been much discussion as to whether the great diabase sheets are intrusive or extrusive rocks. Wilson" summarizes the evidence in favor of extrusiqn as follows: 1. The very widespread occurrence of unconformities between diabase sheets and underlying formations. 2. The occurrence of bowlders of granite and gneiss and schist in diabase, the latter resting on similar rocks in situ in localities where there is direct evidence that before the advent of the trap the underlying rocks were buried beneath the sediments similar to those now present, near by, under the same diabase sheet. 3. The occurrence of old soils in situ at the bases and on the sides of sedimentary ridges, the whole being covered in places with a diabase cap. 4. The nicety of the adjustment by which the diabase sheets have fitted themselves to the underlying topography. MTiile the upper surfaces of the residuals of the capping sheets are everywhere fairly uniform in height, the base of the sheet has adju.sted itself to a topography where the relief was at times as much as 300 feet. 5. The mechanical problem which arises in explaining the numerous unconformities, especially those on the embossed Archean surface, by the theory of intrusion vanishes completely on the theory of surface erosion prior to surface extrusion. 6. The features characteristic of the upper surface of sills — the occurrence of overlying beds or fragments thereof, aphanitic structures, included fragments of an old cover in the upper parts of sheets — are not found. 7. The medium to coarse texture, which characterizes the sheets, would be found at the base of thick surface flows as well as in sills, being dependent not on the nature and thickness of the cover so much as on the rate of cooling. 8. A glassy matrix, amygdaloidal or porous structure, basaltic texture, flow structure, and associated volcanics would not be characteristic features of the under parts of surface flows, and the ujiper parts of these sheets are unques- tionably removed, without a single exception. In favor of intrusive sills are: 1. Entire absence of any of those features that are usually associated with the upper parts of a surface flow — glassy matrix; amygdaloidal, porous, or basaltic texture; flow structure; associated volcanic rocks, either lava breccias or pyroclastic rocks. 2. A medium to coarse crystalline texture, usually indicative of a slow rate of cooling, such as would normally take place only at some considerable distance below the surface. From the evidence presented Wilson draws the following conclusions : * It seems that we have no data relative to the actual character of the upper surface of the trap "caps;" such negative evidence as is available is equally applicable to both theories. With regard to the texture of the residual basal por- tions of the sheets there are no recorded differences which would indicate that it belonged to a flow and not to a sheet. On the other hand, numerous unconformities exist, and the diabases are known to rest successively upon Laurentian, Keewatin, Huronian (possibly middle, certainly lower, and Animikie), and Keweenawan (lower, middle, and upper beds), and these unconformities are very widely distributed. Owing to the mechanical difficulties involved by any other interpretation it seems to the writer that the balance of evidence available is distinctly in favor of considering these capping sheets as the basal residuals of a once very extensive flow or series of flows of a very fluid diabase over the well-dissected topography of a previous cycle. It may be suggested in this case, as in so many others, that the diabases of the Keweenawan sheets are not exclusively intrusive or extrusive. It has heretofore been the prevailmg view that the cappmg diabases, so characteristic of the step topography of the Animikie area and of the Keweenawan area on the northwest side of Lake Superior, are sills down to which erosion has worked. Wilson has held that some are not sills but are flows upon an old erosion surface. His conclusion Ihat the flows are as late as Cretaceous rests on very slender evidence — that is, on the identification of the plane on which the flows rest as of post-Cretaceous age. He presents no evidence to show that the flows are not Keweenawan or some of them even Animilde. The view that they are Keweenawan is favored by their petrologic, areal, and structural relations with known Keweenawan rocks of the northwest and south sides of the Lake Superior basm. RELATIONS OF THE KEWEENAWAN OF BLACK AND NIPIGON BAYS TO OTHER ROCKS. As the sediments of Black and Nipigon bays are at the bottom of the Keweenawan series their relations to the underl_yTJig rocks are important. At the very base of the series occur conglomerates the debris of wMch is derived from the underlying Huronian series, including the a Wilson, A. W. G., Geology of the Nipigon basin, Ontario: Canada Dept. Mines, Geol. Survey Branch, Memoir No. 1, 1910, pp. 94-95. ' Idem, pp. 95-96. 47517°— VOL 52—11 24 370 GEOLOGY OF THE LAKE SUPERIOR REGION. Animikie group, showing that there is an unconformity between the normal sediments making uf) the earhpst Kcweenavvan and the latest Iluronian. One of the best exposures of this uncon- formity is at a cliff adjacent to Surprise Lake, a short distance from Silver Islet village. Here in actual contact with the slates of the Animikie group is a conglomerate about G feet in thickness, which is largely composed of angular fragments of slates from the Animikie with, however, detritus from granites, mica schists, vein quartz, etc., but no fragments of any of the Keweena- wan lavas. The contact between the conglomerate and slate is knifelike in shar[)ness. Locally the matrix of the conglomerate is limestone. The conglomerate grades upward into wliite qiiartzite interstratified with slaty layers, over wliich are bands of red and white dolomite. Here, as is common between the Keweenawan and Animikie, the discordance is shown mainly by the conglomerate and not by an important difference in dip, but in a number of places the conglomerate cuts across the slate bands in a minor way. Other very satisfactory contacts between the Keweenawan and Animikie are those in a cut of the Canadian Pacific Railway about a mile west of Loon Lake and at the south shore of Deception Lake. Here the conglomerate of the Keweenawan resting upon the Animikie con- tains bowlders as much as 2 feet in diameter. At the railway cut the phenomena are very similar to those at Surprise Lake, but at Deception Lake the Animikie rocks have been some- what sharply folded, and the conglomerate rests horizontally upon the truncated beds of the Animikie. The debris of the Keweenawan conglomerate at these localities includes the slates from the underlying Animikie, material from the iron-bearing formation of the Animikie, and granites and scliists from the lower Huronian or Archean. At all these localities the completely indu- rated pebbles of the Animikie as compared with the much less cemented Keweenawan are notable. Tliis, combined \vith actual discordance, would indicate an important time break between the two series, an inference wMch is confirmed by the relations of the two in the Penokee-Gogebic district. According to Wilson, in the Nipigon basin diabases rest unconformably on the Keweena- wan, Animikie, and Archean rocks. NORTHERN MINNESOTA. THE KEWEENAWAN AREA. The Keweenawan rocks of northern Minnesota he in a great crescent-shaped area, opening lakeward, extending from Fond du Lac, on St. Louis River, at the southwest to Grand Portage Bay at the northeast. Both the lower and the middle Keweenawan are represented. This area of Keweenawan rocks is undoubtedly the largest continuous area of the series. It covers approximately 4,500 square miles. "^ As yet tliis great region has been too insufficiently studied to jiermit a satisfactory account of it, and many points remain doubtful. Granites and diabases intrusive into the Animikie of the Cuyuna and St. Louis River areas are probably of Keweenawan age. LOWER KEWEENAWAN. The lower Keweenawan is represented by the Puckwunge conglomerate. AccorcUng to Winchell,'' tliis conglomerate is seen in various locahties at tlic top of the .Vnimikie group from Grand Portage Island, in Grand Portage Bay, as far west as the middle of R. 3 E., a distance of about 20 miles. He states that the basal rock of the Keweenawan is a conglomerate which grades up into sandstone. The thiclcness of the conglomerate is not determined, but tliis forma- tion is just what one would expect between the Animikie group and the Keweenawan series from the character of the lower division of the Keweenawan about Black and Nipigon bays. Winchell "^ also states that a (|uartzite conglomerate which he regards as Puckwunge occurs in a Elftman, A. H., The geology of the Keweenawan area in northeastern Minnesota: Am. Geologist, vol. 21, 1898, p. 175. 6 Winchell, N. H., The geology of Minnesota, vol. 4, 1899, pp. 307, 327, 517-519; vol. 5, 1900, pp. 50-52. c Idem, vol. 4, p. 13. THE KEWEENAWAN SERIES. 371 sec. 1, T. 48 N., R. 16 W., on St. Louis River, and that its total tliickness is nearly 100 feet. There are, however, rare pebhles of Keweenawan rocks in this formation. It is conformable below the younger beds. The pebbles of this conglomerate are largely derived from the quartz veins of the slates of the underlying Animikie, and the conglomerate therefore lies unconform- ably on the Animikie. The formation grades into a white sandstone and then into a shale. Thus the sechmentary formation is seen at the base of the Minnesota Keweenawan at both the northeast and the southwest ends. In the intervening stretch of more than 100 miles the exact base of the sedimentary or volcanic Keweenawan has not been traced because of lack of expo- sures and because of the intrusion of the great Duluth gabbro to be mentioned later. MIDDLE KEWEENAWAN. The middle Keweenawan rocks comprise all of the Keweenawan in Minnesota except the relatively insignificant Puckwunge conglomerate. They represent the volcanic epoch of the Keweenawan. Broadly the middle Keweenawan of northeastern Minnesota may be divided into two great divisions — (1) the effusive rocks and the associated sediments and (2) the intru- sive rocks. EFFUSIVE BOCKS. The effusive rocks occupy the larger part of the Minnesota coast and extend for varying distances inland. The Minnesota coast line, looked at as a whole, presents a flat crescentic shape, with the concavity toward the lake. The same is true of the courses of the effu- sive rocks, but the crescents formed by them have a smaller radius and hence intersect that formed by the coast line, trend- ing more to the north at the Duluth end and more to the east at the Grand Portage end. In following the coast, then, from Duluth to Grand Portage, we ascend in geologic horizon to a point near Two Islands River and descend from a point just east of Temperance River to Grand Portage. These rocks consist dominantly of a well-stratified series of volcanic flows having a gentle lakeward dip, winch commonly is from 8° to 10° but locally is as low as 5° or 6° and as high as 25° or 30°, or rarely even 45° or 60°. Numerous minor bowings and corrugations may be seen in the incUvidual layers and sets of layers, which may be followed for some miles. These may be seen rising into arches, locally of short span, and sinking into synclines to reappear as anticlines a short distance away. The lavas are diabases which are commonly amygdaloidal. Many of these amygdaloids are very scoriaceous. These rocks are softer than the intrusive rocks and are especially Ukely to constitute the" bays. There are subordinate masses of intermediate rocks, wliich usually have not been separated on the maps from the basic flows. At one place, east of Kadonces Bay, tins intermediate rock has a peculiar spheroidal weathering similar to that of the Ely green- stone, a structure which has been regarded as evidence of subaqueous extrusion. Associated with the basic lavas are masses of acidic lavas represented by quartz jaor- phyrites and felsites. One of the more notable locahties for these rocks is the great Palisades (fig. 56). The conglomerates and sandstones interstratified with the lavas are subordinate in amount. In the lower part of the series they are either absent altogether or are represented by very thin beds. In the upper part of the series, especially the portion to which Irving "■ has given the Water line Figure 56. — Section on south cliff of Great Palisades, Minnesota coast. (After Irving.) a, Amygdaloid; 6, columnar diabase-porphyrite; c, mingled amygdaloid and detrital matter; d, quartz porphyry. " Irving, B. D., The copper-bearing rocks of Lake Superior: Mon. U. S. Geol. Survey, vol. 5, 1883, pp. 323-329. 372 GEOLOGY OF THE LAKE SUPERIOR REGION. name Temperance River group, the sandstone and conglomerate beds are numerous. Most of these beds are only a few inches to a few feet in thickness, but there are some beds which arc 100 feet tliick, and according to Elftman" one wliich is 250 feet thick. Lawson** estimates that the sandstones and conglomerates oceupy less than 0.5 per cent of the coast line. INTKUSIVE BOCKS. The intrusive rocks comprise both basic and acidic types. BASIC KOCKS. The basic rocks include the Duluth laccolith, the Beaver Ba)^ and similar laccolitlis and sills, the anorthosites, and the dike rocks. DULUTH LACCOLITH. Area and character. — The Dulutli laccolith is a gabbro. It extends from St. Louis River to the northeast, grtuiually widening until in the center of the belt it is 30 miles wide. From this maximum breadth it narrows toward the east until it makes a point at the Minnesota coast. It is not our purpose here to give anything more than a most general petrographic account of the Didutli gabbro. It is, for the most part, normal gabljro, but it has many facies. Min- eralogically it ranges from a very magnetitic gabbro through olivine gabbro hi wliich the feldspar is subordinate and ordinary ohvine gabbro to olivine-free gabbro, or ordinary gabbro, and finally to a rock m which feldspar is the dominant mineral, the rock beuig a labradorite or an anorthosite. The anorthosite masses vary from those a few feet across to those liundieds of feet in diameter. The anorthosite appears to be but a diH'erentiation phase of the gabbro, there being every gradation between it and both coarse and fine grained phases of the main mass of the rock. These relations are particular]}- well seen at Little Saganaga Lake, where, accordmg to Clements," the anorthosite unquestionably shows gradations into the surroundmg basic masses. Nowhere is there a sharp line of contact between the two rocks. In these respects the occurrences are in sharp contrast with the anorthosite and the diabase of the Minnesota coast, to be later described. Structurally the gabbro is ordinarily massive. However, at manj^ places, especially near its borders, it has a sheeted structure. Some of the sheets are verj' tliui and strongly resemble bedded rocks. This variety may be very well seen in the north bay of Basliitanequeb Lake. In addition to this sheeted structure there is a banded structure, due to the parallel arrangement of the mineral constituents. TexturaUy the gabbro varies from a rock of very coarse grain to one that is almost aphanitic. All varieties, coarse and fine, are granulitic. Relations to other formations. — The structural relations of the Duluth gabbro are veiy interesting. On the north, in passing from St. Louis River to Grand Portage, the gabbro is in contact for a long way with the upper Huronian, then for many miles with the several members of the lower Huronian and the Archean, and finally for many miles again with the upper Huronian. It thus cuts diagonally across the upper Huronian in its northern and southern parts and in passing toward the center of the area goes through the lower Huronian and deep into the Archean. Evidence of its intrusive character is afforded bj' its coarse crystallization; by the presence of numerous subordinate bosses and dikes, offshoots of the gabbro mass, in the Huronian series; by the inclusion of isolated masses of upper Hiu'onian near its margin and the profound meta- morphic effects of the gabbro, the rocks being changed to schists or gneisses or even to com- pletely granular crystalline rocks for distances up to half a mile or a mile from the main gabbro mass, an effect not to be expected from a rapidly cooling extrusive; and finally by the higher density of the gabbro than of the hitruded rocks. a Elftman, A. H., The geolqgy of the Keweenawan area in northeastern Minnesota: Am. Geologist, vol. 21, 1898, p. 185. !> Lawson, A. C, Sketch of the coastal topography of the north side of Lake Superior: Twentieth Aon. Kept. Geol. and Nat. Hist. Survey Minnesota, 1893, p. 190. c Clements, J. M., The Vermilion Iron-bearing district of Minnesota: Mon. U. S. Geol. Survey, vol. 45, 1903, pp. 402-103. THE KEWEENAWAN SERIES. 378 The relations of the gabbro to the hivas of the coast have not been satisfactorily deter- mined. Their contact is mainly in the plateau of the interior, is very poorly exposed, and has not been sufficiently studied. However, it is believed that when these relations are worked out it will be found that the galibro is intrusive and has produced profound metamorphic effects. If this inferred intrusive relation is confirmed, the Duluth gabbro is a great laccolith, which has as a basement the Huronian and Archean and as a roof the Keweenawan lava flows. The relations of the Duluth gabbro to the Puckwunge conglomerate at the base of the Keweenawan and to the earlier Keweenawan lavas have not been established. Until this is done it is impos- sible to gain any definite conception as to how far Keweenawan time had advanced before the appearance of the gabbro. If the Duluth gabbro is interpreted as a laccolith it surpasses in magnitude any other yet described. With a maximum diameter of 100 miles, if its thickness has approximately the ratio shown in the typical laccoliths of the Henry Mountains," the thick- ness would be 75,000 feet. If an average dip of 10° for 50 miles on the north shore is assumed the thickness would figure 45,000 feet. The intrusion of so vast a mass of material must have required a long time. The parts earlier intruded were doubtless solidified long before magma ceased to enter. Thus, offshoots of these later parts would be found as dikes in the earlier solidified parts. There would be great variation in its coarseness of crystallization. Ample time would be afforded for differentiation by fractional crystallization, separation by gravity, and other processes, and thus is explained the structural complexit}^ of the gabbro and its great variation in mmeral and chemica character. THE BEAVER BAY AND OTHER LACCOLITHS AND SILLS. Intruded in the lavas of the Minnesota coast are a great many laccoliths or sills of diabase. These intrusive rocks are especially ])revalent in the lower pait of the lavas, and particularly in the part below the Temperance River group. In textiue these rocks vary from diabases to gabbros and include the so-called black gabbros of Irving.* The diabases in many places show a remarkable luster mottling due to the inclusion of numerous individuals of plagioclase in large individuals of augite. Not uncommonly the augites are several inches in diameter and include hundreds of lath-shaped feldspars. Many of these laccoliths and sills were supposed by the earlier geologists' to be lava flows, but when exammed closely they are found to cut the lava beds by passing gradually across their edges and by sending out dike offshoots. In not a few places they show a distmct columnar structure at right angles to their borders. The local steep dips of the lava beds mentioned in the previous section are apparently all due to the influence of the intrusive masses and thus their exceptional character is explained. A typical illustration of these laccoliths is seen at Beaver Bay. The center of this laccolith extends from a point near Beaver Bay to a point near Two Harbors Bay. In this distance it occupies the entire coast. Neither its top nor its bottom is seen. In this part it is not luster mottled but is the coarse black gabbro of Irving.'' Its central part is sheeted and in general has a coarse or imperfect columnar structure at right angles to the horizon or nearly so. Wliere it is foiuid in association with the lavas farther east and west, as at Split Rock and Beaver Bay, its structure corresponds with the bedding of the amygdaloids, so that it was natural for Irving to regard it as a bedded flow, although even he recognized that at some places it cut the amygda- loids in a curious way. Indeed, locally it cuts the amygdaloids in a most intricate fashion, following the joints, wmding around the blocks, intruding itself as films between the plating of the amygdaloid, but always with sharp contacts. It is a significant fact that near the lavas the laccolith is luster mottled. Very close to the amygdaloid it is locally fine grained. In places it retains its coarse texture, even in narrow strmgers. The laccoliths and sills, being resistant rocks, usually make the major headlands of the coast, just as the lavas usually constitute the bays. a Gilbert, G. K., The geology of the Henry Mountains, 2d ed.: U. S. Geog. and Geol. Survey RockT,- Mtn. Region, 1880, p. 55. 4 Irving, R. D., The copper-bearing rocks of Lake Superior: Mon. U. S. Geol. Survey, vol. 6, 1883, pp. 267-268. 374 GEOLOGY OF THE LAKE SUPERIOR REGION. The Logan sills jiiul capping rocks in the Animikie and Keweenawan of northeastern JVlinnosota should be ])articularly mentioned. These aic doubtless to be correlated with the great gabbro mass. In fact, in the Gunllint Lake district tliey seem to l)e directly connected. To these sills are due the step topography of tiiis region. Wilson " has concluded that farther to the east in the Nipigon basin some of tlie capping steps instead of ])eing sills are really flows. It is possi})le that tliis conclusion may be applied to j)art of the caj)j)ing rocks in north- eastern Minnesota. . ANORTHOSITES. The anorthosites of the Minnesota coast early attracted attention because of their brilliant light color. They may be well seen at Split Rock, BeaVer Bay, and Carlton Peak. At these places a large portion of them are inclusions in the basic laccolitlis and sills, such as the BeaA'er Bay laccolith. Indeed, at many jilaces they form a stucco in this diabase laccolith. In size the inclusions range from those which are minute, being no more than indiviihial ciystals of fclds])ars, to great masses .50 or 60 feet in diameter. In adtlition to these masses, wliicli are plamlj' inclusions, there are other masses which are so large that they can not be a.sserted to be inclusions. These are mantled by the Beaver Bay laccolitli, as described by Lawson,' the relations at the bottom, however, not being exposed. Some of these masses on the Minnesota coast are as large as a cathedral, and the largest masses are found at Carlton Peak, the different points of which are composed entirely of anorthosite. The anorthosite inclusions are not con- tained in tlie central part of the Beaver Bay laccolith, but in its upper part, where it is in contact with or near the anwgdaloids. The relations above described conclusively show that the anorthosite, as a rock, antedated the including rocks. Lawson '^ has interpreted this to mean that the anorthosite marked a pre-Keweenawan terrane, but from our point of view the anorthosite is but a facies of the great Duluth gabbro mass which had been segregated before the diabase intrusions (seep. 372), and therefore has been included in the diabase, as above described. It is conjectured that the very abvmdant diabase laccoliths and sills at Beaver Bay and other localities are but later offshoots of the original reservoir of magma from which the Diduth gal)bro was also derived. The alliance between the diabase intrusives of the coast and the Duluth gabbro is shown by their chemical and mmeralogical likeness. BASIC DIKES. Diabase dikes cut the lavas and sills at numerous places. As a rule they are nearly vertical. Many of them lie approximately at right angles to the coast, and are likely to make projections into the water. Others run approximately parallel to the coast. These dikes conform to the sets of strike and dip fractures which were produced by the deformation. Commonly these diabase dikes are less than 50 or 60 feet across. At some places they have a columnar stracture at right angles to the walls, parallel to the bedding of the lavas, and consequently at right angles also to the coluimiar structure of the laccoliths. ACIDIC KOCKS. Along the northwest shore of Lake Superior and back from the coast are many areas of acidic rocks, collectively mapped as red rock, because of their jirevaUing red color.'' The red rock consists of intrusives, mainly granite and augite sA'enite, and their equivalent elTusives, quartz porphyry. These are later than the associated basic extrusive and intrusive rocks, succeeding the Duluth gabbro and the diabase of the Beaver Bay laccolith. The red rocks range in size fi-om considerable masses to minute stringers. In many places the intrusives intricately cut tiie basic rocks. This is well illustrated at Beaver Bay, where both the amygda- loidal lavas and the diabase are intruded. Dikes of the red rock, great and small, cut the diabase a Wilson, A. W. G., Geology of the Nipigon basin: Canada Dept. Mines, Geol. Survey Branch. Memoir No. 1. 1910, pp. 95-%. i> I.awson, .\. C, The anorthosites of the Minnesota coast of Lake Superior: Bull. Geol. and Nat. Hist. Survey Minnesota No. 8, 1S93. p. 1& (■Idem, p. 19. d Klftmau, \. 11., The geology of the Keweena*an area in northeastern Minnesota: Am. Geologist, vol. 22, 1898, pi. 7. THE KEWEENAWAN SERIES. 375 through and through, and have produced an important exomorphic effect. Where thus altered the diabase grades into a rock of a somewliat more acidic aspect and becomes the ortlioclase gabbro of Irving. ° Wherever we have seen tliis rock it is but a facies of the diabase, produced through the minute penetration of the acidic magma of the red rock. It is clear that the chemical com])osition of the diabase has been affected by minute penetration of the acidic magma and its emanations. KEWEENAWAN ROCKS IN THE CUYUNA DISTRICT OF NORTH-CENTRAL MINNESOTA. Granite, diabase, and gabbro cut the slates of the Animikie in the great north-central area of Minnesota, including the Carlton, Cloquet, Cuyima, and Little Falls areas. Being later than the Animikie, they are probably to be correlated with the Keweenawan intiiisive rocks of north- eastern Mimiesota. They are probably to be regarded as the plutonic equivalents of the Kewee- nawan flows. In the Cuyima district there is also a thin layer of amygdaloidal acidic rock, 15 feet thick, resting upon the eroded edges of the slates and iron-bearing formation of the Animikie group. Drilling in this district discloses many masses of basic and acidic rock intricately asso- ciated with the slates of the Animikie, but the relations are not yet determined. THICKNESS OF THE KEWEENAWAN OF MINNESOTA. Irving,* in his monograph on the copper-bearing rocks of Lake Superior, makes a formal division of the Keweenawan of the Minnesota coast into six groups, for which he estimates thicknesses as follows, from the top down: Feet. Temperance River group 2, 500-3, 000 Beaver Bay group 4, 000-6, 000 Agate Bay group 1, 500 Lester River group 2, 600 Duluth group 5, 000 St. Louis gabbro [now called Duluth gabbro] Thickness uncertain. Excluding the gabbro, Irving'^ estimates the total thickness to be between 17,000 and IS, 000 feet. It is to be remembered that these estimates of thickness include large masses of intrusive rocks, as, for instance, the Duluth gabbro and the diabase of Beaver Bay. Also it is far from certam that the lavas on the Minnesota coast have the regidarity of superposition supposed bj' Irving. Finally, it is imcertain what part of the present dip of the lavas is initial. Elftman, the one other geologist who has made an extensive study of the Keweenawan of the Minnesota coast, gives the following order:'' 1. Later diabase member. 2. Temperance River member. 3. Red Rock member. 4. Beaver Bay diabase member. 5. Gabbro member. This is the structural order. It is clear that the order is only partly one' of age, for before the gabbro and other laccoliths and sills could be intruded in the Keweenawan a certain amount of sediments and lavas must have been buUt up. This succession, as well as that of Irving,* ignores the Puckwimge conglomerate. Elftman supposed that between the "Temperance River, member" and the "Red Rock" member there is a considerable unconformity, because at the bottom of the "Temperance River member" is a conglomerate 100 feet thick. This conglomerate contains fragments of diabase similar to the diabase of Beaver Bay, and also many fragments of red rock, indicating a Irving, R. D., The copper-bearing rocks of Lake Superior: Men. U. S. Geol. Survey, vol. 5, 1883, pp. 50 et seq. (■ Idem, pp. 200-268. <■ Idem, p. 260. ut ion to the petrography of the Keweenawan: Jour. Geology, vol. IS, 1910, pp. 63S-657. THE KEWEENAWAN SERIES. 377 white quartz, some of them being 8 or 10 inches in diameter. FUnt and black hornstone pebbles are also plentiful. This conglomerate gradesup into a coarse quartzite, and this mto a fine-grained compact quartzite. Immediately to the north of the latter formation are the basic flows of the middle Keweenawan, and 400 or 500 feet south of the conglomerate are upper Huronian mica- ceous gravAvackes. The thickness of the conglomerate and quartzite exposed is probably from 300 to 400 feet. The quartzites adjacent to the Keweenawan in Barron. County, Wis., may be in part Kewee- nawan. There are here at least two series of pre-Cambrian quartzites, the upper of which is reddish, feldspathic, and not strongly consohdated, and has comparatively low dips. These facts, together with the position of the quartzites on the southeast side of the Keweenawan syncline, have suggested to Weidman'^ the possibility that they represent lower Keweenawan sediments, but this has not been proved. MIDDLE KEWEENAWAN. The general characters of the middle Keweenawan in this region are substantially the same as those of northeastern Muuiesota. The igneous rocks comprise both plutonic and volcanic masses. The volcanic series covers a much greater area than the plutonic rocks. At the sec- tions which have been studied, Potato River, Tylers Fork, and Bad River, the igneous rocks, accorduig to Irving,' consist dominantly of beds of diabases, diabase amygdaloids, and mela- phyres. With the basic igneous rocks are subordinate masses of felsite and quartz porphyry. Interstratified with the lavas are subordinate beds of conglomerate and sandstone. Along the north side of the Keweenawan of Wisconsiji, in Douglas County, the lower part of the series is coniposed wholly of igneous rocks, but at higher horizons in the southeastern part of the district conglomerates are interstratified with lava flows. On the whole the interbedded detrital rocks of tliis area are apparently less abundant than on Keweenaw Pomt but more abimtUxnt than in Minnesota. The hthology of the interstratified conglomerates and sandstones is in no respect pecuhar. So far as we know, there has been no approximately accurate determination of the entire thickness of the lava flows and interstratified sediments of the middle Keweenawan in Wisconsin. Berkey has estimated the thickness of the Keweenawan emptive rocks exposed along the St. Croix Dalles as 4,000 feet. Hall"^ estimates a thickness of 20,000 feet on Snake and Kettle rivers in Minnesota. On the south side of the synclme at the base of the Keweenawan m Wisconsin is a great basal gabbro, which in every respect is equivalent to the Duluth gabbro described on pages 372-373. Tliis gabbro has been traced from Black River in Micliigan as far west as R. 7 W., but how much farther it extends is unknown. Thus it has an extent northeast and southwest of 60 miles or more. For most of the distance the belt is from, 2 to 5 miles broad. The I'ocks of the mider- lying upper Huronian along most of tliis gabbro belt dip about 75° N. If the thickness of the gabbro mass were calculated at right angles to the dip of the underl;y'ing Huronian rocks, this would give a thickness of 9,500 to 25,000 feet. It has been explamed in connection with the Penokee-Gogebic district that this gabbro cuts diagonally across all the formations of the Huronian series and down into the Archean; also that adjacent to the contact the upper Huronian rocks are profomidly metamorphosed, the Tyler slate into mica slates and mica schists, the iron-bearmg Ironwood formation into actinolite-magnetite schists, the Bad River hmestone into a coarsely crystalline tremolitic lime- stone. Further, witliin the Huronian and the Archean are smaller masses of intrusive gabbro which doubtless are offshoots or necks of the main mass. Thus in every respect the relations of tliis basal gabbro to the underlying rocks are the same as in northern Minnesota. a Personal communication. b Irving, R. D., Tlie copper-bearing rocks of Lalce Superior: Mon. U. S. Geol. Survey, vol. 5, 1883, pp. 230-231. cBerkey, C. P., Geology of the St. Croi-x Dalles: Am. Geologist, vol. 20, 1897, p. 382. d Hall, C. W., Keweenawan area of eastern Miimesota: Bull. Geol. Soc. America, vol. 12, 1901, p. 331, 378 GEOLOGY OF THE LAKE SUPERIOR REGION. Unfortunately the relations between the gabbro mass and the lavas of the Keweenawan have not been closely studied. Irving ° represents this gabbro as feathering out into a series of points to the east, suggesting very strongly its intiusive character. However, iiis descriptions scarcely correspond to that distribution. He says:" The coarse gray gabbros so largely developed in the Bad River country of Wisconsin, at the base of the series, present the appearance of a certain sort of unconformity with the overlying beds. The.se gabbro.s, which lie immediately upon the Iluroniaii ."latos, form a belt which tapers out rapidly at both ends and seem.s to lie right in the course of the diabase belts to the east and west, since these belts, both westward toward Lake Numakagon and eastward toward the Montreal River, lie directly against the older rocks, without any of the coarse gabbros intervening. The coarseness of grain, the perfection of the crystallization, the abrupt terminations of the belts, the complete lack of structure, and the presence of intersecting areas of crystalline grani- toid rocks led Irvrng** to the beUef that these rocks were not ordinary lavas, but had solidified at a great depth. The acidic rocks cutting these coarse gabbros are clearly intrusive. The gabbro in Wisconsin, Uke the Duluth gabbro, is behoved to be a great laccolith, which was intruded in Keweenawan time after a considerable thickness of Keweenawan lava beds had been built up, and, as in Minnesota, it roughly followed the contact at the base of the Kewee- nawan and penetrated diagonally across the lower formations as well as irregularly across the Keweenawan beds themselves. It has since been turned up at angles of 75° or 80° and trun- cated by erosion. Gabijro on the north side of the Keweenawan trough in Douglas County, Wis., is described by Grant," but its extent has not been determined. It dips to the south and its relations to the lavas are similar to those of the gabbro on the south side of the Douglas County syncline. It is faulted on the north against the Cambrian rocks, which are on the downthrown side. It dips in the same direction as the Duluth gabbro, and the displacement of the fault is in such a direction as to show that it may have been originally continuous with the Duluth gabbro. UPPER KEWEENAWAN. The upper division of the Keweenawan in this area consists of red sandstones, shales, and conglomerates, divided, in the eastern part of the district, into several distinct members. Beginning at the base are found conglomerate 300 to 1,200 feet in thickness, black shales up to 400 feet, about 19,000 feet of red arkose sandstone, grading up to more siUceous sandstone, red and green shales, and coarse arkose. Above tliis is quartz sandstone, somewhat feldspathic at the base, nearly 4,000 feet thick, here called the Lake Superior sandstone. These beds appear to thin rapidly toward the west. These figures make no allowance for initial dip. RELATIONS OF THE KEWEENAWAN TO OTHER SERIES. The only places at which the relations between the Keweenawan and lower series are shown are in Wisconsin. Here, as has been seen, the lowest formation of the Keweenawan is made up of conglomerate and coarse sandstone and is overlain by the lava flows of the middle Keweenawan. The coarse conglomerate of Potato River is evidence of the erosion interval between the Keweenawan and the upper Huronian, but the magnitude of the imconformity is realized only bj^ a study of the relations of the two along the strike, which gives evidence of a large amount of erosion of the Huronian series before Keweenawan time. The details proving the greatness of this unconformity are given in the chapter on the Penokee-Gogebic district (pp. 2.34-2.35). As to the relations of the middle Keweenawan %\ith the Upper Cambrian sandstone along St. Croix, Kettle, and Copper rivers (of Minnesota), there is no difference of opinion. The Upper Cambrian sandstone, in horizontal attitude, rests upon the steeply tilted and eroded a Mon. U. S. Oeol. Survey, vol. 5, 1883, pp. 155-156. t> Idem, p. 144. c Grant, U. S , Preliminary report on the copper-bearing rocks of Douglas County, Wis.: Bull. Wisconsin Geol. and Xat. Hist. Survey Xo. 6 (2ded.), 1901, pp. 31-32. THE KEWEENAWAN SERIES. 379 edges of the middle Kewecnawan rocks and bears abundant detriius from them. It is there- fore perfectly clear that before the sandstone was laid down the middle Keweenawan had been placed at its present angles and had been profoundly eroded. The relation is very well illus- trated at Taylors Falls on St. Croix River, where the Cambrian sandstone is fossiliferous and has been certainly determined as of Upper or Middle Cambrian age. The relations between the diabases and the Cambrian here are shown by figure 57. The relation of the upper Keweenawan feldspathic sandstone and the quartz sandstones, here called the Lake Superior sandstone, has long been a subject of dispute, but the discovery by Thwaites in 1910 of outcrops on Fish Creek near Ashland has thrown new hght on the question. At this point the layers are steeply inclined to the north, exposing about 1,400 feet of strata and disclosing a transition between the red shales, arkose sandstones, and conglom- erates of the upper Keweenawan and the Lake Superior sandstone. A deep well at Ashland passes into these red shales at a depth of 2,670 feet. A reexamination of Middle River in Douglas County north of the great fault showed that the sandstone beds are inverted." About 3,100 feet of strata have been turned up by the faulting, exposing mud-cracked and ripple-marked green and red shales and arkose sandstones of the usual Keweenawan aspect, grailing above into the Lake Superior sandstone such as is found in horizontal attitude along the shore of the lake. On St. Louis River, Minnesota,^ a similar transition occurs between red shales and brown sandstones. Clinton Point, where somewhat quartzose sandstones are found, does not belong to the Lake Superior sandstone but is the crest of a minor anti- cline in the lower beds. Nearly 2,000 feet of similar rocks he some distance beneath the red shales on Fish Creek. Carnt The contact with the flat-lying quartz sandstones (Lake Superior sandstone) along the north side of the area of middle Keweenawan in Douglas County has long been known to be a fault. The best exposures are on Black, Copper, Amicon, and Middle rivers. That on Middle River has been described above. At all other points the sandstone is turned up sharply . , ,. ', ii-jii-i i-iT 1 FiGUEE 57.— Sketch showing unconformable tor a short distance to the north Ot the fault, which dips, where contact between Keweenawan diabase por- expOSed, 38° to 45° S. At all places the trap is intensely P^^^y '^'"^ Cambrian sandstone at Taylors , . , 1 , , • 1 1 rp 1 ,^ T.1 1 Falls, Minn. (After Strong.) brecciated, but the sandstone is much less attected. On Black and Amicon rivers the sandstone is conglomeratic for a few feet from the contact. The pebbles are usually small and are not matched in the neighboring igneous rocks. Within the trap breccias are found large blocks of sandstone. The view in the past has been that this contact was an unconformable one along a fault scarp, and that movement had taken place along the fault since the deposition of the sandstone, thus comphcating the simple unconformable relations. An alternative view, supported by considerable evidence, is that the conglomerate has been faulted up by parallel faults from conglomerate found at lower horizons in the sandstone, and in jiart dragged up along the fault plane. The displacement must be at least equal to the thickness of the beds turned up at Middle River — 3.100 feet. The significance of the relations of the Keweenawan to the Lake Superior sandstone is discussed on pages 415-416. KEWEENAWAN GRANITES OF FLORENCE COUNTY, NORTHEASTERN WISCONSIN. The granite along the south side of the Florence district of northeastern Wisconsin is intrusive into green schists which are interbedded with upper Huronian slates. These granites are probably part of the same mass that intrudes the Quinnesec schist of the Menominee district, where the relations are similar. These granites of northeastern Wisconsin, therefore, " Grant, U. S., Junction of Lake Superior sandstone and Keweenawan [raps in Wisconsin: Bull. Geol. Soc. America, vol. IS. 1902, pp. 6-9. ' Winchell, N. H., A rational view of the Keweenawan: Am. Geologist, vol. 16, 1895, p. 150; Geology of Miimesota, vol. 4, 1899, p. 15. 380 GEOLOGY OF THE LMvE SUPERIOR REGION. like those south of the Cuj'una district in central Minnesota, arc to be regarded as the phitonic eciuivalents of igneous flows. In both areas these plutonic masses have greatlj' metamorphosed the invaded strata. NORTHERN MICHIGAN. DISTRIBUTION. The Keweenawan rocks of northern Michigan ()ccu])y a broad belt running continuously from Montreal River, the boundary between Michigan and Wisconsin, along the lake shore to the outer extremity of Keweenaw Point and including Manitou Island and Stannard Rock. Tills belt ranges in breadth from 15 or 20 miles west of Lake Gogebic to about 6 miles at the outer part of Keweenaw Point. Approximately one-half of Keweenaw Point is occupied by rocks of the Keweenawan series. The general strike roughly follows the coast. In passing from the southwest the strikes gradually change from about N. 45° E. to east-west, and at the extreme outer part of the point the rocks swing south of east, here having a northwesterly strike. This curved outer area of the end of Keweenaw Point beyond Portage Lake corresponds almost exactly with the strike of the rocks. Except in one fold in the Porcupine Mountains the dips are always to the north or northwest. The dips of the middle and lower divisions are in general lower toward the east end of Keweenaw Point, the steepest dips ranging from nearly vertical on the Gogebic Range to 27° at the end of the point. There is a somewhat regular decrease in the dip of each of the sections in passing from lower to higher horizons. The best illustration of this is furnished by the section at Black River in Michigan, which shows a continuous succession from the base of the series to and including a part of the upper sandstone. According to Gordon," at the base of the series the dips are from 75° to 78° N., whereas the highest strata show a dip of about 20° N. The change in dip in passing from the lower to the higher members is gradual. Further illus- trations are furnished by the sections on Keweenaw Point; for instance, at the Portage Lake section the dips of the lower beds are as high as 55°, whereas in the lower part of the upper series they have dropped as low as 7°. At the outer part of Keweenaw Point the dips of the lowest part of the series there exposed are from 51° to 57°, but according to Hubbard,* the dips of the higher beds constituting the outer front of the point do not average more than 23°. In this region, as in northern Wisconsin, the lower, middle, and upper Keweenawan are all represented. The general characterization which has been made for these divisions (see pp. 376-379) applies to the northern Michigan area. The Keweenawan of Micliigan will be more specifically discussed below. KEWEENAW POINT. SUCCESSION AND CORRELATION. On account of the occurrence of great and valuable deposits of copper on Keweenaw Point, more detailed studies have been made of this than of any other of the Keweenawan districts, with the possible exception of Lsle Royal. Areas which have been studied with consiiierable detail are the outer part of Keweenaw Point, especially Eagle River, by Mai-^ine ''■ and Hubbard; <* Mount Bohemia, by Wright; <^ and the Portage Lake area, where the important deposits of cop])er occur, by Pumpelly-'^ and Hubliard.'^ Studies of intermediate areas have been less detailed but still suflicient for Irving,'' Seaman,'' and others to attempt to correlate the differ- ent formations for Keweenaw Point. (See PI. XXVIII.) o Gordon, W. C, assisted by A. C. Lane, A geological section from Bessemer down Black River: Rept. Michigan Geol. Survey for 1906, 1907, p. iKS. t Michigan Geol. Survey, vol. 6, pt. 2, 1898, p. 53. c Marvinc, A. U., Ocol. Survey Michigan, 1869-1873, vol. 1, pt. 2, 1873, pp. 47-01, 95-140. "1 Hubbard, L. L., Keweenaw Point, with particular reference to the felsites and their associated rocks: Geol. Survey Michigan, vol. 6, p*. 2, 1898. t Wright, F. E., The Intrusive rocks of Mount Bohemia, Michigan : Ann. Rept. Geol. Survey Michigan for 1908, 1909, pp. 361-402. / Pnmpclly, Raphael, Geol. Sur%'ey Michigan, 1S()9-1873, vol. 1, pt. 2, 1873, pp. 1-46, 02-94. e Irving, R. D., Copper-bearing rocks of Lake Superior: Mon. II. S. Geol. Survey, vol. 5, 1SS3. "Jour. Geology, vol. 15, 1907, pp. 8SO-095. u s. afOLoan:*!. SURVEY JIEQBGE OtiS SUirx. QIBECtOW MONOGmpH ui mie I GEOLOGIC MAP OF KEWEENAW POINT COPPER DISTRICT, MICHIGAN Rovis.-il I'v A !•: Seaman. Mirhigan Coflego oPMmes Sonir lulaaci n KEWICNWflW V^m* WPVIENAWAN SiuiilHlunra AJid lanAlomrrsIrs Lovnn-MioDLE wpwtm*w*w Acidic]' lavDaond ininiai" ])IMIiliily iinciudiH* »oni llUrrbeddoil nin^unurale; niimbcr^d iiccordijil lu Irvuii* ._v.. •"■ii.V,p1S>+- "" Lonaotiii-iaii-li THE KEWEENAWAN SERIES. 381 Below are given the successions of Irving ° for the entire point and of Hubbard'' for the outer part of the point, with their corrchition. Sections of rocks on Keweenaw Point. Irving. Hubbard. 12. Eastern sandstone. Keweenaw scries. Upper division: U. Red sandstone. 10. Black shale and gray sandstone ("Nonesuch belt"). 9. Red sandstone and conglomerate ("Outer conglomerate"). Outer conglomerate. Lower division: S. Diabase and diabase amygdaloid, including at least one conglomerate belt ("Lake Shore trap"). Lake Shore trap (upper). Middle conglomerate. Lake Shore trap (lower). 7. Red sandstone and conglomerate ("Great conglomerate"). Great conglomerate. 6. Diabase and diabase amygdaloid, including several sandstone belts (Mar- vine's " Group C " of the Eagle River section). 5. Diabase and diabase amygdaloid, including conglomerates. 4. Luster-mottled melaphjTes and coarse-grained gabbros and diabases (" Green- stone group"). Ophites and porphyrites witli interbedded conglomerates and sandstones. 3. Diabase, diabase amygdaloid, and luster-mottled melaphyre, including a number of conglomerate beds. Melaphyres and interbedded conglomerates. 2. Quartz porphyry and felsite. (a) Bohemia conglomerate.iLocallv Mount Houghton fel- (6) Melaphyre. I site replaces a and 5. (c) Porphyrite and felsite porphyrite. -;l 1 1. Diabase, diabase amygdaloid, melaphyre, diabase porphyry, and orthoclase '•'■ gabbro, including also conglomerate beds and beds or areas of quartz porphyry and granitic porphyry (■' Bohemian Range group"). Ophite belt. Lac la Belle conglomerate. LOWER AND MIDDLE KEWEENAWAN OF KEWEENAW POINT. ORDER OF EXTRUSION. Hubbard ■= lias studied the order of extrusion for the outer part of Keweenaw Point. He finds the oldest lavas to be melaphyres and these are interstratified with melaphyre conglomer- ates. Following the melaphyres are porphyrites and interstratified with the porphyrites are porphyrite conglomerates. Next come the felsites and interstratified with these and above them are the felsite conglomerates. All these rocks are at very low horizons. Above them lies a great mass of melaphyres, ophites, and porphyrites with their various interbedded conglom- erates and sandstones. Still higher are the "Great" conglomerate and tlie "Lake Shore" trap with the "Middle" conglomerate. Thus Hubbard's studies of Keweenaw Point led him to the conclusion that there was a regular order of extrusion of the igneous rocks — (1) basic melaphyres, (2) intermediate porphyrites, (3) acidic felsites and porphyries, and (4) the upper basic rocks represented by melaphyres, opliites, porphyrites, etc. PRESENCE OF BASIC INTRUSIVE ROCKS. Curiously the descriptions of the basic rocks of Keweenaw Point mention no interstratified intrusive sills, all the basic rocks being assumed to be flows. However, certain groups, as for instance the greenstone group, are described as contrastmg sharply with the rocks above and below them. They contain no mtercalated amygdaloidal beds. They consist of massive laj^ers. In texture they vary from diabases to gabbros. Although this and other masses were not sufii- ciently examined to make any positive assertion possible, it is our impression that a large part of the greenstone is an intrusive sill. The other masses of rocks which have been described as gabbro or orthoclase gabbro, especially those on the southwestern part of the point, are intrusive. o Mon. U. S. Geol. Survey, vol. 5, 1883, PI. XVII. 6 Geol. Survey Michigan, vol. 6, pt. 2, 1898, PI. IV. c Op. eit. 382 GEOLOGY OF THE LAKE SUPERIOR REGION'. On Mount Bohemia the intrusive gabbro has produced contact effects on the invaded ophites. Tlic prolilem of separating the intrusive basic rocks from the extrusives remains partly to be accompUshcd. ' ACIDIC INTRUSIVE ROCKS. Hubbard's'* studies show that the felsites of Bare Hill and West Pond at ver\' low horizons are intrusive. Tiic fclsite of Bare Hill, when mapped in detail, is seen to cut across the beds of other rocks, although in a single section near its center it would seem to be interstratified. The felsite of West Pond has disturbed the beds in its immediate area. They are broken into frag- ments and in places are even changed into typical breccias, some of which are almost undistin- guishable from the conglomerates. These intrusive rocks were perhaps correlative with the extrusive felsites of Mount Houghton and others of approximately the same age found at higher horizons. The intnisive nature of these felsites explains the absence of pebbles derived from them ill the melaphj-re conglomerates interstratified with the melaphyres adjacent to and at horizons above the felsites. Wliile some of the felsites and por[)liyries are extrusive, even these have a very minor extent. This is very well illustrated at Mount Houghton, where the felsite locally replaces the "Bohemia" conglomerate and the melaphyre flow below. (See preceding table.) NATURE AND SOURCE OF DETRITAL MATERIAL. It is well known that the felsite and porphyry pebbles are ver}- prevalent and in places dominant in the numerous conglomerate beds interstratified with the basic rocks at the higher horizons of Keweenaw Point, and even in the "Great" conglomerate, "Middle" conglomerate, and "Outer" conglomerate. There seems to be an enormous amount of felsite and porphj-ry detritus in the sediments as compared with the known original areas from which it may have been derived. Doubtless a part of the acidic detritus of Keweenaw Point may have been derived from porphyries farther east and west than the point, as, for instance, those of the Stannard Rock area to the east and the Porcupine Mountains to the west. But also the lack of large areas of felsites may be due to the exceptional erosion to which they have been subjected because of their viscous and bunchj' character, which raised them and made them the objects of excessive attack. Finally, a considerable portion of the acidic detritus may have been in the form of volcanic fragmental material that was scattered far and wide from the original cones from which it was ejected and therefore never formed a part of any continuous solid intrusion or extrusion. Lane* states that the detritus of several conglomerates, especially of the "Great" conglom- erate, includes numerous pebbles of intnisive red rock and gabbro. He says that if he is correct in his identification of the materials there is evidence of an erosion of sufficient magnitude dur- mg middle Keweenawan time to expose these plutonic rocks at the surface. He also finds agate pebbles which he believes to have formed in the lavas lower in the series, and thus he concludes that extensive metasomatie changes have taken place in this part of the series before the liigher interstratified conglomerates were laid down. VARIATIONS IN THICKNESS OF SEDIMENTARY BEDS. Close studies of Keweenaw Point show rapid variations m the thickness and character of the interstratified sedimentary beds. These have been especially studied in the mineraUzed area. Many illustrations coukl be given, but perhaps one of the clearest is that of the "Great" conglom- erate wliich Hubbard '^ says tliins 400 to 700 feet in passing from Copper Harbor to a pouit 7 miles farther east. Not only do the beds change in their character, but a single sedimentary 1)0(1 may be split into several beds separated by lava flows. Thus m the Bohemia basin a con- glomerate is first split mto two parts by a bed of melaphyre and the lower part is in turn split into two beds by a mass of felsite. The beds are in general lenticular, broadly consitlered, but some of these lenses ma}- be onh* a few miles in length, as illustrated by the Calumet and Hecla conglomerate. a Uuhbaril, L. L., Michi;;an Geol. Survey, vol. 6, pt. 2, 1898, pp. 35, 43. ti Lane, .\. C Geology of Keweenaw Point: Proc. Lake Superior Min. Inst., vol. 12, 1907, p. 93. c Op. ell., p. «4. THE KEWEENAWAN SERIES. 383 FAULTS. Hubbard's " detailed studies of small areas have led also to the conclusion that the middle Keweenawan has been displaced by a very large number of dip faults, the throws of which, how- ever, are of minor extent. These have been worked out in great detail with reference to the melaphyre and melaphyre conglomerates at West Pond. Here are figured no less than twelve cross faults, the throws of which, however, are not sufficiently great to be traced into the thick overlying formations, and hence they do not appear on liis general map. Similarly Lane ^ states that there are a large number of small transverse faults in the mining district. The throws of most of these faults are not more than 2.5 feet and very few exceed .50 feet. However, the presence of many faults at each of the two areas that have been closely studied on Keweenaw Point suggests very strongly that when like thorough studies are made of other areas on this point similar faulting wUl be found. In the mining district there are also many slide faults. According to Lane,"^ the dip of many of these slide faults is somewhat steeper than the bedding, so as to cut diagonally across the beds at acute angles. As to the direction of movement along these dip faults, he thinks it is more commonly down than up on the hanging-wall side, for beds are more likely to be cut out than repeated. Hubbard" described one very important slide fault, the major movement of wliich, instead of being parallel to the dip, is nearly parallel to the strike. Tliis is the fault at the top of the Kearsarge conglomerate, whicli is well illustrated in tlie Central mine. Hubbard makes a calculation of throw and reaches the conclusion that "the part of the Keweenawan series that lies above the Kearsarge conglomerate has moved from its original position, in a northerly direction, horizontally, about 2.7 miles, or along an inclined plane its equivalent dis- tance of about 2.9 miles." Such a slide fault as this approaches the ordinary strike faults, the chief difTerence being that of hade, the bedding fault having such a hade as not to intersect the bedding, whereas ordinaiy strike faults do intersect the bedding. Although the Kearsarge shde fault is nearly in the direction of the strike, it is believed to be probable that the most common direction of movement in the faults of this area is parallel to the dip. In this case the move- ments are largely explained by the natural adjustments which are necessary when a set of beds, is folded. UPPER KEWEENAWAN. The upper Keweenawan consists, from the base upward, of three members — (1) the " Outer" conglomerate, (2) the Nonesuch shale, and (3) the Freda sandstone. The "Outer" conglomerate is found at the north side of the east end of Keweenaw Point as far as Gate Harbor, where it passes under the water; it reappears on the point some miles west of Eagle River and continues along to the point and westward through Michigan into Wisconsm. It is in no respect different from the underlying "Great" conglomerate or other conglomerates interstratified with the Keweenawan, except that, accordmg to Lane,'* it contains near its top fragments derived from the jaspery and other Huronian formations. The Nonesuch formation ranges from a soft, fine-grained, highly argillaceous shale to a sand- stone. The shale is predominant, the sandstones bemg mterbedded. In color the shale is dark purplish gray to nearly black and the sandstone dark gray to black. The thin sections of the sandstones show detritus from the porphyries and other acidic original rocks of the Keweenawan. With these materials in all the sections is mingled more or less basic detritus. Indeed, the basic material is usually abundant and not uncommonly becomes dominant. The basic material is more abundant in the darker-colored rocks. In these rocks there is also a plentiful calcite a Hubbard, L. L., Michigan Geol. Survey, vol. 6, pt. 2, 1898, pp. 87-91. !> Lane. A. C, Geology of Keweenaw Point: Proc. Lake Superior Min. Inst., vol. 12, 1907, pp. 83-84. c Idem, pp. S4-85. d Lane, A. C, Jour. Geology, vol. 15, 1907, p. 690. 384 GEOLOGY OF THE LAKE SUPERIOR REGION. cement filling all interstices between the fragments. The basic detritus appears in the .shape of particles of the basic rocks, showing more or less plainly the several ingredients, always much altered, and of particles of the single minerals — augitc, almost wholly altered to a greenish sub- stance triclinic feldspar, and magnetite. The formation also contains materials which must have been contributed by the Ihironian, Kewcenawan, and Laurent ian rocks. The Nonesuch shale therefore differs from the sediments interstratilied with the Kewcenawan m the greatly decreastid amount of acidic material, the abundance of basic material, and the presence of detri- tus derived from other formations than the Kewcenawan. The Freda sandstone is in no respect different from the sandstone of the liCfger areas in Wisconsin, which are a continuation of the sandstone in Michigan. It need here be only remarked that the materials have the same varieties of sources as the Nonesuch' shale, but the materia derived from the basic lavas seems to be even more prominent. BELATIONS TO CAMBRIAN BOCKS. On the north and west sides of the Kewcenawan the series nowhere comes into contact with the Cambrian. The possible relations between the two are discussed in another place. (See pp. 415-416.) On the southeast side the Kewcenawan is in contact with the Cambrian. Irving and Chamberlin," in their bulletin on this contact, conclude that the sandstone was deposited uncon- formably against an ancient fault scarp of Kewcenawan rocks and that it was subsequently faulted dowTi along the old fault plane. This relation is apparently similar to those observed at the fault on the north side of the Keweenawan syncline in Douglas County, Wis., and thence southwestward into Minnesota. MAIN AREA WEST OF KEWEENAW POINT, INCLUDING BLACK KIVER AND THE PORCUPINE MOUNTAINS. A very detailed study of the entire Keweenawan section at Black River has been made by Gordon.'' According to him, this river shows the following descending succession,*^ the classifi- cation into middle and lower Keweenawan being added by us. Section of Keweenawan rocks at Black River, Mich. Upper Keweenawan: I. Upper sandstone lacking. leet. II. Nonesuch formation 500 III. Outer conglomerate o, 000 5,500 Middle Keweenawan: IV. Lake Shore trap, consisting of five flows, having from the top downward the following thicknesses: 35, 35, 115, 85, 130 feet, respectively 400 V. Conglomerate 350 VI. Mixed eruptives and sedimentaries ■">, ■'>00 VII. Felsite •. ^50 VIII. Eruptives with very few sedimentaries 2G, 000 IX. Mixed eruptives among which are conspicuous labradoritc porphyrites 4, 800 X. Gabbro 200 XI. Melaphyres and labradorite porphyrites that are not conspicuous 4, 500 42, 200 Lower Keweenawan: XII. Basal sandstone 300 48,000 "In'ing, R. D.,and Chimlwrlin, T. C, Observations on the junction between the Eastern sandstone and the Keweenaw series on Keweenaw Point. Lake Superior: Bull. U. S. Ocol. Survey No. 23, 18.S5. i> Gordon, W. C, assisted by A. C. Lane, A geological section from Bessemer down Black River: Kept Michigan Geol. Survey tor 1900, 1907, pp. .197-507. cidem, p. 421. THE KEWEENAWAN SERIES. 385 Throughout the Black River section there is no evidence of a physical break in the Kewee- nawan. Lane," because of the character of the formation, suggests that possibly there might be a slight break at the base of the Nonesuch shale, but Gordon's detailed descriptions give no evidence in support of this view. It is known that in this district dip faults occur. According to Gordon,*" at least four such faults traverse the Trap Range north of Bessemer, the throws of three of which are 80, .350, and 1,500 feet, the throw of the fourth not being determinable. It is very likely that strike faults occur, for great strike faults occur elsewhere in the Keweenawan. (See p. 38.3.) Though such faults have not been detected, they may very readily occur at any of the very numerous stretches of the river where exposures are lacking. The presence of faults at these places is very probable because of • the brecciation and consequent more easily erosible condition of rocks along fault planes. In the Porcupine Mountains the same divisions of rocks occur as in Keweenaw Point, but the order is only in a general way similar to that on the point, the difference being that compara- tively high in the series are large masses of quartz porphyry and felsite, and the acidic rocks at these horizons perhaps largely explain the source of the abundant felsite, quartz porphyry, and augite syenite pebbles in the "Great," "Middle," and "Outer" conglomerates. In the Porcupine Mountains the great synclinal basin of Lake Superior, which controls the general dip of the Keweenawan rocks about the lake, is disturbed by a subordinate fold, so that in a section diagonally northeast and southwest across the mountains the lower beds are regarded by Irving "^ as repeated. He shows a subordinate anticline and sj'ncline between the monoclinal beds north of Lake Gogebic, at tlie south side of tlie middle division and the northward-dippmg beds at tlie lake. This area is a forest-covered one in which the exposures are somewhat imperfect, and it is hinted by Hubbard <^ that possibly the abundant felsite and porphyry here are intrusive, as they are at Bare Hill and West Pond, and that the unusual structure may be explamed by these intrusive masses rather than by exceptional orogenic movement. This suggestion is made because of very considerable disturbances in the regidar bedding of the rocks about the intrusive felsite of Bare Hill. The Porcupine Mountains are now being studied in detail by F. E. Wright for the Micliigan Geological Survey, but the results of his work have not been available in the preparation of this monograph. The upper Keweenawan of this area is the same in all respects as that described for Kewee- naw Pomt. THE SOUTH RANGE. Beginning in T. 47 N., R. 44 W., Michigan, from the lower Keweenawan, which there consists of diabase, diabase amygdaloid, melaphyre, and a few coarse interbedded thin con- glomerates, an arm projects to the east and south nearly to Gogebic Lake and east of this lake again for some distance. This is the so-called South Range. It is separated fi-om the main range of the Keweenawan by the Jacobsville or "Eastern" sandstone. At the eastern point the South Range is 18 miles south of the northern area of the Keweenawan. This range varies from less than half a mile to 2 miles or more in breadth. The rocks of the South Range dip to the north at angles of 30° to 50°. At some places at the base of the Keweenawan series in the South Range there is a coarse sandstone. At other places the lowest rock is a basic lava. Locally sediments are hiterstratified with the lavas. Thus the conditions prevalent in early Keweenawan time, as indicated by the rocks at the base of the Keweenawan of the South Range, are similar to those of other tlistricts. In no respects do these rocks differ from those near the base of the Keweenawan to the west. West of Gogebic Lake the Keweenawan rocks rest directly upon the upper Huronian. The western part of this belt of Keweenawan rests directly upon the Tyler slate. When followed to the east it is seen to pass diagonally to lower and lower horizons, imtil at Sunday Lake it is in contact with the Ironwood formation. These relations have been more fully described in connection with the Penokee district. a Lane, A. C, Jour. Geology, vol. 15, 1907, p. 091. " Irving, R. D., Men. U. S. Geol. Survey, vol. 5, 1883, pp. 209-225. i> Op. oit., pp. 464-465. d Hubbard, L. L., Geol. Survey Micbigan, vol. 6, pt. 2, 1898, pp. 5-8. 47517°— VOL 52—11 25 386 GEOLOGY OF THE LAKE SUPERIOR REGION. It is believed that the separation of the South Range from the mam range is due to a great strike fault between the two which results in a repetition of the beds of the main range in the South Range. ROCKS OF POSSIBLE KEWEENAWAN AGE IN OUTLYING AREAS. Certam reddisli feklspatliic and little-consolidated sandstones of low dip, lying uncon- formably across the end of the upper Iluronian of the Felch Mountain trough, may possibly be classed as Keweenawan. Similar rocks are known also in the Sturgeon trough to the north. THICKNESS OF THE KEAVEENAWAN OF MICHIGAN. Irving" gives an estimate of the thickness of the Keweenawan of northern Michigan at Eagle River and Portage Lake, and Gordon' estimates a section on Black River. EAGLE RIVER SECTION.' Irving's section at Eagle River,'' based largely on the detailed work of Marvine, is as follows: Section of Keweenawan rods at Eagle River, Michigan. Upper division: Feet. Outer conglomerate;, porphyry conglomerate and sandstone; about 1, 000 Lower division : Lake Shore trap; very plainly bedded fine-grained diabases, strongly marked amygda- loids, and one or more thin porphyry conglomerates; about 1, 500 Great conglomerate ; jjorphyiy conglomerate and sandstone 2, 200 Marvine's group "c;" plainly bedded and separable fine-grained diabases, with strongly marked amygdaloids, predominatingly calcitic; and some 850 to 900 feet, in all, of inter- stratified sandstones 1, 417 Marvine's group "b," or the Ashbed group; made up mostly of thin, fine-grained diabases, which vary a good deal in appearance, but are generally provided with distinct amyg- daloids; including some beds of the peculiar tj'pe known as ashbed diabase; also several Bcoriaceous amygdaloids, being intermingled sandstone and amygdaloid; also one thin sandstone seam 618 Marvine's group "a;" made up of relatively heavy beds without strongly developed amygdaloids; including one thin seam of .sandstone 925 Greenstone group; made up of relatively heavy beds, without amygdaloids, of rocks for the most part relatively coarse grained;, these belong mostly to the coarse-grained olivine-free diabases and gabbrosand to the luster-mottled melaphyres, or fine-grained olivine-diabases, the greenstone at the base of the group being of the last-named class. . 1, 200 Subgreenstone group, in which all of the fissure-vein mines are working; having at top a thin conglomerate, the equivalent of the "Allouez" and "Albany and Boston'' con- glomerates in the Portage Lake district; composed of fine-grained diabases, with not very strongly developed amygdaloids; about 1, COO Central Valley beds; the layers not well exposed, but evidently chiefly fine-grained dia- bases and amygdaloids, with a number of thin porphyry conglomerates, in all respects like the overlying group; about 5, 540 Bohemian Range beds; made up chiefly of diabases and melaphyres in all respects like the higher layers, and including .some of the usual porphjTy conglomerates; but also in part made up of quartziferous porphyry, felsite, nonquartziferous porphyry, and coarse- grained orthoclase gabbro; in all. al)c)ut 10, 000 26, 000 Of this thickness, the "Great" conglomerate and the ten conglomerates and sandstones in "group c" together constitute 3,100 feet. These sediments are all in the upper 5,000 feet of the lower Keweenawan. The lower part contains only a few seams of detrital material. The lower five-si-\ths of the lower Keweenawan for this section is therefore almost exclusively vol- canic, and of the total lower Keweenawan somewhat less than one-ninth is sediment arA", the remaining eight-ninths being igneous. " In-ing, R. D., The copper-bearing rocks of Lake Superior: Mon. U. S. Geol. .Survey, vol. 5, 1883. pp. lGf>-I97. b Gordon, W. C, assisted by .\. C. Lane, A geoiogica! section from Bessemer down Black River; Rept. Michigan Geol. Survey for 1906, 1907, p. 421. clrvlng, K. D., op. cit., pp. lso-187. THE KEWEENAWAN SERIES. 387 PORTAGE LAKE SECTION. At Portage Lake the section is as follows: Section of Keweenawan rocks at Portngc Luke. a Upper division: ■ peet. Larijely covered, but apparently for the most part red shales and sandstone; toward the base there is a considerable thickness (upward of 200 feetj of dark-colored, fine-grained sandstone and black shale, in which the usual porphyry detritus is mingled with more or less basic detritus; the lowest layers are also conglomeratic; in all about 9, 000 Lower division: Covered space of some 1,200 feet, in which must be the equivalents of the outer trap of the eastern part of Keweenaw Point, corresponding to a thickness of about 500 The Great conglomerate, including the sandstone and conglomerate at the Atlantic mill and conglomerate 22 on the south side of Portage Lake, with some intervening ex- posures, about I ^ 500- Diabase Og Conglomerate 21 15 Diabase and amygdaloid 51 Conglomerate 20 19' Diabase 100' Conglomerate 19 13- Diabase 94 Conglomerate 18 I55. Diabases and amygdaloids 34O Conglomerate 17 (Hancock West) 32: Diabases and amygdaloids; including the South Pewabic cupriferous amygdaloid at 50 feet below 17 55O Conglomerate 16 (not seen on south side of Portage Lake) lO- Diabases and amygdaloids; including, at 400 feet above conglomerate 15, the Pewabic cupriferous amygdaloid or "lode" so largely worked for copper on the west side of Portage Lake 900 Conglomerate 15 (Albany and Boston conglomerate on the north side of Portage Lake). . 3S Diabases and amygdaloids 33O Conglomerate 14 (the Houghton conglomerate of the north shore) 2 Diabases and amygdaloids 1_ 400 Conglomerate 12 (north side of Portage Lake) 3 Diabases and amygdaloids 680 Conglomerate 11 20 Diabases and amygdaloids 200 Conglomerate 10 gO Diabases and amygdaloids 46Q Conglomerate 9 (sandstone seam). Diabases and amygdaloids; including, at 670 feet above conglomerate 8, the Grand Port- age cupriferous amygdaloid, and at 510 feet the Isle Royal cupriferous amygdaloid, largely worked on the south shore of Portage Lake 2, 05O Conglomerate 8 12 Diabases and amygdaloids 420 Conglomerate 7 24 Diabases and amygdaloids 260 Conglomerate 6 3 Diabases and amygdaloids IgX Conglomerate 5 24 Diabases and amygdaloids 240 Conglomerate 4 12 Diabases and amygdaloids 1 149 Conglomerate 3 5g Diabases anct amygdaloids 37O Conglomerate 2 35 Diabases and amygdaloids 1 140 Conglomerate 1 97 Amygdaloid 14 22, 680 ' Irving, R. D., The copper-bearing rocks of Lake Superior: Mon. U. S. Geol. Survey, vol. 5, 1S83, pp. 194-195. 388 GEOLOGY OF THE LAKE SUPERIOR REGION. In tlio above section of tlio lower Keweenaw an the thickness of tiie c()n<;lomeiates amounts to 2,125 feet, leaving 11,555 feet for the igneous rocks. Thus the lower Keweenawan is about one-sixtli sediment and about five-sixtlis igneous. The Portage Lake section differs in one important respect from the Eagle River section. At Portage Lake the interstratified conglomerates extend to the bottom of the section, whereas at Eagle River the conglomerates and sandstones do not occui' in the lower five-sixths of the section, tlie thickness of which as a whole is al>out the same as at Portage Lake. BLACK RIVER SECTION. In the Black River section the total thickness, according to Goi-don," is 48,000 feet. Irving* estimates the thickness of the upper sandstone at Montreal River, a few miles west of Black River, at 12,000 feet. This part of the section is absent on Black River, and if it were ailded to the Black River section this would give for this district a thickness of 60,000 feet for the entire Keweenawan series. In the middle Keweenawan of the Black River section (p. 384) the sediments are mainly in the upper 6,000 feet, and of this amount sedunents are known to make up 575 feet, distributed as follows: Feet. In V, conglomerate 350 In VI, mixed eruptive and sedimentary rocks: Sandstone .30 Conglomerate 20 Sandstone 2.5 Sandstone 30 Sandstone 20 Conglomerate 100 225 575 As a space corresponding to .3,000 feet is not exposed, doubtless the total thickness of the sediments is much greater than tliis, though a part of this .3,000 feet is certain to be volcanic. However, the addition of all of it would make the maximum possible thickness of sediment 3,575 feet. Thus the sediments at most make up only about one-twelfth of the middle Kewee- naw^an and are largely concentrated in the upper sixth of the division. The question now arises whether this apparent thickness for the several sections repre- sents the real thickness of the series as laid down. It is believed to be probable that the real thickness is less than the apparent thickness. The reasons for this belief apply as well to the estimated thicknesses of other districts, and therefore they are given later. (See pp. 418-419.) RELATIONS OF THE KEWEENAWAN OF MICHIGAN TO UNDERLYING AND OVERLYING FORMATIONS. The only locality in which the relations of the Keweenawan with the underlying forma- tions are shown is in the Penokee-Gogebic district. It has been stated (pp. 234-235) that these relations are those of unconformity, erosion amoimting to several thousand feet having taken ])lace after Iltironian time and before the deposition of the Keweenawan. Still, the strike and di|) of the two series are very nearly the same, and the greatness of the break between the two appears only b}" their stratigraphic relations. The Upper Cambrian ("Eastern") sandstone comes against the lower part of the Keweena- wan from the outer end of Keweenaw Point to the region west of Gogebic Lake. It is agreed by all who have studied this contact that it marks a great fault. The Keweenawan along the contact has its usual steep northern dips. The sanilstone at the contact is bent and locally broken, so that it strikes and dips in various directions, in some places dipping away from the o Gordon, W. 0., assisted by A. C. Lane, A geological section from Bessemer down Black River: Kept. Michigan Geol. Survey for 1906, 1907, p. 421. l> Irving, U. D., The copper-bcnring rocks ol L:iko Superior: Mon. U. S. Geol. Survey, vol. 5, 1S83, p. 230. THE KEWEENAWAN SERIES. 389 Keweenawan and in others apparently dipping under it. A short (Hstance away from the Keweenawan, usually within a few hundred feet, the sandstone assumes its normal horizontal attitude. At only a few localities has the Upper Cambrian sandstone been found in close relations with the rocks of the South Range. Irving '^ concluded that in the South Range this sand- stone rests unconformably against the Keweenawan rocks. However, the particular locality he described as showing unconformable relations has been interpreted differently by Seaman,* who finds there a dike of igneous rock penetrating the so-called "Eastern" sandstone and spreading out above. Seaman regards the "Eastern " sandstone here as probably Keweenawan and believes that there is no way of proving that it is of different age from the "Western" sandstone (upper Keweenawan). ISLE ROYAL. Isle Royal is 45 miles in length and varies in width from 3 to 8 miles. From the Rock of Ages, the farthest outlying reef to the southwest, to the Gull Island rocks on the northeast, the distance is 57 miles. The island lies off Thunder Bay, northwest of the outer part of Kewee- naw Point. The strike of Isle Royal and Keweenaw Point are substantially the same, north- east and southwest. This island has been mapped geologically by Lane.'' His succession in descending order is as follows: Section of Keweenawan rocks on Isle Royal. Sandstone and conglomerate ("the Great conglomerate"?). Ophites dowfl to Island mine conglomerate (Marvine's group C). Intercalated sandstones and conglomerates. Melaphyre porphyrites and scoriaceous conglomerates ("Ashbed" group). "The greenstone" — thickest ophite. Amygdaloids and thin ophites down to Minong breccia (Kearsarge conglomerate). Minong porphyrite and Minong trap. Ophites and conglomerates, including Huginnin porphyrite, down to felsite. It is clear from the general character of the succession that it is like that of the middle Keweenawan of the remainder of the Lake Superior region; that is to say, it consists of igneous rocks and sediments. The igneous rocks are dominantly basic. They are all regarded as extrusive by Lane.*^ However, the same question may be raised with reference to the greenstone, wliich is given a tliickness of 2.33 feet, as was raised concerning that of Keweenaw Point. Is it an extru- sive or is it a later intrusive ? Certainly it has all the characteristics of the diabase of Beaver Bay on the Minnesota coast, which is almost certainly intrusive. The intercalated sandstones and conglomerates, from lowest to highest, contain a much greater proportion of material from acidic rocks than would be expected from the small pro- portion of original acidic rocks. The sandstones and conglomerates are subordinate in amount in the major portion of the section and only become of great volume with the appearance of the "Great" conglomerate. The field terms for the igneous rocks and their relations are expressed by Lane<* as follows: Felsite Melaphyre Trap — nonamygdaloidal dark rocks Amygdaloid Porphyry Porphyrite Ophite Lane' gives one very detailed section based largely on drill records. Its thickness is 9,000 feet. In this section the felsite flows are confined to the lower 150 feet, but at a high horizon one bed of porphyry tuff 10 feet thick is noted. This tuff may be regarded as a confirmation "Irnug, R. D., The copper-bearing rocks of Lake Superior; Men. U. S. Geol. Survey, vol. 5, 1883, pp. 360-361. 6 Personal communication. c Lane, A. C, Geological report on Isle Royale, Michigan: Geol. Survey Michigan, vol. 6, pt. 1, 1898, 281 pp. dldem, p. 53. c Idem, pp. 27 et seq. 390 GEOLOGY OF THE LAKE SUPERIOR REGION. of the suggestion made in another place (p. 382) that volcanic fragmental rocks of the acidic type are nuich more abundant in the Keweenawan than had been supposi'd. Most of the interstratifiod sedimentary beds are conglomerates and, with three exceptions, they range from a knife-edge to 50 feet in thickness. Two of the thicker beds are mainly sandstone. In addition to a number of seams which were too small to be measured, the total number of sedimentary beds in the district is 24 and the total thickness is 430 feet. To the "Great" conglomerate is given a thickness of 2,600 feet, making a total thickness of sediments of 3.030 feet. This leaves 5,970 feet for the lavas. In the matter of correlation. Lane " assumes that the thick conglomerate at the top of the series is a continuation of the "Great" conglomerate of Keweenaw Point, and with this horizon as a starting point he attempts to correlate somewhat closely the h6x\s of Isle Royal with those of Keweenaw Point, as is indicated by the succession given on page 389, the names in parentheses being those of formations on Keweenaw Point. Although it is probable that the top conglomerate corresponds to the "Great" conglomerate of Keweenaw Point, and although it may be possible that the formations are to some extent equivalent, it maj- perhaps be doubted whether the correlation of individual thin beds, such as the interstratified conglom- erates, is justified, especially as. there is so remarkable a likeness in the petrography of the beds of the Keweenawan at difterent horizons in the several districts of Lake Superior. If the bed of greenstone more than 200 feet thick is really intrusive, as suggested, its correlation with the greenstone of Keweenaw Point on a stratigraphic basis is very questionable. In any section on Isle Royal there is a lessening of the dip in passing from lower to higher horizons, just as at Keweenaw Point and at Michipicoten. For instance, at the west end of the island on the north side the dips are 16° S. and on the south side in the "Great" conglom- erate 8° S., a difference of 8°. Toward the east end of the island the dips on the north side are 26° and on the south side 18°, again a difference of 8°. MICHIPICOTEN ISLAND. The folio whig account of Michipicoten Island is taken almost wholly from Burwash,' who alone has made a close study of this area. However, it should be said that Logan's general accoinit of this district " is remarkably accurate. The island is roughly ellipsoidal in shape, about 16| miles long by 6 miles in greatest width. Its longer axis lies east and west parallel to the coast, and its west end is south and a httle west of Pukaskwa River. The Keweenawan rocks occupy the entire island as well as the row of smaller islands off its south shore. They are confined wholly to the middle Keweenawan. The igneous rocks overwhelmingly dominate in mass. They are described as extrusive, no intrusive rocks being mentionetl. Lithologically they include all the varieties of the ordinary extnisive rocks, ophitic and diabasic melaphyres, amygdaloids, porphyrites, felsites, and quartz jiorphyries. The acidic rocks are much more readily eroded than the basic rocks. In consequence they usually occupy depressions, whereas the basic rocks constitute the ridges. In this respect there is a contrast between Micliipicoten Island and Keweenaw Point, where the acidic rocks constitute elevations. It may be suggested that the difference is due to the fact that the- Michipicoten acidic rocks are largely extrusive, while those of Keweenaw Point are largelj' intrusive. No order of extrusion of the lavas is suggested, the acidic and basic rocks both occurring from the higliest to the lowest horizons. As to volume, there does not seem to be much difference between tiie basic and acidic varieties. Selwyn and Coleman state that pyro- clastic rocks occur on Michipicoten Island, but such rocks were not observed by Burwash <* and if ])resent are certainly extremely insignificant in amount. The lava ])eds attain their maximum thiclaiess in the eastern and central parts of the island and are thinner toward the a Lane, A. C, Geological report on Isle Royale, Michigan: Geol. Survey Michigan, vol. 6, pt. 1, 1898, pp. 99 et seq. 6 Bnrwa.sh, E. N., The geology of Michipicoten Island: Univ. Toronto Studies, Geol. ser., No. 3, Toronto, 1905, with map. « Logan, W. E., Report of progress to 18G3, Geol. Survey Canada, 1863. dOp. clt., pp. 27, 47. THE KEWEENAWAN SERIES. 391 west, where they are interstratified with the conglomerates. The lower beds strike approxi- mately northeast and southwest. In passing to higher horizons the strike approaches east and west. Thus there is an appearance of minor unconformity between the lower and upper beds. The debris of the conglomerates is as usual derived largely from the acidic rocks, but with them are included granites, greenstones, and biotite gneisses derived from pre-Keweenawan formations. Abundant material derived from the basic rocks is also recognized. The sedi- mentary rocks occur mainly at lower horizons, although one conglomerate is foimd at a com- paratively high horizon. These conglomerates are confined to the nortliwestem part of the island, being thickest at the west and thinning out to the northeast. These facts suggest that the central and eastern parts of the island formed a center of volcanic dispersion, that the lavas flowed toward the west, and that in the part of the area somewhat removed from the main volcanic outbursts there was opportunity to build con- glomerates between the successive lava flows. The dip of the beds on the north and northwest sides of the island is 55° S. From this there is a steady decrease in dip until on the islands off the south shore of Michipicoten the dips are about 14° S., the lessening of dip across the series being therefore 40°. Burwash" gives the following descending succession: Section of Keweenawan rods on Michipicoten Island. Feet. 1. Felsite of islands off the south shore 1, 000 2. Pitchstone bed 530 3 . Quartzless porphyry of Quebec Harbor 695 4. Melaphyre porphyrites of Channel Lake 1, 660 5. Quartz porphyries ; 1 355 2 • 1,160 3 1,493 6. Beds exposed at lake on road 1, 575 7. Felsite 513 8. Diabase porphyrite •. 463 9. Beds underlying farm (three) 1, 140 10. Several beds at mine 645 11,230 This result, obtamed by accurate measurement of three sections and by careful studies, is a remarkable confirmation of the judgment of Logan,'' who states that the thickness of tlie formations developed in Michipicoten Island, at the most moderate dips observed, would not fall far short of 12,000 feet. It is stated that on the mainland near the mouth of Pukaskwa River there are rocks of Keweenawan age, and this leads to the suggestion that the Keweenawan constitutes a mono- clinal succession from the north shore of Lake Superior to the south side of Michipicoten. For the intervening distance between the mainland and the island an estimated thickness is given of 34,000 feet, and thus a suggested thickness for the entire Keweenawan series of 45,000 feet. But it seems to us more probable that between Michipicoten Island and the main shore there is a strike fault and that therefore the Micliipicoten rocks may be near the bottom of the Keweenawan series. This idea is perhaps confirmed by the presence in the conglomerates of the Michipicoten district of material from pre-Keweenawan sources. EAST COAST OF LAKE SUPERIOR. Several prominent points along the east coast of Lake Superior exliibit Keweenawan rocks. While none of these areas are large, they are significant, extending along nearly the entire east coast of Lake Superior from Cape Choyye, near Micliipicoten Harbor, to Gros Cap, intervening locaHties being Cape Gargantua, Pointe aux Mines, and Mamainse Peninsula. At all these local- n Op. lit., pp. 40-41. 6 Logan, W. E., Report of progress to 1803, Geol. Survey Canada, 1863, p. 82. 392 GEOLOGY OF THE LAKE SUPERIOR- REGION. ities the rocl« belong to the middle Keweenawan. They consist of basic lavas, including mela- pliyres, porphyritos, and amygdaloids, and interstratified sandstones and conglomerates. The sandstones and conglomerates tlillVr from the ordinary sculimentary rcjcks interstratified with the lavas in that they contain a consitlerable amount of detritus derived from the subjacent Archean rocks. This is particularly noticeable at Mamainse. For the most part the masses exposed are small, but Logan" estimates the tliickness of the series at Pointe aux Mines to be 3,000 feet. At Mamainse Peninsula the Keweenawan rocks occupy much the largest area along the east coast. Macfarlan(\ * calculates a total thickness in this locality of 16,208 feet, of whicli inter- stratified conglomerates make up 2,138 feet. Macfarlane's section, from the base upward, is as follows: Section of Keweenawan rods on Mamainse Peninsula. Feet. 1. Granular melaphyre, coilsisting of a small-grained mixture of dark-brown feldspar with angular grains of a dark-green chloritic mineral. It varie.s frequently in its structure, and in the upper part contains amygdules of calc spar and delessite (iron chlorite) 3, 930 2. Brown argillaceous sandstone, striking N. 20° W. and dipping 35° SW 12 3. Compact greenish-gray melaphyre, with grains of feldspar, iron chlorite, and hematite; strike N. 10° W.; dip 32° SW 1,787 4. Conglomerate holding granitic or gneissoid bowlders 852 5. Granular melaphyre, containing feldspar, which weathers white, and dark-green chlorite. 426 6. Sandstone 20 7. Dark-brown compact trap 71 8. Conglomerate 70 9. Dark-green melaphyre, slightly amygdaluidal 710 10. Conglomerate 43 11. MelaphjTe, striking N. 5° W., dip 30° W. ; fine grained and of a dark-brownish color 1, 207 12. Conglomerate 71 13. Granular melaphyre, containing brownish-red feldspar and abundance of delessite 355 14. Conglomerate 35 15. Fine-grained greenish-red melaphyre, becoming amygdaloidal in the upper part of the bed. Strike N. 20° W., dip 35° SW., where it adjoins conglomerate N. 15° W.>45° SW. 489 16. Conglomerate, with a small layer of sandstone, the latter striking N. 17° W., dip 40° SW.. 163 17. Compact dark-brown crystalline trap 340 18. Conglomerate 170 19. MelaphjTe 100 20. Conglomerate, striking N. 5° W. and dipping 42° W. at junction with overlying rocks 204 21. Melaphyre 240 22. Conglomerate, striking N. 12° W. In this bed the bowlders are smaller than in those hitherto mentioned 34 23. MelaphyTe, striking N. 23° W. and dipping 37° SW 682 24. Conglomerate and sandstone, striking N. 14° W. and dipping 44° SW 12 25. Melaphyre; strike N. 33° W.; dip 28° SW 250 26. Measures concealed 160 27. Melaphyre, granular and of a reddish-green color, striking N. 30° W. and dipping 18° SW. 25 28. Measures concealed 125 29. Melaphyre; strike N. 33° W.; dip 28° SW 272 30. Mea-sures concealed 180 31. Melaphyre, amygdaloidal in part 436 32. Measures concealed 400 33. Conglomerate, consisting of bowlders of Laurentian rocks in matrix of red sandstone 330 34. Measures concealed 172 35. Melaphyre, strikingN. 35° W. and dipping 20° SW 100 36. Conglomerate, in which the bowlders consist to a much greater extent than heretofore of amygdaloidal and other varieties of melaphyre. Strike N. 20° W.; dip 25° SW. at the junction with the overhang rock 50 37. Reddish-gray granular melaphyre, becoming amygdaloidal in the upper part 200 38. Sandstone, strikingN. 30° W. and dipping 24° SW : 12 39. Conglomerate, containing here and there layers of sandstone, striking N. 40° W. and dipping 15° SW 30 u I.ognn, W. E., Report of progrc-is to lSf.3, Oeol. Siin-ey Canada, 1S63, p. 82. 1> Mactarlaiie, Tlioimis, Report of progress from 1SG3 to 1860, Geol. Sur\'ey Canada, 1866, pp. 132-134. THE KEWEENAWAN SERIES. 393 Feet. 40. Dark-green glittering nielaphyre, striking, at its junction mth the underlying conglom- erate, N. 50° W. and dipping 30° SW 114 41. Measures concealed 137 42. Melaphyre, striking N. 50° W. and dipping 29° SW ;. 16 43 . Measures concealed 114 44. Melaphyre, dark reddish green, striking N. 50° to 55° W. and dipping 21° to 25° SW 300 45. Dark-green and glittering melaphjTe; N. 25° W.>20° SW 250 46. Compact fine-grained trap, containing geodes of agate, in which calc spar frequently occupies the center 350 47. Porphyritic conglomerate and sandstone; N. 8° W.>21° W 30 48. Compact fine-grained trap, containing agates in many places 72 16, 208 This thickness does not inchide the basal sandstone to be mentioned below. It is to be noted that in the 5,729 feet at the bottom of the section there is only one layer of sediment — a sandstone 12 feet thick. In the remainder of the section, 10,479 feet, conglomerates and sand- stones are interstratified at several places, the thickest bed being S.52 feet thick and lying at the bottom of the part containing sandstones and conglomerates. Thus the lower third of the middle Keweenawan is essentially igneous and the upper two-thirds consists of igneous and sedimentary rocks. At Mamainse, Pointe aux Mines, and Cape Choyye the lower Keweenawan beds are con- glomerates and sandstones. At Mamainse these basal beds of sandstone, according to Mac- farlane," seem to have a very considerable thickness. At Pointe aux Mines, according to Logan,* there are sandstones at the base of the series nearly in contact with the gneiss. At Cape Choyye the basal bed is a red sandstone of considerable thickness. However, at Cape Gargantua and at Batchewanung Bay the amygdaloidal trap rests unconformably upon the Archean, and thus at these points igneous rocks are at the lowest horizon of the Keweenawan series. Thus for eastern Lake Superior the Keweenawan may be divided into lower Keweenawan and middle Keweenawan, the former being represented by the sediments at the bottom of the series and the latter by the lavas and interstratified sediments. The dips at Mamainse are 20° to 30° lakewartl, and fi'om these amounts on the east coast they range up to 60°, as at Gros Cap. In general direction the strike of the strata of the Kewee- nawan of the east coast curves in and out, corresponding to the minor folds of the synchnorium, but the average strike is somewhat west of north, corresponding with the general direction of the east coast, and the dips are to the west, varying from as low as 10° at Cape Choyye to as high as 45° or even 60° at Gros Cap. The usual dips, however, run between 20° and 35°. From the general relation of the Cambrian sandstone (Sault Ste. Marie, "Eastern" or Pots- dam sandstone of several wTitcrs) and its extensions adjacent to the Keweenawan, Logan con- cluded that there was an unconformity between the two. He says:'' The contrast between the general moderate dips of these sandstones and the higher inclination of the igneous strata at Gargantua, Mamainse, and Gros Cap, combined with the fact that the sandstones always keep to the lake side of these, while none of the many dikes which cut the trappean strata, it is believed, are known to intersect the sand- stones (at any rate on the Canadian side of the lake), seems to support the suspicion that the sandstones may overlie unconformably those rocks which, associated with the trap, constitute the copper-bearing series. GENEEAL CONSIDEEATION OF THE KEWEENAWAN SERIES. LOWER KEWEENAWAN. In reference to the lower Keweenawan, it need here only be remarked that these sediments are in no way pecuHar. They are derived from the preexisting Huronian and Archean precisely as similar detrital formations are built up. At the bottom are conglomerates ; over these lie sandstones; and in the Black and Nipigon bay districts above these are interstratified marls, hmestones, shales, and sandstones. a Report of progress from 1863 to 1S06, Geol. Survey Canada, ISGO, p. 134. t> Report of progress to 1803, Geol. Survey Canada, 18G3, p. 82. c Idem. p. 85. 394 GEOLOGY OF THE LAKE SUPERIOR REGION. Thoui^li it is not known that sediments were everywhere deposited at the base of the Kewee- nawan, it is a remarkable fact that in most places where the actual contact between the non- intrusive ])arts of (ho Keweenawan and the next underlyinfj rocks can l>e seen such sediments occur. Those deposits liave their greatest vohnne and widest extent in tlie n-gion about Black and Nipigon bays, where the tliickness is variously estimated from 550 to 1,400 feet. In north- eastern Minnesota, at the base of the series is the Puckwunge conglomerate. In Micliigan, at Black River, at the bottom of the succession is a basal sandstone known to be 300 feet thick, and it may be considerably thicker than this, occupying a part of the unexposed area to the soutli. How far this sandstone extends east and west is not known, as the formations next underlying tlie Keweenawan are not usually exposed. However, the formation is known to be present north of Ironwood and also in sec. 11, T. 45 N., R. 1 W., near Potato River, in Wisconsin, more than 20 miles west of Black River (Michigan). At the latter place the conglomerate and quartzite below the lavas are probably as tliick as at Black River. On the east side of Lake Superior the actual contacts between the pre-Keweenawan and the Keweenawan are found at a number of localities, and at the more extensive of these exposures the lowest formation of the Keweenawan is a conglomerate, although at other locahties the lavas he directly against the gneiss. Where the lowest Keweenawan rock is an intrusive, as for instance the Duluth gabbro, this must of course be excluded from all consid<>ration in connection with the oldest formation of the Keweenawan. Also there must be excluded from consideration the localities, such as Keweenaw Point and western Wisconsin, where the base of the Keweenawan is not exposed. MIDDLE KEWEENAWAN. The middle Keweenawan was the great epoch of combined igneous and aqueous activities. There are two divisions of its rocks — original igneous and derived sedimentary. IGNEOUS ROCKS. VARIETIES. The igneous rocks constitute a province of rather remarkable uniformity. The different kinds anil their relations are substantially the same in each of the important districts. Chem- ically the igneous rocks include basic, acidic, and intermediate varieties. The basic materials overwhelmingly dominate, the acidic rocks are considerable in quantity, and the intermediate rocks are few and local. Each variety of rocks includes both intrusive and extrusive facies, so that the basic, acidic, and intermediate gi'oups all have textures characteristic for plutonic and volcanic rocks. Barring the work of KIoos and Streng," which was limitetl in scope, Pum- pelly* made the first careful petrographic study of the Keweenawan rocks. In general Irving' followed PumpeUy in the use of terms, but his studies were more extensive and disclosed new variations. According to Irving, the basic plutonic igneous rocks comprise olivinitic and nonolivinitic gabbros, olivinitic and nonolivinitic diabases, and "anorthite rock." Tiie surface varieties include melaphyres, porphyrites, and amygdaloitls. The coarser-grained melapluTes have often been called dolerites, diabases, or ophites, depending on their texture. The deep-seated phase of the acidic rocks is granite, augitic, or Jiornblendic, and the extrusive phase is made up of porpliyry, cjuart ziferous and nonquartziferous, and felsite. The intermediate rocks occur in subordinate amounts. The most important intrusive phases of them are described by Irving as augite syenites and orthoclase gabbros, and the extrusive varieties as porjiliyrites. The term "trap" is used by Irving in its usual sense to include both basic and intermediate fine-grained rocks. <■ Slreng, A. , and KIoos, J. H. , tjber die krystallinisohcn Gcsteine von Minnesola in Nord-Amerilta: Neiies Jahrb., 1877. ii Pnmpelly, Rapliael, Copper-bearing rocl, 1883; Geology of Wisconsin, vol. 1,1883, p. 340. >■ Wright, F. E., Science, vol. 27, June, IMS, p. 892. . i Winchell, N. n., Proc. Am. Assoc. Adv. Sci., vol. 30, 1881, p. 100. i Wadsworth, M. E., Bull. Geol. and Nat. Hist. Survey Minnesota No. 2, 1887. 398 GEOLOGY OF THE LAKE SUPERIOR REGION. gabbro by Wadswortli may belong to any one of a dozen types as commonly recognized. Never- theless, Wadsworth's names as actually applied in tliis case may be correlated approximately with the names of otlier writers, as shown in the tal)le on page 400. Wadswortli huiorsed Irving's protest against using the distinction between augite and diallage as a basis of rock classification, and yet, like Irving, he used it. He did not discrimi- nate sharply between the ophitic and the poikilitic textures, both of which may be found, sometimes together, m jMimiesota diabases. Bayley," in 1889-1897, described the gabbro batholith of Minnesota in considerable detail and also studietl tlie peripheral phases of the ga})bro. To emphasize the close connection in origin between the peridotite and the gabbro of tlie district, he called the former nonfeldspathic gabbro. Although some of the peripheral phases described by Bayley may be of later date than the gabbro, if we assume that they all belong in the Keweenawan, we fbid that Bayley recognizes not only the augite syenite of Irving, but also a porphyritic equivalent which he calls quartz keratophyre on account of the presence of anorthoclase. He speaks of ohvine-pyroxene aggi'egates which should apparently be correlated with wehrhte, dimite, and pyroxenite. In tlie peripheral phases he finds a texture which he considers somewhat characteristic; it consists of the presence of many rormded grains of the more important constituents inclosed by other minerals. Bayley calls it the granulitic texture. It has been called the contact structure by Salomon and the globular by Fouque. It is well described by the term globular or globuhtic. Grant,' m 1893 and 1894, described gabbro, diabase, granite, and fuie-grained rocks pre- viously called muscovadites in the Minnesota reports. Grant's granite is the equivalent of Irving's augite syenite, later called soda-augite granite by Bayley. (See table on p. 400.) The fuie-grained rocks, called muscovadites, mclude border facies of the gabbro mass of various types, but especially norite, fuie-graiiied gabbro often with hypersthene, ohvuie norite, cordierite norite, etc. Hubbard,<^ in 1898, described various types of the Keweenawan of Keweenaw Point. His melapliyre is cliiefly andesite or basalt; Ms doleritic melaphyre is a coarser basalt or a gabbro; his ophitic nielaphyre is a poikilitic and luster-mottled diabase; and liis porphyrite is cliiefijr andesite and trachyte. Lane,"^ in 1898-1906, described the Keweenawan rocks of Isle Royal and northern Micliigan. His melapliyre porphyrite is the equivalent of Pumpelly's "Ashbed" diabase and Ii-ving's diabase porphyrite. Lane's melapliyre ojjhite is an olivine diabase, luster-mottled by means of poikilitic textures; his doleritic melaphyre is a basalt porphyry. Lane would confine the name diabase to dike rocks. His augite syenite is said to be at least in part an equivalent of Bayley's quartz diabase. He uses the term ophitic in a narrow sense, not justified b}'' the original defi- nition of Michel Levy,* nor by his usage./ He applies it to those luster-mottled rocks in which single pyroxene individuals inclose several plagioclase crystals, usually lath-shaped and irregularly placed. It denotes thus, for Lane, a variety of the poikilitic texture. In its original meaning, still commonly used by many and adopted here, it refers to that texture of a basic igneous rock produced when the plagioclase crystallizes in lath-shaped forms before the pyroxene solidifies. A. N. Winchell,^ in 1900, described in detail a few samples of the Keweenawan rocks of Minnesota. He used the new term jdagioclasite for the rocks previously known usually as anorthosites. a Bayley, W. S., Am. Jour. Sol., 3ci ser., vol. 37, 18S9, p. 54; vol. 39, 1890, p. 273; Bull. U. S. Oeol. Surrey No. 109, 1893; Jour. Geology, vol. 1, 1893, p. 433; vol. 2, 1894, p. S14; vol. 3, 1895, p. 1 ; Mon. U. S. C.eol. Survey, vol. 2S, 1S97, p. ,il9. ft Grant, U. S., Twenty-first Ann. Rept. Geol. and Nat. Hist. Survey Minnesota, 1893, p. 5: Twenty-second Ann. Rept., 1894, p. 70. I- lluhbani, L. L., Oeol. Survey Michigan, vol. (1, pt. 2, 1898. d Lane, A. C, Geol. Survey Michigan, vol. 6, pt. 1, 1898; Bull. Geol. Soc. America, vol. 14, 1903, pp. 369, 385; Jour. Geology, vol. 12, 1904, p. 83;. Ann. Kept. Geol. Survey Michigan for liKB, 1905, pp. 205, 239; idem for 1904, 1905, p. 113; I'roc. Lake Superior Min. Inst., vol. 12, 1906, p. 85. c Bull, Soc. gSol. I'"rance, vol. 0, 1878. p. 158. / Min(?raIogie micrographuine. 1879, ?1. XXXVI. See also p. 153. Winchell, .\. N., Am. Geologist, vol. 20, 1900, pp. 151 (197), 261, 348. THE KEWEENAWAN SERIES. 399 N. H. Winchell and U. S. Grant" publislied in 1 noo hy far the most complete accounts of tlie peti'ogiaphy of the Keweenawan igneous rocks. Theii- nomenclature varies very little from that commonly in use at present. They described practically all the petrographic types of the Keweenawan ])reviously known and added some half dozen new varieties. They used diorite-porphyrite or diabase-porphyrite to designate more or less ophitic types of andesite porphyry or augite andesite porphyry. They used Wadsworth's name zirkelite for a devitri- fied basalt, basaltic tuff, or tachylyte; devitrified obsidian they called an apobsidian, and a devitriftod rhyolite an apoi-hyolite, as suggested by Bascom. Wadsworth's quartz-biotite dio- rite is called syenite by Grant. It is an intermediate type corresponding to a monzonite. o Winchell, N. H., and Grant, U. S., Final Rept. Geol. and Nat. Hist. Survey Minnesota, vol. 5, 1900. 400 GEOLOGY OF THE LAKE SUPERIOR REGION. Rl i •a (9 i s 2 1 1 i .a •a i w 5 3 2 2 g. i a S o s 1 Pi a 3 § 1 o eg .1 p. S3 2 o 1 o 1 1 05 U M C3 a 0* 2 a B 1 2 5 OS II H i 2 i ■3 ill a a ft 2 S -< 2 1 < ■o li 1 >> o p. 1 2 1 3 O (X 03 1 — .2 o 2 t 1 2 1 i >> Eh o a eg U) o 9 1 >» a a 1 1 S 1 5 2 .£3 II s .1 •3 i 2 ! 2 1 «- g . a 6 2 ^ g| g a a 1 z 1 i «; ■C a O § S g 1 b 9 >> & a 3 li III C ag'S-S sfsg o ft 1 (3 3 2 0) «-. 1 it 3.2 (-. G?0 .2 ■3 2 1 2 •E a >> ft If < c? -.JcJcc ^ciw-v G? woi o ^cJ H C? 1-4 C4 CO o £ 2 CO >» tJ ^ li 1-. 03 « K ^ i p->> 1 ^ b 2S a e; . d H 3-2 j' 1 i +^ SI i4 p & (S Q -^ci s d 1 3 i ' 1 03 00 . 5t C OS 2 ■3 - o ^" 1 2 5 ^" o 9 i S9 o . d ■T- ^s 2 t- s •a-S . ^ c c « >-t^ o t £3 2|s:. Is 1 0) 03 3 oaf li-g. S <§" < <-4« -^oi .1 03 .u J3 1 a g o 1 O. o 1 a 1 a 11 ■a . 1. M a 2 2^ 3 a> Ph 1 1 o o Im t « 3 2 1 ■5 2 i a: a 'C p P '.E a; P X 2 1 1 Oh on n 2 a a "> o 1 o Q ■m 1 s a o'3 P-. fp 3 .1 2 bC 0) P s 2 .2 ■3 C 1 1 a 6 a 2 p £3 6 ' 1 2 Ph Oh e O 0} S X! ta bD CD 9 > o '2 § m 2 1 8 1 •3 i .2 o & sS . lis. 2 1 § 'So < eg s mm "» "33 'ji! :i: tj SS .-H c-i CO S s •§ £ p. .2 a ^ 1^ p. ° £ ■- • x -^ ^£2 TD't. ■1t.2|| x; a i "0 1 i t •c V o £ 1 a 5 P. s 1 1 1 1 1 £ 1 a s 1 1 Ml .3 > s u 00 p 1 1 K a Ph _o c g£ s •i P. p 2; 5 1 if GO -4 01 , i «a ll 52 fee .-"'ct li rH(N li ID t e- o Ph H if •<.2 : a ^c4 is C P o5 is 00a .— 'cicc 1 P 1 & . d p, — ■ C3 OS .1 ■5-3 — (S ii i o n s i Ml ca XJ M 47517°— VOL 52—11- -26 402 GEOLOGY OF THE LAKE SUPERIOR REGION. ■'S> ex S S5, 3 CO s o Is §1 3 a ^1 ^ 1 ^ 3 si =11 CO 2 cm' g '1 "^ o ^1 J Mug -J-— •J- c '^* o ^rt . 0.3 ■ s .^ • a . » 03 .-^ C-o >-;a -;►-) r-;3 .-;h s.< Ke- ==> 5<: «-*; Sn «^ sj, .TiJ=« — O ■ p « 0<3 — cj III i"" Ph o* •o" i« * A fl •ij- ■lia' (D & R, Sa. 1.2 "Wo °%a oW-3 •^ «';d .^ o z f^ Z,''- 0. 1^ T Si IE «^. -51 = W o 2.11=. .iJ s^* ojcc c:j3 ffirf , ^ — (- aj:: P-S.9 ^2 0) u ^ S S t> !« >, 0) ^x ^iai .^i^ pa Sw£ h3 s fe £ fe '^ >s . wtn.n qT CJ ■3^ i3 . s-^ 1>-S .« = .«. aT" og''. i c'- ar»J sic-o 2 So = 3 3^ ^-^'^ 2|g — a SI ^^1 1^ 0'-. « n? C a o? ■ «3i Bn •^ S f- r^ td = . a> !5-g _'-^ ci^: «g K- 00 doT M ^X id ac a?" ■a tj Sf a 22.:: OS ^-^ £^^ «^.^ 2g Sil s s la'"' O 5=^ ^. V V *". me E^" w ts lsgi.2 Is 3« tf D._ tX-J © u • w. xtxa" .li i 5| ft.g «:> — a^S a,e<^ CO £5a -e^ < K THE KEWEENAW AN SERIES. 403 "1 da " OJ =? = § 5S" 2 OS SO "1 • ai,3 o-£x-|S^--fc. >-C- O Q O So of •a ft" o o A nr o ^ <» • +^ = »|-5gao.g o^-|;5go._g.g S25Si2o5&a wo3 si —Co) 03 t- — r o o o.d d o2h a .saas -a o o ^ffl.^o'o C3 .22«Ooa g £ Q cJja =* -S <»2 S £■ a oj .-s Q ="". roo a S a-'O • acj &£+^ So « K « a o~ ■sswgsa a • • o . £!'->■ tJ CO O^ .BJ S=2MO-g ?■ ^^ tt-T « o '^ -■E-h o" .-=0> •- O -*• C3 = a . g 't> a (B a ~~ ^ ^ _~ o .03; a eft: M £: o 3 fe^Sfeg: c5 "^ _r p, . . 00 t3 c,r bc^ 'to 00 ^ CD O) ^ -k^ ■- fe .- • iffc. • £•■- r; o oiS OS -av' p.oft>aj^aciooiHHffi ^<; O^goo^i>gcoo3g ^S==:Da .3oo>73'» ^ 3 S = 2 • " O a a .sat- .^aiMM .< ^ ■^ r- r-" r-*0 2; ■ • - •,'Z>^^i~'^ r..- .55 - >t">>^.a .'■•NiN =^ 00*— ..._--_,i-ap. a a oi2-ii oi ji; j.i: o L.^ a a a is 3 fe.?i°' a— a a a ;! a-i 2 ^ o.5f -CT ..^ a o a< . "3"^ o ^-'^ o o • - . ."^ S tn'o f^o-ag?.;^ 'n;5 >>. a ft --•g&ao a 3 o o "S— ^^ ^ J o» - r"- a o wi '^''<;<;-^ ft. !er division of the Kewee- nawan. The apparent thickness of the entire formation is not less than 19,000 feet. Irving " gives the thickness of the sandstone exposed at Montreal River as 12,000 feet, and 7,000 feet of overl3dng beds are seen near Ashland. Accortling to Irving it is a characteristic feature of this sandstone tjiat quartz is ver\' subordinate. Indeed, in jilaccs it is nearly quart zless. The detritus has therefore been derived tlominantly fi'om the basic igneous rocks and only subordi- nately from the acidic igneous rocks of the Keweenawan, and apparently the pre-Keweenawan rocks have contributed but small amovuits of material. However, Lane* states that pebbles of banded jaspery hematite and other .iron-bearing rocks occur abundantly in the "Outer" conglomerate and further that the detritus of the sandstones themselves is derived predomi- nantly from the Iluronian and Keewatin rocks. Proljably the statements of Irving and Lane were made with different areas in mind, and more ex,tensive studies of the upper Kew cenawan are perhaps necessary in ordjer to make exact general statements concerning the sources of its detritus. As the upper Keweenawan is confined to Michigan and Wisconsm, it, like the middle and lower Keweenawan, fails to be regional in extent, although it has a greater linear and surface extent than the other two divisions. It is probable, however, that the upper Keweenawan origmally occupied a large part of the Lake Superior basin. It is the softest division of the series and was therefore more deeply eroded than the others. At present the area once prob- ably covered by tliis sandstone is occupied by the Cambrian sandstone or the waters of the lake. RELATIONS TO UNDERLYING SERIES. The Keweenawan rests unconformably on all of the lower series with wliich it comes into contact. This unconformity is so perfectly cle^ar for the Archean gneisses that it has been recognized since the days of Logan, "^ that great geologist having noted tliis relation at Granite Island, on the north side of I^ake Superior, and at several points on the east shore of the lake. The Keweenawan has unconformable relations vritli each of the Iluronian liivisions with which it comes into contact, but in earlier days the unconformity between the Keweenawan and the upper Iluronian was not recognized. The relations of the Keweenawan series and the Animikie group have been especialh' studied north of Thunder Bay, and here the Animikie was indurated and yielded well-rounded fragments to the Keweenawan basal conglomerate at many points. Details as to these rela- tions are more fuUy given on pages 207—208. In the Penokee district the Keweenawan extends for many miles along the upper Huronian, and here there is evidence of even a greater erosion interval between the two series than on the north shore. It has been noted that the Duluth gabbro at its bottom is in contact at many places with the Iluronian and with the Archean. Near its bolder, in areas occupied by the rocks of these periods, are numerous dilces and bosses which are identical in chemical composition and even correspond very closely in mineralogical character with the Duluth gabbro. Indeed, some of the masses may be actually connected with the Duluth gabbro. There can scarcely be any doubt that these intrusive rocks in the lower series are of Keweenawan age. The Keweenawan ago of the great dikes and sills of diabase, which are so abundant in the Arumikie group, is scarcely less clear. These ilikes and sills are Identical in their chemical and o Mon. U. S. Geol. Survey, vol. 5, 1883, p. 230. c Logan, W. E., Report oJ progress to 1863, Geol. Survey Canada, 1863, p. "S. 6 Jour. Geology, vol. 15, 1907, p. 090. THE KEWEENAWAN SERIES. 415 mineralogical composition and in their structural and textural characters with those which are found in the Keweenawan itself east of the Animikie at Thunder anil Black bays and west of the Animikie in ]\Iinnesota. Some of the capping diabases of the Nipigon basin may be flows resting unconformably upon lower Keweenawan, Huronian, and Archean rocks. In the Penokee-Gogebic district numerous diabase dikes cut the iron-bearing formation. These have attitudes at right angles to the dips and in chemical composition are like the basic lavas on the overlying Keweenawan traps. It can hardly be doubted that these are the pipes through which the lavas issued. The Animikie group, including the latest Huronian formations, is cut by acidic intrusive rocks which are almost certainly Keweenawan. The largest of these that has been recognized is the Embarrass granite of the Giants Range, the granites south of the Cuyuna district of Minnesota, and the granite intrusive into the Quinnesec schist of northeastern Wisconsin. Dikes of granite are known to cut the Animikie group along the Giants Range. RELATIONS TO OVERLYING SERIES. The lowest fossiliferous Cambrian rocks in the Lake Superior region are of Upper Cambrian age. These rest unconiormably upon the middle Keweenawan in the St. Croix Valley and on the southeast side of Keweenaw Point. In the former locality an actual unconformable contact is observed, but in the latter the relations are complicated by faulting. The middle Keweenawan throughout is considerably tilted, wliile the Upper Cambrian beds are uniformly flat-lying. These facts prove only that the middle Keweenawan is pre-Upper Cambrian. The upper Keweenawan is in contact only with the Lake Superior sandstone (supposedly Upper Cambrian), a red, quartzose sandstone outcropping along the southwest shore of Lake Superior. The feldspathic sandstones and shales of the upper Keweenawan grade conformably up into the red quartzose Lake Superior sandstone. Exposures of the gradation are observed on Fish Creek, on Middle River, and on St. Louis River. The only possible doubt about the gradation is the fact that the feldspathic sandstones and mud-cracked shales have not been absolutely proved to be Keweenawan, although from their character, distribution, and rela- tions to the Keweenawan there is every reason to believe that they are the uppermost Kewee- nawan. At no place are there fragments of the Keweenawan sandstone within the Lake Superior sandstone. Finally, the upper Keweenawan sandstone and the Lake Superior saml- stone are closely related in their deformation, for whUe the upper Keweenawan as a whole is folded, and the Lake Superior sandstone as a whole is flat-lying, along the axis of the synclino- rium in the vicinity of Asliland and eastward, both are tUted. The western Lake Superior sandstone seems to be areally connected with the known Upper Cambrian of the St. CroLx River valley and has been correlated with the Upper Cambrian. However, it is nonfossiliferous, areal continuity with the known Cambrian is not established, and it is entirely possible that the western Lake Superior sandstone as a whole may be older than the Upper Cambrian. If the Lake Superior sandstone is Upper Cambrian, as it is now correlated, then the upper Keweenawan is pre-Upper Cambrian. In the absence of the Middle and Lower Cambrian, it is difficult decisively to prove that the Keweenawan is pre-Cambrian rather than Middle or Lower Cambrian. It has seemed to us, as it has to Irving," to ChamberUn,'' and, in fact, to most of tlie geologists who have studied this area, that in hthology, lack of fossils, deformation, and separation of the middle Keweenawan from the Upper Cambrian by unconformity the Keweenawan series as a whole is much more closely allied to the pre-Cambrian than to the Cambrian. Another group of geologists, while admitting all these differences, nevertheless hold that the Keweenawan is probably Cambrian. Our reasons for assigning the Keweenawan as a whole to the pre-Cambrian rather than to the Middle or Lower Cambrian are summarized below. While we assume the Upper Cambrian "Irving, R. D., Mod. U. S. Geol. Survey, vol. 5, 1883. tChamberlin, T. C, Bull. U. S. Geol. Survey No. 23, 1885. 416 GEOLOGY OF THE LAKE SUPERIOR REGION. ao-o of tlio Lake Superior sandstone, these conclusions are no^ wiiolly dcnendcnt upon such interpretation of age of the Lake Superior sandstone. The Cajnl)rian is fossihferous-. the Keweenawan is not. The Canihrian is largely a subacpieous deposit; the Keweenawan is largely subaorial. The Cambrian contrasts with the Keweenawan in lacking volcanisni. The known Upper Cambrian is almost flat-lying. The same is true for niost of the Lake Superior sandstone. The Keweenawan as a whole is tilted. In the few localities where the Lake Superior sandstone and upper Keweenawan are tilted together, this may be due partly to movements as late as the Cretaceous. Also, as already noted, there is possible doubt about the Upi)er Cambrian age of the Lake Superior sandstone. It is agreed by all that the known Upper Cambrian rests unconformably upon middle Keweenawan beds. The Cambrian rests upon a peneplain of continental extent, over which the Paleozoic sea swept and deposited Paleozoic sediments, with overlap relations to the pre-Cambrian rocks. This sea did not reach the Lake Superior country until Upper Cambrian time, and parts of Canada were not reached until Ordovician time. If the Keweenawan is Cambrian it constitutes a marked local variation from the general uniform conditions of overlap. The upper Kewee- nawan sediments rest on a plane which cuts the pre-Cambrian peneplain at a considerable angle, as is well shown on Keweenaw Point. (See p. 97.) If the Keweenawan were to be regarded as MidiUe or Lower Cambrian, it would be necessary to conclude that the Middle or Lower Cambrian in this district had taken on remarka])le local characteristics different from those of the Middle and Lower Cambrian elsewhere. On the other hand these local character- istics are accordant with those of the pre-Cambrian rocks of this area. The similarity of lithology and accordance of structure between upper Keweenawan and Cambrian are the natural sequence of transgression of a sea over fiat-lying sediments. The conditions are not different from those that would prevail if the ocean were to transgress to-day from the Gulf of Mexico across the flat-lying and little-consolidated Paleozoic sediments of the upi)cr Mississippi Valley. It would be extremely difflcult to prove the unconformity in any limited area, especially where exposures are not numerous. In fact, it is known that the Lake Superior basin was formed during Keweenawan time, and it is entirely probable that local sedimentation within this basiii would merge upwards into tlie sedimentation from the overlapping Upper Cambrian ocean, while upper Keweenawan beds may locaUy have uncon- formably overlapped the lower-middle member, from whose detritus they are in large part built up. It is concluded that the Keweenawan is mainly pre-Cambrian. Our view of the sequence of deposition is this: The main portion of the Keweenawan was put down in pre-Cambrian time. During and subsequent to its deposition folding developed the Lake Superior basin. In late Keweenawan time erosion of the lower beds near the rim of the basin and deposition of the upper beds within the basin were going on simultaneously. The deposition within the basin continued nearly or quite to the time that the Paleozoic sea, encroaching from the south, reached the basin. The Paleozoic sea then deposited its beds with marked structural discordance upon the lower-middle Keweenawan, and with substantial accordance upon upper Keweenawan beds in parts of the Lake Superior basin in wliich deposi- tion was continuous up to the time of the arrival of this sea. CONDITIONS OF DEPOSITION. The r|uestion now arises as to the i)hysical conditions under whicli the Keweenawan was laid down. According to the standard interpretation the widespread sandstones and con- glomerates at the bottom of the Keweenawan would be taken a,s evidence that at the beginning of Kc'weenawan time this region was submerged. Under this interpretation the occurrence of sandstones and conglomerates between the lavas has been taken as evitlence that the ellusive rocks were largely submarine. The persistence of sedimentary beds such as those that occur at the up])er iiorizons and es|)ecially the "Great" conglomerate of the middle Keweenawan has usually been taken as decisive evidence of this conclusion. However, work by Medhcott and THE KE WEEN A WAN SERIES. 417 Blanford," Walther,'' Passaige/ Davis,'' Huntington/ Johnson/ Barrell/ Chamberliii and Salisbuiy,^ and others has emphasized the importance of continental sedimentary deposits. As yet the criteria for discriminating continental and submarine deposits have not been fully worked out, and therefore there must be considerable uncertainty as to our conclusions upon this matter concerning the Keweenawan, especially as the Keweenawan sediments have never been studied with reference to this particular point. The following evidence we take to favor the terrestrial oi-igin of at least a part of the Keweenawan : 1. The thickness of the sediments. 2. The repetition of conglomerate beds at many horizons through several thousand feet. This would involve too rapid fluctuation of water level for the beds to be satisfactorily explained as aqueous deposits. The continuity of thick beds of conglomerate also is in accord with ter- restrial sedimentation, for subaqueous sedimentation is more likely to develop thick beds over only local areas, as about steep shores. 3. The feldspathic, poorly assorted, and almost completely oxidized character of the Keweenawan sediments, as shown by their prevailing red colors and lack of graphitic material. They also show locally alternating beds of red, yellow, and purple, suggestive of seasonal varia- tions. 4. Many ripple marks in the Freda sandstone are of the horseshoe shape made by rills of water at the surface. These contrast with the ripple marks made by wave action. 5. The fact that except for alterations, the basic flows are in all essential respects like the subaerial basaltic lava flows of Tertiary time. Their upper and lower surfaces are amygdaloidal. Although in places their surfaces have a broken or pseudoconglomerate appearance, they usually lack the peculiar ellipsoitlal structure wliich is "characteristic of the Keewatin and Huronian basic lavas described in another place (pp. 510-512) and which has been shown to be especially characteristic of subaqueous basic lava flows. 6. The fact that the matrix of the basal conglomerate on the north shore is in places a lime- stone, suggesting deposition of evaporation under surface arid or semiarid conditions, as may be observed to-day in the Bighorn Mountains and elsewhere in the West. 7. The lack of fossils. 8. The general contrast with the underlying Huronian sediments, in which evidence of water deposition is faii'ly good. 9. Mud cracks are common in some shales. 10. The rapid alternation of thin beds of coarse unweathered debris with fine red mud- cracked and ripple-marked shales. We are therefore inclined to believe that terrestrial deposition has played an important part in the development of this portion of the Keweenawan, but with the information now avail- able we are unable to say how much of a part it has played. The truth probablj' lies between the two extremes of the subaqueous and subaerial Iiypothe- ses; that is, the Keweenawan lavas and sediments were neither exclusively terrestrial nor exclu- sively subaqueous, though too little is known to warrant definite statements concerning their origin. For the middle and upper Keweenawan it is believed to be largely subaerial, but also in considerable measure subaqueous. When the orogenic movement and the period of volcanism of middle Keweenawan time were well under way it would be very natural that the areas where oMedlicott, H. B., and Blanford, W. T., Geology of India, 2d ed., revised by E. D. Oldham, 1S79, pp. 149-150, 391-458. 1) Walther, Johannes, Das Gesetz der Wiistenljildung, Berlin, 1900. c Passarge, Siegfried, Die Kalahari, Berlin, 1904. d Davis, W. M., The fresh-water Tertiary formations of the Eocky Mountain region: Proc. Am. Acad. Arts and Sci., vol. 35, 1900, pp. 345-373; Bull. Geol. Soc. America, vol. 11, 1900, pp. 590-COl, 603-604; A journey across Turkestan: Carnegie Inst. Washington, Pub. 26, 1905. « Huntington, Ellsworth, Pulse of Asia, 1907. /Johnson, W. D., The High Plains and their utilization: Twenty-first Ann. Eept. U. S. Geol. Survey, pt. 4, 1901, pp. C09-741. 9 Barrell, Joseph, Origin and significance of the Mauch Chunk shale; Bull. Geol. Soc. America, vol. 18, 1907, pp. 449-476; Belations tetween climate and terrestrial deposits: Jour. Geology, vol. 16. 1908. pp. 159-190, 255-295, 363-384. liChamberlin, T. C, and Salisl)ur>-, E. D., Geology, vol. 2, 1906. . 47517°— VOL 52— 11 27 418 GEOLOGY OF THE LAKE SUPERIOR REGION. tlie flexures were large and where the lavas were issuing rapidly, that is, along the border (jf the lake, should be above the water. However, the movement producing the synclinal Imsin would certaiidy make a depression in the center of the lake whicli would naturally be Idled with water. Thus along the borders of the Keweenawan the conditions may have favored terrestrial deposits and in the basin of the lake the conditions may have favored subaqueous deposits, and at the shore zone there were various combinations of the two. If these tentative conclusions are correct, the question still remains open as to whether the water-deposited parts of the Keweenawan were submarine or continental, for deposits laid down in great lakes are usually classed as continental. Wliethcr tJiis basin connected with a sea or was inclosed there is now no means of knowing, unless the possible extension of the Keweenawan into central Minnesota, cited on ])ages 376-379, may indicate such a connection. THICKNESS OF THE KEWEENAWAN ROCKS. In the descriptions of the individual districts the estimated thicknesses of the Keweenawan have been given. Wherever there is a fidl section the estimated thickness is verj' large. For northern Minnesota it is 17,000 or 18,000 feet exclusive of the gabbro laccolith, for northern Wisconsin and Michigan a maximum of 60,000 feet, and for Mamainse, at the east end of Lake Superior, 16,000 feet. Only relatively small parts of these thicknesses are made up by the sediments. There are a number of factors which make all these estimates of very uncertain accuracy. The more important of these factors are faults, intrusive rocks, arid initial dips. It has been seen that during the formation of the Lake Superior syncline strike, dip, and bedding joints and faults were produced, and that some of the strike faults are of great magni- tude. The different conglomerates and lava beds of the middle Keweenawan are very similar litliologically and it is therefore extremely difficult, indeed usually impossible, to recognize the individual beds except those of large size, like the "Great" conglomerate. Hence, it has only been in the vicinity of the mining areas, where studies of the most detailed nature have been made, that the extent of the faulting is appreciated. There can be no doubt that strike faults have repeated the beds at numerous localities. It is to be said that the close studies of Hubbard ° on Keweenaw Point, those of Gordon'' at Black River, those of Lane*^ on Isle Royal, and those of Burwash"* at Michijjicoten have not discovered faults which have repeated the beds of these areas to any considerable extent. It has been seen, however, that the strike fault between the north and south ranges of Keweenaw Point reproduces the lower parts of the rocks of the Keweenawan in the south range. Similarly it is probable that 1)etween Isle Royal and Black and Nipigon bays is a great strike fault which results in the repetition of the Black and Nipigon bays Keweenawan on Isle Royal. In the estimates of the thickness of the Keweenawan the intnisive rocks have been ignored. It is certam that in northern Minnesota the intrusive lavas constitute a considerable proportion of the igneous rocks of the Minnesota coast. Also it is suspected that closer studies will show that the intrusive rocks are more extensive in other areas, as, for instance, at Keweenaw Point, than has been supposed. Indeed, the recent studies of Hubbard" have shown tlus to be true for the acidic rocks, but as yet studies have not been made along the same lines for the basic rocks. In estimating the thiclcness of these rocks no account has been taken of initial dips. It is well known that the initial dips of basic lavas and all coarse conglomerates are in many places higher than 10°, and they may be more than 20°. This statement applies both to sub- aqueous and to subaerial deposits. oHiibbard, L.L., Keweenaw Toiat, with particular reference to the felsites and their associated rocks: Gcol. Survey Michigan, vol. 6, pt. 2, 1S98. l> Gordon, W. C, assisted by A. C. Lane, A. geological section from Bessemer down Black River: Rept. Geol. Survey Michigan for 1906, 1907, pp. 397-507. cLanc. .\. C. OeoloKical report on Isle Royale, Michigan: Geol. Snrvey Michigan, vol. R, pt. 1, 1S9S. li Burwash, E. N., The geology of Michipicotcn Island: Univ. Toronto Studies (Geol. scr.), No. 3, 190S; with map. THE KEWEENAWAN SERIES. 419 _^i1_^^ r '^V;vv:l"v„v^^^^^W ^^ ^'vv???:-\ c' FiGUBE 58.— Diagrammatic section illustrating the assigned change of attitude of a series of beds, like the Keweenawan, from an original depositional inclination (B-C) toa more highly inclined attitude (B'-C), a comparatively simple change. If the beds were laid down horizontally in a sinking basin, as illustrated at the right ( F-G), it is obvious that a greater and a more com- plicated movement would be necessary to bring the Ijcds into the attitude represented in the lower figure at the left, which represents the present attitude of the Keweenawan beds. (After Chamberlin, T. C, and Salisbury, R. D., Geology, vol. 2, 1900, fig. 110.) There thus arises, in connection with the middle Keweenawan especially, the same problem that arises in determming the thiclaiess of a delta deposit, the larger portion of which (the foreset beds) in a great delta has rather steep initial dips. If such a delta coulil be truncated through its central part and the thickness of the beds determined on the basis of dii> it might be calculated that the delta represents many thousands of feet of strata, although as a matter of fact the deposit might not be vertically more than a few hundred feet thick. (See fig. 58.) Plowever, there are reasons for believing that a large angle of dip is due to erogenic movements, and such an angle is sufficient to allow a large thickness. Because of the factors named above it is extremely probable that aU the esti- mates of the thickness of the Keweenawan based on appearances are excessive. To what extent they are ex- cessive is a matter of con- jecture, but we suspect that the vertical thickness of the Keweenawan at the tune it was formed was probably not more than half and possibly only a third of the apparent thickness. AREAS OF KEWEENAWAN ROCKS. The areas of the different phases of the Keweenawan in square miles are as follows: North shore: Basic intrusive rocks 2, 170 Acidic intrusive rocks 550 Basic extrusive rocks 1, 950 4, 670 Sediments 752 5, 422 South shore: Basic intrusive rocks 95 Acidic intrusive rocks 145 Basic extrusive rocks 4, 500 4, 740 Sediments ' 2, 070 6, 810 East shore: Basic extrusive rocks 145 Grand total 12, 377 Total area of basic intrusive rocks 2, 265 Total area of acidic intrusive rocks 695 Total area of basic extrusive rocks 6, 595 Total area of sediments 2, 822 VOLUME OF KEWEENAWAN ROCKS. From the foregoing figures of tliicltncss and area it is apparent tliat the volume of the Keweenawan rocks is very large. For the extrusive rocks an area of 6,000 square miles and a thicliness of 4 miles would give a volume of 24,000 cubic miles. For the sediments an area of 2,800 square miles and a thickness of 4 miles would give a volume of 11,200 cubic miles. These figures leave out of account the enormous masses of intrusive rocks. If the gabbro has a circular outline, as indicated by the convex border of Minnesota, and if its southern border is indicated by the Gogebic district, the diameter would be about 100 miles. With the 420 GEOLOGY OF THE LAKE SUPERIOR REGION. ratio of thiclmcss to diameter given by Gilbert " for the Henry Mountains the maximum tliick- ness would be 15 miles. On calculating the thickness in another way, by assuming an average dip of 10° for a distance of 50 miles on the north shore, si maximum tluckncss of 8t miles is obtained. With a thickness of 8^^ miles at the center and a diameter of 100 miles approximately 30,000 cubic miles may be figured for these intrusive rocks. Althougii these figures merit little consideration as actual measurements, it is beheved that they are of value in showing the enormous donunance in volume of tiie igneous rocks over the sediments and of the mtrusive igneous rocks over the extrusive igneous rocks. Reduced to terms of mass, these figures would be somewhat changed, but the essential conclusions would not be altered. LENGTH or KEWEENAWAN TIME. Because of the facts discussed in the foregoing section on thickness it is of course impossible to give any estimate of the time involved in the deposition of the Keweenawan series, but allowing a wide margin for overestimates of thickness we can hardly escape the conclusion that the Keweenawan probably required as long a time for its formation as the average geologic period, such as the Silurian, Devonian, and Carboniferous, and it may have been as long as the Cambrian. JOINTING AND FAULTING. Commonly, where the dip of the lava beds is considerable, the beds are cut by two sets of joints, one of strike joints and the other of dip joints. Both sets are approximately at right angles to the beds, but the plane of the strike joints contains or does not vary greatly from the line of strike, and the plane of the dip joints contains or does not vary greath" from the line of dip. These positions for the joints have been noticed by Grant ^ for northern Wisconsin and by Hubbard '^ for northern Michigan. In many places there are also joints parallel to the beds or between them, and these may be called bedding jomts. Where the intrusive rocks have dis- turbed the lava beds the jomtmg is very much less regular. As would be expected in a fractured series of rocks, there is also somewhat extensive faulting. Indeed, faulting has been discovered in almost every locahty where close studies have been miade, but usually the greater number of the faults are not of sufficient magnitude to be an important factor in the stratigraphy. Like the joints, the common faults may be divided into strike faults and dip faults, there being a general correspondence between the planes of the faults and those of the joints. Most of the dip faults have no great throw, although locally the displacement may be very considerable. A beautiful illustration of the dip faults is fur- nished by Hubbard '' for the West Pond area on the south side of Keweenaw Pomt. (See p. 383.) F. E. Wright's detailed mapping of the Porcupine Mountains and vicinity* discloses a large number of both strike and dip faults. Some of the strike faults are of great magnitude and extent. The greatest of these known is that at the southeast side of the Keweenawan series, extending from the end of Keweenaw Point along the border of the Keweenawan to Gogebic Lake. Another great strike fault is known in Douglas Coimty, m northern Wisconsin, and in Minnesota along the northern border of the Keweenawan. Both of these faults are at the contacts of the Keweenawan and the Lake Superior sandstone, a,nd it is beheved that the newer series represents the downthrow side. If so, this downthrow was to the south of the Keweenawan at Keweenaw Point and to the north of it in Douglas County. The latter fault plane dips 38° to 45° S. and in Wisconsin at least has aspects of an overthrust fault. Martin (see pp. 112-115) concludes on physiographic groimds that there is a fault along tiie Minnesota coast havmg a throw of at least 1,000 feet. There is notliing to show that the throw a Gilbert, G. K., The geology o( the Uenry Mountains, 2d ed.: U. S. Oeog. and Geol. Survey Rocky Mtn. Region, 1880, p. 55. 6Granl, U. S., Preliminary report on the copper-bearing rocks of Douglas County, Wis.: Bull. Wisconsin Geol. and Kat. Hist. Survey Xo. 6, 2ded., 1901, p. 21. cHubbard, L. L., Keweenaw Point, with particular reference to the felsites and their associated rocks: Geol. Survey Michigan, vol. 6, pt. 2, 1898, pp. 19, 2>1, 35. didem, pp. 87, 91. « Ann. Rept. Geol. Survey Michigan for 1908, 1909, PI. I. THE KEWEENAWAN SERIES. 421 is not much greater than tliis amount. The evidence given by Martin confirms what was before a behef as to the existence of this fault, based on the fact that if there were not such a fault between Isle Royal and the mainland, repeating the beds, it would be necessary to accept an almost iiicredible tliickness for the Keweenawan. The faults in the zone between Isle Royal and the Miimesota coast are probably an extension of that in Douglas County, Wis., or, if not, they accomphsh for the jVIumesota area correspomling adjustment of the Keweenawan during deformation. Just as there are bedcUng joints there are also bedding faults. These are especially likely to occur between the diiferent beds of lava or of lava and conglomerate. In many of them the dip is slightly steeper than the beddmg. The direction of movement along these beddmg faults may be parallel to the strike, parallel to the dip, or at any angle between them. Although this is true, it woukl be natural to expect that the most common movement along the beddmg faults would be approximately parallel to the dip, this being the natural direction of differential movement between beds in a folded series. As to the direction of movement along the dip, by differential movement in a fold of ordinary magnitude the higher bed moves upward as compared with the lower bed, but it is far from certain that tliis rule would hold in a great simple syn- clinorium like that of Lake Superior. It might be that gravity would be more important than the strength of the beds and that the upper members woidd move downward as compared with the lower. Hubbard " and Lane * conclude from their close study of the Keweenawan district that bedduag faultmg or slide faulting is very common. Hubbard finds that at least one slide fault substantially parallel to the dip has a very large movement. Lane'' says that many of the shdo faults have a slightly steeper hade than the dip. The details of these occurrences are given in the section on Keweenaw Point (p. 383). Along any of the faults there may be slickensides or even brecciation. Such brecciation is especially prevalent at the bedding faults, wliich follow an amygdaloidal lava surface, one of their most common positions, because the amygdaloidal belts are planes of weakness. It will be seen on pages 575-576 that the several classes of fractures and faults have a very important bearing on the development of ore bodies. The time of the fracturmg is partly contemporaneous with the folding of the series and partly later; how much later is not known. Some of the faults, notably the great faults bounding the Keweenawan series on Keweenaw Point and in Douglas County, Wis., are partly post-Cambi'ian. It has been suggested by Wilson"^ and Weidman,'' from work m other areas, that faulting may have affected these rocks as late as Cretaceous time. THE LAKE SUPERIOR SYNCLINAL BASIN. It is little short of certam that the great Lake Superior synclinal basin beg> n to form during middle Keweenawan time. The general character of tlris syncline is admirably exhib- ited in figure 59, from Irving, and by the sections on the general map, Plate I. This synclinal basin is rather remarkable for its simplicity. Indeed only at one place does Irving figure a subordinate fold, that at Porcupine Mountains. The strikes and dips of the rocks show several prominent flexures, however, as, for instance, along St. Croix River of Wisconsin, near Ashland and Clinton Point at the head of Lake Superior, and at Michipicoten Harbor. Later strike faults have considerably modified the syncline. Doubtless future close studies will show that the Lake Superior synclmorium has a greater complexity in detail than has been supposed. Cer- tainly one very important subordmate basin, that of Lake Nipigon, must be attached to the major synclinorium. It is to be remembered that along the main shore line and outer islands of Black and Nipigon bays the middle Keweenawan is found with lakeward dips at angles of nOp. cit., pp. 87-91. t> Lane, A. C, Geology of Keweenaw Point, a l)rief description: Proc. Lake Superior Miu. Inst., vol. 12, 1907, pp. S3-S4. c Geol. Soe. America, winter meeting, December, 1908. d Personal communication. 422 GEOLOGY OF THE LAKE SUPERIOR REGION. ^ s 'A c _o > £ _H rs f/l ^f o 'k !rr § \n •^ TJ ■-1 1 (^ ^ QJ s > b. t; Si r/. o a % s 3 §3=§ -MO ■ ^ .2 K '5 -^ « 2 c ^^! iSS t«l r- ?; o « -2 5 5 a ° S S a P g o .a 3 ^ f II S « E B2 S 5-E| ■3 o c S fj en a* 3 ^ 1.1 .si g p. ■" aj 2-S 2 .^5 a a THE KEWEENAWAN SERIES. 423 8° to 10°. In the peninsulas between Thunder, Black, and Nipigon bays the lower Keweena- wan hes substantially flat. Farther to the north the middle Keweenawan reappears, overlying the lower division \\'ith northern dips. It thus appears that at Black and Nipigon bays there is a subordinate anticlinal arch,' which separates the great synclinal fold of Lake Superior from the subordmate synclinal fold of Lake Nipigon. The latter lake is in a subordinate basin of Keweenawan rocks, just as Lake Superior is in a great basin of that series. Similarly Batchewanung Bay, at the east side of the Lake Superior basin, is a subordinate synclinal fold. A part of the shore is Archean. Inside of this is a fragmentary border of Huronian almost cut away; inside of this a partial border of Keweenawan, and the center of tlic basin is fiUetl wath Cambrian. In short this bay is a miniature of the Lake Superior basin, containing the four great divisions of rocks of the region — the Ai'chean, Huronian, Keweenawan, and Cambrian — in a synclmal basin. It has been seen that in general in any one section the dips are much steeper at the lower horizons than at the higher horizons. It is certain that the present dips at the lower horizons are largely due to the folding wliich formed the Lake Superior basin. To illustrate: The Keweenawan lava flows and sediments north of the Gogebic range have the same dip as the upper Huronian sediments, and therefore the main dips of both must have been produced by orogenic movements. Indeed it is thought probable that in general the major portion of the dips of tlie most steeply' mclined lavas is due to orogenic movements, for the natural position of repose for basalts, such as those of Ivilauea, is with dips of 10° to 18°. It is reasonably certain that if 15° is subtracted from the lakeward dip of the basic lavas the remainder of the dip is due to orogenic movement. The steadily lessening dips of the lavas at liigher horizons are therefore to be largely explained by the progress of the orogenic movement which pro- duced the Lake Superior basin, although they are doubtless in part explained by the natural lessening of the dip toward the center of a syncUnal fold. To Ulustrate again: In the Black River section the dips at the base are from 75° to 78° N., and at the highest strata exposed on the "Outer" conglomerate only 20°. In the Keweenaw Point section the lavas at the south side dip 55° N. and those of the middle division at the north side cUp 25°, and it may be supposed that during the time in wliich the lavas and conglomerates of the middle Keweenawan in this area were built up the synclinal movement had tilted the lower beds 30° as a maximum, but from this amount to obtain the actual tilting there must be subtracted the unknown amount which is due to the normal decrease in dip toward the center of a syncline. Similarly at IMichipicoten, on the northwest side of the synclinorium, the basal beds have a dip of 55° SE., and at the top of the exposed sections on the islands south of Micliipicoten the dip is 14°, a inaximum difference of 41°, wliich may be attributed to orogenic movement during the formation of the middle Keweenawan in this part of the region. The same thing is illustrated at Isle Roj^al, where at the southwest end of the island the dips on the north side are 16° and on the south side 8°, and at the east end of the island the dips on the north side are 26° and on the south side 18°. It thus appears that the decrease in dip from the north to the south side is 8°, without reference to the steepness. Tliis fact strongly suggests that the steeper dips at the northeastern part of the island as compared with the southwestern part are to be explained by greater orogenic movements in that part of the island, and thus gives a confirmation to the suggestion made that the steep dips are mainly due to orogenic movement rather than to the original angle of deposition. The foldmg of the basin was practically complete at the end of Keweenawan time, but in post-Cambrian time and possibly in post-Cretaceous time the region suffered the great strike faulting already noted. METAMORPHISM. For the most part the metamorphism of the Keweenawan igneous rocks is that of the zone of katamorphism. The alterations, fully described byPumpelly" and Irving,* have pro- duced very extensive changes in the lavas, especially those wliich were scoriaceous. The a Pumpelly, Raphael, Metasomatic development of the copper-bearlBg rocks of Lake Superior; Proc. Am. Acad. Arts and Sci., vol. 13. 1878, pp. 253-309. I> Irving, R. D., Mon. U. S. Geol. Survey, vol. 5, 1883. 424 GEOLOGY OF THE LAKE SUPERIOR REGION. important secondary minerals produced in the basic rocks are the zeolites, epidotes, chlorites, calcite, quartz, laumoutito, prclinite, datoUte, etc. Man}' of tlio thin vesicular })eds are largely transformed to these substances and the vesicles have been filled with them, forming amygdules. Although the porous beds are extensively altered, the massive centers of the thick lava flows, the dike rocks, and the sills and laccoliths are ver}' fresh; indeed some of them are almost as little altered as similar rocks of Tertiary age. The felsitc and (|uartz {)orphyries have undergone the usual metasomatic alterations for ancient acidic lavas. The glasses have devitrified. A wide variety of secondary minerals have formed, but they occur usually in such minute particles as to be determinable with difliculty. The alterations of the Keweenawan lavas doubtless began as soon as they were consohdated. The process continued tlirough Keweenawan time and the great erosion period between the Keweenawan and Cambrian, and indeed is still going on. The alterations of the sedimentary rocks vary greatly in degree. The lower and middle Keweenawan sediments are much more changed than those of the upper Keweenawan. In the sandstones and conglomerates interstratified with the lavas the same metasomatic change took place as in the lavas, resulting in the formation of a hke group of secondary minerals. The filling of the openings between the grains and pebbles, strictly analogous to the filling of the openings in the vesicular lavas, has been nearly complete, thus thoroughly indurating the rocks. The cementing materials in the sandstones and conglomerates interstratified with the lavas are much more varied than those of ordinar}^ cementation. It was in these rocks tliat the senior author first noted the secondary enlargement of detrital feldspar. So thoroughly have the clastic materials been cemented that where the rocks have not been weathered fractures commonly pass across both pebbles and matrix. The sandstones are intermediate between sandstones and quartzites in their cementation. Though these sediments are well indurated they certainly are less metamorphosed than similar secUments of the Animikie group. Inasmuch as the conditions since they have been laid dowTi have been practically the same as those that have affected the Animikie beds, upon which they rest, this difl'erence in metamorpliism confirms the conclusion as to a considerable time break between the two series. The cementation in the sandstones of the upper Keweenawan has not proceeded so far as m the detrital rocks of the middle tlivision. Indeed these sanilstones are very similar to those of Cambrian age. The individual particles of these sandstones, bemg largely basic, are usually much altered, but it is difficult to say what part of these changes have taken place since they were deposited as sandstones and what part took j^lace before they were broken from the lavas from which they came. The segregation producing copper ores was an incident of the metasomatic changes above summarized, and the details of it are considered in another place (pp. 580 et seq.) The intrusive rocks, especially the great basal gabbros, and the large masses of acidic rock, as has been noted in another place (p. 411), produced profound anamorphic changes in the pre-Keweenawan rocks which they cut. It is believed that later studies will show that in con- nection with the deep-seated bathohths of Minnesota and Wisconsin anamorphic changes will be found in the intruded Keweenawan lavas and sediments, but as yet studies have not been made along the border of the gabbros in order to ascertain whether or not this conjecture is correct. Tliis suggestion gains much probability from the fact that along the borders of the much smaller laccolith of Black River in Michigan F. E. Wright has found the intruded Keweenawan lavas and sediments to be greatly metamorphosed. RESUME OF KEWEENAWAN HISTORY. From the facts which have been presented we ma^- make the following general statements: After the great epoch of upper Iluronian deposition the Lake Superior region was raised above the sea and was sul)jected to denudation for a long time, duiing which the erosion amounted to thousands of feet. The Keweenawan period was begun by the deposition of sediments, con- sisting of conglomerates, sandstones, shales, and limestones, now found generally at the base of THE KEWEENAWAN SERIES. 425 the known intrusive part of the Keweenawan where it has been looked for. These may be subaeiial deposits. After the deposition of sediments of very moderate tliickness occurred the events of the middle Keweenawan, which especially characterize the series. The chief event was the out- break of regional volcanism in the larger part of the Lake Superior basin. In a large part of the region, and perhaps all of it, igneous rocks practically excluded sedi- ments in the lower portion of the middle Keweenawan. Igneous rocks, with an almost inappreci- able proportion of sediments, constitute the Minnesota coast, the lower eight-ninths of the Eagle River section, nine-tenths of the Portage Lake section, all of the Douglas County range of Wis- consin, all of the 4,000 feet of the Taylors Falls section, more than eleven-twelfths of the section at Black River, and about 4,000 feet, or one-fourth of the section, at Mamainse. It does not follow that the time represented by the sediments may not be as long as or even longer than that represented by the lavas. After the period of dominating volcanism had continued until thousands of feet of lava had been built up, there was a decrease in volcanic activity and the sediments again became of sufficient importance to be recognized in the section. This was the later part of the middle Keweenawan. The change in conditions in the niiddle Keweenawan by wliich the sediments, insignificant in the lower part, became important in the upper j)art is not supposed to have occurred at the same time over the entire Lake Superior basm. Indeed, it seems extremely probable that the change was not simultaneous in all parts of the region. This niay be illustrated by the Portage Lake and Eagle River sections on Keweenaw Pomt. The alterations of notable masses of sediments with the lavas seem to have become important in the Portage Lake section before they did in the Eagle River section, for at Eagle River lavas, to the practical exclusion of sediments, constitute all but the upper 5,000 feet of the middle Keweenawan, whereas at Portage Lake the portion containing sediments is much thicker. As the middle Keweenawan epoch neared its close igneous activity ceased. In northern Michigan the longest cessation of volcanism was marked by the deposition of the "Great" conglomerate, which is locally more than 2,000 feet thick. After this conglomerate was laid down there were further outbreaks of volcanic activity, which resulted in the "Lake Shore" trap. But tlie outbreaks represented by tliis formation were relatively feeble, as is indicated by the fact that the lava beds are separated by conglomerates of considerable thickness. For Michigan tliis "Lake Shore" trap represents the last dying effort of the epoch of regional volcanic activity. Thus middle Keweenawan time witnessed a sudden begmning of volcanic activity, which was dominant for a long time, then intermittent volcanic activity, then total cessation. Evi- dence has been presented which seems to favor the view that the midtUe Keweenawan was deposited largely imder subaerial rather than subaqueous conditions. The present distribution of the middle Keweenawan shows that much, if not all, of the Lake Superior basm must have been covered by volcanic flows, for the igneous material, besides occurring along the rim of the lake, constitutes Isle Royal, Micliipicoten, and Stannard Rock, off Marquette. During middle Keweenawan time there were at least two alternations of basic and acidic rocks, and locally between basic and acidic rocks of the first cycle there were intermediate rocks, as on Keweenaw Point and Isle Royal. Whether these cycles were general for the Keweenawan over the Lake Superior region and whether there were more cycles than two is as yet undetermined. As already stated (p. 410), during middle Keweenawan time, contemporaneous with and followmg the extrusions of the lavas, there were also mtrusions, and these mtrusive rocks are of very great quantitative importance. In many places m the lava series the mtrusions in the form of beds and dikes compose a considerable percentage of the mass. Although the mtrusives to a large extent rose into the middle Keweenawan beds, still greater masses spread out approximately along the contact between the Keweenawan and the lower I'ocks, and also 426 GEOLOGY OF THE LAKE SUPERIOR REGION. between the laj^ers of the lower formations. The vastest intrusive body of this class is the great Duluth laccohth, which extends from Duhith to the international boundary and has a breadth reaclung 30 miles. Another of these great intrusive masses is that at Bad River. The bodies intruded between the beds of the Animikie group are so prominent that they have been called the Logan sills. The so-called crowning overflow of Thunder Cape may fall here. The peculiar topography of the steep cUfFs about Thunder Bay and Pie Island is due largely to these mtrusive flat-lying sills. The acidic rocks intrusive in the lower Keweenawan are also important. Granite bosses of considerable size intrude upper Huronian rocks in central Minne- sota and northeastern Wisconsin. During middle Keweenawan time progressive folding of the Lake Superior basin went on, with the result that the upper beds have a lower dip than the lower ones. Conformably upon the rocks of the middle division were built up the sediments of the upper Keweenawan. These sediments consist, in ascending order, of the "Outer" conglomerate, havmg a maximum thickness of 5,000 feet; the Nonesuch shale, having a maximum tliickness of 500 feet; and the Freda sandstone, having a maximum tliickness of 19,000 feet. As the "Outer" conglomerate hes directly upon the basic lavas and in its main mass is lithologically like the conglomerates interstratified with the lavas there is no reason to suppose that the conditions at the time tliis conglomerate was deposited were in any way different from those prevailuig at the time of the earher conglomerates, except that late m the epoch detritus from pre-Keweenawan rocks appeared. ^Beginning with the Nonesuch shale, the sediments are of a different character from those lower in the Keweenawan series. This formation and the Freda sandstone are largely and in places mainly composed of detritus derived from the basic lavas. ^Vlso, they contain contributions from the Huronian, Keewatin, and Laurentian rocks. This means that by the erosion of the basic lavas, or by tliis cause combined with uplift, the pre-Keweenawan became the subject of attack by atmospheric agents. The relative lack of abundance of material from the acidic lavas may also mean that the volcanic mountains composed of acidic rocks had by late Keweenawan time become so reduced as to yield only a small amount of material. As the change in the nature of the materials of the sediments from those mterstratified with the lavas to the Freda sandstone was gradual, there is no reason to place a break at any defuiite horizon. Volcanic activity gradually died out, orogenic movement and erosion con- tinued, and these afford sufficient explanations for the increasing variety of the detritus of the upper Keweenawan. As the Nonesuch shale and Freda sandstone together are of very great thickness and are made up of fine-grained sediments, there must have been steady and long-continued subsidence of the basin where these formations were deposited. Also, their volume is so great as to indicate steady upfift in some other part of the region, exposing the lavas and other rocks to erosion. The development of the Lake Superior syncline continued to the end of Keweenawan time and w'as then substantially complete. The basm was modified afterwards only by post-Cambrian faulting. Keweenawan sedimentation was largely subaerial, but it may have become subaqueous toward the close of the period in the water-filled Keweenawan syncline and may have ultimately merged into Upper Cambrian subaqueous deposition. CHAPTER XVI. THE PLEISTOCENE By Lawrence IMartix. ' THE GLACIAL EPOCH. PLAN OF PRESENTATION. The statement that the Lake Superior region has been invaded and profoundly modified by a continental glacier or ice sheet docs not require proof. It will- suffice to name some of the locahties in wliich the proofs are found and to describe the glacial phenomena and their effects on the present topography and the Ufe of the region. ° The ice, wliich advanced from two centers, one east and one west of Hudson Bay, in a series of lobes, oscillated so that glacial deposits thought to be of two or more ages were produced. The latest of these are called the deposits of the Wisconsin stage of glaciation and cover the greater part of the area here discussed. In advancing, the ice produced striae, roches moutonnees, cirques, broadened, deepened, and hanging valleys, etc. It transported great quantities of the materials eroded in producing these forms. As the ice melted, these materials were deposited as an overmantle of glacial drift. The drift, which is partly stratified, was formerly known as modified and unmodified drift. Later studies show, however, that the largely unstratified (unmodified) drift, including terminal or recessional moraines, ground moraine, and drumhns, was deposited directly by the ice. The drift deposited by rumiing' water either under or in front of the ice or in standing water is stratified, though not essentially modified, and includes out- wash deposits, lake deposits, loess, kames, eskers, etc. Most of these varieties of drift are found both in the older and in the latest glacial drift, as will be discussed. In all the glaciated area the drainage was greatly modified by the erosion and deposition due to the ice. During deglaciation there was a great series of marginal glacial lakes, the ancestors of the present Great Lakes. Since the glacial period there has been warping in the region, resulting in tilting of the shore Unes of the former lakes. Streams have made sfight modifications of the glacial drift and of the topography of the land. The lake shores, especially those of Lakes Superior and Michigan, are the seat of active work, and in these lakes the detritus carried from the land by the rivers and from the shores by waves and currents is being deposited. ICE ADVANCES. The scratches and grooves upon the ledges in the Lake Superior region aH'ord the ])rincipal evidence of the direction of movement of the glaciers, and the sketch map (fig. 60) is a generaliza- tion based on these marks. It will be seen that in general the ice moved in a series of lobes of which those in the Lake Michigan basin, the Lake Superior basin, and the valley of Red River were the most important, the lobes between these, especially one extending from the highland region of northern Wisconsin, known as the Chippewa-Keweenaw lobe, and one extend- ing from the highland region of northern Minnesota, known as the Rainy Lake lobe, being less extensive. a The author is indebted to Messrs. Frank Leverett aad W. C. Alden, of the United States Geological Survey, who have more recently done detailed work on the glacial features of the south coast of Lake Superior and in eastern Wisconsin, respectively, for critical suggestions concerning this chapter. The author, however, assumes responsibility for any errors in interpretation. 427 428 GEOLOGY OF THE LAKE SUPERIOR REGION. The ice whicli overspread the Lake Superior region came from two principal sources, one in the highlands of eastern Canada, generally called the Labrador glacier, and one in the region west of Hudson Bay, usually kno%\Ti as the Keewatin glacier. It seems probable that fully two-thirds of the ice which covered the Lake Superior basin came from the Labrador glacier. It is supposed, however, that this glacier was not the first to spread over the region, but that the Keewatin glacier, while largely synchronous and confluent with the Labrador glacier, arrived earher and stayed longer, probably advancing over parts of the region formerly covered by lobes of the Labrador glacier after these lobes had retreated to the northeast. WTiether or not the ice advance from the northwest covered all the Lake Superior region is unknown. In the area covered by this monograph the glacial lobes were profoundly affected by the areas of highland and lowland, and, as would naturally be expected, the ice was the thickest and moved fastest in the tleepest depressions; consequently, the Lake Michigan lobe of the Labrador glacier (figs. 4, p. 87, and 60) extended farther south than any of the others, and the FiGVRE GO. — Sketch map showing the glaciation of the Lake Superior region, giving names of lobes and probable directions of ice flow. There may have been an earlier stage with ice advance from the northwest through a large part of tbe area. Green Bay lobe of the Labrador glacier, also having a deep axis of flow, extended nearly as far south as the Lake Micliigan lobe. The Keweenaw and Chippewa lobes of the Labrador glacier, being obhged to advance over the highland region of upper Micliigan and northern Wisconsin, cUd not advance as far south as the lobes to the east, though the Chippewa lobe over- rode the i)art of Keweenaw Peninsula west of Ontonagon River and advanced farth(>r south than the atljaceut Keweenaw lobe. The Lake Superior lobe of the Labrador glacier, turned west- ward by the topography, advanced to the west end of Lake Superior, where it escajied from the confining walls of the rift valley or trough near Duluth and spread out in a much broader lobe (fig. 60), part of which advanced nearly westward in the region south of Leech Lake, probably moAnng southwest in the region of Mille Lacs and swinging round to the south, and even to the southeast in the vicinity of St. Croix Falls. The Rain}' Lake lobe, which seems to have come partly from the Labrador and partly from the Keewatin center, moved south and southwest THE PLEISTOCENE. 429 over the hills of northern Minnesota (fig. 4). The Red River lobe, the principal division of the Keewatin glacier, often referred to as the Minnesota lobe, advanced southward in the valley of Red River. Although these lobes are described and discussed as somewhat separate glaciers, too much emphasis should not be placed on their separate existence. It would naturally be true that as the Labrador glacier advanced from the northeast it would project farthest where the deepest valleys existed and would have reentrants where the hills caused obstruction to free glacial advance. It therefore seems probable that the Lake Michigan and Lake Superior lobes actually did advance independently over the regions described ; but it must also be remem- bered that with farther advance to the south the lobes in the Great Lakes basins and those on the hills would coalesce until the hilly region was completely covered by one confluent ice sheet. For example, after the Lake Superior lobe had advanced westward from Duluth and the Rainy Lake lobe had advanced over the highland area of northern Minnesota and the international boundary, their farther advance would cause these short lobes to become confluent and form one great ice cap. DRIFTLESS AREA. If there was not time enougli for two lobes to become confluent before the retreat of the ice, there would be left between them an area where the soil, the ledges, and the drainage bore no evidence of the glacial advance. Such an area might have been formed in northern Minne- sota if the Lake Superior lobe and the Ramy Lake lobe had never coalesced. They did coalesce, however, but in one small area at the extreme northeastern part of Minnesota, described by N. H. Winchell " and U. S. Grant, * the drift is so thin that, although the topography, striated rock surfaces, and scattered foreign bowlders definitely prove glaciation of the area, the fact that the residual soil has not all been removed and the absence of nearly all glacial deposits have led to the description of the locality as " a possibly driftless area." Part of the Marquette district is an area of very thin drift, as near the Mansfield mine, on Michigamme River. Similar areas of tliin drift are described as occurring in Canada. In western Wisconsin and the adjacent parts of Minnesota and Iowa there is a true driftless area, and this was recognized in 1852 or earher by D. D. Owen" and has been studied and fully described by Chamberlin and Salisbury.'* A portion of the Driftless Area (fig. 68, p. 45.3) is included in the southwestern part of the region described in this report. Recent studies by Weidman^ and by Leverett and Alden are somewhat modifying the ideas previously held as to the shape and boundaries of this area, although the main fact of its existence and the assign- ment of its cause to insufficient time for the reduced supply of ice from the north, retarded by the higlilands, to reach this driftless region still stand approved. RETREATING ICE. The so-called retreat of the ice sheet was not an actual backward motion, the opposite of the forward motion of the advance, but a melting back of the front of the ice sheet. Wliile the front of the continental ice sheet was retreating from tliis region, the highlands first emerged from the ice cover because the ice was thinnest above their tops, and valley glaciers or lobes lingered longest in the valleys because it was there that the ice was thickest and, after tliinnuig by ablation, most protected by the load of soil and stones which it was carrying. Accordingly during the retreat the ice front was always lobate. The lobes in the Lake Michigan and Lake Superior basins were much more extensive than those in the northern Minnesota and northern Wisconsin liiglilands, as the glacial deposits that have been left in the region prove. There were probably slight readvances during the retreat of the ice sheet south of Lake Superior. a Filteenth Aim. Rept. Minnesota Geo!, and Nat. Hist. Survey, 1S87, p. 350. 6 Am. Geologist, vol. 24, 1899, pp. 377-381; Final Rept. Minnesota Geol. and Nat. Hist. Survey, 1899, pp. 421, 437-438. c Geological survey of Wisconsin, Iowa, and Minnesota, 1S52. d Chamberlin, T. C, and Salisbiu-y, R. D., Tlie driftless area of the upper Mississippi Valley: Sixth Ann. Rept. U, S. Geol. Survey, 1884, pp. 199-322. « Bull. Wisconsin Geol. and Nat. Hist. Survey No. 16, 1907, pp. 548-565. 430 GEOLOGY OF THE LAKE SUPERIOR REGION. A study of thcso deposits also suggests that the ice lobe whic^h advanced down the Red River valley, moving southward and southeastward in the area discussed in this monograph, came after the Lake Superior and Lake Micliigan lobes had retreated for some distance," perhaps into the basins of the present lakes. Moreover, glacial grooves and stria; on the ledges seem to show the same thing. In the St. Croix Dalles region glacial scratches on the rock are associated with the deposits made by the Lake Superior lobe of the Labrador glacier in such a way as to suo'i-est that they wore made during a first glacial advance, while striations associated with overlying glacial deposits made by the Red River lobe of the Keewatin glacier differ in direction and were probably made after the first set.*" The relation of moraines of red and of gray drift near the south boundary of the upper peninsula of Michigan, west of Crystal Falls, suggested the possibility to I. C. Russell <^ that the Chippewa (or Keweenaw) lobe of the Superior glacier was still advancing after the Green Bay lobe of the Lake Michigan glacier had partly retired from the- area. That there were slight rcadvances of the ice during its general recession is indicated in several places, as in eastern Wisconsin, where red till moraines of the Green Bay and Lake Michigan lobes overlie the earlier moraines of the Wisconsm glaciation. Certain stages of the marginal glacial lakes discussed later also indicate a halt in Lake Michigan in the latitude of Manistee and a subsequent slight readvance. These readvances durmg the deglaciation of the region, however, do not seem to have been very many or very great, so far as the preliminary studies thus far made give evidence. CONTRASTED GENERAL EFFECTS OF GLACIATION. In general the glacial invasion stripped the peneplain of its soil in the area north of Lake Superior, while south of the lake, in the highland region of northern Wisconsin, it removed the soil but left a heavy mantle of glacial deposits. Nevertheless, tliroughout this area the influence of glaciation on topography was minor, while the effects on soU, drainage, forests, and the subsequent pursuits of man were most profound. WTiat was a hill in this upland area north of the lake before the glacial advance is still a hUl; what was a vaUey is almost without exception still a valley, but it may be marsh or lake, or stony soil, and so useless for agriculture. It may have had a fertile soil before glaciation, or may have contained some evidence of an adjacent body of u-on ore, and this the glacier has taken away, leavmg as compensation per- haps a sandy soil supportmg a splendid pine forest, possibly a ledge from which the location of the ore body may be inferred, perhaps only a clogged valley, a chain of lakes, and broad, loiter- mg stream courses along which the prospector or geologist may travel by canoe, and so reach regions of mineral wealth that otherwise might have lain hidden to this day. Quite in con- trast to tliis pre-Cambrian area, the horizontal Cambrian rocks of the south shore of Lake Superior near Duluth and Ashland and eastward from Marquette to Sault Ste. Marie, the belted plain of Wisconsin and Michigan, and the flat-lymg Cretaceous deposits of east-central Minnesota are deejDly obscured by glacial drift. Throughout nearly all these areas the rocks were so readily abraded by the ice and the hills were so little higher than the adjacent valleys that the glacial deposits have entu-ely covered the preglacial topography ami molded a new topography of their own. Moreover, the draming of the glacial lakes which occupied the basin of the present Lake Superior and overlapped its shores has permitted streams to produce a peculiar topograpliy of scidptured lake clays.'* DESTRUCTIVE WORK OF THE GLACIERS. Removal of weathered rock. — Glacial erosion removed quantities of weathered rock, includ- ing the nonrcsistant iron ores, perhaps truncating the iron-bearing roclvs to a lower level ui Canada than in tlie United States, and hence maldng the Canadian mines less productive, as a Weidman, Samuel, Bull. Wisconsin Geol. and Nat. Hist. Survey No. 16, 1907, fig. 21, p. 434, and map in pocket. Berkey, C. P., Jour. Geology, vol. 13, ISIII.'i. pp. .15, 39. 6 rhamberlin, R. T., Jour. Geology, vol. 13, 190), pp. 249-251. c Ann. Kept. Cieol. Survey Michigan for 1906, 1907, pp. 47-52. d Irving, R. D., Geology ot Wisconsin, 1873-1879, vol. 3, 1880, p. 69. b THE PLEISTOCENE. 431 Van Hise° has suggested. Lawson,* however, brmgs evidence to show tliat there was no very material reduction of level. Striee and roches moutonnees. — The scratches or strias and the smoothly polished surfaces which were made by the ice advancing over the region are to be found throughout the Lake Superior region wherever there are ledges of hard rock which will preserve them. On the Archean and Algonkian ledges these striae are exceeduagly common. The advance of the glaciers over these ledges modified them, producmg the rountled forms known as roches mouton- nees. Some of these have longer axes in the direction of ice movement and steep or even pre- cipitous slopes on the lee side, due to a process called plucking, in which large blocks of ice are rasped or torn away by the glacier. In the pre-Cambrian areas these roches moutonnees are exceedingly common, although in the main rather low and not very prominent. • Broadened and deepened, valleys. — In certain favorable localities, either where the ice flow is very strong or where the rock is exceptionally weak, glaciers broaden and deepen their valleys. Clements'' suggests the possibility that in the Vermilion district "glacial erosion was also active in widening and deepenmg these preglacial valleys, changing V-shaped into U-shaped valleys." The overdeepening of certam parts of the bottoms of valleys results in the production of basins in the solid rock, and these are afterward occupied by lakes. (See PI. VI, p. 118.) The rock basins of this description are very common m the Lake Superior region, and glacial erosion has probably caused the deepenmg of many of the lakes in the granite area of northern Minnesota, where it is possible to go all around the lake shores on ledges, demonstratmg that the lake basins are lower than the surrounding country. Lake Superior was somewhat deepened by glacial erosion at the time when the ice was advancing through it (PI. II, p. 86), and Lake Michigan and Green Bay,'^ like the Wumebago Valley, were also somewhat deepened in this way, although, as previously stated, these depressions must have existed before the glacial ice advanced through them. Glacial erosion also broadened and rounded out the great transverse valley of Portage Lake, which crosses Keweenaw Pomt at Houghton, as well as many other valleys in the region, especially in the more hilly areas. The overdeepening may be seen west of Houghton, where Huron Creek occupies a hanging valley (PI. XXX, B, p. 434). The effect of glacial erosion on the Duluth escarpment northwest of Lake Superior, where Thimder, Black, and Nipigon bays occupy submerged hanging valleys, has already been dis- cussed (p. 114). Glacial rock basins. — The rock-basLn lakes occupymg depressions produced by glacial erosion are numerous in the areas of pre-Cambrian rocks. (See PI. VIII, in pocket.) Their character and origm may be inferred from one specific illustration. In the Michipicoten district a series of lake basins entirely rimmed by rock has been studied by Coleman,^ who concluded that these basins have been formed by chemical action and are not due to glacial erosion. The writer visited the Michipicoten district durmg the summer of 1907 and after a study of these rock basins came to a conclusion different from that of Coleman. For a number of reasons it seems probable that Hematite Mountain, at whose base is the Helen iron mine and one of the rock basins, was the seat of a local glacier that probably came into existence as the ice was advancing over southern Ontario and lingered as the ice sheet was retreating, because of the height" of the hill (1,700 feet). The north and northwest slopes of the hill would receive less sunlight and heat than the south slope and the snow and ice would therefore Imger there longest. The local glacier would naturally be on that side. The shape of the depression in which the Helen mine is situated is such as to suggest that it is a glacial cirque (fig. 61), and the rock basin is of exactly the kind which is made by small glaciers m tlieii' cirques. A a Van nise, C. R., Twenty-flrst Ann. Kept. U. S. Geol. Survey, pt..3, 1901, pp. 411-412. 6 Laws»n. A. C, Bull. Geol. Soc. America, vol. 1, 1890, p. 169. c Clements, J. M., Men. U. S. Geol. Survey, vol. 45, 1903, p. 43. d Winchell, N. H., Am. Jour. Sci., 3tl ser., vol. 2, 1871. pp. 15-19. ' e Coleman, A. P., Rock basins of Helen mine. Michipicoten, Canada: Bull. Geol. Soc. America, vol. 13. 1902, pp. 293-304; Univ. Toronto Studies, 1902, pp. 5-6, 26; Rept. Bur. Mines Ontario, vol. 15, pt. 1, 1906, pp. 187, ISS; Econ. Geology, vol. 1, 1906, p. 522. 432 GEOLOGY OF THE LAKE SUPERIOR REGION. ledge separates it from an adjoining rock basin a little farther down (a normal glacial rock basin relation of which many examples are known) and a rather marked hanging valley (PI. XXIX, ^) connects the depression in which these two lakes are situated with a lower trunk valley in wliich lies still another lake (fig. 61). The existence of this hanging valley indicates glacial erosion in the region. The glacial striae in the upper part of the valley, which occasioned one of Cole- man's difficulties in believing this a glacial rock basin, are oblique to the trend of the valley, as would be natural during the higher stages of the continental glacier, but the lowcrstriajrun in the proper direction for the later stages of a local glacier. Ice would naturally excavate HefTjatitc Mtn Glacial rock basins I700 Hanging SnvrsZ- Grade .of main valley to which hanging valley is tributary valley ^ ^-■ISSS'' Talbot Lake BOO' 1000 2000 3000 -WO O FEET FiGUKE 61.— Sketch showing the glacial cirque, the rock basins, and the hanging valley near the Helen mine, Michipieoten. along the zone of weak iron-bearing rocks, which were possibly somewhat prepared for the excavation by chemical action of the sort that Coleman suggests. ° The real crux of the determination of these lake basins as of chemical or glacial origin lies in the fact that the iron ore remaining in the basins is found in just that locality where a small glacier in a cirque would protect it, although removing the rest of the iron ore, whereas if a chemical origin is thought plausible, the selective chemical action in preserving the ore at just this point and removing it elsewhere in the basin must be accounted for. TRANSPORTING WORK OF GLACIERS. It is well established that the deposits carried by the glaciers have been worn by the ice from the .ridges over which the ice sheet advanced and that in any place where glaciers have been the rocks brought by them are apt to be of an entirely difjFerent sort from the ledges which underlie them, although a large part of the material in the drift may be of local derivation. This transportation of foreign material was early observed in tliis region, though explained by Bigsby '' as due to "an earthquake sea wave" or "loaded icebergs." When rocks of a distinctive kind are found in an area where no similar rocks normally occur and the striae indicate that the glaciers moved in the proper direction to carry these rocks, it may be con- sidered demonstrated that glacial ice has moved the material from one place to the other. The early students lacked this conception of moving glaciers. Devonian limestone Math fossils was thus brought into the Micliipicoten district from a locality some 150 miles to the northeast, and iron ore was thus transported in the upper peninsula of Micliigan."^ Cambrian or Silurian limestone pebbles "* from ledges in Manitoba seem to have been brought to the Lake of the Woods region of old crystalhne rocks bj" a later movement of the Keewatin glacier after the chief northeast-southwest movement of the Labrador ice sheet. Many fragments of the granites anil gneisses of the Archean and the por]:)hyrites and quartzites and jaspers of the Lake Superior region were transported by the glaciers and are now foimd in the region of horizontal Paleozoic rocks to the south, fragments of tliis kind coming from both the north and the south shores of Lake Superior. It is sometimes an aid to the iron prospector to study the stones in the glacial drift in order to determine where possible ledges of iron-bearing formations may be found. The most notable case of glacial transportation of iron ore is that of the 30,00()-ton mass south of the Fayal mine, on the Mesabi range, which Leith « describes as being entirely inclosetl in the glacial drift and hence evidently transported bodily from the ledges to the north. <■ Coleman, .\. P., Rept. Bur. Mines Ontario, vol. 8, pt.2, 1899, pp. 156-157 l> Bigsby, J. J., On the erratics of Canada: Quart. Jour. Geol. Soc., vol. 7, 1S51, pp. 215-238. ' c Brooks, T. B., Geol. Survey Michigan, vol. 1, 1873, pp. 76-79. dLawson, A. C.. Gcnl. anrt Nat. Hist. Survey Canada, vol. 1. 1885, p. 132cc. ' Leith, C. K., The Mesabi iron-bfearing district of Minnesota: Men. U . S. Geol. Survey, vol. 43, 1903, p. 263. U. S- GEOLOGICAL SURVEY MONOGRAPH Lll PL. XXIX A. HANGING VALLEY NEAR HELEN MINE, MICHIPIGOTEN. Talbot Lake in foreground. See page 432. B. LAKE CLAY OVERLYING STONY GLACIAL TILL IN MOUNTAIN IRON OPEN PIT, MESABI RANGE, MINN. See page 443. THE PLEISTOCENE. 433 Among the distinctive materials wliich are found in the glacial drift are diamonds and native copper. The copper is of course traceable to the copper-bearing rocks of northern Wisconsin and Michigan and Michipicoten Island, but the source of the diamonds is not loiown." CONSTRUCTIVE WORK OF GLACIERS. GROUND MORAINE. Much of the material carried by the ice sheet is ground finer and finer until it is reduced to clay, and this clay with the included stones of various sizes which were not ground up so fine forms the most widespread of the deposits left by the glaciers. It is generally called till or bowlder clay and was formerly known as unmodified glacial drift. It reached its present position simply by being dropped from the melting ice, and forms the great mantle of ground morame and parts of the ridges of terminal or recessional moraines. The present thickness varies with the former tliickness of the ice, the amount of such debris which was contained in the ice, and the amount of erosion by running water either in connection with the melting ice or subsequently. Tliis glacial till is found with varying tliicknesses in every part of the Lake Superior region, overlymg the Archean, Algonldan, Paleozoic, and Cretaceous rocks, being entirely absent or represented only by scattered stones in some rock ledges, and covering other areas and completely obscuring the bed rock by an overburden 200 to 300 feet thick. The type of topography produced by the glacial till in the ground-moraine areas depends largely on whether enough of it accumulated to bury the preglacial topography or not. Many hills in the glaciated area still have the form of their bed-rock cores or are merely thiidy veneered with the bowlder clay. Many vaUeys also are only partly filled by the tdl (fig. 55, p. 364) and remain as vaUeys, though not now as deep as before the glacial advance. On the other hand, more commonly the topography was so mild before the glacial advance and the accumulation of glacial deposits was so thick that an entirely new topography is modeled by the ice. (See Pis. XI, p. 180, and XXXI, A, p. 436.) This topography is generally of the "moderately rolhng," "undulating or rolling," and "flat or undulating" types described by Warren Upham and others.'' DRUMLINS. A class of till, or unassorted ground moraine, which deserves special mention is the drum- lin. Drumlins in only one or two areas within the field of this report have yet been described, but they doubtless exist at numerous other points. The drumlins of the Lake Superior region are lenticular hills of bowlder clay or till, varying m shape from that of half of an egg that has been bisected lengthwise to that of half of a cigar cut in two in the same way. They character- istically have one rather steep side and one gentle slope, the steep slope being on the side from wliich the ice came. The long axis of the drumlin is invariably parallel to the direction of the latest ice movement. Three areas of drumlins in Micliigan have been described. The first is in the Menominee district,'^ where the drumUns are found over an area of about 150 square miles and have an average height of about 40 feet. The second area is also in the upper peninsula of Michigan, including Les Cheneaux Islands and a portion of the adjokdng mainland on the north shore of Lake Huron.*^ The tliird drumhn area is in the Grand Traverse region,^ in the northM'estern part of the southern peninsula of Michigan. oSalishury, R. D., Notes on the dispersion of drift copper: Trans. Wisconsin Acad. Sci., Arts and Letters, vol. 6, 1SS6, pp. 42-50. Ilobbs, W. H., Emigrant diamonds in America: Ann. Rept. Smitlisonian Inst., 1901, pp. 359-366; Am. Geologist, vol. 16, 1894, pp. 31-35; Jour. Geology, vol. 7, 1899, pp. 375-388. Farrington, O. C, ( orrelation of distribution of copper and diamonds in tbe glacial drift ol the Great Lakes region: Proc. Am. Assoc. Adv. Sci. vol. 58. 190S, p. 288. i> Final Rept. Geol. and Nat. Hist. Survey Minnesota, te.xt accompanying county maps. "•Russell, I. C, The surface geology of portions of Menominee, Dickinson, and Iron counties, Mich.: Ann. Rept. Geol. Survey Michigan for 1906, 1907, pp. 8-91. d Russell, I. C, A geological reconnaissance along the north shore of Lakes Huron and Michigan: .\nn. Rept. Geol. Survey Michigan for 1904, 1905, pp. 39-150. elxverett, Frank, Science, new ser., vol. 21, 1905, p. 220; Water-Supply Paper U. S. Geol. Survey No. 183, 1907, pp. 333-335. 47517°— VOL 52—11 28 434 GEOLOGY OF THE LAKE SUPERIOR REGION. Geologists have, not tliorouglily agreed as to the origin of ch-umlins. Two theories have been held. One holds that the drumlins are constructed under the ice by the accumulation of material there, the material being derived by the ice sheet from the land from wliich it is advancing and the drumlins being built somewhat like bars in a river. The alternate hypoth- esis ascribes drumUns to a dcstnactive action, the ice sheet being supposed to carve drumlins from a preexisting mass of tUl laid down by a previous ice sheet. The drumlins of the first two areas described seem to have been formed by the destructive process, as very decisive evidence by Russell proves, but Leverett tliinks that some of the drumlins in the Grand Traverse region are constructional rather than destructional. ESKERS. Another glacial feature to be described, the esker, is a fossil stream course formed in or under the ice by a stream flowing in a tunnel and depositing its load of sediment, wldch is preserved on the surface as a low winding ridge after the ice has melted away. Eskers in many parts of the Lake Superior region, as in northeastern Miimesota " and the Menominee district,* have been described. Russell describes them as low, serpentine gravel ridges in the valleys between the drumlins. They are doubtless also present in many other areas. They are mentioned here rather than with the other stratified drift dejjosits, like outwash plains, because in this area they are commonly associated with the ground moraine rather than with the outwash of the valleys. TERMINAL MORAINES. The deposit piled up at the end of the ice tongue or lobe is called a terminal moraine, and the name ds applied not only to the deposit made at the farthest advance of the ice but also to those made at any point where the ice halts. The latter are also sometimes called recessional moraines. The only terminal moraines in the Lake Superior region wliich mark the farthest advance of the ice lie aroimd the borders of the Driftless Area, but recessional moraines are more abundant. Some of them, so far as mapped, are shown in figure 68 (p. 453). These recessional moraines may be made up of two rather different kinds of material — the glacial till, or unmodi- fied drift, and the drift which is assorted and stratified by running or standing water. A termi- nal or recessional moraine in the Lake Superior region usually consists of a series of ridges or knolls (PI. XXX, A), in general constituting a long, narrow zone of hilly country, which may be in a single ridge, but is more commonly an irregular belt of ridges and valleys. The charac- teristic terminal moraine is made up largely of laiobs and kettles. The belts of terminal morauie range from several hundred yards to several miles in width but are rarely over 4 or 5 miles wide and generally a mile or less. A great terminal moraine of course indicates that the edge of the ice remained at one point for a considerable length of time. During this time, if the glacier was moving, it would be constantly bringing material up to tliis point, dropping the material there, and perhaps, by slight readvances, shoraig ahead the material wliich had prcA-iously been deposited by the melting ice, and all this material would be subject to constant removal or rearrangement by the running water that issued from the ice as the glacier was melting. These terminal morames are therefore made up of a mixture of unmodified till and stratified sand, gravel, and clay deposited by running water, with variations of the two as the ice may have advanced, or as the water may have cut chamiels in the deposits, or as portions of the ice may have been buried beneath the deposits made by the melting of the upper ice layers or laid down by the streams. The subsequent melting of these buried ice blocks has caused the glacial drift to slump down, forming broad hollows and steep-sided pits. This is the general origin of the kettles which are found in terminal moraines. o Elftman, A. H., Am. Geologist, vol. 21, 1898, p. 97. b Russell, I. C. The siirface geology of portions of Menominee, Dickinson, and Iron counties, Mich. : Ann. Kept. Geol. Survey Michigan for 190(5, 1907, pp. 8-91; .Vm. Geologist, vol. 35, 1905, pp. 177-179; Science, new ser., vol. 21, 1905, pp. 220, 221. (5 IV^o O « THE PLEISTOCENE. 435 KAMES. Karnes, or irregular hummocks of waterworn sand and gravel, are present throughout the morame belts of the Lake Superior region, many of them at the borders of valleys, as if formerly at the margin of an ice sheet whose melting has caused the edges of marginal terraces to slump down into irregular hummocks and kettles. Russell describes irregular hillocks of rounded kame gravels in the Menominee area and ascribes them to accumulation beneath wells, or moulins, in the ice sheet, where streams on or in the glacier fell vertically and deposited their load. BECESSIONAL AND INTERLOBATE MORAINES. The recessional moraines formed at temporary terminal points of the ice sheets during the Wisconsin stage are seen from the map (fig. 68, p. 453) to be definitely related to the larger lowland and highland areas, and it is by a study of these moraines that some of the conclusions as to the behavior of the different ice lobes in the Lake Superior region have been reached. As the ice retreated from the maximum stage of a confluent ice cap and once more resolved itself into lobes, some very distinctive deposits were formed between the adjacent lobes, and these are called interlobate moraines. An example of the moraines of this kind is found in the interlobate (kettle) moraine of eastern Wisconsin, which was accumulated between the Green Bay lobe and the Lake Michigan lobe. Other interlobate moraines were formed between the Chippewa lobe and the Superior lobe in Bayfield and Douglas counties, Wis., west of Ashland, and between the Superior and the Rainy Lake lobes m northeastern Minnesota. DRAINAGE OF DRIFT-COVERED AREAS. The accumulation of till over this great area has modified the drainage, and one of the most prominent effects of this accumulation is the destruction of mature or submature preglacial drainage and the superposition of young drainage on the drift, causing gorges, waterfalls, and the great numbers of lakes and swamps for which the region is noted. (See PI. XXII, in pocket.) These lakes and swamps are due to a common cause — mterference with the free run-off of rain by the irregular deposition of the drift. Among the most common kinds of lakes and swamps or muskegs (PI. XXXI, B) are those wliich are produced by the accumulation of water m shallow depressions in the undulating or mildly irregular till sheet. As the material of the till was largely clay, it would naturally be difficult for the water to escape tlirough it. Another com- mon cause of lakes is the accumulation of a greater thickness of the glacial till in one part of the valley than in another, producing an obstn.iction to drainage. Many of the streams were also forced out of their preglacial courses by the deposits of glacial till, and numerous rapids and waterfalls are due to this cUsplacement. Clements " has described Deer River, Michigan (PI. XXII), as typical of a stream with associated swamps and lakes in a till-covered area and has outHned the life history of such a dramage system. The normal type of preglacial drain- age of the entire Lake Superior region is illustrated in Plate XXXI, A, showing part of the Driftless Area. Plate XXXI, B, shows the young drainage of the glacial drift which now covers the greater part of the region. DIFFERENCES BETWEEN YOUNGER AND OLDER DRIFT. There is evidence in the central United States which has been interpreted as indicating that the glacial period, instead of being simple, was decidedly complex. It is thought that the ice did not advance from the Labrador and Keewatin centers once and retreat once, but that instead it underwent a series of oscillations so that glacial deposits were laid down under or in front of the ice, the ice retreated from them, and then the weathering and erosional agencies acted upon these deposits. This is the reason why the lakes among the older glacial deposits are largely either filled or drained, the till-veneered liillsides are cut by streams, the stones in the drift are weathered and disintegrated, and the soluble constituents have been leached out of the soil by percolating water. After all this had taken place the glaciers are thought to have read- vanced and covered the older drift with a sheet of new till, etc., wliich in some places extends ■■ Clements, J. SI., Mon. U. S. Geol. Surrey, vol. 30, 1899, pp. 32-30; Am. Geologist, vol. 17, 1890, pp. 120-127. 436 GEOLOGY OF THE LAKE SUPERIOR REGION. farther out tliaii tlio older drift and in olliers has left a Ijroad zone of it exposed. Tliis fresh, unweathered, young till forms a decided contrast to the older drift. Only the extreme southwestern part of this region contains any of what has been inter- preted as older drift. In the greater part of the region the drift seems to be solely the work of the Wisconsin ice sheet. The drift near the borders of the Driftless Area has been ascribed to two or three earlier glacial epochs, but most of the Lake Superior region furnishes no evidence whatever of more than one glacial advance, eitiier in the deposits or in the topograph}'. EFFECT OF NUNATAK STAGES ON DISTKIBUTICN OF DRIFT. In spite of the lack of detailed studies in a large part of this region, it seems probable that the behavior of the ice in retreating can be somewhat discriminated, ^^^len an ice sheet covers an irregular land surface, there are two ways in which it may retreat. It may disappear grad- ually from the lowlands and linger longest in' the upland regions, as is the case in the Rockies, in Norway, in Alaska, and in Switzerland to-day. It does this, however, only where the elevated areas are high enough to become centers of local glaciation and to supply new ice. The con- trasting condition is found where the highland areas are not sufficiently elevated to retain snow through the summers and therefore to supply ice. Where the latter condition prevails, the glacier does not continue to be active up to the very time of its extinction, as in the Rocky Mountains at present, but becomes stagnant because there is no fresh supply of ice. When an ice sheet becomes stagnant, the high areas are first exposed by melting, because over them the ice is thinnest, and they rise out of the ice sheet as nunataks. These nunataks gradually increase in size, and eventually the ice shrinks until it is found only in the valleys, where it was thickest. The conditions just described seem to have prevailed in parts of the Lake Superior region. Northwest of Lake Superior the Giants Range was a nunatak (figs. 60 and 6;?), emerging in the interiobate area between the Rainy Lake glacier and the Lake Superior glacier. These lobes gradually retreated to the Lake Superior basin and to the valley of Red River, respectivclvj marguial lakes being formed as described in another section (p. 441). North and northeast of Lake Superior, in Ontario, the conditions maj^ possibly have been similar, the ice shrinking away from an interiobate area near the Height of Land and occupying the basin of Lake Superior largely as a stagnant mass. South of Lake Superior, however, the highland area seems to have had a sonjewhat different history. The ice from the Chippewa and Keweenaw lobes, which advanced over the highland region of northern Wisconsin and somewhat down its southward slope, probably retreated northward over the same slope without the emergence of the northern Wisconsin highland as a nunatak area, although the Porcupine Mountains were probably uncovered as a nunatak region about the time the glacier became lobate in the valle3's east and west of Keweenaw Point. The Huron Mountains aeem also to have first emerged as a nunatak area," lying between the Kewee- naw and Green Bay lobes. Some of the earliest drift deposits were developed about these emerging nunataks. VARIATION OF DEPOSITS WITH SLOPES. When a glacier is retreating — that is, melting back faster than the ice advances, or melting back with no advance, as in a stagnant. ice sheet — two rather different kinds of deposits are made in association with two diverse topographic conditions. One kind is formed where the land slopes away from the ice, allowing a free run-off of the glacial streams which are fed by the melting ice. The other kind is formed where the land slopes toward the ice and the drainage from the ice is detained in a glacial lake imtil it rises to a sudiciently high level to flow over a neighboring divide. The first condition was well exemplified by the t'hippewa-Keweenaw lobe as it retreated from the highland region of northern Wisconsin, when its streams flowed freely awaj-, carrj'ing great quantities of gravel, sand, and cla}'^ that were deposited in outwash plains » Davis, C. \.. N'inth Rept. Michigan Acad. Scl., 1907, pp. 132-135. \:^M*H>'ii> .♦i'.'.V-'" +. X . — r-r .i.\ /j-. .: I > 1 1 X — ^ — -^^ pqi o '"^m D X. 3 iJ THE PLEISTOCENE. 437 or valley trains, a number of which cross the Driftless Area of Wisconsin. At later stages such outwash gravels are likely to be so dissected by stream erosion that terraces are formed at higher levels than the present stream. This is believed to be the origin of the terraces in the valley of Wisconsin River near Wausau, in that of the St. Croix near the Dalles, and along several other stream courses of the region. OUTWASH DEPOSITS. Wlien several streams flowing out side by side build up a broad plain of the same kind as the valley trains, but not confined to a valley, the deposit is called an outwash plain (PI. XXX, A). Outwash plains of this type are found in the Upper Peninsula of Michigan, in Ontario, in Minne- sota, and in northern Wisconsin. Weidman " has described some of them as "alluvium" and believes that these deposits are associated with (a) uplift of the land, rejuvenating the streams and causing intrenchment; (6) lowering of the land, permitting aggradation, during which these so-called alluvial deposits were laid down; and (c) later uplift, permitting reintrenchment of the streams, and terrace cutting. The age of this alluvium he is- inclined to place as perhaps pre-Iowan, between his "Second" and "Third" drift sheets. It may be pointed out that the alluvium is in places directly associated with terminal moraines, and Weidman has not brought forward evidence to show that it extends beneath them or is plowed up by them. After short field studies by the writer it seems more probable that nearly all of this material is normal outwash. In view of some of the most recent conclusions concerning the conditions that determine stream work, it may be conceived that the volume and load of the streams have varied, rather than the grade. The advance of ice sheets, with increased supply of water, would perform the same work of intrenchment as the uplift postulated by Weidman, if indeed this intrenchment is not preglacial. Later the increased load of the streams, supplied with debris from the melting ice, would necessitate aggradation and the formation of outwash deposits, exactly similar to Weidman's alluvium and such as are knowTi in association with existing ice fronts the world over. Still later the diminution of the debris furnished to the streams by melting ice would result in their relief from overloading and in a return to processes of intrenchment and terrace cutting. More than this, Weidman's alluvium, where supposedly overriden by the ice depositing the "Third" drift in the Wisconsin River valley, seems to lack entirely the broad truncation and grooving characteristic of gravels overridden and eroded by ice, as they are known in Alaska. Again, Weidman has not shown that the terraces are gullied or the drift in them weathered and leached as it should be if they are pre-Wisconsin in age. Lastly, if these so-called alluvial deposits are not outwash and mostly of Wisconsin age, it may be asked. What became of the water and debris from the melting Wisconsin ice sheet ? I. C. Russell *> described a series of interesting outwash deposits in the valley of Menominee River (PL XXVI, in pocket). They he in a series of steplike levels associated with moraines, marking recedmg stages of the border of the Green Bay lobe. The angular turns of Menominee River seem also to be related to these receding stages. At Grantsburg, Wis., in the valley of the St. Croix, C. P. Berkey" has studied a series of laminated red and gray clays, judged to have been formed in a glacial lake whose deposits over- lie Wisconsin till. He reaches the conclusion that the clays were derived fi-om the melting of an oscillating ice sheet and estimates a greater length of time than is usually thought of since the retreat of the ice, on the theory that each of the laminae represents a year of melting interrupted by freezmg and supply of finer sediment. He has also compiled an excellent sketch map<^ showing the relation of recessional moraines west and south of the end of Lake Superior in Wisconsin and Muinesota. 1 Weidman, Samuel, Bull. Wisconsin Geol. and Nat. Hist. Survey, vol. 16, 1907, pp. 418-421, 425, 477, 497-498, 5Dl, 504, 506, 514-547, 569-571, 609-610, 622-624. b Ann. Rept. Michigan Geol. Survey for 1906, 1907, p. 65. c Jour. Geology, vol. 13, 1905, pp. 35-44. dldem, flg. 1, p. 43. 438 GEOLOGY OF THE LAKE SUPERIOR REGION. PITTED PLAINS. There is one phase of the btiikling of outwash gravel deposits or valley trains which deserves special mention. In numerous places these gravel deposits arc deeply pitted. Such pits or kettles are well dcvel()])e(l, for example, near Negaunee, m the Marquette Winchell, N. H., Final Kept. Geol. and Nat. Hist, Survey Minnesota, vol. 4, 1899, pp. 2-3, 18-20. cT. B. Taylor (A short history of Oie Great Lakes: Studies ip Indiana geography, 1897, chapter 10, pp. 1-21) has written a review of the various lake stages and the outlets, etc., associated with the different positions of ice fronts and levels of the land. 444 GEOLOGY OF THE LAKE SUPERIOR REGION. consequence of the retreat of the ice barrier would he tliat lower valleys across the hills to the south or east might h(; exposed, and as a result of tliis the waters of the lake would fuid a way out throujfh the new divide and the lake would fall to a new level. The earliest glacial lakes in northern Wisconsin, hkc; tlie predecessor of Lake Gogebic and the great marginal lake in the Ontonagon Valley," probably began to exist before or tluring the Lake Xemadji stage. 25 50 75 100 125 150 MILES Figure C3.— Glacial Lake Nemadji. LAKE DTJLUTH. As the ice retreated northeastward, after the Lake Nemadji stage, it soon retired to a point far enough to the northeast to expose the col now crossed by the Chicago, Minneapolis, St. Paul and Omaha Railway. As a result the outlet near Carlton was abandoned and the waters of this lake outflowed directly southward through the St. Croix to the Mississippi (fig. 64) through a channel ''419 feet above the present Jjake Superior, between the headwaters of the Brule and those of the St. Croix. Exactly where the ice front of the Lake Superior glacier stood at this stage can not be stated, but it probably halted at several points east of the Apostle Islands and perhaps as far east as Keweenaw Point, the other margin resting against the north shore of Lake Superior at several points in Minnesota, smaller marginal lakes being held on each shore between the ice and the land in Mmnesota, Wisconsin, and Michigan. The great glacial lake of this stage is called Lake Duluth,'= although Upham "^ had previously named it the West Superior glarial lake. It is evident that this lake existed for a longtime, and there are three kinds of dejiosits which indicate that this was so. One kmd consists of the elevated beaches which are still found along the liillsides at the level of the St. Croix outlet and which are so broad and well developed on the escarpment face above Duluth that the Boulevard « Lane, A. C, Summary of the surface geolopy of Michigan: Ann. Kept. Geol. Survey Michigan for 19(17. 190.S, pp. Hl-l-l?. & The elevation of this channel is (liven as 1,070 feet by Warren I'pham (Twenty-second .Vnn. Kept. Oeol. and Xat. Hist. .Purvey Minnesota. 1893, p. 55: Final Kept. Geoi. and Nat. Hist. .Purvey Minnesota, vol. 2, 1SS8, pp. (>-I2-G43). The aUitiule of the stunmit in this channel is stated by I.cverett to be 1,021 feet, as shown in a profile in House! Doc. 330. ,Wth Cong., 1st scss., 1890. c Taylor, F. B., Studies in Indiana geography, 1897, fig. 1, p. 10. i Twenty-second Ann. Rept. Geol. and Nat. Hist. Survey Minnesota. 1894. pp. 54-55. THE PLEISTOCENE. 445 Drive follows one or two of them for miles. This aliore can be traced from a point east of Ashland westward to Brnle River and on the other side around the head of the lake to a point some distance east of Dulutli. Similar beaches or terraces in the Lake Superior basin were observed early in the ex])loration of the region " and were explained as wave-wroutrht forms. The second class of deposits indicating that the glacial lake at Duluth existed for a long time comprises the deltas that were l)uilt where streams flowed into the lake at the level of the Boulevartl beaches, as at Thomi)son east of St. Ijouis River, on Tischers Creek, and on Chester Creek at Duluth. » The third class of these deposits consists of the lake clays, which without question accumu- lated in later periods as well as in this, but which would of course have formed to a considerable depth when the ice front stood across the lake and was discharging icebergs with glacial material, and when streams from the hills to the north, south, and west contributed their load of sediment. 25 50 75 100 125 150 MILES Figure 04.— Glacial Lake Duluth. INTERMEDIATE GLACIAL LAKES. As would naturally be expected, with the continued retreat of the Lake Superior and Lake Michigan ice lobes, the lake levels were falling lower and lower. One of the next levels at which there was a notable stand of the ice was when the waters of the western Lake Superior basin escaped past Chicago through Illinois River to the Mississippi. This was probably some time after the early Lake Duluth stage (fig. 65). Whether there were intermediate outlets between the two stages referred to is not known, but it seems probable that the ice in retreating northeastward gradually exposed the highland of northern Wisconsin and Micliigan so that " Logan, W. E., Report on the geology of the north shore of Lake Superior: Geol. Survey Canada, 1847, p. 31. Hubbard, Bela, House Ex. Doc. No. 1, 31st Cong., 1st sess., pt. 3, 1849, pp. 910-911. Foster, J. W., and Wlutney, J. D., Report on the geology and topography of a portion of the Lake Superior land district, vol. 1, 1850, pp. 194-197, 211-213. Desor, E.. idem, vol. 2, 1S51. pp. 248-255, 268-270. Whittlesey, Charles, idem, pp. 270-273. Agassiz, Louis, Lake Superior, 1S50, pp. 00, GO, lOO-lOl, and frontispiece. i Upham, Warren, Twenty-second Ann. Rept. Geol. and Nat. Uist. Survey Minnesota, 1893, pp. C5-06. 446 GEOLOGY OF THE LAKE SUPERIOR REGION. eventually the waters from the enlarged Lake Duluth abandoned the St. Croix outlet for some lower ones in northern Wisconsin and Michigan, and still later outflowed southward along the margin of the ice sheet into Lake Jean Nicolet, in eastern Wisconsin, which drained into Wis- consin and Mississippi rivers. Still later the drainage went into the enlarged Lake Chicago. It is Icnown that there were a number of intermediate stages due either to lowering of the ice barrier, to discovery of lower outlets, or to tilting of the land, because the beaches preserved on the hillsitles below the upper Lake Duluth beach indicate other stands of the lake waters for considerable periods of time. The beaches associated with these intermediate stages are found at several levels below the Boulevard Beach, as shown in the ta])le (p. 4.51). It seems likely that some of the intermediate stages, like the Lake Duluth stage, were of considerable duration, because the beaches that were built are broad, the cliffs that were cut are well marked, and good-sized deltas were formed at the mouths of the streams. Of these deltas that of Dead River at Forestville near Marquette and those of Swedetown and Huron creeks near Houghton are good examples." The fine material carried beyond the deltas into Figure 65.— Hypothetical intermediate stage with the expansion of glacial Lalce Chicago and the later stage of glacial Lake Duluth; part of glacial Lake Agassiz near the northwest corner. Xn isolated stagnant ice block is shown in the Lake Superior basin. the lake formed thick deposits of glacial clays, of which some are now exposed and others are still below lake level. LAKE ALGONQUIN. After the episode of the Chicago outlet the glacial barrier continued to retreat to the north- east, and the glacial lake, which came into existence gradually, occupied all of the basin of the present Lake Superior, its waters covering parts of the peninsula of upper Michigan west of Marquette and being confluent with those in the basins of the present Lakes Michigan and Huron (fig. 66). This is called the Lake Algonqum stage. . At this time the ice barrier stood east of North Bay in the Ottawa Valley, and had retreated from Lake Superior north of the Height of n Lane, A. C, Sunuuary of the surface geology of Michigan: .Vmi. Kept. Michigan Geol. Surrey (or 1907, 190S, p. 142. THE PLEISTOCENE. 447 Land. Possibly there was a stagnant isolated ice block in the Lake Superior basin at this time or just before. During the Lake Algonquin stage, which of course came after a series of inter- mediate stages in which Lakes Chicago and Duluth were enlarged as recorded by the successive beach levels one below the other, the waters deserted the outlet past Chicago to Illinois and Mississippi rivers because lower outlets were uncovered to the east. Lake Algonquin had two such outlets. The first led past Port Huron through the present Lake St. Clair and Lake Erie into glacial Lake Iroquois, which covered more than the basin of the present Lake Ontario; the second outlet also led into Lake Iroquois tlirough the Trent River valley from Georgian Bay. There were several oscillations with one or both of these outlets running during the Algon- quin stage. The Lake Iroquois waters flowed eastward through Mohawk River to Hudson River and New York Harbor. All around the Lake Superior basin the strongest Lake Algon- quin beaches are well-marked shore lines elevated high alcove the waters of the present lake. At this stage glacial lakes probably occupied the Kaministikwia and Nipigon River valleys, including all the basin of the present Lake Nipigon. Figure 66.— Glacial Lake Algonquin. NIPISSING GREAT LAKES. With the continued retreat of the ice sheet to the northeast, a still lower outlet than that tlu'ough Mohawk and Hudson rivers was exposed. This was along the present Lake Nipissing near North Bay and down Ottawa River to the lower St. Lawrence. This is called the stage of the Nipissing Great Lakes. With the uncovering of the Ottawa River outlet the waters of the Lake Superior basin fell to a considerably lower level than that occupied before and accord- ingly regions about the shores of Lake Superior which had been submerged or had groups of islands were wholly uncovered. The largest area of this sort was the lowland east of Mar- quette, in the Upper Peninsula of Michigan (fig. 67). Romiuger, who described the superficial deposits of this region," was somewhat at a loss to explain the mi^iture of ground moraine, reces- o Kominger, Carl, Geol. Survey Michigan, vol. 1 1873, pp. 15-20. 448 GEOLOGY OF THE LAKE SUPERIOR REGION. sional moraines, assorted drift, and lake clay witli wliicli tlie region is covered as a result of its occupation first by ice, then by melting ice fronts, and later by glacial lakes. One notable change was the temporary abandonment of the outlet from Lake Iluion jjast Detroit to Lake Erie. Lake Erie continued to drain into Lake Ontario, which may have been an arm of the sea, while Lakes Superior, IMichigan, and Huron (the Nipissing Great Lakes) drained independently to the Ottawa. Another marked change was the disconnection of the Lake Nipigon basm so that Lake Nipigon at tliis time first assumed somewhat its present form and was independent of Lake Superior. Isle Royal, the site of several small islets at the Algon- quin stage, assumed form as one large island of nearh' its present area. All about the lake shore the waters stood at lower levels. The beaches built at the Nipissing stage seem to be the largest that were formed at any time in the history of the Lake Superior basin. The.se beaclics are so broad and the chfTs cut by the Nipissing waves are so high that it has been inferred that this stage of the lake was continued for a very long time — longer, in fact, to judge from the strength FiGUEE 67.— Part of Nipissing Great Lakes. of the shore lines, than the present level of Lake Superior has been maintained as yet, though postglacial gorges are cut back much farther at the present level than they were at the Nipissing stage. EFFECT OF TILTING ON GLACIAIi LAKES. Up to this point in the histoiy of tlie Lake Superior basin the lake waters fell every time a lower outlet was exposed by the northeastward retreat of the ice sheet. For some time before this there had been going on a broad warping winch was producing an uphft of the region to the north or a sinldng of the region to the south. The e^adence of this disturbance is found in the fact that the beaches of the glacial lakes, wliich must have been originally horizontal in jjosition, for the waters of the lake were hoi-izontal, are now inchned from north to south at a sUglit angle. It was not until after the clo.^e of the Nipissing stage that this war])ing of the lake basin had any very profound efiects, except to produce a fanhke splitting of glacial-lake hhore lines and to THE PLEISTOCENE. 449 cause temporary oscillations in the outlets of the Algonquin and Nipissing stages. During and after the Nipissing stage, however, the tilting became sufficient to bring about a new and rather dramatic change in the history of the glacial lakes. It has been stated that the lake levels had fallen because lower and lower outlets toward the northeast were exposed by the ice sheet (figs. G3-0()). The normal result of such a series of changes would be the establishment of a per- manent outlet of the Great Lakes along the line of greatest depression between the uplands of New England and the Adirondacks on the one hand and the Height of Land of Canada on the other. The Lake Nipissing and Ottawa River outlet was so situated; but after the occupation of this outlet for what may have been a longer time than the present St. Lawrence outlet has been occupied, to judge from the strength of the beaches, as already stated, the uphft of the land toward the north became sufficient to raise the Nipissing-Ottawa Valley to a liigher level than another valley farther south, and the latter valley became the outlet of the Great Lakes. The three upper Great Lakes at this time, instead of draining through Lake Nipissing to Ottawa River or through Trent River and Georgian Bay to Lake Ontario, were once more turned south- ward and drained through Lake St. Clair past the present site of Detroit into Lake Erie, whence the waters of the four upper lakes once more passed over Niagara Falls to Lake Ontario and down the St. Lawrence by the present route. The amount of tilting necessary to accomplisli this result was not very great, although that it was greater than the i)revious tilting is proved by the fact that in places these lower beaches are more liighly inchned than any above them. That it did not affect the whole region is shown by the horizontality of some of the beaches. This tilting has continued up to the present time and is still going on, as is proved by several kinds of evidence. One proof is found in the fact that on the south side of Lake Superior and the other Great Lakes the waters are being canted into bays and river mouths, so that what were formerly valleys are now becoming bays and estuaries (PI. II, p. 86), as noted in northern Wisconsin by the land surveyor G. R. Stuntz" in 1869. In these southern rivers the lake water extends backward far enough to make river navigation possible for some distance, as from Duluth 17 miles up St. Louis River to Fond du Lac; but in all except the largest rivers on the north side of the lake the water cascades down in falls and rapids almost directly into the basin of the lake itself. The lower courses of many rivers on the south side of Lake Superior are so broad that it requires a double line to represent them on the map, whereas on the north side of the lake practically all the rivers are so narrow that they are represented by a single line. Tliis canting of the lake waters into the river valleys on the south side of the lake has had a very important effect in connection with man's occupation of the region, by producing good harbors, and of such harbors that at Duluth and Superior is the best (figs. 69, p. 457, and 70, p. 458; PI. V, A, p. 112), having been protected by the sub.gequent building of great sandbars. To the submergence of old stream valleys during this tilting are due the Apostle Islands, wliich have been briefly described by Whittlesey ' and Irving. "= PKESENT POSITION OF RAISED BEACHES. The effect of the tilting of tliis elevated shore line has been to sid)merge some of the beaches of the former lakes, so that the Nipissing shore line, for example, is elevated many feet above the level of Lake Superior on the north shore of the lake, whereas on the south shore it is now submerged in places b}^ the lake waters. It has been estimated that the shore line of the Nipis- sing stage in Lake Superior is 25 feet below the present water surface at Duluth and that this shore line appears above the present water surface at Beaver Baj-, beyond which it rises with an average slope of about 7 inches to the mile."* Numerous observations and notes on these abandoned strands were made by pioneers in the region. Some of these by Sir William Logan, Foster and Wliitney, Bela Hubbard, a Stuntz, G. R., Some recent geological changes in northeastern Wisconsin: Proc. Am. Assoc. Adv. Sci., vol. 18, 1870, pp. 205-210. & Whittlesey, Charles, Geological sur\-ey of Wisconsin, Iowa, and Minnesota, 1852, pp. 437-438. c Irving, R. D., Geology of Wisconsin, 1873-1879, vol. 3, 1880, pp. 72-76. d Taylor, F. B., Am. Geologist, vol. 15, 1895, p. 307. 47517°— VOL 52— 11 29 . 450 GEOLOGY OF THE LAKE SUPERIOR REGION. W. A. Burt, Agassiz, Desor, Whittlesey, and others have already been alluded to. None of these furnish very specifu; data or contain more than scattered observations. A. C. Lawson," how- ever, made a very painstaking stiuh' and instrumental measurement of these elevated shore lines on the northern shore of Lake Superior, and concluded that these strands were horizontal and wore formed in a great lake, held in bj^ a land barrier that was progressively lowered by warping. He rejected tlie idea of an ice barrier. Subsequently F. B. Taylor ** pointed out that Lawson and also Warren LTpham,<= who supported Lawson's conclusion as to the hori- zontality of these shore lines, though recognizing the glacial-lake condition, had not sufficiently considered the possibility that the sliore lines observed from point to point along the shore of Lake Superior were inclined instead of being horizontal. By field study Taylor demonstrated that the shore Imes Avhich Lawson interpreted as horizontal were indeed inclined at a small angle, and pointed out conclusively that they were formed in a glacial lake wliose barrier was an ice dam to the east."* These raised beaches on the north shore of Lake Superior, especially in the Michi])icoten district, have also been studied by A. B. WiIlmott<^ and by A. P. Coleman,/ who has noted very many more shore lines than were measured by Lawson. Near Lake Nipigon Coleman has also measured many new shore lines, ^ and a number were noted by C. R. Van Ilise and J. M. Clements'' in a trip around northern Lake Superior in 190L Observations on the raised beaches in northern Lake Michigan, Green Bay, and western Lake Huron have been made by Taylor,* Russell,^' Goldthwait,* and others. The writer took a hasty trij) around the north shore of Lake Superior from Duluth to Saidt Ste. Marie in 1907 and visited a number of the localities described by Lawson. Although feeling that Lawson's observations in general were most thorougli and accurate, he believes that the conclusion suggested by Ta^'lor is fully warranted and that at least the lower beaches of this region show a decided tilt to the south and southwest. In evidence of the tilting and the long duration of the Nipissing stage established by Taylor, he found that near Duluth and northward from that city to Beaver Bay the mouths of the small postglacial gorges contain no bed rock but are uniformly either filled with gravel deposits or occupied by the waters of the lake, as at Lester Creek, north of Duluth. Northeast of Beaver Bay most of the small stream valleys are found to have no gorges extending down to or below the present lake level, but instead the streams flow over the bare rock surface of the hillside. An especially good illustration of this is Current River, northeast of Port Arthur, Ontario. Good ev-idence was found that the Nipissing shore line dips under the lake at Beaver Ba}', Minnesota. It has been shown by G. K. Gilbert ' that the canting of the lake basins is still in progress, and his estimate of the rate of tilting is that the north end of a south-southwest line 100 miles long in the Great Lakes region would in a centurj' be tilted 0.42 foot above the south end. This amount of tilting, of course, is small, but it would be sufficient to divert the waters of Lake Superior again, just as they were once diverted from the Nipissing Valley to the St. Law- rence Valley, turning them southward to Chicago River, where the waters would once more flow southward rather than over Niagara and through the St. Lawrence. More recent studies o Sketch of the coastal topography of the north side of Lake Superior; Twentieth Ann. Kept. Geol. and N'at. Hist. Survey Minnesota, 1893, pp. 230-282. Ii The Nipissing Beach on the north Superior shore: Am. Geologist, vol. 15, 1895. pp. 304-314. c Am. Jour. Soi., 3d ser., vol. 49, 1895, p. 7; Twenty-second Ann. Kept. Geol. and Nat. Uist. Survey Minnesota, 1S94, pp. 54-66; Bull. Geol. Soc America, vol. 6, 1895, pp. 21-27. i Taylor, F. ]!., .\m. Geologist, vol. 15, 1895, pp. 304-314; vol. 20, 1897, pp. 111-128. e Rept. Bur. Mines Ontario, vol. 7, 1898, p. 193. / Idem, vol. 8, pt. 2, 1S99, pp. 150-158; vol. 9, 1900, pp. 175-170; vol. 11, 1902, p. 181; vol. 15, pt. 1, lOOC, pp. 193-199. c Idem, vol. 10, pt. 1,1907, p. 135. * Unpublished MS. ■ Taylor, F. B., The abandoned shore lines of Green Bay: Am. Geologist, vol. 13, 1S94, pp. 316-:i27; A rcconnalsfsance of the abandoned shore lines ol the south coast of Lake Superior: Idem, p. 3fi5; The highest old shore line on Mackinac Island: \m. Jour. Sci., 3d ser., vol. 43, 1892, pp. 210-218; The Munuscong Islands: .\m. Geologist, vol. 15, 1895, pp. 24-33; The great ice dams ol Lakes Mauniee, Whittlesey, and Warren: Idem, vol.24. 1S99 pp. (■)-38. J Ru.'iseli, I. C, Ann. Rept. Michigan Geol. Survey, for 1904, 1905, pp. 83-93; idem for 1906, 1907, PI. III. * Goldthwait, J. W.. .\handoned shore lines ol eastern Wisconsin: Bull. AViscon.sln Geol. and Nat. lli-st. Survey No. 17, 1907, pp. 43-119; Jour. Geology, vol.14, 190fi np. 411-124; Bull Illinois Geol, Survey No. 7. 1908, pp. M-6S; lour Geology, vol. ir.. 19I1.S, pp. 4,i9-476. ' Modification ol the Great Lakes by earth movement, Nat. Geog. Mag., vol. 8, 1897, pp. 233-247; Recent earth movement In the Great Lakes region: Eighteenth Ann. Rept. 1'. S. Geol. Survey, pt. 2, 1898, pp. li01-(>47. THE PLEISTOCENE. 451 by J. W. Goldthwait" indicate that the abandoned shore Hues in the southern ))art of the Lake Aiichigan basin are horizontal. The axis of tilting runs south of Green Bay. The effect of the presence of this hinge Ime will be to postpone very much the time before the tilting can be sufficient to divert the drainage of Lake Superior and the other Great Lakes to the Chicago outlet. A series of observations as to the fluctuating level of Lake Sujierior have been made by Capt. J. H. Darling, of the L^nited States engineer office, at Duluth, who comes to the conclu- sion that so far as evidence from two stations nearly on an east-west line, Duluth and Marquette, for eighteen years indicates, there is no adequate proof of a cliange in the level of the present water surface. It seems possible to the writer, however, that this fact of no variation at two points, one almost directly west of the other, would indicate that the axis of tilting runs nearly east and west in the Lake Superior basin, as it seems to run in Lake Michigan. One of the great unsolvetl problems of the glacial-lake history in the Superior and upper Lake Micliigan basins concerns the stages intermediate between Lake Duluth and Lake Cliicago or Lake Algonquin. Between the time of the St. Croix outlet and the Hudson River outlet Lake Duluth must have had an outlet to Lake Chicago through a series of lakes and straits, including Portage Lake on Keweenaw Pomt, the possible marginal channel east of Manjuette in the Au Train and Wliitefish valleys to Green Bay,'' and perhaps a channel through Sturgeon Bay in the Door Peninsula of Wisconsin. Nothing conclusive can yet be said as to tlie halts of the ice front or the time of shifting from the St. Croix outlet to the temporary initial Lake Algonquin outflow past Chicago, a stage which preceded the double outlets to Lake Iroquois and thence to the Hudson. Further observation, however, will settle these questions. Another interesting possibility, at present merely a hypothesis, is that which supposes stagnant ice in the deep eastern part of the Superior basin with retreat southward from the Height of Land, instead of northward toward it as has always been inferred. No evidence known to the writer disproves this possibility and certam unusually high beaches in the Mar- quette and Michipicoten districts suggest it. It is jiossible that this stagnant mass may have become completely detached from the retreating ice sheet. At the beginnmg of this withdrawal marginal lakes of high level were formed in the Micliipicoten district, just as Lakes Omini and Kaministikwia were formed earlier on the northwest side of the lake. The following table shows some of the ju-esent altitudes of the abandoned shore lines, the discrepancies in elevation in the same beach proving the tiltmg and indicating how the warping varied from the earlier to the later stages and from one part of the region to another. Not all the higher isolated beaches are listed, and some of the correlations are tentative. Elevations above Lake Superior {602 feet) of some of the abandoned beaches. Glacial Lake Duluth (or highest early lake recorded). Glacial Lake Algonquin. Nipissing Great Lakes. Satilt stage. Duluth 632-535, 510-515, 470-475 410-415 314(?) 203-315.380 400-4.50 -25 30 60 90 105-110 110-115 Beaver Bay 467-498 482 Port Arthur Nipigon 28 JaokHsh 418 410 Peninsula Harbor 40-45 Michipicoten 728,843.470 315,534,543 Old \v Oman River Root River 212-266 414 412 85-148 10-34 Sault Ste. Marie 49 35 25 Grand Marais Munising ■ Marquette . . . 590 338 260,240.236,3.35 338 Huron Mountains 25 20 L' Anse 590 718 Ontonagon Valley Houghton 410 25 40 Lac La Belle Pnrnnpine Mniintain«! 561 570 510 535 419 470-535 Iron River Maple Ridge Brule-St. Clair outlet Duluth <• Bull. Wisconsin Geol. and Nat. Hist. Survey No. 17, 1907, p. 42; Jour. Geology, vol. 14, 1906, pp. 411^24, vol. 16, 1908, pp. 459-476. t Winchell, N. II., Am. Jour. Sci., 3d ser., vol. 2, 1871, p. 19. 452 GEOLOGY OF THE LAKE SUPERIOR REGION. Elevations above Lake Michigan {580 feel) of some of the abandoned beaches. Glacial Lake Algonquin. Nlpissing Great Lakes. Sault Sle. Marie Detour Oediirviile St. Icnace Miiiiiiscon^ Islands. Mackinac Cooks -Mills KnsiKii Ganlen liluH Fayette Burnt, liluir Escanaha Hlver Gladstone , Ford River Pine Kidf-'e Birch ('reck Rock Island WasIiinf:lon Island. Dcatlis Door Bluff.. p^phraini Egg Harbor Graceport Sturgeon Bay Wilco.v Sawyer Clay Banks Little Suamico Dykesville Cormier Two Rivers 412-434 280 200 170 120 120 130 125 140 120 ino 110 50 99 95 79 62 51 40 61 40 45 30 30 24 20 GLACIAL LAKE DEPOSITS. Tlie deposits laid dowii in the glacial lakes differ from the deposits now being made in the Great Lakes in the rapidity of accumulation and in the character of materials laid dow-p in water which was fed by melting ice and in which icebergs floated. The deposits made in these glacial lakes were predominantly clay, although sands and gravels were laid down near the lake shores. Great thicknesses of these clays were accumulated at the west end of Lake Superior during the Nemadji, Duluth, and Algonquin stages and acquired a prevailing red color by derivation from the Keweenawan rocks. These clays form a distinctly different soil from that found in the region not covered by marginal lakes. Well boruigs near Ashland and Sujjerior, Wis., show thicknesses of 100 to 150 feet or even more of red clay, in places with a little blue clay, generally without any stones, overlying what is reported as sand and "hard])an." the latter possibly glacial till. The total thickness of clay and sand in one boring is 193 feet and in others is over 200 feet. West of Duluth and Superior and extending eastward from Superior on tlie south shore of the jn-esent lake, these thick lake clays, overlying the horizontal Cambrian sandstone, form a plain which ajipears horizontal though sloping imperceptibh" northward." This ])lain has been cut by ])ostglacial streams into a series of rather deep, steep-sided gullies, which necessitate the buikling of a great number of ])ri(lgos l)y the railroads: for oxamjile. the Duluth, South Shore and Atlantic between Ashland and Duluth and the Northern Pacific and Great Northern between Duluth and Carlton, Minii. The highways extending east and west across this region, where the streams generally flow from south to north, are continually going up and down hill in crossmg ridges and valleys. West of Duluth and south of Fond du Lac, ^linn., tlicse gullies are of very great depth, some as deep as 200 feet, so that the railroads swing far southwaixl in order to cross the gullies near their heads, reducing the number and height of the bridges which must be built. The bridges on the Great Northern Railway are in striking contrast with those on the Northern Pacific, both in their number and in their height above the streams, tlic latter railway crossing nearer the headwaters of the streams. The flat ])lain of these clays is not es])ecialh' suited for agriculture and has not been cleared. The clays were covered with timber, but have been devastated by fire and at present constitute a rather deso- late countiT that is traversed in the first hour of the ride from Duhitli to St. Paul. o Grant, U. S., Bull. Wisconsin Geoi. and N'at. Hist. Survey Xo. il, 19U1, p. ti. THE PLEISTOCENE. 453 North of Duliith, as previously indicated, tlie ice retreated southward toward the Lake Superior basin, and between the Mesabi range and Lake Superior the area of flat-lying lower Huronian rocks was the bed of a great glacial lake, called Lake Ui)luuu, which gradually increased in size as the ice retreated, and in which great quantities of clay were accumulated: The inter- ference with drainage in this lake-clay ]>lain has brought about the great prevalence of muskegs along the Duluth, Missabe and Northern Railway, which pursues an almost mathematically straight course for over 25 miles because of the levelness of the lake-bottom plain. Nearly all of tins distance is through muskeg swamps, mterrupted here and there by low gravel ridges, whicli are believed to be portions of recessional moraines built at temporary halts of the ice during this southward retreat and later jtartly submerged by tlie accumulation of lake clay. The bed of glacial Lake Agassiz is similar in nature. THE FOUE PLEISTOCENE PROVINCES. GROUNDS FOR DISTINCTION. In review of the conditions prevaUmg in the Lake Superior region as regards minor topog- raphy and soil, it may be stated that this region includes four distinctive ])rovinces — (1) the Driftless Area, (2) the area of the older drift sheets, (3) the area overlain by tlic till and the 25 50 75 100 125 150 MILES m %. ^ -r:;.v:^^''' ^^ Driftless Old drift Last drift Lake deposits (tv/th known moraines) Figure 08,— Sketch map showing Driftless Area and regions of older drift, last drift, and lake deposits. assorted glacial deposits of the last (late Wisconsin) stage of glaciation, and (4) the area where glacial-lake deposits predominate. (See fig. 68.) These provinces are bounded respectively by the terminus of the outermost of the older drift deposits, by that of the glacial lobes of the Wisconsin stage, b}' the border of the highest shore lines of the great glacial lakes in the Lake Superior and Lake Michigan basins, and by the highest shore of Lake Agassiz. Not all the 454 GEOLOGY OF THE LAKE SUPERIOR REGION. glacial lobes, and by no means all the glacial lakes, were contemporaneous, so the map should not 1)0 understood as roprcscnting conditions that were ])roducpd at any one time. It merely rc])roscnts four groups of areas witlun each of which tlie average conditions are strikingly similar and wliich contrast vnth one another. DRIFTLESS AREA. In the Driftless Area the minor topographic conditions are intimatel_v related to the undcr- Ij^ing rock. The drainage is mature (PI. XXXI, A). The valleys are cut almost entirely' by streams. Resistant rocks make prominent ledges with castellated forms, and weak rocks arc worn to insignificant relief. Waterfalls and rapids in the streams are rare. Lakes are absent. The soil consists of the materials of the underlying rock or of some adjacent material from a source uphill from its ])resent location. It usually grades downward with coarser and coarser fragments to the undecayed ledge from which it has l)een derived by disintegration. It is a typical local or residual soil. AREA OF OLDER DRIFT. The province of older drift includes the regions adjacent to the Driftless Area where deposits were left by one or more of the earlier glacial advances before the Wisconsin. The topography and soil of this proN-ince are contrasted with that of the Driftless Area on one hand and with that of the area of Wisconsin glaciation on the other. The preglacial topography is partly obscured. The valleys are due in part to ineriualities in glacial accumulation as well as to stream cutting. The streams may have rapids or waterfalls, though these are rarer than in the region of latest drift. Lakes are rather rare, and many lakes and swamps have been filled and drained. The glacial topography has slumped down to a softened outline. The soil is distinctly a transported soil, containing foreign fragments quite different in composition from the un fidouard Desor was the first to dcscrihe the Lake Superior shore features ( Foster, J. W., and Whitney, J. D., Geology of the Lake Superior and district, vol. 2, 1851, pp. 2.W-268), as Charles Whittlesey (idem, pp. 270-273) did for Lake Jtichigan. In 1880 R. D. Irving dcscrilwd the coast in the .Vshland region (Geology of the eastern Lake Superior district: Geology of Wisconsin, 1873-1879, vol. 3, 1880, pp. 70-72). I. C. Russell has described some recent changes on the north shores of Lakes Huron and Michigan (.\nn. Rept. Michigan Gcol. Survey for 1904, 1905, pp. 102-105). .V. C. Law- son has described the modern cliffs, beaches, etc., of the north shore of Lake Superior (Twentieth .\nn. Rept. Minnesota Geol. and Nat. Hist. Snr\ey, 1893. pp. 197-230), discussing the shore contours and the coaslal profiles in the various kinds of rocks. G. L. Collie (Bull. Geol. Soc. .\merica, vol. 12, 1901, pp. 197-211'.) has done some work on the modern shore lines of tlie soulh const of Lake Superior in Wisconsin. G. K. Gilbert used many Illustrations from Lake Superior and northern Lake Michigan in his Topographic Features of Lake Shores (Fifth .\nn. Rept. W S. Geol. Survey, 1885, pp. 75-123). c Foster, J. W., and Whitney, J. D., op. cit., pp. 124-129, plates. THE PLEISTOCENE. 457 and caves, as well as isolated stacks and, still farther out in the lake, reefs. The attack of the waves upon Cambrian sandstone, upon the Keweenawan lavas, and upon the Algonkian and Archean rocks has produced different styles of coastal topogra{)hy, and the cHffs cut in the glacial drift are different from all others. On the north shore of Lake Superior the relative position and resistance of certain dikes and sills have modified the shore topography, as was long ao'o described by Agassiz." Logan * carried the idea of coast control by dikes still further — further, indeed, than Irving'^ thought justified. The bold north coast forms a striking scenic contrast to the mild south shore of Lake Superior, as Irving <* lias pointed out. Between the headlands beaches have been formed, and these beaches are of the usual sand and gravel and bowlder type, associated with spits, hooks, bars, and sand dunes. In places where such beaches have been built across the mouths of valleys or bays and separated them from the lake, ponds have been held in, as on the south shore of Lake Superior or the 4 Miles FiGUKE 70.— The present St. Louis River, which has been converted into an estuary by post-Nipissing lilting. of sand spits which have been buHt. The figure also shows the two sets east shore of Lake Micliigan near Grand Traverse Bay, where some very large ponds of this sort are found. Elevated examples of these ponds were observed by C. R. Van Hise and J. M. Clements in 1901 on the north shore of Lake Superior, along the Black Bay coast, form- ing a pecuUar type of lakes associated with the raised beaches." In the Micliipicoten district a bar of this kind was thrown across the bay now occupied by Wawa Lake at the time of one of the higher lake stages, as described by A. P. Coleman.-'' The modern and abandoned beaches, chffs, caves, and skerries on Isle Royal have been described by Lane,? and the older and modern beaches at Pigeon Point, Minn., by Bay ley.'' lAgassiz, Louis, Lake Superior, its physical character, vegetation, and animals, 1850, pp. 420-425. b Logan, W. E., Geology ol Canada, 1803, p. 72. c Irving, R. D., Mon. V. S. Geol. Survey, vol. 5, 1883, pp. 336-337. e furnace, as it is necessary to figure on the phosphorus in the flux and fuel as well as that in the ore itself. THE IRON ORES. 479 recent rapid development of the open-hearth pr>_.pess has allowed shipment of ores higher in phos])horus. The tlevelopment of the basic open-hearth process depends ultimately on the availability of large reserves of non-Bessemer ore, but in turn the develojmient of the open hearth reacts upon and determines the grade of ore shipped from any district or for any period. MINERALOGY OF THE ORES. The iron-ore minerals in general are as follows: Magnetite: Magnetic oxide (FejOi), including titaniferous magnetite. Theoretical iron content of the pure mineral, 72.4 per cent; generally containing some feematite. Hematite: Anhydrous sesquioxide (FCnOj), including specular hematite, red fossil ore, oolitic ore, etc. Theoretical iron content of the pure mineral, 70 per cent. Brown ore: Hydrous sesquioxide (FejOj.nH^O), including turgite, limonite, goethite, or a mixture of these minerals, known locally as brown hematite, bog ore, gossan ore, etc. Theoretical iron content of iron minerals, 59.8 to 66.2 per cent, depending on degree of hydration. Carbonate: Siderite, iron carbonate (PeCOj), known locally as spathic ore, black band ore, etc. Theoretical iron content of the pure mineral, 48.2 per cent. The Lake Superior iron ores are (1) soft, brown, red, slaty, hydrated hematites; (2) soft limonite; (3) hard massive and specular hematites; (4) magnetites; and (5) various gradations between the other classes. The proportions for the entire region of these different classes shipped in 1906, as calculated from average cargo analyses, are as follows: Total production of iron ore in Lake Superior region, by grades, for 1906. Class of ore. Soft brown, red. slaty, hydrated hematite Soft limonite ores Hard massive and specular hematite Magnetite (less than 1 per cent; included with hard ores) 35,652,174 2,741,323 38,393,497 Per cent of total. 93 7 100 The approximate mineral composition of the average ore of the entire region for the years 1906 and 1909, calculated from the average analyses, is as follows: Approximate mineral composition of average iron ore of Lake Superior region for 1906 and 1909. Hematite 1 (more or less hydrated), with some magnetite (SFejOs.HjO). Quartz Kaolin Chlorite (and other ferromagnesian siUcates) Dolomite Apatite (all phosphorus figured as apatite) Miscellaneous 100.00 a The iron minerals may be expressed in terms of hematite and limonite as follows: 1906, hematite 66.60, limonite 22.00; 1909, hematite 66.75, limonite 19.70. These minerals do not, in fact, exist in these proportions, there being a number of hydrates between hematite and limonite. The mineral compositions above given are necessarily only approximate, as ferric and ferrous iron are not separated in the chemical analysis, and water, carbon dioxide, and pos- sibly a small amount of organic matter are all included under loss on ignition. The mineral compositions were calculated from the average analyses, as follows: All phosphorus was figured as apatite; the remaining lime was combined with the proper amount of magnesia and COj to form dolomite; the remaining magnesia was combined with the proper amounts of alumina, silica, and water to form chlorite; the alumina not -used for chlorite was taken with sufficient silica and water to form kaolin; the remaining water, combined with the iron figured as ferric oxide, was figured as hydrated hematite. The proportions of the different minerals for the individual districts calculated in the same way are given in the discussion of these districts. 480 GEOLOGY OF THE LAKE SUPERIOR REGION. In the above table are mentioned the abundant minerals associatetl with the iron, such as quartz, kaolin, and chlorite. Many of the minerals termed miscellaneous ' in the table arc present in small aniounts at a few places. Some of these minerals arc apatite, adularia, wavellite, calcite, dolomite, siderite, pyrite, marcasite, chalcopyrite, tourmaline, masonite, ottreUte, chlorite, mica, garnet, rhodochrosite, manganite, pyrolusite, barite, gypsum, martite, aphrosidcrite, analcite, goethitc, and turgite. Though many of the Lake Superior ores are slightly magnetic, there are only two mines in the region which ship ores classed as magnetite ores, the Republic and Champion, and even these ores are largely specular hematite with considerable quantities of magnetite. There are in the region, however, great quantities of lean nontitaniferous magnetic iron-bearing rocks, as at the east end of the Mesabi range and in the Gunflint district, where the Duluth gabbro cuts and overlies the iron-bearing formation; at both the east and west ends of the Gogebic range, where Keweenawan intrusive rocks cut the iron-bearing formation, and in parts of the Marquette district. The magnetite ores consist of coarse-grained magnetite-quartz rock caiTving a considerable variety of metamorphic silicates, including amphiboles, pyroxenes, garnets, chlorites, olivines, cordierite, riebeckite, dumortierite, etc. (See pp. 545 et seq.) Locally pyrite, pyrrhotite, and iron carbonate are pj-esent. The minerals show greater variety and more complex chemical constitution than those of other phases of the iron-bearing formation. Where altered at the surface the magnetite may be locally coated with limonite and the silicates may have gone over to chlorite, epidote, and calcite. The yellowish-green colors so develoj)ed are extremely characteristic of the surface of the exposures. PHYSICAL CHARACTERISTICS OF THE ORES. GENERAL CHARACTER. The ores range from the massive and specular hematite and magnetite tlu-ough ores which are partly granular and earthy and partly in small hard chunl 01 o PJ 6 z X 70 60 50 40 30 20 o o X) > o 6 z c o ■o o 00 > o o o z o ■a > S o o 6 z ■00 10 \y % \0> ■•o \'£. ■.(0 ^ / \ A' \ Me_sa,^i.djst.__. / /' / \ X / \ ^\'^" '~~~~-^^^:^^ ■- -'^-: FiGUKE 72.— Textures of Lake Superior iron ores as shown by screening tests. Biweekly samples, representing 43 grades of ore and an aggregate of 22,376,723 long tons, were taken by the Oliver Iron Mining Company during 1939, and tests were made on the average year's sample. The results of mine tests are averaged for each district in proportion to the tonnage mined to give the figures shown on the diagram. CtTBIC CONTENTS OF ORE. RANGE AND DETERMINATION. The cubic content per ton ranges from 7 cubic feet for the hard ores to 17 cubic feet for the soft ores. It depends on the density, the jjore space, and the moisture and may be cal- culated directly according to the methotl following. 47517°— VOL 52—11- 482 GEOLOGY OF THE LAKE SUPERIOR REGION. The cubic content of an ore is a direct function of (a) true specific gravity of the material — that is, the specific gravity unaffected by porosity or moisture; (6) porosity of the material, in terms of percentage of volume occupied by pore space or voids; (c) percentage of moisture in the material — that is, the percentage loss in weight on drying at 110° C. To facilitate the determination of the cubic content of ores the diagram or graphic equation shown in Plate XL was devised, expressing the relation between these tliree factors and the number of cubic feet per ton. Actual determinations in the ground are unsatisfactory in that they do not show the individual effects of the three factors mentioned, especially moisture content, which may vary widely at different times and jilaces. By use of the diagram the three factors are considered separately and their individual, relative, and net effects may be obser^'ed. The use of the diagram is not confined to iron ores but is also applicable to other ore or mineral substance in the ground. USE OF THE DIAGRAM. The operation of the diagram may perhaps be made clear most easily by applying a con- crete problem as an illustration. Given an ore with a specific gravity of 4. ,5, |)orosity .30 per cent, and moisture 7 per cent. Select a point on the upper edge of the diagram indicating the given specific gravity (4.5) ; from this point move downward, as indicated by the dotted line, to the line representing the given porosity. (There are two sets of inclined lines crossing the upper part of the diagram; the less steeply inclined set, numbered at the left side of the dia- gram, indicates degree of porosity.) From this point move upward to the right along the more steeply inclined lines to the edge of the diagram. This point (3.1.5) indicates the specific gravity as corrected for porosity. From this point move directly downward! to the lower edge of the diagram, where the number of cubic feet per ton is indicated. This shows 11.4 cubic feet per ton of dry material. The factor of moisture has not yet been considered. Wlien moisture is present tlie material is heavier and consequently the volume per ton smaller. To introduce tliis factor of moisture, move directly upward from the last point (11.4) to the horizontal line indi- cating the given percentage of moisture (7), and from tliis point down the inclined Une to the lower edge of the diagram, where the number of cubic feet per long ton is found to be 10.6. At the lower edge of the plate is a transformation table sho^ving the relation between cubic feet per long ton (2,240 pounds) and cubic feet per short ton (2,000 pounds). For example, 10.2 cubic feet per long ton is equivalent to 9.1 cubic feet per short ton. CONSTRUCTION OF THE DIAGRAM. The following discussion of the derivation of the diagram is given with the idea that one desir- ing to make use of it would first wish to be assured that it rests on a rational mathematical basis. The top and bottom Imes of the diagram proper, labeled respectively "Specific gravity" and "Cubic feet per ton" and connected by parallel vertical lines, constitute a transformation table by means of which the number of cubic feet per ton of a material of a given density may be at once determined (or \'ice versa) by moving vertically between the upper and lower edges of the diagram. Immediately below the edge of the diagram proper is a scale of pounds per cubic foot, wliich may be used by moving vertically downward from any point on the ' ' specific gravity" or "cubic feet per ton" scales. Effect of porosity. — The effect of porosity is to decrease the density of a substance, hence rock specific gravity is less than mineral specific gravity m proportion to the degree of porosity of the material considered. To introduce the factor of porosity in the diagram, the upper line was extended to the right to the point indicating a specific gra^^ty of zero (not shoMii on the diagi-am). The Une at the left edge of the diagram was dra\ni perpendicular to the upper edge and divided into 100 equal divisions, representing percentages of pore space. Each of the points of the vertical "porosity" fine was then connected with the point mdicating a specific gravity of zero. Hence on moving vertically downward from any point on the ■"specific grav- ity" line, a succession of equally spaced lines are crossed indicating percentages of pore space. To enable the diagram to show automatically the change in specific gravity resulting from a given porosity of a substance of known mineral specific gravity, a set of parallel lines was drawn, properly connecting pomts on the "porosity" and "specific gravity" fines. These lines were THE IRON ORES. 483 drawn parallel to the line connecting 100 per cent porosity with zero specific gra%dty and agree with the following formula: G, = G^{\ -P) where G^ = rock specific gravity, G^ = mineral specific gravity, and P = porosity. The diagram then automatically shows the relation between mineral specific gravity, porosity, and cubic feet per ton. To illustrate, a certain ore with a mineral specific gravity of 5.0 has 40 per cent of pore space. Beginning at the point 5.0 on the upper edge of the tliagram, move downward to the line indicating a porosity of 40 per cent; from this point move along the parallel inclined lines upward to the right, to the edge of the diagram, where the specific gravity as reduced by pore space (rock specific gravity) is found to be .3.0; immediately below this point, on the lower edge of the diagram, it is seen that the ore runs 11.95 cubic feet per ton and 187.25 pounds per cubic foot. Effect ofmmsture. — The diagram so far takes no account of moisture and hence is applicable only to perfectly dry material. Moisture when present in an ore or similar substance occupies the pore space. When the pore space is filled with moisture the material is said to be saturated. As the moisture occupies the natural openings in the ore, its presence affects the weight of the ore and not its volume, hence its effect is to increase the density and ilecrease the number of cubic feet per ton. Moisture is expressed in percentage of total weight. Let D= density as affected by porosity; then, as a cubic foot of water weighs 62.5 pounds, Cubic feet per ton = jp~^^ When moisture (M) is present the above equation becomes — ^ , . , , ^ 2,240 (l-M) C u Die reet per ton = r>^ „„ - — . r x>X62.5 The lower part of the diagram is crossed by a set of parallel horizontal lines indicating per- centages of moisture, as showTi at the right-hand edge of the diagram. Follo^\'ing the above equation, a set of inclined lines were drawn, properly connecting points on the "moisture" and "cubic feet per ton" lines. Given the numlser of cubic feet occupied by a ton of any porous material when dry, the effect of any percentage of moisture is indicated automatically by the diagram. For example, a certain ore when dry occupies 12 cubic feet per ton; it is desired to know the effect of 10 per cent of moisture. From the point 12 on the lower edge of the diagram move vertically upward to the horizontal line indicating 10 per cent moisture; from this point move downward along the inchned line to the edge of the diagram, where it is found that the moist material occupies 10.8 cubic feet per ton. Moisture of saturation. — Up to this point it has been shown that, given the mineral specific gravity, porosity, and moisture content of an ore or similar substance, the diagram automatically indicates the number of cubic feet per ton. In many classes of ore the factor of moisture is the most variable of the three named above. The mineral specific gravity and porosity of an ore determine the amount of moisture which it can hold. This maximum, or moisture of saturation, may be calculated as follows: 6-'„j = mineral specific gravity. D = density of dry porous material. P = porosity. M = moisture of saturation. D =GM-P). D P from which P=l — rr- and M = Substituting the value above given for P — M- 1-^ Gm. 484 GEOLOGY OF THE LAKE SUPERIOR REGION. By substituting values for D and G^ in the above equation tlie curves for moisture of satura- tion were constructed across tlie lower part of the diagram. Those, curves enable one to determine at once the moisture of saturation of any material, given tlie mineral specific gravity and porosity. Each cui^ve corresponds to a certam mmeral specific gravity, and the moisture of saturation is found by moving vertically from the point indicating the number of cubic feet per ton of the dry material to the proper curve for moisture of saturation. For example, an ore with a mineral specific gravity of 4.0 and a porosity of 36.0 per cent occupies 14 cubic feet per ton if dry; its mois- ture of saturation is found by moving upwanl from the point 14 to the curve (/ = 4.0, and reading the indicated moisture — 12.2 per cent; that is, 12.2 per cent of moisture would fill the pore space of tliis ore. Excess of moisture Tmndled in mining. — It frequently happens in mining tliat ore as hoisted to the surface contams a larger percentage of moisture than it did before it was mined; m fact, it may contain a percentage of moisture greater than the moisture of saturation of the unmined ore. This may be caused by the handling of broken ore on undrained mine floors. The ore after being broken doAVTi has a much larger percentage of voids than before and hence a greater ability to absorb and retain moisture. The diagram is useful in this connection in showing, from determuiations of specific gravity and original porosity of hand specimens, the moisture of saturation of the ore in place. Tliis figure compared with tlie percentage of moisture of ore as it leaves the mine teUs at once whether or not an unnecessary amount of water is being hoisted with the ore, owing to improper drainage. EXPLORATION FOR IRON ORE. The location of explorations within the areas of the iron-bearmg formations is determined by outcrops, by magnetic Imes, by mining, and by general geologic structure. It has been possible to confine most of the exploration to the area of the iron-bearing formations, but in certain districts, notably the Cuyuna, Florence, Crystal Falls, and Iron River districts, the distribution and limits of the iron-bearing formation are so uncertain that much exploratory work has had to be done even to locate the formation. All the facts bearing on the distribution of the iron-bearing formation discussed in this monograph are taken into account in choosing areas for exploration. Some of the larger mining companies employ their own geologists to make special reports on the geology of given areas as a preliminary to underground explora- tion, and nearly all the explorers make liberal use of all the geologic information available in localizing their work. As the few ore deposits exposed at the surface were found years ago, explorations are now largely conducted by drilling and sinking test pits and shafts. The large size of the iron-ore deposits makes it possible to find and outline them by drilling to an extent not possible in smaller ore deposits, with the result that the greater number of ore bodies, especially in recent years, are thorouglily explored by drilling before mining begins. It has usually been assumed that if drilling does not locate an ore body it is useless to sink a shaft for tliis purpose. Mming operations have necessarily disclosed much ore which hail not previously been found by drilling, especially in certain districts like the Menominee or the Gogebic, where the structural conditions are such as to make the location of ore by drillmg extremely dillicult. In the region as a whole mining operations have almost evervwhere disclosed greater reserves of ore than the drilling had indicated. The great dependence placed on drill work has resulted in enormous expenditure for this purpose. Accurate estimates of the amount of drilling done so far in the region can not be made, but a rough estimate compiled from tentative estimates of engineers of the several districts is as follows: THE IRON ORES. Drilling done for iron ore in the Lake Superior region. 485 District. Number of drill holes. Average deptii of drill iioies (feet).o Mesabi Vermilion Cuyima Marquette Otber Michigan ranges and Wisconsin ranges 15,000 1,000 1,500 5,000 4,000 175 600 250 500 300 26,500 1 Estimates probably low. This totals 7,200,000 feet, or about 1,363 miles of drilling. At an average cost of $3 a foot, which is a low estimate, the total expenditure has been roughly $21,600,000. It is estimated that at the present time there are 400 drills in operation in the region. In the earlier days of exploration test pits were relied upon to a large extent, especially in areas where the surface drift is thin and the water level below the rock surface. This method of exploration, however, is unsatisfactory because of the great depth of the drift at many places, the difficulty of handling water, and the difliculty after finding the ledge of penetrating it by this method. In later years the use of test pits has been largely superseded by drilling. Both diamond and churn drills are in use. Through surface and soft-ore formations the churn drill is used. Much of the Mesabi district may be so explored. The cost of churn drilling has ranged from -fl to $3.50 and averaged about $2.50 a foot, varying from district to district according to accessibility and cost of transportation ai)d other factors. The cost of diamond drilling has ranged from $2.25 to $8 a foot and averages at present about $3.75, but varies from district to district. Test pits are cheap, averaging perhaps $1.25 a foot. The necessity for the most careful study of the structural geology in drilling is illustrated by the frequent failure of drills to locate ore deposits even after what seemed to be careful drilling and the subsequent discovery of the deposits either by further drilling or by mining operations. Indeed, as one comes to realize the variety and complexity of underground structural conditions, he is likely to become more and more disinclined to submit a negative report on any property, no matter how extensively it has been drilled. This difficulty is illustrated by the ore shoots in the Gogebic and Menominee districts, many of which have been missed by drilling and picked up in mming operations. Many of the ore shoots m the Vulcan member of the upper Huronian slate of Michigan pitch beneath the surface, following the axes of drag folds, and it is easy for drills to pass one side or the other, or, if the drill hole is inclined, to go above or below them. On examination of drilling plats of exploration areas it is easy to see where linear shoots of ore might pass through at places not penetrated by the drilling. In fact, drilling in some of these localities is almost as uncertain as shooting a bu'd on the wing. There are many ways of missing the ore. As knowletlge of structural conditions increases, however, adverse chances diminish, with the result that in certain areas after the local structural problems are solved, it is possible to drill with a high degree of success. A higher average of success in drillmg would unquestionably result if greater care were taken in the interpretation of drill records. The drill runner is often allowed to report the character of the drillings and the samples are not kept, with the result that many valuable inferences that might be drawn from the lithology, the dip and strike of beilding and cleavage, and other features are lost. Not infrequently also failure to plat drill records in such a maimer that they may be considered in three dimensions may cause promising chances for ore to be overlooked. There has been a considerable tendency to generalize the principles of ore occurrence and in exploration to carry such principles fi'om one district to another. As a matter of fact, although some of the basic principles are general for the region, the local variations of structure require the most careful study of each area to prevent mistakes in interpretation. When explorers of the Gogebic district, where the ores lie in regular, impervious, pitching basins, went to the 486 GEOLOGY OF THE LAKE SUPERIOR JtEGIOX. Mesabi district, where the rocks are of tlie same age, tiiey naturally attempted to use the same methods in exploration. But here the flatter dip of the formation, the shallowness of basins, the effect of overlying slates in ponding waters, and the unusually large influence of joints in localizing the concentration of ore made the finding of ore largely a new j)rob]em, which was solved at much expense and trouble. Recognizing the danger of carrying the method of explo- ration of one district into another, certain explorers have gone to the other extreme and have attempted to disregard all guides derived from the study of tlie structural geology, with results even more unsatisfactory than if they had used principles developed for other districts. Much the greater part of the exploration of the region has been conducted without taking the fullest advantage of all geologic knowledge available, but there has been a rapidly increasing tendency to follow geologic structure and therefore an increasing demand for geologic informa- tion, as shown by the cordial support that the mining men have given to the efforts of the United States and State surveys in this region and by their considerable expenditures for private geologic surveys. Certain of the drilling companies doing contract work now have geologists on their staff to aid in the interpretation of records, notwithstanding the fact that such inter- pretation is primarily in the hands of their clients. The problems of underground exploration are followed keenly, intelligently, and energetically by a large number of skilled men in the employ of mining companies, with the result that advances are being made ^^^th a rapiditj' which is sometimes almost bewildering. Six months may see the development of new facts requiring changes in the interpretation of the drilling of a district. The statements as to struc- tural conditions ]>resented in another chapter of this book may require some modification b}' the time the book is given to the public, because of the amount of rapidly accumulating information in the interval between the writing and the printing. MAGNETISM OF THE LAKE SUPERIOR IRON ORES AND IRON-BEARING FORMATIQNS. All ores of iron are found to be magnetic when tested by sufficientlj^ delicate means. Ordi- narily magnetite is the only iron mineral which causes conspicuous disturbance of the magnetic needle. Practically all the Lake Superior iron-bearing formations contain at least minute quantities of magnetite, and hence all exert an influence on the magnetic needle, but in ^\-idely varying degree. The iron-bearing formation of the Vermilion district and other Keewatin areas is strongly magnetic. The same is true of the formation in the east end of the Mesabi district, the Gunflint district, the Cuyuna district, and the east and west ends of the Gogebic district, and of most of the Negaunee formation of the Marquette district. Less magnetic parts of the iron-bearing fonnations are those producing principally hematite and limonite, as the central and western parts of the Mesabi, the central part of tha Gogebic, and parts of the Menominee and Crystal Falls districts. The iron-bearing member of the Iron River district of Michigan affects the magnetic needle onl}' 'jcaUy and slightly. Every known iron-bearing formation ' i the Lake Superior region, Anth the exception of that in part of the extreme west end o'' the Mesabi district, has been outlined partly as a result of magnetic surveys. In some of t/ie districts, as, for instance, the Iron River district, the magnetic variation is slight, but careful observations will detect it. In addition several magnetic belts are known in wliich exploration has not yet showTi the character of the iron- bearing formation. On the general map (PI. I, in pocket) magnetic belts are not indicated over all of the iron-bearing formations. They are showTi only in places where the formation is not naturally exjiosed or uncovered by exploration. Strong magnetic disturbance does not necessarily mean ore, and, vice versa, ore does not necessai'ih' cause strong magnetic disturbance. Lean amphibolitic schists may be highly magnetic, while rich hydrated soft ore has but little effect on the needle. Although magnetic disturbance is usually caused by an iron-bearing formation, it is also caused by certain basic igneous rocks, like the ellijjsoidal basalts of tlie Keewatin or gabbro intrusives. There is Uttle THE IRON ORES. 487 difficulty in ascertaining the cause of tlie attractions, however, for somewhere along most of the magnetic belts in the Lake Superior region there are outcrops which indicate the nature of the rock causing the disturbance. If the rock is entirely covered, it may still be possible to deter- mine whether the disturbance means iron-bearing formation or some other rocks. The iron- bearing formations are sedimentary deposits with certain linear characteristics of distribution, giving even lines or "belts'" of magnetic attraction, whereas the basic igneous rocks are likely to cause a much more irregular magnetic field. Because of the conditions above outlined, it is seldom practicable in the Lake Superior region to draw from magnetic observations inferences with regard to the shapes of the iron- ore deposits themselves as distinguished fi-om the rest of the iron-bearing formation — such inferences as have been drawn by magnetic surveys of deposits in eastern Canada, Sweden, and elsewhere. In those regions the ores consist of magnetite associated with relatively non- magnetic wall rocks, and the magnetic disturbances are produced by the iron ore itself, not by u-on ore and wall rock; hence it is possible to draw satisfactory inferences as to the shape and attitude of the iron-ore deposits. In the Lake Superior region the magnetic attractions are useful in locating iron-bearing formations and thus ultimately the iron ore by underground exploration, but do not directly point out the iron-ore deposits themselves. The highly- devel- oped Swedish methods of determining both the intensity and the direction of the magnetic pull are therefore unnecessarily detailed and slow for use in the Lake Superior region, and when attempts have been made to locate ore deposits by them the results have been disappointing Although the iron ores may not be discriminated by means of the magnetic disturbances, it is possible under some conditions to draw useful inferences frona them as to the dip or folding of a buried iron-bearing formation. A sharp, narrow belt of magnetic attraction leading up to a definite maximum usually means a liighly tilted formation presenting a narrow erosion edge at the rock surface, as in the Gogebic or Vermilion district. A wide, more irregular, and less well defined belt of attraction is ordinarily associated with a flatter dip, exposing a greater area of iron-bearing formation to the erosion surface. The producing part of the iron- bearing Biwabik formation of the Mesabi district illustrates tliis. Unequal magnetic gradient on two sides of a maximum may indicate the direction of dip of the iron-bearing beds. The outward dip of the iron-bearing formation about the Archean ovals of the Crystal Falls district is so indicated. Several roughly parallel, more or less discontinuous magnetic belts, here and there converging and joining, may indicate repeated pitcliing folds, as in the Cuyuna district. General laws of interpretation of magnetic attraction require much local mochfication. It is usually riecessary to ascertain for each locality the magnetic character of the iron-bearing formation, to correlate tliis with known facts fi'om outcrops or underground workings, and from the knowledge thus obtained to interpret the results of the magnetic formations in covered parts of the area where the magnetic reachngs alone are available. H. L. Smyth," in connection with much magnetic field work in the Lake Superior region, has developed mathe- matical relations between magnetic fields and various attitudes of the rock beds which may serve as a useful guide in detailed surveys. The instruments wliich have been used in Lake Superior magnetic surveys are the dip needle and the dial compass. The dip needle determines the vertical component of the mag- netic pull, as well as the direction of the horizontal pull; the dial compass determines only the direction of the horizontal pull. Methods of using and interpreting these instruments are discussed in detail by Smyth. The dial compass is essential in most of the work because it affords means of keeping accurate directions necessary-for location and of reading the horizontal component of the magnetic variation. It may be used only on sunny days, and thus mag- netic work in the Lake Superior region is likely to be slow and expensive. The dip needle may be used at any time, but in a disturbed field it affords no means of keeping horizontal directions, and hence location. This is an essential defect in a country in wliich the roads and other works of man afford little aid in keeping location. In theory the use of the magnetic needle is simple, but much practice is required to insure uniformly accurate observations. The unskilled observer finds many pitfalls in the mechanism cMon. U. S. Geol. Survey, vol. 36. 1899, pp. 335-373. 488 GEOLOGY OF THE LAKE SUPERIOR REGION. of the instrument, in the manner of holding it, in the effects of temperature, in electrification from rubbing the glass, etc. There is much opportunity for the exercise of good judgment in the determination of the intervals at wliich reachngs shall be taken, the direction and number of runs, etc. These should be varied for different areas, depending on the structure found or suspected. Finally, the interpretation of the results calls for consideration and careful bal- ancing of a great variety of^ factors, capacity for wliich is acquired only by wide experience and painstaking observation. MANGANIFEROUS IRON ORES. All the Lake Superior iron ores contain minute quantities of manganese, and certain ores carry as high as 20 to 25 per cent. In the Cuyuna district of Minnesota a drill hole in the iron-bearing member averages 13 per cent for the upper .35 feet and about 2 per cent below. Another hole, in sec. 28, T. 47 N., R. 29 W., has an average of 11.33 per cent for the upper 30 feet. Similar results have been obtained from drilling in the Baraboo district. The larger part of the manganiferous ores shipped so far have come from the Gogebic district. Manganiferous ores are often not discriminated from the iron ores in figures of slupment, and tliis makes it difficult to estimate the tonnage of manganese iron ore and the average percentage of manganese in so-called manganiferous iron ores. E. C. Eckel ° estimates that during 1906 the Lake Superior region produced about 1,000,000 long tons of low-manganese iron ore with an average manganese content of about 4 per cent and ranging as sliipped from 1 to 8 per cent. According to Burchard,* the total production of manganiferous iron ore in the Lake Superior region from 1885 to 1909, inclusive, has been 8,968,449 long tons, or about 77 per cent of the total production for the United States during that period. The percentage of manganese in the manganiferous ores of the Lake Superior region is so low that the ore may not be classed either as a manganese or a liiglily manganiferous iron ore hke those of Arkansas and Colorado. It produces a basic pig. None of the ore shipped from the Lake Superior region has been liigh enough in manganese to be available for ferro- manganese or spiegeleisen, which require at least 15 per cent of manganese. Mineralogically the manganese is mainly in the form of pyrolusite (MnOj). In the Cuyuna district this has been found at the surface to be mixed with rhodochrosite (MnCOj). The psilomelane so commonly associated with pyrolusite in the Appalachian manganese ores has not been especially looked for in the Lake Superior region but is probably present. The conspicuous association of manganese with the upper parts of the iron-ore deposits seems to prevail in the Lake Superior region, as in deposits of manganiferous iron ore in other parts of the United States. IRON-ORE RESERVES. DATA AVAILABLE FOB ESTIMATES. Up to 1910, 335 mines have been in operation in the Lake Superior region, and many thou- sands of test pits and churn and diamond drill holes have been sunk. The mines and explora- tions, together with natural exposures, afl'ord data for a fair estimate of ore reserves in the producing areas. There are considerable areas not yet explored. AVAILABILITY OF OKES. Evitlently the question of the present and future availability of the iron ores is one of costs — in mining, in transportation, and in the furnace. The costs are determined — (1 ) By the character of the ore itself, its percentage of iron and deleterious constituents, and the nature of its principal ganguc material. (2) By the cost of mining, whether, for instance, by open pit or underground mothoil. (3) By whether or not the ore must be concentrated, as, for instance, the sandy taconites of the western Mesabi. o Mineral Resources U. S. for 1906, U. S. Geol. Survey, 1907, p. 106. SBurcbard, E. F., The production of manganese ore In 1909: Extract from Mineral Resources U. S. for 1909. U. S. Geol. Survey, 1911, p. 10. THE IRON ORES. 489 (4) By the cost of transportation to the furnace. Between Vermihon and Marquette ores there is a difference of about 75 cents a ton in the cost of transportation to the lower lakes. Viewed in another way, the cost of transportation is the amount necessary to bring together the coal, limestone, and iron and to transport the finished product to consuming centers. This introduces another set of costs for ores smelted at the upper lakes. (5) By the cost of reduction in the furnace, depending on the character of the ore and on the success in modifying and applying furnace practice to local conditions. For instance, the use of by-products from coke in certain furnaces in the Lake Superior region makes approxi- mately the difference between profit and loss for the combination of conditions there existing. (6) By the nature of the ownersliip. A large corporation holding a variety of ores and equipped to assemble the raw material under the existing conditions can handle ore which would not be available to a smaller company not equipped to control the situation in a large way. In recent years the average percentage of iron in the ore shipped has varied between 60 and 54 per cent for the ore in the natural state (see pp. 477, 493), the grade on the whole low- ering. These grades may be regarded as approximately the lowest average grades available under the conditions prevaihug in those years. Low-grade, high-sihca ores, running as low as 40 per cent in iron, favorably located for cheap mining and transportation, have been used to a small extent for mixtures, as, for instance, ores in the Palmer area of the Marquette district and in the Menominee district. In most of the region at the present time ore ru nnin g 50 per cent (natural) in metaUic iron is considered of about as low grade as is at present available, and estimates are made accordingly. Locally ores of lower grade are included as available ores, either because of favorable conditions of niirdng and transportation, because of differences in the policy of the companies making the estimates, or because they may be concentrated by washing, as in the western Mesabi. The table of production (see pp. 49-69) shows what has been the relative availabihty of ores of the different districts, all factors considered. BE SERVES OF ORE AT PRESENT AVAILABLE. ESTIMATES. • The authors have made no independent detailed estimates of Lake Superior iron-ore reserves for this monograph. "They have, however, had access to the detailed estimates of the principal mining companies and to the records of the Mnnesota Tax Commission and are from their field study famihar with most of the large deposits or groups of deposits. The estimates here given represent their judgment as to the approximate tonnage of ore now avail- able, based on the above information. The variations in the independent estimates of mining companies and the difference of opinion as to how low a grade of ore in any given place is to be included in the available ores give latitude for considerable variations in estimates. The authors can claim no finality for the figures published. They are what seem to them reason- able approximations. Estimates of the available pre-Cambrian iron ore of the Lake Superior region. Long tons. Marquette district 100, 000, 000 Gogebic district 60, 000, 000 Menominee and Crystal Falls districts 75, 000, 000 Mesabi district 1, 600, 000, 000 Vermilion district 30, 000, 000 Cuyuna district 40, 000, 000 1, 905, 000, 000 The reserve reported includes about 1.30,000,000 tons of washable ores from the western Mesabi, averaging 46 per cent of iron (dry) of non-Bessemer character. Of the remainder of Mesabi ores, approximately 40 per cent are Bessemer. There is a further low-grade reserve in the CUnton ores of Wisconsin which may be of con- siderable magnitude. (See pp. 566-567.) 490 GEOLOGY OF THE LAKE SUPERIOR REGION. LIFE OF ORE RESERVES AT PRESENT AVAILABLE. Figure 73, prepared hy II. M. Roberts, shows the total production of ore from the Lake Superior region for 30 years before 1907 and the rate of increase of production. To the close of 1910 20.5 per cent oi' the known reserves had been consumed. If the above ostiniates of 100 90 80 70 60 to Id Z Id U 50 40 30 20 10 / / / « / / 1 / / / / / / / / ' i / / / / / / / / / / / / / / / / / / / / / 1650 I860 1870 1860 1890 1900 1910 YEARS I9Z0 1930 1940 1950 I960 FlGtntE 73.— Diagram showing relation between estimated ore reserves of the Lake Superior region and rate of production. The estimated reserve, 1,905,000,000 tons, plus the total amount of ore miued to the end of 1910, is represented as 100 per cent on the vertical line. For each year there is shown the percentage of this total which had been removed to the end of that year. For example. 15.9 per cent of the kno«-n ore was removed to the close of 1907. For the last five years, 1905 to 1910, the curve is practically a straight line. If this line is projected at a uniform slope, it indicates complete exhaustion of the known reserves in 1960. Reasons are given in the text, however, for the belief that the date of exhaustion will be later. reserves at present available are even approximately correct and the rate of jiroduction remains the same as that in 1910, the hfe of the ore deposits as now estimated will be 45 years— that is, to 1956. If the rate of production increases in the future this time will obviously be shorter. As some increase in the rate of pi-oduction seems Hkely in spite of the ])robable temporary recessions due to business dei)ressions, if onty liigh-grade ores are mined the exhaus- tion of the existing deposits, or, if not these, of the amount of high-grade ore equivalent to that now in sight, will probably occur earlier than this date. But even this conclusion must be modified by the fact that In proportion as the inadequate supply of liigh-grade ores becomes THE IRON ORES. 491 depleted there wall be an increased use of lower-grade ores with the high-grade material, whose life will be thereby prolonged. This factor is regarded as so important as to rendoi- it probable that the use of the high-grade ores will be distributetl through a much longer period than 45 years, just as there will be first-growth white pine remaining uncut long after the date when all the white pine would be gone at the present rate of use. Also new discoveries of ore of jires- ent commercial grade are made yearly. Prior to 1911 the discoveries have kept well ahead of the sliipments. The region is now so well known that there is httle likelihood of discovering another Mesabi range. Though it is not impossible that in the next few years the reserves may be sufficiently increased by discovery to keep pace with the sliipments, this is rather unhkely. Still less Ukely is it that the increase of reserves will keep pace with an acceleration of production. If, for instance, the increase of production for a year amounts to 2,000,000 tons, and it is estimated that the present reserves will last 20 years at the lower rate, it will be necessary in that year of increase to discover 40,000,000 tons of ore in order that the life of the reserves may not be lessened. RESERVES AVAILABLE FOR THE FUTURE. ESTIMATES. Reserves available for the future must be considered as having a present small and intangible value, for the reason that the estimates of ores at present available include all ores wliich can be immediately mined or wliich will be taken out in the normal course of development of present mines. When we remember that iron is one of the most widely disseminated metals of the earth's crust (by actual analysis constituting 4 per cent of all the rocks of the earth), it is apparent that only the most arbitrary limits can be placed on future reserves. In the following estimates of future reserves are included rocks containing a per- centage of iron lower than the percentage in the reserves at present available but sufficiently liigher than that in the common rocks of the earth's crust to give them future priority in use as iron ores over the average rocks of the earth's crust. It will probably be many hundreds of years before any but an insignificant portion of these reserves available for the future are utihzed. The additional discovery of liigh-grade ores — as, for instance, those of the great field in Brazil — the enormous quantities of low-grade ores now available from Alabama and Cuba, the extension of the known high-grade reserves of Lake Superior, and the increased use of scrap iron and steel will postpone the use of the bulk of the low-grade Lake Superior reserves available for the future. On the other hand, the diminution in supply of the reserves at present available will lead gradually, and probably in the not far distant future, to the drawing on minor amounts of these future reserves for mixtures. It is to be remembered that the available ores are associated with iron-bearing formations, which differ from the ore mainly in having more silica and which show all gradations to the ore. The character of these formations, so far as iron is concerned, is best shown by the fol- lowing table of analyses from drill sections compiled by the Oliver Iron Mining Company: Character of iron-bearing formations in Lake Superior region (not including available ore). Diamond-drill Averages. Range. Number of holes. Total number of feet. Number of analyses. Average percentage of iron. Gogebic 15 30 32 30 24 5,890 4,814 11,025 5,287 5,400 490 1,517 1,726 1,681 1,094 36 65 Baraboo . 36.40 35.12 Menominee , . 37. 93 Mesabi 3S. 00 Other Sources. Marquette Trenches Levels 975 94 905 41.53 Menominee. Zfi 40 492 GEOLOGY OF THE LAKE SUPERIOR REGION. These analyses include both the lean and the partly concentrated parts of the iron-bearing formations, but do not include the available ore. If tlic partly concentrated parts of the forma- tion are left out of consideration, the average would be 2.5 per cent of iron. In the following table column 4 contains a rough estimate of the tonnage of all iron- bearing formations outside of "available" ore to a (U-pth of 1,2.50 feet for the steeply dipping formations and to a depth of 400 feet for the Mesahi district, where the thickness ranges from a knife-edge to 900 feet. Column 5 contains a rough estimate of the tonnage of the part of the iron-bearing formations which will run above 35 per cent in iron. Total tonnage of iron-bearing formations to given depths and tonnage estimated to run 35 per cent or mare in iron. District. (1) .\roa. (2) Depth. (3) Volume. (4) Quantity of iron formation. (5) Quantity con- tainini; 35 per cent or more of iron. Michifran; Crystal Falls Sq. mi. 7.8 28.5 5.6 5.8 1.0 127.0 15.6 .7 5.S 11.0 10.0 6.6 30.0 Feet. 1,250 1,250 1,250 1,250 1,000 400 1,250 1,250 1,250 3.i0 100 1.250 1,250 Cu. mi. 1.85 &75 1.30 1.40 .20 9.M 3.70 .16 1.40 .70 .19 1.57 7.10 Tons. 24.100.000.000 sr.SIXI.UIM.l.lMMJ 16. !«KI. I«K1.I«)0 IS, 200. QUI 1.01)0 2,600,000,000 125, 000. 000. 000 4,S, 100, 000, 000 2,1.50.000,000 IS, 200. 000. (KK) 9,100,000,000 2.500,000.000 20.400.000.000 92,400,000.000 Tons. I,.5or).ono.nno 10,000.(Ml. (HKI.OIIO l,2.'i0.l-K10.l»« Swanzy 260, iKlO.noO Minnesota: Mesabi 30,000,flf)i).OOO 1,023,000,000 Wisconsin: 215, OM. 000 Penokee 1, 250, (KH 1.000 910,000.000 Ontario: 250.000.000 2,040.000.000 North shore of Lake Superior 9,240,000,000 200.000,000 255. 40 35 92 467,4.50,000,000 67,640,000,000 We may conclude, therefore, that while the ores at present available would probabl_y be exhausted within about 50 j-ears if they alone were drawn from, the increasing use of lower- grade ores, already begun, will lengthen this jjeriod many times. COMPARISON OF LAKE SUPERIOR RESERVES WITH OTHER RESERVES OF THE UNITED STATES. For comparison a table showing ores available at present and in the future in different parts of the United States is given below. The figures, with the exception of those for Lake Superior, are those of the National Conservation Commission." Iron-ore reserves of the United States available at present and in the future. Commercial district. Ore at present available. Ore available in the future. Long tons. •29S.000.000 53S. 440. 000 1,905.000.000 315,000,000 57. 760, 000 68,950,000 Long Ions. 1 . 095. 000. 000 2. Souttieastern 1 276 5'X) 000 67, 640, tmO. (XX) 4. Mi.ssis.sippi Valley 570 000 (XX) 120. 665. (XX) 6. Pacific slope - 23,905,000 3,183,150,000 70,726,070,000 1. Vermont. Massachusetts, Coimecticut, New York, New Jersey. Pennsylvania. Maryland, Ohio. 2. VirEinia. West VirRinia. eastern Kentucky, North Carolina, South Carolina, Georgia, .Vlabama, eastern Tcimessec. 3. Micnijian. Minnesota, Wisconsin. 4. Northwest .\labama, western Tennessee, western Kentucky. Iowa, Missouri. .Vrkansas. eastern Texas. 5. Montana, Idaho, Wyoining. Colorado, Utah, Nevada, New Mexico, western Te.xas, .\rizona. C. Washington, Oregon, California. It appears from this table that the Lake Superior region contains approximately 60 per cent of the reserves at present available and 96 per cent of the future reserves, as figured in tons. If measured in units of iron, the Lake Superior reserves form a still larger proportion of the total. oBull. U. S. Geol. Survey No. 394. 1909, p. 103. THE IRON ORES. 493 LOWERING OF GRADE NOW DISCERNIBLE. Lower and lower grades of ore are being included in successive estimates of available ores. A comparison of the iron-ore tonnage of the United States with the production of pig iron for the last 20 years shows a distinct increase in the number of tons of iron required to make a ton of pig iron, and thus a lowering of the grade of iron ore mined. Figure 74, prepared by H. M. Roberts, compares the pig iron and tons of ore used and shows an average annual drop in grade of tiie ores for the last 20 years of 0.3.5 per cent in iron. Each temporary increase of 100 90 80 70 60 10 o Sso o a: LJ 40 30 20 10 ^'^ l'' ''. / V \ \ i -I 1 1 r-A / . ~~~1' w^ V V A V ■ — 1886 1890 IS94 1898 1902 1906 YEARS 1910 1914 1916 1922 I9E6 Figure 74.— Diagram representing decline in grade of Lake Superior iron ore since 1889. The light black line represents the approximate average percentage of metallic iron in the total production for the United States for each year. The heavy black line is the average slope computed liy method of least squares, from the variations of the light continuous line. It represents the average decline of grade since 1S89, which amounts to about 0.35 per cent per year. The broken line shows the percentage of the entire production of the United States which comes from the Lake Superior region. As this proportion has steadily increased, it is apparent that the drop in grade of the iron ores, figured for the entire United States, is shared by the Lake Superior ores. production has been followed by a lowering m grade, and decrease of production has meant raising of the grade in about the proportion that might be calculated from the general drop in grade with mcrease in production for the last 20 years. It is not likely that the grade will lower as rapidly in the future as in the past, for as successively lower grades of ore are utilized the amounts available are larger. As the Lake Superior region produces nearly 80 per cent of the iron ore of the United States, the conclusion as to lowering of grade drawn from the diagram may be taken to apply conspicu- ously to this region. 494 GEOLOGY OF THE LAKE SUPERIOR REGION. Tlie present marked tcudeucy toward the use of lower-grade ores docs not necessarily mean that the higher-grade supplies are exhausted, but simply that they are being conserved for the future. In working a series of deposits ranging from the highest to a low grade, in strong financial hantis, it is regarded as the best business policy not to rob the deposits of their liighest grade, as was formerly done, but so to mix the high and low grades as to give the maximum tonnage of an ore just rich enough to be commercially available. The prospective short life of the highest-grade ores, probably not more than .50 years, is undoubtedly influencing the present conservative action. The conservation of the higher-grade supphcs is favored by the marked concentration of control of the industry in a few hands. When the ore was held by many owners the range of grade available to each owner was necessarily limited; when it is in few hands the range is greater and correspondingly greater care can be taken in the proper mi.xing of grades in order to yield a maximum amount of the lowest grade which the market will stand. In the discussion of mining methods (p. 498) some reference is made to the care taken in getting out the proper grades from kny individual deposit. The same general methods are apj)lied by the United States Steel Corporation in apportioning the desired ores among the different deposits available. EFFECT OF INCREASED USE OF LOW-GRADE ORES. If it is established that the high-grade ores have a limited life and that the direction of the development of the ore industry is now toward the use of lower-grade ores and is likely to be more so in the future, and that this tendency will lengthen greatly the life of the ore deposits, there are certain consequences which may be expected. • 1. The distribution of the production of iron ore is likely to be modified and the relative importance of iron-mining centers will vary somewhat. As the grade falls, new "low-grade" districts will come into existence and some old districts which have had a somewhat precarious existence in competition with higher-grade districts will be enabled to meet them on less unequal terms. This will be the effect not only locally within the Lake Superior region, but also in the relations between the Lake Superior and other regions. Western iron ores not now mined will come into the market. Appalachian ores, which can even now, in spite of their low grade, compete with Lake Superior ores because of favorable conditions of transportation and prox- imity to smeltmg materials and consuming centers, may in the future attain an even stronger position, for the difference in composition of the ores marketed is sure to become less, in view of the fact that the change toward low grade in the Lake Superior region is likely to be much more rapid than it is on the large low-grade supplies of the SoutheJist. The same general arguments will apply to the large Cuban reserves. This increased use of lower-grade southern Appalachian ores is further favored by the distribution of the population of the United States and the prevailing freight rates. In a personal communication Judge E. H. Gary, chairman of the board of directors of the United States Steel Corporation, says : Under the existing freight rates for the cruder forms of steel products, if the freights from Birmingham be taken to a series of points extending approximately east and west, so selected that the rate from Birmingham to each point is the same as the rate from Chicago to that point or the rate from Pittsburg to that point, and a line be drawn connecting these points, more than 30 per cent of the population of the United States lives in the territory south of the line so formed, and the rail freight rates from Birmingham to all points in this territory are lower than the freight rates from either Pittsburg or Chicago to these points. If a line be located approximately north and south by selecting the points reached at equal freight rates from Chi- cago and Pittsburg, about 32 per cent of the population of the United States lives in the territory west of this Pittsburg- Chicago line and north of the Birmingham line, and about 38 per cent of the population of the United States lives east of the Pittsburg-Chicago line and north of the Birmingham line. The preeminence of the Lake Superior region is due to the riclmess of its ores, wliich offsets relatively adverse conditions of distance and transjiortation. The lowering of the grade of ore will undoubtedl}' for a time favor other regions more than the Lake Superior region, but it would be rash to assume that the preeminence of the Lake Superior region will be lost. The lower-grade su])plies of the Lake Su))erior region will not bo called into use until long after those from other districts, and this will make it possible to maintain for a long time a liigher grade of output in the Lake Superior region tlian in other ilistricts. THE IRON ORES. 495 2. As a result of the increasing use of low-grade ores, tne distribution of blast furnaces and steel plants may be changed. At present the higher transportation charges on ores to lower lake ports as compared with upper lake ports are just about counterbalanced by increased cost of fuel and flux for smelting at upper lake points as compared with lower lake points. As the grade of ore is lowered this equilibrium will be disturbed. 3. As a result of decrease in reserves of low-phosphorus ores, the change from the acid Bessemer process to the open-hearth process of steel making will continue. The amount of high-grade Bessemer ore now in sight is scanty. Attention should be called, however, to the fact that the low-grade ores which may be drawn upon in the future are not necessarily high in phosphorus. In fact, the ratio of phosphorus to iron remains substantially the same whether the ore is lean or rich, the difference between grades of ore being mainly in the percentage of silica present. Lowering of grade may call into use new methods of smelting iron. 4. The. lowering of the grade of the ore may favor combination of capital in the mining industry if such combination will make possible additional economies and the use of a wider range of ores. COMPARISON WITH PRINCIPAL FOREIGN ORES. The large deposits of low-grade limonite in Cuba have already been mentioned. These will doubtless be largely developed for use of the iron industry along the east coast of the United States. The local ore supplies of England and Germany are of low grade. Both countries import high-grade ores for mixture, partly hematites from Bilbao, Spain, and partly magnetites from northern Sweden and Lapland. The high-grade Bilbao deposits are nearly exhausted. Sweden limits the exports of its magnetite ores. Bessemer hematites of the highest grade are known in enormous quantities within 300 miles of the coast in Minas Geraes, Brazil. vSteps are now being taken to develop these deposits. They are likely to be an important factor in the future in the British and German markets, and it is not improbable that they may be used on the east coast of the United States, especially for mixture with the Cuban ores. TRANSPORTATION. The transportation of tlie Lake Superior ores is one of the most important factors determin- ing their availablity. They 'have been able to stand high transportation charges because of their high grade. MINE TO BOAT. The following table shows the principal ore-carrying railways, distances, rates, and the total tonnage hauled to December, 1 90S : Ore-carrying railroads of the Lake Superior region. Railroads. Ranges supplying traffic. Principal range shipping points. Lake termini at which ore docks are located. Average haul. Approximate average cost per ton from mine to dock. Total iron ores hauled to December, 1908. fVerrailion Tower, Ely {■Two Harbors, Minn. . . Duluth Minn Miles. f 70-90 \ 66 SO 120 40 70 45 45 80 S3 03 12-15 $0.90-11.00 .80 .80 .80 .40 .25 .25 .40 .40 Tons. } 75,153,936 79,118,051 '■40,268,854 Eveleth,Sparta.Biwabik. Virginia, Hibbing, Cole- raine. Virginia, Hibbing, Nash- wauk. Hurley , Tronwood , Besse- merj Wakefield. Michigamme, Negaunee. . DuluthjMissabe and Northern Mesabi Great Northern . Mesabi Sunerior Wis fGogebic Ashland Wis Chicago and Northwestern Marquette Escanaba, Mich Marquette, Mich Menominee Crystal Falls Iron River Iron Mountain, Norw'ay.. Crystal Falls, Amasa 6131.219,397 Duluth, South Shoreand* Atlantic. Marquette /Marquette /Ishpeming, Negaunee 28,493,359 Lake Superior and Ishpeming Negaunee, Ishpeming Marquette, Mich { ^~-\f. 17,420.583 Wisconsin Central Gogebic Menominee Bessemer, Hurley, Iron- wood. Crystal Falls, Iron Moun- tain. Ashland Wis 50 40-60 16.592,713 Chicago, Milwaukee and St. Paul. Escanaba, Mich a Since January 1, 1897. ti Since ISSO. Includes ores other than iron ores between June 1, 1888, and July, 1903. 496 GEOLOGY OF THE LAKE SUPERIOR REGION. Eighty-five per cent of the tonnage has been hauled .'iO miles or more and 15 per cent has been hauled less than 50 miles. The average cost for hauhng the ore to the lake has been 60.42 cents a ton. Four of the railways hauling the ore are controlled directly hy the companies owning or mining the ore. The United States. Steel Corporation owns the Duluth, Missabe and Northern and the Duluth and Iron Range railroads; J. J. Hill controls the Great Northern Railway; and the Cleveland-Clill's Company the Lake Superior and Ishpcming Railway. DOCKS. The docks antl their capacities are as follows: Record of ore docks on the Great Lakes. [Revised to May 1, 1909. Table furnished by Oliver Iron Mining Co.] Kaikoad. Location. Dock No. Num- ber of pock- ets. Storage capacity. Height from water to center hinge hole. Height from water to deck of dock. Width of dofk from out- side to outside of partition posts. Length of spouts. Length dock. Angle pockets Capacity per pocket to bot- tom of stringers. Chicago and Northwestern . . . Do Escanaba, Mich do I 3 4 5 6 1 2 1 2 3 4 5 be 2 3 4 1 2 3 4 5 1 1 1 2 1 cl 184 226 250 202 320 234 234 Tons. 21,143 28, 792 34.923 29,310 69, 760 42,120 42, 120 Ft. in. 28 10 31 2 36 6 28 6 40 40 40 Ft. in. 48 G 52 8 59 2 53 3 70 70 70 Ft. in. 37 37 37 37 50 2 50 2 50 2 Ft. in. 21 27 30 21 8 30 30 30 Feet. 1,104 1,356 1.500 1.212 1,920 1,404 1,404 , 39 30 45 43 40 45 45 45 Cubic/eet. 1,918 1,969 Do do 2.191 Do do 2,8.?2 Do ... .do 4.114 Do Ashland. Wis 3,915 Do do 3,915 Two Harbors, Mum. . . do 1.630 268, 170 Duluth and Iron Range Do. . 202 208 170 108 168 148 40, -100 41.600 34, 000 36,960 35,450 43,246 35 5 33 5 40 37 39 40 39 6 57 6 66 62 66 9 73 49 49 49 49 49 53 27 27 27 29 30 32 4 iil.,38.S 1,280 1,054 1,042 1,050 920 38 42 38 42 43 32 38 42 43 32 45 3,006 3,006 Do...: do 3.006 Do do 3.270 Do do 3,126 Do do 4.272 Duluth, Minn . 1.0B4 231,656 Duluth, Missabe and North- 384 384 384 69,120 80.640 119,274 32 40 7 41 9J 57 6 67 J 72 6 49 59 57 27 9 27 9 30 IJ 2,336 2,304 2,304 45 45 45 2,363 ern. Do So 2.782 Do . ..do 3,867 1.152 269.034 374 . 350 326 100,980 94.500 88.020 40 40 40 73 73 73 62 8 62 8 62 8 32 4 32 4 32 4 2.244 2,100 1.956 4S 45 45 4,972 Do do 4.972 Do .do 4.972 Marquette, Mich . .do 1,050 200 200 400 283,500 Duluth, South Shore and At- lantic. Do . 28,000 50,000 27 9 40 47 3 70 10 36 8 51 21 1 32 4 1,200 1,236 39 45 45 1.839 3,848 do 78,000 LakeSuperiorand Ishpemlug. WiSGonsui Central 200 314 36,000 48,356 30 9 40 54 66 2 50 36 27 7 27 1,232 1.908 38 40 50 45 2.713 Ashland, Wis 2.435 Escanaba, Mich do ■ Chicago, Milwaukee and St. Paifl. Do 240 240 50,400 63,500 40 2i 40 Hi 66 6 69 2 52 54 120 27 120 29 30 4i 1,500 1,500 45 45 2,900 3.150 Michipicoten, Ontario. Key Harbor, Ontario. . 480 113,900 Algoiiia Central and Hudson 12 20 34 43 4 61 9 ■ 25 28 22 6 30 311t 240 44 Bay. 2,000 o312 feet single pockets: 1.070 feet double pockets. t> Steel superstructure on concrete. c Pockets ailed by belt conveyor from stock pile trestle 30 feet high. The cost of unloading from train to dock and from dock to boat aggregates 4 cents a ton. Most of the structures up to the present time have been made of wood and are so inilammable as to require almost prohibitory insurance rates, are easily choked in cold weather by the freezing of the water in the ore, and are easih^ lied up by strikes. Tlic destruction or tying up of a dock is a most serious setback to the iron-ore industiy and one which can be less easily (. S. GEOLOGICAL SUfivE O^JOGt'sPH LI' fl_ 1 A ORE DOCKS AT TWO HARBORS, MINN. See page 497. Jl EXCAVATIONS AT STEVENSON, MINN. See page 497, THE IRON ORES. 497 avoided and less quickly remedied than any other of the misfortunes affecting the industry. Steel is used in new docks at Two Harbors, Minn. (PI. XLI, A), and this may be the beginning of a revolution in dock building. The docks have undergone little stractural modification since they were first used in the Lake Superior region. There "is still room for mechanical improve- ment to make the movement of ore more certain and continuous between the train and the boat. BOATS. The ore is carried on the Great Lakes by a fleet of vessels numbering 660 in 1907. Of the total tonnage which has gone dowm the Great Lakes much the largest percentage has gone to Cleveland and a small percentage to Chicago. The proportion going to Chicago is constantly increasing. The following table shows tonnage and rates from upper Lake ports in 1907: Quantity of ore shipped from upper Lake ports in 1907, with rates per ton. ERRATUM. The rate stated on page 497, in the last sentence under the heading "Dock to furnace," is incorrect. In 1910 the rate per ton from Lake Erie docks to the Youngstown district was 64 cents, to Pittsburgh $1.04, and to Philadelphia $1.53. DOCK TO FURNACE. Still another transportation charge to be added to the ore is that of unloading at the Lake docks and short rail transportation to lower Lake furnaces. From Conneaut and Ashtabula to the furnaces the distance is 50 miles and the charge 50 cents a ton. TOTAL COST OF TRANSPORTATION. The average cost of transporting Lake Superior ores to the furnaces during 1907 was $2.14 a ton. When it is remembered that approximately three-fourths of the transijortation is done by companies controlling the ore and that this transportation charge contains a con- siderable profit for the mining companies, the real cost of carrying ore to the furnaces is seen to be considerably lower. Although the cost of transportation for the ore has been high, on the other hand the furnaces have been located fairly close to the distributing centers for finished materials, so that transportation of the finished material has been correspondingly less. As the center of population has moved westward, the smelting in the vicinity of Chicago has become propor- tionally more important and the cost of transportation of the ore proportionally less. METHODS OF MINING. It is the purpose here mere.ly to mention some of the most elementary features of the mining methods used in the Lake Superior region. The ores in general are taken from the ground by open-pit and underground methods or some combination of them. By far the larger number of mines are underground mines. Most of the open-pit mines (see Pis. XI, p. ISO; XLI, B) are in the Mesabi district, where, in 1908, 63.7 per cent of the ore was so produced. The pro- duction of the Mesabi open-pit mines is so large that, notwithstanding their small number as com- pared with the total number of mines in the region, they produced, in 1908, 42 per cent of the 47517°— VOL 52—11 32 • THE IRON ORES. 497 avoided and less quickly remedied than any other of the misfortunes affecting the industry. Steel is used in new docks at Two Harbors, Minn. (PI. XLI, A), and this may be the beginning of a revolution in dock building. The docks have undergone little stnictural modification since they were first used in the Lake Superior region. There is still room for mechanical improve- ment to make the movement of ore more certain and continuous between the train and the boat. BOATS. The ore is carried on the Great Lakes by a fleet of vessels numbering 660 in 1907. Of the total tonnage which has gone down the Great Lakes much the largest percentage has gone to Cleveland and a small percentage to Chicago. The proportion going to Chicago is constantly increasing. The following table shows tonnage and rates from upper Lake ports in 1907: Quantity of ore shipped from upper Lake ports in 1907, with rates per ton. Port. Shipped in 1907. Percent- age of total. Rate per ton to lower lakes. Rate times peroent- age carried. Escanaba Tms. 5,7(3,988 3,013,826 3,437,672 8,188.906 7,440,386 13.445,977 13.95 7.30 8.43 19.79 18.00 32.00 SO. 60 .70 .75 .75 .75 .75 837 511 Marquette Ashland 1,485 Superior Duluth 2,400 41,288.755 99.47 7.216 72.16 The average cost per ton of transporting all the ore shipped in 1907 from the upper to the lower Lake ports was 72.16 cents. DOCK TO FURNACE. Still another transportation charge to be added to the ore is that of unloading at the Lake docks and short rail transportation to lower Lake furnaces. From Conneaut and Ashtabula to the furnaces the distance is 50 miles and the charge 50 cents a ton. TOTAL COST OF TRANSPORTATION. The average cost of transporting Lake Superior ores to the furnaces during 1907 was $2.14 a ton. Wfien it is remembered that approximately three-fourths of the transportation is dcfne by companies controlling the ore and that this transportation charge contains a con- siderable profit for th^ mining companies, the real cost of carrying ore to the furnaces is seen to be considerably lower. Although the cost of transportation for the ore has been high, on the other hand the furnaces have been located fairly close to the distributing centers for finished materials, so that transportation of the finished material has been correspondingly less. As the center of population has moved westward, the smelting in the vicinity of Chicago has become propor- tionally more important and the cost of transportation of the ore proportionally less. METHODS OF MINING. It is the purpose here merefy to mention some of the most elementary features of the min ing methods used in the Lake Superior region. The ores in general are taken from the ground by open-pit and underground methods or some combination of them. By far the larger number of mines are underground mines. Most of the open-pit mines (see Pis. XI, p. 180; XLI, B) are in the Mesabi district, where, in 1908, 63.7 per cent of the ore was so produced. The pro- duction of the Mesabi open-pit mines is so large that, notwithstanding their small number as com- pared -with the total number of mines in the region, they produced, in 1908, 42 per cent of the 47517°— VOL 52—11 32 • 498 GEOLOGY OF THE LAKE SLTPERIOR REGION. entire Lake Superior shipments. Stripping operations in the Mesabi district, taking into account the removal of ore, are far more extensive than the work conducted at the Panama Canal, the total material removed during 1909 in the Mesabi district being 49,750,000 cubic yards as compared with 35,100,000 cubic yards at the Panama Canal." The underground metliods have in common the general use of gravity in milling the ore to lower levels fi'om which it may be trammed to the shaft and then hoisted to the surface. The ores are taken out by square-set rooms running up from sublevels, or by top and side slicing downward from the upper parts of the deposits, or by milling through untimbered chutes to levels below after the surface material has been taken from the top. The cost of this work has ranged from 40 cents to $1.60 a ton, or even higher. An average figure would be perhaps $] a ton. The essential feature of open-pit mining is the removal of the surface material and the transfer of the ore directly to railway cars wthout the intermediate use of the tram or shaft, and without the loss due to leaving pillars. The thickness of drift removed ranges up to 100 feet or more. The general method of work is much more scientific than would at first appear, for it is not a matter of shoveling ore at random onto cars. The character and physical conditions of the deposits are determined by drilling, and the steam-shovel cuts and tracks are distributed so as to reach the desired grades of ore by handling the least possible amount of waste. The possible grades which the mine may produce are ascertained, and when a certain grade is desired by the market the greatest care is taken to extract this grade from the ore body without leaving undesirable ores which must be later moved at a loss. It would be obviously undesirable to take out a high-grade ore and leave a low-grade ore adjacent which could not be sold because of its low grade when by mixing a high and low grade it would be possible to get a medium gi'ade which could be sold. Extreme care is taken to match the different grades in such a manner as to leave them accessible at proper times. The prob- lem is primarily an engineering problem and is worked out by engmeers from most careful measurements and calculations. "Wlien a request for a certain grade of ore comes to an open- pit mine, orders are sent out to load so many cars from a certain cut and so many cars from another cut, or to make a steam-shovel cut in a certain position; and it is kno'mi in advance that the analysis of the ore thus ordered mil run very close to that required. The grading of the ore is becoming closer every year. In the utilization of expert engineering help the open-pit mines are fully as far advanced as any other form of mining. In connection with grading the ore accurate analytical chemical work on a very large scale is necessary. The work of sampling and analyzing the ores, both at the mines and at the works, has been developed to a remarkable degree of accuracy. An illustration of tliis is shown by the following pairs of analyses, representing the total average of 21,030,909 tons of ore shipped by the Oliver Iron Mining Company from the Lake Superior region in 1909. The average from mine analyses was iron 59.19, phosphorus 0.068^ moisture 12.22, silica 6.38; and the average of the same ore as analyzed at the smelting plants was iron 59.04, phosphorus 0.068, moisture 12.33, silica 6.66. This is an exceedingly close check on perhaps the largest piece of quantitative chemical work recorded. The cost of open-pit work depends primarily on the amount of overburden to be removed and the ratio of tliis to the size of the ore body. The average cost of loading on the car may be only 4 or 5 cents a ton. The average cost of stripping, however, to uncover a ton of ore may run from 20 to 30 cents. It is obvious that the figure would be small where the drift is thin or where the amount uncovered is large in proportion to the thickness of the cover, so that the cost of surface removal may be charged against a large number of tons. In general the cost of steam-shovel mining has probably averaged less than 30 cents a ton. With this great difference in cost in favor of the open-pit method of minin g, the question may naturally be asked why any of the Lake Superior ores are mined by underground methods. For many of the deposits the answer is obvious. Their larger dimensions are vertical rather aMin. and Sci. Press, vol. 101, 1910, p. 769. THE IRON ORES. 499 than horizontal, requiring hoisting apparatus to get them to the surface. But even in the Mesabi district 37 per cent of the ores are mined by underground methods and for such mines the reason is perhaps not so obvious. It may be that the drift is too thick; that the topog- raphy does not afford a sufficiently gentle slope for the approach of the track; that adjacent land for a proper approach is owned by others; that the deposit may have a considerable amount of low-grade material on top which must be moved before the material of better grade can be obtained. It may be that the company has insufficient financial resources to make the large initial expenditure necessary for the open-pit method before ore is mined or sold, or it may be that the deposit is not sufficiently large in proportion to the expense of preparing it for the open-cut method to warrant piHng up this great advance charge against the ore deposit. It may be noted that the percentage of ore uncovered by open-pit methods is being rapidly increased and that conditions which a few years ago were regarded as insuperable obstacles to open-pit handling are now easily managed. It may be pointed out further that tliis change in methods has accompanied the combination of mining capital, strong concerns being able to do what the weaker concerns could not attempt. RATES OF ROYALTY AND VALUE OF ORE IN THE GROUND. The ores of the Lake Superior region are leased at royalties ranging from 10 cents to .fl..35 a ton. The average for the region is somewhere between 30 and 50 cents a ton. The liigher figures appear in the later leases. The Mesabi range has the highest general average of royal- ties. Here the Oliver Iron Mining Company pays the J. J. Hill ore interests a royalty of 85 cents a ton on a muiimum of 750,000 tons for 1907; this minimum to be increased by 750,000 tons annually until it reaches 8,250,000 tons a year, after which it remains constant, the royalty to increase 3.4 cents a ton per year for ore carrying over 59 per cent in iron. The royalty rate practically measures the value of the ore in the ground to the fee oWTier. The fee owner demands on an average as high a price as the leaseholder can afford to pay for the ore. On tliis basis the value of the ore in the ground is between 10 cents and $1 a ton. The value is liigh in proportion as grade is liigh and costs of mining and transportation are low. The Minnesota State Tax Commission has adopted an excellent classification of ore reserves, based on compulsory returns from the mining companies, and has valued the ores for purposes of taxation at 8 to 33 cents a ton, this valuation being 40 per cent of what is regarded as the real value. The tax-commission figures would therefore indicate that the value of the ore in the ground is from 20 to 75 cents a ton in Minnesota. The present cash value of a ton of ore is obviously less than the value which will ulti- mately be reafized from royalty after a period of years. If it be assumed, for instance, that the ore must lie in the ground 15 years before the royalty is received, its present cash value would be roughly 42 per cent of its ultimate royalty value. ORIGIN OF THE ORES OF THE LAKE SUPERIOR PRE-CAMBRIAN SEDIMENTARY IRON-BEARING FORMATIONS. OUTLINE OF DISCUSSION. Under the above heading are included all the productive pre-Cambrian ore deposits of the Lake Superior region. It is proposed to sliow in the following discussion — That these iron ores are altered parts of chemically deposited sedimentary fonnations, originally consisting mainly of cherty iron carbonate and greenahte. That a few of the iron-ore deposits represent originally rich layers of iron formation, in which secondary concentration has made only minor changes. That in by far the greater number of deposits, mcluding all the larger deposits, the second- ary concentration has been the essential means of enrichhig iron-formation layers to iron ores. That the conditions of sedimentation of the iron formation may be roughly outlined. 500 GEOLOGY OF THE LAKE SUPEKIOK REGION. That tlic weatlicring and erosion of bed-rock surfaces of average composition would be iniido()uate as a source of tlic materials of tlie iron-bearini; sedinipnts, and lliat the materials fortliose formations have been derived largel}' from basic igneous rocks. That some parts of the sedimentation accompanied or immediately followed the several introductions of jjre-Cambrian l)asic igneous rocks into the outei- zone of the earth and another part came under ordinary weathering concUtions later than tlie extrusions of the parent basic igneous rocks. That the chemistry of deposition of the iron-bearing fomiations under such conditions may be approximated and that original jjhases of the sedimentary iron-bearing formations may be synthesized in the laboratory. That the subsequent oxidation of the iron-bearing formations, the transfer of iron salts, and the leaching of sihca by agents carried in the meteoric waters have secondarily concen- trated the ores and developed all but an insignificant portion of the ore deposits now mined. That tliis second concentration has been localized by a considerable variety of structural and topographic conditions. That in some places before and in other places after concentration the iron-bearing formations have been extensively modified by mechanical deformation or by igneous intrusions, with contact effects such as to prevent the further concentration of ore deposits. That the sequence of events developing the present features of the ore deposits may be outlmed for each district and for the region as a whole. That the development of the ores in general represents a partial metamorphic cycle. THE IRON ORES ARE CHIEFLY ALTERED PARTS OF SEDIMENTARY ROCKS. The iron-bearmg formations are bedded and locally cross-bedded. The Iluronian iron- bearing formations are conformable to other sedimentary formations — quartzite, conglomerate, slate, and limestone — and are not diiTerent from those of the Keewatin, which are associated with but little fragmental sediment. They contain recognizable sedimentary material, such as iron carbonate, greenalite, shale, sand, and conglomerate. We may anticipate our discussion of the secondary alterations of the ores by stating that the original constituents of the iron- bearing formations were domuiantly cherty iron carbonate and iron silicate (greenalite), with minor amounts of hematite and magnetite and with varyuig amounts of the constituents of the mechanical sediments — mud, sand, and gravel. In tracing the development of the iron-bearing formations we must therefore inquire principally into the derivation of the cherty iron carbonate and greenalite. These two substances are nonclastic, though locally some clastic material appears in them; as will be shown later, they are chemical sediments. The sedimentary nature of the iron-bearing formations scarcely needs more elaborate proof. It is so obvious in the field that it has been doubted by only three geologic observers. Whitney, Wadsworth, Winchell, and Hille have held these formations to be of surface igneous origin (see pp. 569-570), but as these views are not now regarded seriously by most men who have studied the subject, and as they liave been abandoned b\' Wadsworth, it will be unnec- essary here to marshal evidence against them. CONDITIONS OF SEDIMENTATION. IRON-BEAKING FORMATIONS MAINLY CHEMICAL SEDIMENTS. The iron-bearing formations are regarded mainly as chemical sedipients (1) because they consisted originall\' of iron carbonate and ferrous silicate and possibly some iron oxide, similar to substances known elsewhere to be tieposited as chemical sediments; (2) because they may be synthesized in the laboratory by the simple chemical reagents which were probably ])resent where the iron-bearing rocks were formed; and (3) because they usually lack fragmental ])ar- tides. To a minor extent they are fragmental sediments derived from the erosion of earlier iron-bearing ami other formations. THE IRON ORES. 501 ORDER OF DEPOSITION OF THE IRON-BEARING SEDIMENTS. The greater mass of the Keewatin iron-bearing rocks, as exhibited in the Vermihon dis- trict, lies above the Keewatin basalts and porphyries and is infolded with them. Another part is interbedded with the basaltic flows. This general association is believed to hokl as a rule for the Keewatin of the pre-Cambrian shield of North America. The Keewatin iron-bearing formations are in beds of limited and irregular extent and thickness. It is concluded that the dej)osition of a few feet of iron-bearing sediments directly in shallow depressions bottomed by basalt was followed by the superposition of another lava flow, and this in turn by more iron- bearing sediments, and so on. Later, when the outflow of lava practically ceased, the main mass of the iron-bearing formation was deposited. Locally a little fragmental material went down immediately upon the basalt basements before iron deposition began. The deposition of the middle Huronian, containing the iron-bearing Negaunee formation of Michigan, began with a coarse conglomerate and sandstone (Ajibik cpiartzite), changing somewhat gradually into a mud (Siamo slate), and this in turn into a chemically deposited iron- bearing formation (Negaunee). In the Cascade or Palmer portion of the Marquette range fragmental quartz sand and ripple marks are conspicuous in the iron-bearing formation. South, of the Marquette district the fragmental beds untlerlying the Negaunee formation are thin or lacking. In certain districts the iron formation is replaced over large areas by basic volcanic rocks (Clarksburg and Hemlock formations and perhaps others unknown). In general, then, during middle Huronian time local sedimentation of sand and clay was followed by more widespread deposition of chemical iron-bearing sediments lacking fragmental material and by simultaneous igneous flows. The iron-bearmg formations of the upper Huronian are the most widespread of the pre- Cambrian. Quartz-sand deposition (Pokegama quartzite. Palms formation, and Goodrich quartzite) was followed suddenly by the widespread deposition of chemical iron-bearing sedi- ments (Biwabik, Ironwood, Bijiki, Vulcan, etc.), with very msignificant amounts of clastic material, and this in turn gave way somewhat gradually to the deposition of mud of probable delta origin (see pp. 612-614) in masses so thick that the thin iron-bearing formations and quartzites previously deposited may be regarded as forming the lower selvage of a mud forma- tion. Thin slate layers and a few quartzite layers are interbedded with the upper Huronian iron-bearmg formations, especially in their upper portions, and the formations locally show a tendency to be replaced along the strike by slate, as in the Mesabi, Gogebic, and Menommee districts. In the Menominee district slate divides the iron-bearing formation, and in addition there are considerable quantities of fragmental quartz sand, iron oxide, and ferruginous slate near the base of the iron-bearing formation. In the Crystal Falls, Florence, Iron River, and Cuyuna districts the ore is in siderite lenses in the upper Huronian slate, and the basal fragmental quartzite has been only locally recognized. These occurrences are apparently farther from the base of the formation than those in the Mesabi, Gogebic, Felch Mountain, and Menominee districts, where quartz sand, iron-bearing formation, and slate were successively deposited as distinct formations. On the south side of Lake Superior, in the western Marquette, eastern Gogebic, and north- western Menominee districts the deposition of the upper Huronian iron-bearing formations was interrupted by the contemporaneous extrusion of great masses of submarine ellipsoidal basalts. These extrusions may have been more extensive than now appears, because evidence of them may be buried or may have been removed by erosion. ARE THE IRON-BEARING FORMATIONS TERRESTRIAL OR SUBAQUEOUS SEDIMENTS? It is beUeved that the iron-bearing formations are subaqueous for the following reasons: 1 . They were originally ferrous compounds in major part. Terrestrial sedimentation usually produces ferric oxides — hematite or limonite and laterite, except in bogs — and reasons are advanced elsewhere to show that only a part of the Lake Superior iron-bearing formations may be so developed. 502 GEOLOGY OF THE LAKE SUPERIOR REGION. 2. Tlio middlo and upper Iluroniau ir<)ii-l)Paring formations arc parts of sedimentary groups containing quartzites and slates of probable subaqueous origin. The slates are essentially delta dojiosits. 3. All the iron-l)('ariag formations are associated with basalts having conspicuous ellipsoidal structures, which can be best explained as developed by flowing out under water. They contrast in this regard with tiio basic lavas of the Keweenawan series. 4. Between the underlying basalts, which are probably subaqueous extrusions, and tlie iron-bearing formations in. the Keewatin series neither weathering nor erosion has taken place except very locallj'. The two are conformable. BOG AND LAGOON ORIGIN OF PART OF THE IRON-BEARING ROCKS. The iron-bearing members of the Crystal Falls, Iron River, and Cuyuna (Ustricts are asso- ciated with slates of probable delta origin, which near the iron-bearing rocks arc so uniformly black and graphitic and generally pyritiferous that black slate is usually regarded as a favorable intlication in jirospecting for ore. Much black slate in the upper Huronian is not associated with the iron-bearing formations, but ore is almost never found without the black slate. The iron-bearing rocks in such associations with black slate are originally carbonate. Smaller amounts of graphitic slates are found also in connection Avith the Keewatin iron-bearing forma- tions. The thicker iron-bearing formations of the Mesabi, Gogebic, Marquette, and Menominee districts are associated with black slates to a less degree. It is suggested elsewhere that some of the slates most abundantly associated with the iron-bearing formations may represent delta deposits, and that the carbon content of the iron formations is probably to be explained as organic. So far as direct evidence is concerned, the organic origin of the graphite and sulphides in the black slates, notwithstanding its probabilit}', should not be regarded as proved, although there is no reason to doubt such an origin. Similar associations elsewhere, as in the Carboniferous, have been shown to be truly of organic origin. On the other hand, in the Lake Superior black slates, as in all other Lake Superior pre-Cambrian formations, no organic forms have been found. These facts raise the question whether the carbon of the slates may not have been effective in the original deposition of the iron-bearing formations, as bog or lagoon deposits, in the manner of Carboniferous and Cretaceous carbonates — that is, by the progressive burial of ferric oxide with organic material, resulting in the reduction of the oxide and the formation of iron carbonate. The way in which reducing organic substances aids in dissolving and transporting iron salts is discussed on pages 519-520. This is probably the origin of the discontinuous carbonate lenses in the carbonaceous slates of probable delta origin in the upper Huronian, but difficulties appear when we attempt to exj)lain in the same waj' the main, tliick, continuous masses of iron-bearing formation of the Keewatin, middle Huronian, and upper Huronian. HYPOTHESIS OF BOG AND LAGOON ORIGIN NOT APPLICABLE TO THE MAIN MASSES OF THE IRON-BEARING SEDIMENTS. The main masses of the iron-bearing sediments are not closely associated with carbonaceous slates; they are not characteristically discontinuous or lens-shaped, but are extensive and tliick; they rest with sharp contacts on quartzite, conglomerate, or basalt. The Lake Superior iron- bearing formations also carry more chert than deposits of known bog origin of the carbonate type. The bog theory of origin involves the assumption that the Lake Superior region may have been, during each of the iron-depositing periods, covered by great bogs or lagoons in wliich vege- table matter could grow at or near the surface of the water over great areas, as in lagoons in ai-lvance of barriers thrown up b}" the sea encroaclung over a gently sloping surface, or under delta conditions. As a process necessarily confined to a shallow zone near the surface, its con- tinuous operatiiju would involve continuous and uniform su])sidence at a rate connnensurate with tlie deposition of the iron salts in t)rder to j)roduce the thicknesses now Icuown. iUthough THE IRON ORES. 503 this theory is probably' applicable to some of the thin lenses of small extent associated with car- bonaceous slates, it is not clear how this process could produce a thousand feet of iron-bearing sediments sliowing uniformity of lithology and bedding and having so little extraneous material through hundreds of square miles. HYPOTHESIS OF GLAUCONITIC ORIGIN NOT APPLICABLE. The greenaUto of the iron-bearing formations of the ]\Iesal)i and other districts is so similar to glauconite as to suggest similarity in conditions of origin — that is, as filUngs of cavities in or replacements of Foraminifera in deep-sea deposits. Dredgings have brought up glauconite from deep and quiet waters but not from places of rapid sedimentation. No glauconite is laiown with so little foreign material as the greenalite beds of the iron-bearing formations. The thick- ness of the deep-sea glauconite beds is not known. In geologic sections the thickest known deposit is 35 feet. The deposition of 1 ,000 feet of greenalite beds in the same manner as glaucon- ite is known to be deposited would require a development of Foraminifera in the prc-Cambrian not known in any other geologic period. IKON-BEARING SEDIMENTS NOT LATERITE DEPOSITS. In many parts of the world, especially in tropical climates, there are bedded iron ores of the laterite type, presumed to develop from the katamorphism of basalt or other basic igneous rock in place. They are characteristically associated with bauxite, clay (lithomarge or bole), usually resting on it. Gradational tyj^es between lateritic iron ore and igneous rock have been described. The Lake Superior iron beds associated with basalts can not in any considerable part be referred to decomposition of the basalt in place after the manner of laterite deposits; the almost complete absence of clay associated with the iron ores and the presence of abundant chert preclude this explanation. Although lateritic decomposition of basalt surfaces may have been an ultimate partial source of the iron ore, transportation and sorting have eliminated the clay, which would be present if the iron beds resulted from lateritic decomposition. The principal impurity in the Lake Superior iron is silica. Tliis could not have developed from decomposition of the basalt in place. In reading accounts of the origin of iron beds associated \vith basalts in different parts of the world," one notes a tendency to ascribe a lateritic origin to the iron beds, even in places where the iron lacks the associated clay to be expected from such a mode of origin. It would seem necessarj' at least to introduce the factors of sorting and transportation to explain these ores. Clay is as stable as iron oxide under surface conditions, and so far as quantitative evidence goes, it remains with the residual iron oxide in a more or less uniform proportion throughout a cycle of decomposition. Finally the evidences of water sedimentation and physical separation of most of the iron formations and basalts are not in accord with the hypothesis of lateritic origin. IRON-BEARING SEDIMENTS NOT CHARACTERISTIC TRANSPORTED DEPOSITS OF ORDINARY EROSION CYCLES. The oxidized carbonate lenses associated with the grapliitic slates (see p. 501) may be regarded as one of the mcidental results of a normal erosion cycle. The fragmental bases of the Vulcan formation in the Menominee district, of the Bijiki scliist in the Marquette district, and of the Cretaceous rocks in the Mesabi distiict contain a great deal of detrital ferruginous chert and iron ore derived from the breaking up of iron-bearing rocks that he unconformably below, but all these phases of the iron-bearing rocks are of minor importance as compared with the thick masses of iron-bearing formation derived from the alteration of iron carbonate and greenalite rocks. "Cole, G. A. J., The red zone in the basaltic series of the county ot Antrim: Geol. Mag., decade 5, vol. 5, No. 530, 1908, pp. 341-344. 504 GEOLOGY OF THE LAKE SUPERIOR REGION. L It has long been recognizctl tlmt tliere are dilliculties in tlie way of explaining the thick and uniform masses of chemical sediments constituting the thicker iron-bearing forma- tions, accompanied by so Uttle mechanical sediment, on the assumption that the kon-bearing formations have been derived from the weathering of average hind areas. If the pecuHar character of chemical sediments depends on depth of water and distance from the shore, then the great thickness of the formations involves uniform subsidence over a great area to keep the conditions uniform. 2. The iron-bearing formations may or may not be associated with ordinarj' clastic sedi- ments. In the Keewatin they usually are not. The middle Huronian consists, from the base up, of quartzite, slate, and iron-bearing formation. The upper Iluronian where best exposed consists of quartzite, iron-bearing formation, and slate. The association of the Keewatin iron- bearing formations with extrusive basalts and not with other sediments shows that the iron ores of the Keewatin, at least, are not the result of dejjosition in any ordinarj^ cycle of erosion and deposition, and tliis strongly suggests that the variety of succession in the sedimentary iron-bearing formations of the Iluronian is also not due to ordinaiy cycles of erosion and depo- sition, and that the deposition of the iron-bearing formations probably was not uniformly related to sea transgression or recession or any other one phase of a topographic cycle. The fact that in many places the sediments above and below the Huronian iron-bearing formations are different is the only feature which suggests that the deposition of iron-bearing sediment is a part of a cycle of erosion and deposition, though it is conceivable that volcanism itself would cause this change, either by efl'ecting changes of levels of land and water or by introducing new rocks for erosion to work upon. Until investigation has disclosed all the different combinations of factors wliich may pro- duce a particular order of sedimentation, it is unsafe to be too positive in concluding that the varied relations of the iron-bearing formations to the order of sedimentation indicate their deposition under exceptional conditions. The conditions producing alternations of iron-bearing sediment with other sediments in varying succession may not be necessarily difi'erent from those favoring the deposition of limestone with a variety of associations — for instance, the Paleozoic Hmestones, which in some places overlie sand and in others mud and are in turn fol- lowed by sand or mud. But the lack of uniformity in the relations of the iron-bearing forma- tions above noted is taken to indicate a probability that conjunction of their deposition with a certain phase of a topographic cycle is not an essential condition to their development. 3. Were the iron-bearing formations derived from the weathering of the older rocks against which they he, it would be diflicult to explain the complete absence of weathered material between certain bands of Keewatin iron-bearing formations and the associated basalts, or of erosion irregularities in the underlying surface. 4. The surface streams are only locally carrying iron in quantity at the present tune. All available analyses of river waters show a lack of iron, with the exception of minute quantities in Ottawa and St. Lawrence rivers. Many of the springs carry iron, but this is conspicuously deposited at the point of escape and does not join the run-off. These facts are correlated with known observations of the maimer of weathering of rocks. The ferrous iron becomes oxidized and, next to alumina, is the most stable of all substances under surface conditions. In fact, so little iron is lost by weathering that Merrill, Watson, and others have used both iron and alumina as a basis against wliich to measure the loss of other constituents. 5. If it is regai'dod as possible that the iron-bearing formations are derived from the weath- ering of ordinary land surfaces, why should the ii-on-bearing formations not be reproduced on the same scale in the Paleozoic rocks, which were deposited on pre-Cambrian rocks similar to those beneath the iron-bearing formations ? The deposition of the Paleozoic rocks was preceded by perhaps the longest period of weathering of which there is record in the Lake Superior coun- try. In many parts of the United States Paleozoic and later sediments contain thin beds of sedimentary iron-bearing formation, but these beds are at their maximum insignificant in thick- ness as comiiared with those of the Lake Superior region. THE IKON ORES. 505 6. A comparison of the composition of the iron-bearing series with the possible sources from which they might be derived by ordinary weathering further shows that the iron is present in higher ]:)ercentage in the iron-bearing formations than in the rocks from which they may have been so derived. The jaspers of the Keewatin series of the Vermilion district average between 28 and 38 per cent in iron, but the associated basalts average 0.56 per cent. The jaspers have little other sedimentary material with them to be figured in this comparison. Therefore the jaspers pi'obably derived their iron from some other source than the weathering of the adjacent basalts, or the complementary fragmental detritus was washed away. The middle Huronian, containing the iron-bearing Negaunee formation, has an average iron content of 11.72 per cent, as indicated by the available figures of composition of the three formations of the middle Huronian and their relative tliickness. Because of the unconformity at the top there is a question as to what factor should be added for materials that have been eroded, but there is no evidence that any large amount of material has been taken away, and as part of the material which has been removed belonged to the iron-bearing formation, this factor can not be assumed to cause much change in the figures given. The composition of the rocks of the ancient land area from which the middle Huronian may have been derived by weathering is not definitely known, but it may be supposed to be not far from the average given by Clarke" for igneous and crystalhne rocks, in which the iron content is 4.46 per cent. Were the shore made up of basic rocks such as the Kitchi schist or Mona schist the iron content would be about 9 per cent. It is thus apparent that, whether we regard average igneous rocks or basic rocks as representing the original land from which the middle Huronian may have been derived by weathering, the sediments contain a considerable excess of ii'on not accounted for. The iroii content of the upper Huronian of the Mesabi and Gogebic districts ranges from 6 to 9 per cent, depending on the thickness of slate which is chosen for the calculation. The smallest percentage is liigher than that of the average igneous rock that may be supposed to represent the land area from which these sediments were derived. The highest is about equal to the percentage of iron in the greenstones. In general, then, if it is assumed that all of the iron of the ancient land areas was trans- ferred and contributed to the iron-bearing formations that were being deposited in neighboring submerged areas (which, as above shown, it was not), this would not be enough to account for the iron in the iron-bearing rocks when the associated sediments are taken into account and allowance made for complementary secUments deposited elsewhei'e. The major part of the iron of the iron-bearing formations was originall}' deposited as a chemical secUment from solution. In view of the fact that in weatheiing only a small propor- tion of the iron present is observed to be carried off in solution, the rest remaining as insoluble ferric oxide, it becomes even more apparent that the iron-bearing formations were not derived by chemical solution and deposition of the materials of average land areas. A similar conclu- sion is to be drawn from the silica content in the iron-bearing formations. ■ Sihca of course is derived abundantly from the weathering of rocks in cold solutions and is precipitated principally in the form of chert in limestones. The part mechanicall}' carried is deposited as c^uartz sand, differing in texture from the chert. The latter mode of derivation is practically excluded for the iron-bearing formations of the Lake Superior region because they contain only small amounts of fragmental quartz at a few localities and horizons. If we attempt to ascribe the cherts of the iron-bearing formations to weathering, we ma}' look only to the sihca carried in solution. To have produced the thick iron-bearing formations contain- ing an average of about 70 per cent by volume of chert, the. solution of silica must have pro- ceeded on an enormous scale, probably too large to be explained by ordinary weathering. That some chert was so derived, just as some iron and some fragmental quartz were so derived, is altogether likely, and it would be difficult to prove the contrary. The percentage of chert in the iron- bearing groups described on page 461 ranges upward from 63 per cent in weight, o Clarke, F. W., The data of geochemistry: Bull. U. S. Geol. Surrey No. 330, 1908, p. 26. 506 GEOLOGY OF THE LAKE SUPERIOR REGION. wliile Clarke's average of igneous and crystalline rocks, which jpoight represent the composition of an average surface under weathering, is a little less than 62 per cent in silica and the basic wreenstones contain less than 50 per cent in silica. Hence, even if all the siUca had been leached (together witli the iron) from these rocks (which never happens), it would not j-icld a percent- age of silica as large as that known in the iron-bearing groups. Organic agencies might locahze precipitation of silica in certain areas, but not enough to account for existing pro- portions over tlu^ entire region. The calcium-magnesium content furnishes still another argument. In the average crj^stal- line or igneous rocks or in the basic igneous rocks or in sediments derived from the igneous rocks, calcium preponderates over magnesium, but in the iron-bearing formations the average proportion of magnesium to calcium is over 5 to 1 . It appears m general that the composition of the pre-Cambrian sedimentary groups con- taining the iron-bearing formations (Uffers from that of the average crystalline rocks wliich formed the shores at those periods in having a liigher content of iron and sUica and in having a tlifTcrent calcium-magnesium ratio. It might be that the extensions of these iron-beaiing sedimentary groups outside of the Lake Supeiior region would be of such different composition as to brmg the average more nearly down to what would be expected from derivatives of the crystallme rocks. Yet it is beUeved that the excess of certain constituents in the Lake Superior sedimentary groups that carry the iron-bearing formations over those wliich seem to have been probably available from ordinary weathering is not counterbalanced by corresponding defi- ciencies elsewhere, for the reason that the sections on which these figures are based are taken through a wide area in the Lake Superior region, and for the further reason that this peculiar composition is repeated over this wide area in the rocks of three successive geologic epochs. If the occurrence of iron-bearing formations in the Lake Superior region is simply a matter of areal segregation,and concentration of the normal products of weathermg, it is verv remarkable that this areal concentration should always have resulted in bringing these peculiar iron- bearing phases in the same region. We conclude, therefore, that the excess of iron and silica and the reversal of the calcium-magnesium ratio in the sedimentary groups carrying the iron- bearing formations, as compared with the average crystalline rocks from wliich they might have been derived by erosion, is probably to be regarded as evidence that some unusual source of material was available. 7. It appears, then, from the foregoing paragraphs that there are objections to regarding the iron-bearing formations entirely as sediments produced by weathering of the rocks that were most abundant in the adjacent lands. It is not meant to imply that ordinary erosion and katamorpliic processes which are known to segregate iron-bearing sediments were set aside in this region. Indeed, as already indicated, there is definite evidence that some of the kon-beaiing sediments were so produced. But it seems that these processes are not adequate to explain the facts. In character and size the iron-bearing formations are unique as chemical sediments and differ from other chemical sediments derived by normal weathering processes. Some unuSual and additional factor seems to be required to explam them. Such a factor is discussed under the following headings. ASSOCIATION OF IRON-BEARING SEDIMENTS WITH CONTEMPORANEOUS ERUPTIVE ROCKS. • All the Lake Superior iron-bearing formations are more or less closely related in time and place to basalt Hows, usually rich in iron at j)resent and giving evidence of having exudetl iron salts at the time of their consolidation. The iron-bearkig formations of the Keewatin series have such relations to the associated ellipsoidal basalts as to point to their do]iositionin the short periods separating the successive flows of basalt or inimediatel}- followmg (ho prui- cipal extnisions. Detailed evidence of this has been noted in a number of places and especially in the Vermilion distiict. (See pp. 126-127.) The Negaunee fornialion of the middle Iluronian is associated with abundant contemporaneous igneous activity, iiroducing ellipsoidal basalts of submarine origin and other extrusive rocks similar to those in the Keewatin series in many THE IRON ORES. 507 places in the Marquette district, especially at the west end (the volcanic Clarksburg forma- tion), and in the Crystal Falls and adjacent districts (the Hemlock formation). The iron- bearing formations of the upper Huronian (Animikie grou]^) are associated with igneous activity similar to that of the preceding periods in the Marquette district (the Clarksburg formation), in the Gogebic district (the volcanic rocks of the east and west ends of the district) , and in the Menominee, Florence, and Iron River districts. The iron-beaiing formation of the Animikie group on the north shore of Lake Superior is not associated with basic greenstones of known contemporaneous development, but as shown on pages 213-214 there is little doubt of its direct contmuity with the rocks of the Cuyuna district and the upper Huronian of the south shore, wliich are associated with basic volcanic rocks. Especially remarkable are the evidences of the close association of iron-bearing sediments and basaltic flows in the upper Huronian of Michigan. Here ellipsoidal basalt, basalt tuffs, and ashes are so intermuigled with the iron-beariiig formation and stained by secondary alter- ation that there is difficulty in discriminating them. Recent work has shown the existence of more of the igneous rocks than had before been suspected. Drill holes in tlie Iron River and Amasa areas of Michigan pass through igneous beds from 2 to 50 feet thick in the midst of the iron-bearing formation. In these places the eye can scarcely detect the break between the grayish and greenish carbonate slates of the iron-bearmg sediments and the fine-grained greenish basalts and tuffs. Under the microscope the surface of contact is seen to be an extremely irregular one, the carbonate apparently irregularly replacmg part of the greenstone. This replacement has not been accompanied by any oxidation. It is found in drill holes hundreds of feet beneath the surface, apparently in an association determined at the time of the deposition of the iron- bearing formation. In the Keewatin of the Vermilion district of Minnesota similar close asso- ciation may be observed between the jaspers and the basalts. (See PI. XLVIII, p. 564.) The significance of the apparent gradation of carbonate of iron and siUca and their altera- tion products into the greenstone is not yet fuUy apparent. It can scarcely be doubted that tliis relation was developed at the time of the deposition of the iron-bearing formation, prob- ably soon after the extrusion of the igneoiis rocks. It is suspected that these phases represent a transition between reactions associated with the hot igneous masses and the normal precipi- tation of a sedimentary formation. Attempt has been made in the laboratory to reproduce these remarkably close relations by some combination of igneous and sedimentaiy processes, but thus far without successful results. Probably of significance in connection with the derivation of the iron-bearing formations is the fact that in many places acidic intrusive and extrusive rocks of the porphyry type closely foUo^v extrusive basalts and are locally even more closely associated with the iron-bearing for- mations than the basalts themselves. This relation is well illustrated in the Vermilion district, where, in a series of mterbedded basalt flows, jaspers, and amygdaloidal porphyries, the igneous rock immediately next to the jaspers is commonly porphyry as weU as basalt. (See fig. 13, p. 123.) Similar conditions appear in the Woman River district of Ontario" and elsewhere. It is suspected that this relation is more general than is yet known. (See p. 513.) The amount of igneous material extruded is not measured by the areas of upper Huronian volcanic rocks now exposed, for extensive extrusive rocks were undoubtedly present in parts of the formation that have been removed by erosion and exist in parts not yet uncovered. It is suggested in the chapter on the Keweenawan (Chapter XV) that the present shore of the Lake Superior basin was the locus of the extrusion of the Keweenawan igneous rocks. If the basin began to form in Animikie time, as is thought possible (see pp. 622-623), a siiaular suggestion, for similar reasons, might be made for the Animikie group, in which case the north shore Ani- mikie may really not be so distant from igneous rocks as now appears. The iron-bearing formation of the Animikie group of the north shore is thus associated in time with igneous extrusions, but may be somewhat distant in place. a Allen, R. C, Iron formation of Woman Elver area; Eighteenth Ann. Rept. Ontario Bur. Mines, pt. 1, 1909, pp. 254-262. 508 GEOLOGY OF THE LAKE SUPERIOR REGION. The deposition of the lower Huronian was not accompanied by basic flows, and it does not contain a well-developed iron-bearing formation. The Paleozoic of the Lake Superior region lack.s basic igneous rocks and also lacks iron-bearing formations like those of the pre-Cambrian. ASSOCIATION OF IRON-BEARING SEDIMENTS AND ERUPTIVE ROCKS OUTSIDE OF THE LAKE SUPERIOR REGION. The derivation of the iron-bearing formations from the associated igneous rocks is sug- gested by the close association of these rocks not only in the Lake Superior country but in other ])arts of the world. Practically all the numerous iron-beaiing sediments extending through the Height of Land country of Canada, as far east as the Quebec boundary, are interbedded with basalt flows. Most of these belts, in the writers' judgment, belong in the Keewatin. On the east coast of Hudson Bay there are younger Algonkian rocks containing an iron- beaiing formation, interbedded wdth fragmental sediments and elhpsoidal basalts. As Low " had called attention to the similarity of these iron-bearing sediments to those of the upper Huronian or Animikie of the Lake Superior region, the junior author visited them in 1909 and found a veiy close similarity, even to the possession of carbonate and greenalite phases. Freedom from vegetation and precipitous shores afford fine exposures for study. Fragmental sediments of the type- now bemg formed along the shores are interbedded with extiiisions of elhpsoidal basalt which give evidence by their textures and associations of having been extruded along tidal flats, and by their high content of jasper and magnetite of having been rich in iron salts at the time of their extrusion. Immediately following the basalt comes the iron-bearing formation, closely associated with volcanic muds. It requires no preconceived hypothesis to lead the observer to the view that the extrusion of the igneous rocks was the variant in the normal conditions of sedimentation necessary to produce the iron-bearing formations. The story is so clear that it is possible to outhne the probable conditions of sedimentation in some detail." Geikie "^ remarks concerning lower Carboniferous basalts of the Fife coast: These lavas are thin sheets, often not more than 15 or 20 feet in thickness, and they, as well as the associated tuffs are intercalated among shallow-water deposits, such as cyprid shales and limestones, coal seams with fire clays, thin sandstones, and ironstones. Some of the basalts have caught up portions of the mud on the sea bottom, but in others the muddy, sandy, or ashy sediment of the next deposit has fallen into the interspaces between the pillows. He also says'* concerning the basaltic lavas of County Tyrone, Ireland: These greenish lavas are occasionally interleaved with gray flinty mudstones, cherts, and red jaspers, which are more particularly developed immediately above. In lithological character, and in their relation to the diabases, tliese siliceous bands bear the closest resemblance to those of Arenig age in Scotland, but no recognizable Kadiolaria have yet been detected in them. Describing the Carboniferous volcanoes of the Isle of Man, Geikie e says : Pauses in the succession of eruptions are marked by the intercalation of seams of limestone or groups of limestone, shale, and black impure chert. Such interstratifications are sometimes curiously local and interrupted. They may be observed to die out rapidly, thereby allowing the tuff above and below tliem to unite into one continuous mass. They seem to have been accumulated in hollows of the tuff during somewhat prolonged inter\-als of volcanic quiescence, and to have been suddenly brought to an end by a renewal of the eruptions. There are some four or five such intercalated groups of calcareous strata in the thick series of tuffs, and we may regard them as marking the chief pauses in the continuity or energy of the volcanic explosions. Again, Geilde ^ states that in the Carboniferous volcanoes of Devonshire — Bands of black chert and cherty shale are interpolated among the tuffs, which also contain here and there nodular lumps of similar black impure earthy chert — an interesting association like that alluded to as occurring in the l"ar- lK)niferous volcanic series of the Isle of Man, and like the occurrence of the radiohirian cherts with the Lower Silurian volcanic series. o Low, .V. p.. Report on an e>qiloration of the east coast of Hudson Bay from Cape Wolstcnholme to the south end of James Bay: .\nn. Kept. Geol. Survey Canada, vol. 13, new stT., pi. L), l'ju;i, pp. 45—10. <> I.cith, C. K., .\n .Mgonkitin liusin in Hudson Hay — a comparison with the Lake Superior basin: Econ. Geology, vol. 5, 1910, pp. 227-246. c .\bstracts I'roc. Oeoi. Soc. London, session 19()T-S, London, 1908, p. 42. d Geikie, j\jchil)ald, Ancient volcanoes of Great Britain, vol. 1, London, 1897, pp. 240-241. <• Idem, vol. 2, 1897, p. 24. / Idem, vol. 2, 1897, p. 36. THE IRON OKES. 509 The following section in Tertiary volcanoes of the Antrim Plateau of Ireland is described by the same author:" Upper basalt, compact and often columnar sheets. Brown laminated tuff and volcanic clays. Laminated brown impure earthy lignite, 2 feet 3 inches. Brown and red variegated clays, tuffs, and sandy layers, with irregular seams of coarse conglomerate composed of rounded and sul>angular fi'agments of rhyolite and ba.salt, 3 feet 4 inches. Brown, red, and yellowish laminated tuffs, mudstones, and bole, with occasional layers of fine con- glomerate (rhyolitic and basaltic), pisolitic iron-ore band, and plant beds, 8 feet 10 inches. Lower basalt, amygdaloidal. The pale and colored clays that occur in this marked sedimentary intercalation have doubtless been produced by the decomposition of the volcanic rocks and the washing of their fine detritus by water. Possibly this decay may have been in part the result of solfataric action. * * * * * * The original area over which the iron ore and its accompanying tuffs and clays were laid down can hardly have been less than 1,000 square miles. This extensive tract was evidently the site of a lake during the volcanic period, formed by a sulisidence of the floor of the lower basalts. The salts of iron contained in solution in the water, whether derived horn the decay of the surrounding lavas or from the discharges of chalybeate springs, were precipitated as peroxide in pisolitic form, as similar ores are now being formed on lake bottoms in Sweden. For a long interval quiet sedimentation went on in this lake, the only sign of volcanic energy during that time being the dust and stones that were thrown out and fell over the water basin or were washed into it by rains from the cones of the lava slopes around. Concerning the Tertiary volcanoes of the plateau of Small Isles, Geikie* writes: It is a noteworthy fact that the sedimentary intercalations among the Canna basalts generally end upward in carbonaceous shales or coaly layers. The strong currents and overflows of water, which rolled and spread out the coarse materials of the conglomerates, gave way to quieter conditions that allowed silt and mud to gather over the water bottom, while leaves and other fragments of vegetation, blown or washed into these quiet reaches, were the last of the suspended materials to sink to the bottom. The Arenig eruptions in the Silurian of North Wales contain interesting sediments, described by Geikie '^ as follows: Many of the tuffs that are interstratified with black slates (? Lingula flags) at the foot of the long northern slope of Cader Idris consist mainly of black-slate fragments like the slate underneath, with a variable proportion of gray volcanic dust. * * * One of the most interesting deposits of these interludes of quiescence is that of the pisolitic ironstone and its accompanying strata on the north front of Cader Idris. A coarse pumiceous conglomerate with large slaglike blocks of andesite and other rocks, seen near Llyn-y-Gadr, passes upward into a fine bluish grit and shale, among which lies the bed of pisolitic (or rather oolitic) ironstone which is so widely diffused over North Wales. The finely oolitic structure of this band is obviously original, but the substance was probalily deposited as carbonate of lime under quiet conditions of precipitation. The presence of numerous small Lingidie in the rock shows that molluscan life flourished on the spot at the time. The iron exists in the ore mainly as magnetite, the original calcite or aragonite having been first replaced by carbonate of iron, which was subsequently broken up so as to leave a residue of minute cubes of magnetite. Radiolarian cherts are characteristically associated with sandstones and basalts, partly ellipsoidal, at Point Bonita,'' Angel Island,^ and at many other points in the Coast Ranges of California. In describing the eruptive rocks of Point Bonita, Ransome saj's: ^ Spheroidal basalt, apparently similar to that described, has been noted by the writer at Tiburon, Marin County at Port Harford, San Luis Obispo County; and on the summit of the north peak of Mount Diablo. It is noteworthy that in these widely separated occurrences the rock is always associated with the red jaspers, and with what is apparently the San Francisco sandstone. These cherts were called "phthanites" by Becker ff and regarded as due to secondary silici- fication. Lawson '' and Ransome,^' on the other hand, regard them as original siliceous deposits a Geikie, Arcliibald , Ancient volcanoes of Great Britain, vol. 2, 1897, pp. 204-205. t Idem, vol. 2, 1897, p. 223. c Idem, vol. 1, 1897, pp. 180-lSl. dRansome, F. L., The eruptive rocks of Point Bonita: Bull. Dept. Geology, Univ. California, vol. 1, 1893, pp. 71-114. e Ransome, F. L., The geology of Angel Island; Bull. Dept. Geology Univ. California, vol. 1, 1S94, pp. 193-240. / Ransome, F. L., The eruptive roclcs of Point Bonita: Bull. Dept. Geology Univ. California, vol. 1. 1893. pp. 109-110. e Becker, G. F., Geology of the quicksilver deposits of the Pacific coast: Mon. U. S. Geol. Survey, vol. 13, 1SS8, pp. 10.5-108. * Lawson. A. C, Sketch of the geology of the San Francisco peninsula: Fifteenth Ann. Kept. U. S. Geol. Survey, 1895, pp. 420-426. < Ransome, F. L., The geology of Angel Island: Bull. Dept. Geology Univ. California, vol. 1, 1894, p. 200. 510 GEOLOGY OF THE LAKE SUPERIOR REGION. which are changed into red jaspers and glaucophanic jaspers here and there at igneous contacts. These cherts locally pass into iron ore and are characteristically associated with njanganese beds." The cherts are characterized by mbiute oval spots found in part to represent radio- larian remains, but in part of unknown origin. Lawson'' discusses their origin as follows: It thus seems to the writer that the bulk of the silica can not be proved to be the extremely altered d(5bris of Radiolaria. The direct petrographical suggestion is that they are chemical precipitates. If now we accept this hj^poth- esis, it becomes apparent that there are three possible sources for the silica so precipitated, \'iz, (1) siliceous springs ■in the bottom of the ocean, similar to those well known in volcanic regions; (2) radiolarian and other siliceous remains, which may have become entirely dissolved in sea water; and (3) volcanic ejectamonta, which may have become similarly dissolved. The last is the least probable, because we are not actually familiar with such a reaction as the solution of volcanic glass by sea water. Our ignorance is, however, no proof that such solution may not take place under special conditions. * * * The hypothesis of the derivation of the silica from siliceous springs and its precipitation in the bed of the ocean in local accumulations, in which radiolarian remains became embedded as they dropped to the bottom, seems, there- fore, the most adequate to explain the facts, and there is nothing adverse to it so far as the writer is aware. The abun- dance of the Radiolaria may be due to the favorable conditions involved in the excessive amount of silica locally present in the sea, or simply to the favorable conditions for preservation afforded by this kind of rock. If the springs were strong, the currents engendered might in some places have been sufficient to deflect sediment-laden counter-currents, and this may serve to explain the general absence of clastic material in the chert. The Pilot Knob deposits of Missouri are interbedded with porphyry flows, tuffs, and ashes, suggesting close genetic relation between igneous rocks and sediments. Illustrations could be multiplied, but enough have been cited to show that basalts, espe- cially the ellipsoidal phases, are characteristically interbedded with more or less graphitic slates, clays, cherts, jaspers, volcanic tuffs, iron ores, and in places sandstone. Practically all the features of the association of basalt with sediments described for the above-mentioned districts are to be seen in the Lake Superior region. The explanations of these associations in other regions therefore become significant in the study of the origm of the Lake Superior ores. In general there seems to be little doubt that some genetic relationship e-xists among surface basalts, carbonaceous slates, cherts, and jaspers, to which attention has been called by several writers worldng from different standpomts."^ They agree that most of the carbonaceous materials are organic, that the deposition is largely subacjueous, and that some of the associated iron is deposited partly through the agency of weathering assisted by organic means. Lawson suggests that the cherts and jaspers may be the result of inorganic chemical deposition by hot solutions. In the Lake Superior region the iron-bearing formations are much thicker and they have certain phases, notably the greenalite or ferrous silicate phase, which are not common elsewhere, all these features seeming to favor the hj-pothesis that the iron formations are in part related to the more or less direct contribution of the iron-bearing materials by hot concentrated solutions from the igneous roclis. SIGNIFICANCE OF ELLIPSOIDAL STKUCTURE OF EBtTPTIVE ROCKS IN RELATION TO ORIGIN OF THE ORES. The basalts associated with the iron-bearing formations have so commonl}' the peculiar ellipsoidal or pillow structure that one is led to assume that conditions favorable to the develop- ment of the ellipsoidal structure may be also favorable to the deposition of the iron ore in this district. Clements ** has described the structure in some detail for the Crj'stal Falls district, and from comparison with occurrences elsewhere concludes it to have been probably a submarine extrusive, similar to the aa lavas of Hawaii described by Button.'' Dah^ f reaches the same a Lawson, A. C, op. cit., pp. 423-424. b Idem, pp. 425-42C. c Wo have received too late for discussion a paper on British pillow lavaa and the rocks associated with them, by nenrj- Dewey and J. S. Fleet (Gcol. Mag., vol. 8. Dec. 5, 1911, pp. 202-209, 241-24S), emphasizing the genetic association of cherts and ellipsoidal iHisalts. .Vlbilization of thefeldspars of the basalts is regarded as evidence of pneuraatolytic emanations, containhig soda and silica in solution and possibly other sub- stances. The cherts are deposited by those emanations. This independent conclusion is remarkably in accord with the inferences drawn in this monograph. Geikie, .\rchibald, Ancient volcanoes oJ Great Britain, London, 1897. i cReid, Clement, and Dewey, Henry, The origin of the pillow lava near Port Isaac in Cornwall; .\bstracts Proc. Geol. Soc. London, session 1907-8, London, 1908, p. 42. rfldem. eFenner, C. N., Featuresof trap extrusions in New Jersey: Jour. Geology, vol. ll'i, 1908, p. 320. /Cole, G. A. J., and Gregory, J. W., On the variolitic rocks of Mont Gen(>vre: Quart. Jour. Geol. Poc, vol. 40, 1890, p. 310. ffRansome, F. L., The eruptive rocks of Point Bonita, California: Bull. Dept. Geology, Univ. California, vol. 1, 1S93, p, 112. ^ Russell, I. C, Geology and water resources of the Snake River plains of Idaho: Bull. U. S. Geoi. Survey, No. 199, 1902, pp. 82 e( seq. ildem, p. 98. ; Dana, J. D., Characteristics of volcanoes, New York, 1890, pp. 242-244. i- Cited in .Vbstracts Proc. Geol. Soc. London, session 1907-8, London, 190S, p. 42. Udem, p. 44. 512 GEOLOGY OF THE LAKE SUPERIOR REGION. ellipsoidal structure of the Lake vSuperior basalts is largely of subaqueous origin. It should not be a.ssumed, however, that all the ellipsoidal basalts of the Lake Superior region are neces- sarily subaqueous. The region is a large one, the conditions arc varied, the ellipsoidal struc- tures are locally associated with structures ordinarily regarded as of subaerial origin, ellipsoidal structure is known elsewhere to develop subaerially, hence it is rather likely that a part of the structures in the Lake Superior region are of subaerial origin. There is little prospect that evidence will be forthcoming to determine exactly the quantitative importance of the subaerial deposit as compared with the subaqueous deposit; indeed, there seems to be little need of such determination wlien it is recognized tliat both are present. Qualitativel}' the evidence favors the subaqueous origin of the major part of the ellipsoidal basalts. ERUPTIVE ROCKS ASSOCIATED WITH IRON-BEARING SEDIMENTS OF LAKE SUPERIOR REGION CARRY ABUNDANT IRON. Abundant sulpliides and associated magnetite are disseminated in quartz veins and irregular quartz masses through the ellipsoidal greenstones of the Lake Superior region antl of much of the pre-Cambrian shield of Canada. The abundance of these sulphides through all parts of these greenstones has been noted by many observers. They are exceptionally conspicuous in the Canadian part of the region, where erosion has cut down into the fresh rocks and exposed sulpiride veins that have not had time to be deeply oxidized at the surface since the glacial epoch. That certain of the sulpliides and the associated magnetites of the basic igneous rocks crystal- lized soon after the crystallization of the igneous roclcs, and are not later secondary replacements of such rocks, is shown by evidence of several kinds, as follows: 1. They are minutely disseminated tlirough the greenstone and grade into pegmatitic veins. 2. The sulphides and the greenstones of this t^i^e are colimital, and the sulphides are not found so abundantly in any other rocks, a fact wliich would be difficult to explain were the sulphides the result of later introduction by percolating meteoric waters or by later extrusions. 3. The matrix of the ellipsoidal basalt flows is in places so higlily charged with magnetite as to disturb the magnetic needle greatly, and the amount of magnetite is much less at the ellipsoids. Illustrations of this are found in the Hemlock formation in the vicinity of the Armenia and Mansfield mines, in the Crj'stal Falls district of Michigan, and in the Keewatin basalts associated with jaspers southwest of Elyj in the Vermilion district of Minnesota. The matrix being the last part of these masses to crystallize, the magnetite is obviously introduced late in the extrusion of the mass. Sulphide of iron is present in the same relations. 4. Many of the amygdules in the basalts are wholly or partly filled \\dth magnetite or jasper, or both. Near the Gibson mine, south of Amasa, in the Crystal Falls district of ilichi- gan, red jasper fillings in amygdaloids are verj' conspicuous. The amygdule fillings in general are characteristic of hot solutions such as would accompany the extrusion of the mass and not of cold meteoric solutions. (See PI. XXXYI, A.) 5. Plate XLVIII (p. 564) shows a Keewatin basalt with gradation phase through siliceous basalt into banded sihceous iron-bearing formation. In the area from wluch these specimens were taken, as well as in other parts of the Iveewatin, it is practically impossible to draw a line between unaltered basalt and the iron-bearing formation. Tliis gradation seems to be one developed on the original sohdification of the mass. The fi-eshness of the basalt, the lack of katamorphism along the contact with the quartz, and the extremely vague surfaces and general lack of vein structures are not characteristic of later introductions of the quartz after weathering. 6. Some parts of the magnetic iron-bearing formations are so related to the associated basalts as to suggest that the iron represents pegmatitic vein material wliich developed directly from the igneous rock. Such instances are cited for the Atikokan and Vermilion districts. Evidence is everywhere to be found that these various iron salts associated with the surface extrusive rocks represent remnants of outpourings of concentrateil iron solutions after the main mass of the basalt had crystallized. Deep-seated equivalents of the basaltic extrusive rocks are believed to bo the gabbros which carry large masses of titaniferous magnetite representing iron salts that diil not have an opportunity to escape at the surface. THE IRON ORES. 513 The fact that in some places the iron-bearing formation seems to be related to late acidic phases of extrusions, us has been noted on page 507, suggests the extrusion of the iron and the acidic phases as extreme differentiation products from the magma. The association of extremes of this type is not uncommon. GENETIC RELATIONS OF UPPER HURONIAN SLATE TO ASSOCIATED ERtTPTIVE ROCKS. The iron-bearing formations of the upper Huronian are so closely associated with slate that evidence bearing on the origin of the slate throws light on the origin of the associated iron-bearing formations. In figure 76 (p. 612), prepared by S. II. Davis, the miner-ilogical com- position of the upper Iluronian slate, calculated from chemical composition, is compared graphi- cally with that of a variety of other claj^s and soils. It appears from this comparison that the slate as a whole gives evidence hj its composition of bemg less leached of its bases than average slates or residual clays and that it has been derived from basic rocks. It may be due partly to weathering of the greenstones, to direct contribution of volcanic ash and muds, and possibl}' even to direct reaction with sea water. (See pp. 610-614.) MAIN MASS OF IRON-BEARING SEDIMENTS PROBABLY DERIVED FROM ASSOCIATED ERUPTIVE ROCKS. The close association of iron-bearing sediments with contemporaneous basic eruptive rocks in the Lake Superior region and in other parts of the world, the riclmess of these eruptive rocks m iron salts, and the probable derivation of the upper Huronian slates associated with the iron-bearing formations from the eruptions make it a plausible hypothesis that these iron- rich eruptive rocks were the principal source of the iron in the iron-bearing sediments. As to the manner in which the iron was transferred from the eruptive rocks to the place of sedi- mentation, there are several possible hypotheses. (1) It may have been transferred in hot solutions migrating from the eruptive material during its solidification, carrying iron salts from the interior of the magma which had never been crystallized; (2) so far as the lavas were subaerially extruded, iron may have been transferred by the action of meteoric waters working upon the crystallized iron minerals in the magma, either hot or cold; (.3) the iron may have been transferred by direct reaction of the hot magma with sea water, in which the iron-bearing sediments were deposited. DIRECT CONTRIBUTION OF IRON SALTS IN HOT SOLUTIONS FROM THE MAGMA. That the igneous rocks contributed some of their iron solutions directly to the water in which the iron-bearing sediments were being deposited is suggested by the fact that basic extrusive rocks have a widely developed ellipsoidal structure, which has been ascribed by many observers to submarine extrusion. (See pp. 510-512.) If these lavas are submarine, then any iron salts extruded must have been contributed directly to the ocean. It will be shown in the following pages that if the salts were so contributed simple and probable chemical reactions would develop the original greenalite or iron silicate phases of the iron-bearing formations. Such phases largely lack the carbonaceous slates so closely associated with the carbonates. It was found in the laboratory that the precipitation of the greenalite phase of the iron-bearing formations required heat in the presence of carbon dioxide and the probable presence of salt water, in both contrasting with the precipitation of iron carbonate, wliich goes on in cold solution, favored by the presence of reducmg organic agencies. Direct contribution would favor the deposition of the iron salts in a ferrous condition in the absence of reducing carbonaceous material and would avoid the oxidation and precipitation which they woiUd undergo if partly carried subaerially. Further, the fact that iron-bearing formation seems to be lacking in association with certain similar greenstones in the Lake Superior region and Canada may be evidence that the iron-bearing formations derive tlieir materials by direct magmatic contributions. Such con- 47517°— vol. 52—11 33 514 GEOLOGY OF THE LAIvE SUPERIOR REGION. tributions are known to be local and variable in composition, and this may explain the localized distribution of the iron-bearing formations. If derived entirely by weathering of basic igneous rocks, iron-bearing formations should be more abundant in association with igneous rocks outside of the Lake Superior region. The percentages of both iron and silica in the iron-bearing formations seem to be too high for direct derivation from crystallized basalt by weathering. Tiioy soeni to accf)r(l better with the hyi)othesis that the iron and silica, especially tlie silica, were precipitated from concentrated solutions coming directly from the magma. The local presence of acidic igneous rocks between the lavas and the basalts and tlie fact that the acidic rocks are slightly later than tlie basalts suggest that the development of the iron-bearing formation came at a time when acidic phases of the extrusion were coming out. The iron salts and the acidic phases then might represent the extreme differentiation products of a primary magma of which the basalt was the first extrusion. Favoring the hypothesis of direct contribution of the iron salts from the lava to the sea water into which it was poured is the lack in many places of any fragmental material between the ia'on-bearing formation and the contemporaneous lava on which it rests, the mutual con- formity at these places, and the absence of any erosion channels in tlie greenstones. In the Vermilion district of Minnesota bands of iron-bearing formation have been traced for consider- able distances resting directly upon the amygdaloidal upper surface of a lava How, showing no evidence of intervening erosion and having a contact like a knife edge. The subacpieous extrusion of igneous rocks would mean the sudden destruction of any organic material m the near-by sea, to judge from results observed near present-day extrusions. It has been shown that after an eruption the sea floor has been covered to a depth of several feet off Hawaii by dead fish and other organic material. It is entirely jjossible that this may explain the origin of some of the carbonaceous materials so closely associated with the iron- bearing formations, especially in the Keewatin, where seams of rich graphitic slate are locally associated with the iron-bearing formation and the basalt. It is possible also that this material might be a source for the carbon dioxide necessary for the formation of the iron carbonates. Quantitatively it is probably inadequate to explain either the amount of carbon dioxide neces- sary for the formation of the iron carbonates or the amounts of carbon to be seen in the associated slates. It is mentioned merely as a possible source of a part of these substances. Its importance can not be quantitatively demonstrated. So far as the parent igneous rocks were extruded subaerially, the escaping iron solutions would be mingled with meteoric waters, perhaps deriving additional iron salts from the breaking up of crystallized minerals described under the next heading. CONTRIBUTION OF IRON SALTS FROM CRYSTALLIZED IGNEOUS ROCKS IN METEORIC WATERS. Some of the basaltic extrusive rocks have textures indicative of subacrial crystallization. Atmospheric agencies, therefore, have been applied during the transfer of the iron solutions to the ocean. Weathering agents would effectively attack sulphif acidic solutions of tliis kind in decomi)osing the rocks antl segregating the iron. The marked softening and disintegration of the rocks nuij^ furnish a source for the unusually large amount of basic mud associated with o Maxwell, Walter, Lavas and soils ot the Hawaiian Islands! Rept. Exper. Sta. Uaivaiian Sugar Planters' Assoc, Div. Agr. and Chem., Special Bull. A, Honolulu, 1905, pp. S-22. THE IRON ORES. 515 the iron-bearing formation. It is entirely conceivable that some of the thin bands of the iron-bearing formation interbedded with basic flows, with little other sedimenta,ry material, may be essentially residual iron oxide or laterite deposits developed in this way. This seems especially likely where the iron-bearing formation is high in alumina, as, for instance, in some of the hornblendic Keewatin belts or in the iron ranges near Lake Nipigon, where E. S. Moore" has found dumortierite. However, the generally low percentage of alumina in the iron-bearing formations seems to show that for the most part they may not be regarded as metamorphosed residual products of rock alteration. Vegetation is known to develop on basic extrusive lavas with great rapidity, as indicated by the cultivation of the slopes of Vesuvius and Hawaiian volcanoes m an incretUljly short time after eruptions, and hence organic agencies may have aided in the transfer. The chemistry of the transfer of iron salts through these agencies is discussed elsewhere (pp. .519- .520). Favor- ing the view that weathering is a factor in the process is the fact that parts of the original rocks of the iron-bearing formation are made up of iron carbonate associated with black car- bonaceous slates, such as may have developed in delta deposits. (See p. 502.) There is no more reason to doubt the organic origin of the carbon in these slates than that of the carbon in the carbonaceous slates, iron-bearing formation, and basalts in County Antrim, Ireland, and elsewhere, except that definite organic forms are lacking. The iron-bearing formations grade locally into phases rich in calcium and magnesium carbonates, as at Guniiint Lake and in the east end of the Gogebic cUstrict. It is usually assumed that calcium antl magnesium carbonates are ordinary products of weathering and sedimentary deposition. It may be asked why weathering did not also deposit iron abundantly in the Paleozoic sea when it advanced later on these same rocks. To some slight extent iron was so deposited at the Chnton horizon. The answer is believed to lie partly in the essential contemporaneity of the basic extrusive rocks with the associated iron-bearing formations, indicating that the process of derivation of the iron salts and deposition went on soon after the extrusion of the igneous rocks, very rapidly at first owing to juvenile contributions and to leaching during the residual heat, but slowly later when the rocks were colder and the easily accessible sulpliides had been reached. Still later, when the Paleozoic sea came over the area, while it derived some iron fi-om these rocks, it was unable to do the work on the same scale as was accomplished immediately after their extrusion. Since glacial time alteration of pyrites in the pre-Cambrian sliield has penetrated only a fraction of an inch or at most a few inches below the striated glacial surfaces, indicating a relatively slow alteration of these substances under ordinary weathering — probably too slow to account for the heavy and rapid chemical deposition of iron-bearing formation without admixture of fragmental material. Powdered Keewatin rocks containing abundant iron sulphide have been treated with oxygenated waters and kept agitated for a period of six weeks. A slight amount of sulphuric acid was also introduced to accelerate the alteration. At the end of this time barely enough iron had gone into solution to be detected by the most refined methods. The slate that is so abundantly present with the ujiper Huronian iron-bearing formations gives evidence in its composition of derivation from the greenstone. (See p. 612.) It is in part doubtless derived by weathering of the type here described. In part also the slate repre- sents volcanic dust and mud directly deposited from the volcanic extrusions, and in part it may result from reaction between the hot lavas and sea waters described below. CONTRIBUTION OF IRON SALTS BY REACTION OF HOT IGNEOUS ROCKS WITH SEA WATER. When basaltic magmas are extruded into the ocean there is reaction with the salt water. The behavior of basic lavas when extruded into salt water has not been carefully observed. There seems to be a tendency in Hawaii and Iceland for rapid powdering and disintegration at these contacts. What the chemical results are is not apparent. When pottery is sprayed » Geology of Onaman iron range area: Ann. Rept. Ontario Bur. Mines, vol. 18, pt. 1, 1909, pp. 212-215. 516 GEOLOGY OF THE LAKE SITPERIOR REGION. with salt water wliiie hot, a glaze of sodium siUcatc (water glass) is formed, which is more or less soluble. In connection with the present study fresh basalts were heated in a muffle furnace to a temperature of 1,200° C, a temperature sufficient to fuse the exterior, and then i)lunged into salt water of the composition of sea water, the result being a violent reaction, producing princi- pally soilium silicate (see p. 525) but also bringing a small amount of iron into solution. From the available evidence it seems likely that such a process may account for part of the sodium silicate wliich, by reaction with ferrous salts, produces the greenalite with excess of silica. (See pp. 521-523.) The experiment does not seem to suggest an adecjuate source for the iron in this reaction. There was also during tliis reaction a tendency toward disintegra- tion. Tliis may indicate ar partial source for some of the muds so closely associated with the iron-bearing formations. CONCLUSION AS TO DERIVATION OF MATERIALS FOR THE IRON-BEARING FORMATIONS. Ordinary processes of weathering, transportation, and deposition of iron salts from terranes of average composition were as effective in the pre-Cambrian of the Lake Superior region as in other times and ])laces, but these processes account for only thin and relatively unimportant phases of the iron-bearipg rocks; for instance, the lenses of iron carbonates associated \\ith graphitic slates of the upper Huronian, probably deposited in lagoons and bogs of a delta. For the derivation of the unique thick and extensive iron-bearing formations of the Lake Supe- rior region it is necessary to appeal to some further agency. This is bebeved to be furnished by the large masses of contemporaneous basic igneous rocks. The association of sedimentary iron-bearing formations and basic igneous rocks is known in mam^ localities outside of the Lake Superior region. The iron salts have been transferred from the igneous rocks to the sedimentary iron-bearing formations partly b}' weathering when the igneous rocks were hot or cold, but the evidence suggests also that they were transferred jjartly by direct contril)ution of magmatic waters from the igneous rocks and perhaps in small part by direct reaction of the sea waters upon the hot lavas. VARIATIONS OF IRON-BEARING FORMATIONS WITH DIFFERENT ERUPTIVE ROCKS AND DIFFERENT CONDITIONS OF DEPOSITION. The basalts contributmg the iron being both subaerial and subaqueous in their extrusion, it is to be expected that the contribution of iron to the bodv of water in which the iron-bearing forma- tions were being deposited was both direct and indirect. Evidence is not available which wUl clearly discriminate iron-bearing formations contributed to the ocean in these two ways. In general the parts of the iron-bearing formations originally consisting of carbonate seem to be related to the indirect contribution from the igneous rocks through the agencies of weathering, and the parts of the iron-bearing formations originally consistmg of greenalite or iron silicate seem to have been contributed in the main directly to the waters without intervening atmos- pheric or organic agencies. The locally close association of these two types of the original iron-bearuig rocks indicates the close association of direct and iiidirect methods of contribu- tion of iron-bearing materials. The fact that the upper Huronian iron-beaiing formation in the ilesabi district was largely greenalite, while the upper Huronian iron-bearing formation of the Gogebic district was lai'gely carbonate, might therefore signifj'^ simply that in one district the salts had been derived primarily from subaerial weathering and in the other from sul>- acjucous contribution, but in each district partly in both ways and in both districts essentialh' from the same I'ocks. It is noted elsewhere that in many places where the greenahtc anil carbonate occur together the greenalite occupies the lower horizon. Tliis might be explained .not only l)y conditions of sid)aerial contribution succeedmg subaciucous contribution, but, as explauied elsewhere, by the more rapid settling of the greenalite when i)recij)itated simul- taneously with the carbonate. The iron-1)oaring lavas extnided at three widely separated jierioils could scarcely be expected to produce iron-bearing formations of exactly the same character, even were the conditions of THE IRON ORES. 517 deposition the s;ime, for in so far as the ores were directly contributed by magmatic soUitioiiSj they were subject to extreme variations in composition. The conditions of deposition of iron salts were also different during these three periods of volcanism. The Keewatin lavas were extruded in larger quantities than at any later time and the associated iron-bearing formations constituted only discontinuous beds between the hot extrusives, but in the middle and upper Huronian the extrusions were much less abundant and sedimentation proceeded on a larger scale and less directly under the influence of igneous rocks. Although some of the differences between these three formations are explained by later alteration, it is believed that the highly amphibolitic and magnetitic character of the Keewatm was partly determined at the time of, or soon after, its deposition, in contrast with the prevailing deposition of ferrous carbonate and ferrous silicate at the later periods. In the discussion of the secondary concentration of the ores it will "be shown that the ores of the Keewatin have under- gone far less secondary concentration than the later ores. This is certainly in part due to anamorphic changes before the katamorphic agents had an opportunity to work, but possibly in part also to original differences in texture and composition, possibly because the Keewatin as a whole seems to contain a lower percentage of iron than the succeeding formations, and partly because of the small area of the formations exposed to concentrating agencies. (See pp. 474-475.) The Keewatin series had produced only 6.5 per cent of the total shipments to the close of 1909. The Keewatin seems to occupy the same subordinate position in Canada, and as the area of Keewatin in Canada is relatively greater than that of later iron-bearing forma- tions, the chances of finding ore- there are relatively smaller than in other parts of the Lake Superior region. It would be expected also that the iron salts closely associated with the igneous rocks would be less regular in their thickness and more generally separated into different belts by intercalated igneous rocks than those at a distance from the areas of extrusion. The latter seem to be illus- trated by the Animikie ores, which attain their maximum development on the north shore of Lake Superior, tlie nearest Icnown extrusive rocks being west of the lake or possibly under the lake. The remarkably uniform character of the iron-bearing formation and the rest of the Animilde group, distinguishing it from all other pre-Cambrian iron-bearing formations, may well be due to its distance from the contemporaneous volcanic activity, for, in view of the con- nection of the ores with igneous rocks above outlined, it would seem to be more than a coinci- dence that the most uniform and widespread of the iron-bearing formations should be the farthest removed from volcanic activitj'. Variation in the iron-bearing formations with varying distance from the igneous rocks is more definitely shown by the iron-bearing formation of the Gogebic district, which at the east end of the range, where associated with extrusive rocks, is extremely varied in its composition and is broken into different belts by other sediments and by igneous beds. The material of this portion of the formation may also originally have been deposited in small part as magnetite or hematite rather than sideriteor greenalite. The irregularity diminishes toward the west, though still existing at Sunday Lake. For many miles west of Sunday Lake the iron-bearing formation was deposited as a continuous thick formation with less amounts of other sediments. These differences may be j^artly due to varying condi- tions of temperature and materials present, as discussed on page 526, and are undoubtedly due in part to the fact that near the exti-usions there were sudden and violent oscillations in level, requiring frequent alternations of sediments, while farther away these oscillations were less marked and the movement was a comparatively uniform one of sinking, perhaps due to the general extrusion of the lavas from the region. Moreover, shore conditions of deposition may well have been different from those offshore. It has been noted that the upper Huronian iron-bearing formations in the Mesabi, Gogebic, ilenominee, and Felch ^fountain districts are clearly defined formations originally contaming greenalite and carbonate between quartz sand below and shale above, and that in these districts they come relatively close to the older rocks, suggesting a possible shoi-e comlition. In the Cuyuna, Florence, and Iron River districts the iron-bearing members, originally sideritic, are 518 GEOLOGY OF THE LAKE SUPERIOR REGION. in numerous layers and lenses in the slates. These are probably higher in the series and may also represent offshore conditions. It may be argued that similar basic igneous rocks elsewhere extruded near or under the sea are not accompanied by deposition of iron-bearing formation on such a scale. That iron-bearing rocks are present on a smaller scale in such association elsewhere is shown on pages 508-510. It should be remembered that only veiy exceptionally do igneous rocks of any sort carry ores with them. There are many areas of Tertiary eruptive rocks and but few Goldfield camps. So far as the Lake Superior iron-bearing formations derive their materials from direct magmatic contrilni- tion of igneous rocks, they are likely to be localized by reason of these exceptional contributions. Tliis may explain why all of the similar pre-Cambrian basalts in Canada or elsewhere in the Lake Superior region are not associated with iron ores, though the geologic conditions are apparently the same. It follows from the foregoing statements that the ores are not derived from basic igneous rocks in general but from certain ones. It may be further argued that wliile the iron-bearing formations of the Keewatin may have readily been derived from the relatively abundant associated greenstones, the iron-bearing formations of the Huronian are so extensive as compared with the contemporaneous volcanic rocks that they could scarcely have been derived from those rocks. Such an argument would be without definite basis, however, because there is no known quantitative relation between volume of igneous rock and volume of materials derived from it as igneous after-effects. The iron ores of the Iron Springs district of Utah show a wide range in abundance as compared with the parent igneous rocks. The contemporaneous volcanic activity in the midille Huronian was extensive, being represented by the Hemlock, part of the Clarksburg, and other volcanic formations. That in the upper Huronian was less in amount, but is represented by most of the Clarksburg, in the eastern part of the Gogebic district, and some of the greenstones of the Menominee district; moreover, it may well be that the present Lake Superior basin was the locus of much more abundant upper Huronian flows, for reasons wliich are mentioned on pages 507-508. CHEMISTRY OF ORIGINAL DEPOSITION OP THE IRON-BEARING FORMATIONS. NATTJKE OF THE PROBLEM. The experiments specifically described in the following paragraphs, if not otherwise cred- ited, have been made in the geological and chemical departments of the Universit}^ of Wisconsin, principally by M. E. Diemer, in cooperation with W. J. Mead, R. D. Hall, and others, to meet conditions specified by the authors. The problem is to explain the original deposition in tliick formations of greenalite ([FeMg] SiOj.nHsO), siderite (FeCOj), chert (SiOj), and perhaps some hematite, magnetite, and limonite, in intercalated layers of varying proportions, under conditions, if our preceding conclusions are valid, ranging from ordinary cycles of weathering, transportation, and deposition to direct con- tribution of iron solutions from the hot igneous extrusives to the water in which the sediments were deposited. Obviously a wide range of chemical processes has been involved in the development of the iron ores. It is unlikely that all are known. It is the aim of the following paragraphs to indicate as definitely as possible certain processes wMch seem likely to have been important, without impHcation that these are necessarily the only ones contributing toward the observed results. The iron may have been carried as a ferrous salt of silicic, carl)onic, sul]>huric, hj-drochloric, or other acids present, or as FeO in presence of IIjO at liigh temperatures it may have been in excess of the available acid radicles. It appears now as an origmal constituent of basic extru- sive rocks in the form of sulpludos, magnetite, hematite, chlorite, and the pyroxenes ami amplii- boles. The absence of greenalite and ferrous carbonate as such among these ^original constitu- ents, and also the absence in the ferrous silicate and carbonate of alkalies, which are associated THE IRON ORES. 519 with the iron as original constituents in the igneous rocks, seem to prechide the direct contri- bution of the iron as ferrous sihcate or ferrous carbonate from tlie igneous rocks and to require certain sifting ami simplifying reactions by outside agencies to explain the composition of the original iron-formation rocks. It will be assumed in the following discussion that the iron is carried as a ferrous salt. From the abuntlance of iron sulphides in the original igneous rocks and in their pegmatitic after-effects it will be assumed further that the acid radicle is sulphuric. This is done also for convenience in experimenting. It is not meant to exclude other possible combinations of tlie iron above mentioned. Carbonic acid was doubtless present. Other combi- nations than these would serve fully as well in the essential steps of tlie process below outlined. FORMATION OF IRON CARBONATE AND LIMONITE. The close association of man}- of the thinner carbonate ))ands of the upper Huronian with black carbonaceous and pyritiferous slates, an association similar to those found in the "Coal Measures" and elsewhere, suggests that the iron carbonate may be the result of reduction of ferric hydrate by organic material buried with it in deltas, bogs, or other similar places. Hydrogen sulphide characteristic of these conditions would react upon part of the carbonate of iron, pro- ducing the iron sulphide, thereby giving both iron carbonate and ii'on sulphide in association with carbonaceous rocks. Van Hise " says : As to the form in which the iron salts enter the seas, we can judge only by analogy, but if the present be a guide to the past, the iron was chiefly as a carbonate and to a subordinate extent as a sulphate, although it might have been in part in the form of other salts. When the iron salts reach the lagoon, they are precipitated under favorable condi- tions as ferric hydrate or possibly in part as basic ferric sulphate. Supposing the iron salt to be carbonate, it would be precipitated according to the following reaction: 4FeC03+3H20+20=2Fe,03.3HoO+4C02. Where this process goes on, on an extensive scale, limonite bodies are built up. It was formerly supposed that this reaction took place as a result of the work of oxygen and moisture alone, and this is true to some extent. But recent observation has shown that where in lagoons iron carbonate is abundant the oxidation is largely performed through the agency of a class of bacteria called the iron bacteria. It has been found thatthese bac- teria are unable to exist without the presence of iron carbonate or manganese carbonate, but the iron carbonate is the chief compound used. This material they absorb into their cells. There the iron carbonate is oxidized and the limonite is precipitated. Says Lafar: "The decomposing power of these organisms is very great, the amount of ferrous oxide oxidized b>' the cells being a high multiple of their own weight. This high chemical energy on the one hand, and the inexacting demands in the shape of food on the other, secure to these bacteria an important part in the economy of nature, the enormous deposits of ferruginous ocher and bog iron ore, and probably certain manganese ores as well, being the result of the activity of the iron bacteria. "6 Evidence is furnished of the precipitation of the limonite of bog iron-ore deposits in this manner by the discovery in some of them of large numbers of the sheaths of the iron bacteria. <^ Further evidence of the importance and activity of these bacteria is furnished by their partly or completely closing water pipes of cities where the water con- tains a considerable amount of iron carbonate. ^ The iron part of the salts carried down to the sea as a sulphate would be likely to be thrown down as basic ferric sulphate,'' according to the following reaction: 12FeS04+60-|-(x-F9)H,0=Fe,(SOj3.5Fe,,03.xHoO-|-9H.,S04. The material thrown down as a hydrated ferric oxide and basic ferric sulphate is mingled with more or less of organic material, and a deposit of considerable thickness may thus be built up. This depcj.sit is below the level of ground water and is therefore in the zone of incom])lete oxidation, or is under the conditions of the belt of cementation. The oxygen required for the partial oxidation of the organic material is derived in part from the ferric oxide, and the iron is reduced to the ferrous form; but probably this reaction does not take place on an important scale at the surface. The reducing agent may be regarded as carbon, carbon monoxide, or some of the hydrocarbons, such as methane. The result is the same in any case. The oxygen and the carbon produce carbon dioxide, and thus the conditions are reproduced for the production of iron carbonate. A representative reaction may have been as follows: 2Fej03-3H,0+3C02-fC=4FeC03-l-3H20. u Van Hise, C. R., k treatise on metamorphism: Mon. V. S. Geol. Survey, vol. AT, 1904, pp. 825-827. ii Lafar, F., Teclinical mycology, vol. 1, Lippinrolt Travers, II. ^ .'., Proc. Roy. Soc., vol. 04, 1S98, pp. 130-1-12. cChamberlin, R. T., The gases in rocks: Pub. Carnegie lust. So. lod, 1908. 528 GEOLOGY OF THE LAKE SUPERIOR REGION. to weathering, transportation, and deposition at ordinary temperatures, aided Ijv organic reducing materials, to conditions of direct contribution of iron-bearing salts from the hot igne- ous rocks to the locus of deposition. Carbonate or greenalite might develop under either set of conditions, but on the whole tlie. former set seems to be more favorable chemically to the development of iron carbonate associated with the carbonaceous slates and the latter set more favorable to the development of greenalite. It also appears that iron carbonate may develop from reactions of greenalite with carbon dioxide, and this is regarded as an adequate though not necessary means of precipitation of iron carbonate knowTi to be more or less free from carbonaceous material and in close association with greenalite. Iron carbonate secondary to greenalite is commonly observed. It maj^ be noted that when carbon dioxide reacts upon greenalite, carbon dioxide is introduced and nothing is taken awaj'. The percentage of silica in the cherty carbonate is therefore less than the percentage of free silica in the original cherty greenalite. The exact difference in percentage will dej)end on the proportion of greenaUte to free sihca chosen for the reaction. An average of all the iron carbonate analyses available from the Lake Superior iron formations gives iron 24. .56 per cent, silica 4L15 percent. An aver- age of all the greenalite-chert analyses gives iron 2.5.05 per cent, silica 55.80 per cent. These figures are derived from a sufficiently large number of samples to make them fair averages. Their validity is strengthened also by their accordance with the comjiosition of the alteration products, the ferruginous cherts and jaspers, the average composition of which has been closel}' ascertained. The lower relative silica content of the iron carbonates is thus suggestive though not decisive evidence of the derivation of some of the carbonates from the greenalite rocks. A condition also pointing to reactions between iron silicate and carbon dioxide to produce iron carbonate is the conspicuous absence in the greenalite and in some of the iron carbonate of the bases which form soluble compounds with the silicates, especially calcium and the alkalies, and the presence in the greenalite and iron carbonate of magnesia, a substance wliich forms an insoluble compound with the alkaline silicates. The average content of these minor con- stituents in the greenalite and carbonate is as follows: . Average magnesium, calcium, sodium, and potassium content in greenalite and cherty carbonate. Greenalite rock. Cherty carbonate. Magnesium Calcium Sodium and potassium. Per cent. 4.20 .08 None. Percent. S.20 .86 None. The above argument will not apply to the exceptional iron carbonates which show gradations into limestones and ferrodolomites, as, for instance, at Gunflint Lake and at the east end of the Gogebic district. Wlaether u-on carbonate develops by reaction of greenahte upon carbon dioxide or under the ordinary surface weathering conditions in the presence of organic material, when we look into the probable sequence of events following the extrusion of the original iron-bearing igneous rocks and leading up to the deposition of the iron formations, we note that in either case the probable tendency would be to develop greenahte first and then carbonate. Also so far as the two arc precipitated at the same time, the higher density of the greenalite would make it settle first, the carbonate following later, as shown by laboratoiy c.xperuuent. ^^^len the ingretlients of the upper Iluronian (quartz sand, mud, greenalite, and iron carbonate) are shaken up together m a vessel of water and allowed to settle, a clean layer of sand is fonned at the bottom, showmg a most distinct contact with the la3'er next above. Then follows greenalite with some carbonate and mud, then carbonate and mud with some greenalite, and finally mud with some carbonate. Thus, whatever emphasis is put upon the different ways of producmg iron carbonate, it seems probable that in any iron-bearing formation greenalite materials would be more abundant near the bottom of the formation, or near shore, and the carbonate higher up, or offshore. THE IRON ORES. 529 The distribution of the greenalite and carbonate rocks in the upper Huronian is remarkably in accord with inferences drawn from the chemistry of tiieir deposition. GreenaHte is as yet known only at the lowest horizons of the upper Huronian and is exposed in tlie Mesabi, Felch Mountain, and Menominee districts and to a slight extent in the Gogebic district. In the upper part of the iron formation of the Mesabi district iron carbonate becomes relatively more abun- dant, and just beneath the overlying Virginia slate forms a layer up to 20 feet in thickness. la higher parts of the upper Huronian associated with the slate in the Cuyuna, Crystal Falls, Iron River, and Florence districts the iron formation consists dominantly of iron carbonate. The presence of the carbonate near the base of the series in the Gogebic district would imply under the above principles a proportionally greater abuntlance of carbon dioxide there than in the Mesabi district, for unknown reasons. SECONDARY CONCENTRATION OF THE ORES. GENERAL STATEMENTS. The secondary alteration of the iron formations to ore has been accomplished by both chemical and mechanical processes, under conditions of weathering, with modifications due to folding, deep burial, and proximity to igneous intrusions. All the ores are partly the result of secondary concentration, but some have suffered more and some less concentration. Layers of iron formation origmally rich in iron hav become iron ores by less concentration than liave layers of iron formation originally poor in iron. In a few places in the region, as in the east end of the Gogebic district and in parts of the Mesabi district, there is evidence that certain layers of iron formation were originally nearly rich enough in iron to be mined as iron ores, after only a slight amount of secondary alteration. In such places the shape and dimensions of the original layers determine essentially the shape and dimensions of the iron ore deposits. Wliere secondary concentration has been largely effective ill producing the iron ore, as it has in most of the larger deposits of the region, the shape and distribution of the ore bodies are determined by the structural conditions which localize the secondary concentration, rather than by the ]>rimary bedtling of the iron formation. The essential secondary changes in the development of the ores have been effected by weathering. The ores once formed, alterations effected by dynamic action, igneous intrusion, or redeposition as fragmental sediments may be regarded as for the most part subsequent and modif3'mg factors, tendmg to change somewhat the character of the ores and ore deposits, but adding little to their size or richness. Dynamic and igneous metamorphism actmg before the concentration of the ores tends to inhibit ore concentration by making the iron forma- tion refractory to weathering agencies. In the following treatment emphasis will be placed accordingly. CHEMICAL AND MINERALOGICAL CHANGES INVOLVED IN CONCENTRATION OF THE ORE UNDER SURFACE CONDITIONS. OUTLINE OF ALTERATIONS. It requires only the most general field observation to bring out the fact that the iron forma- tions are being and have been rapidly altered by percolating waters carrying oxygen, carbon dioxide, and other constituents from the surface and that the present characteristics of the formations are considerably different from those they had when they first became consolidated. Now they consist mainly of ferruginous chert and jasper, uith subordinate quantities of iron ore, paint rock, greenalite, iron carbonate, amphibole-magnetite rock, etc. Formerly they were more largely cherty iron carbonate or greenalite. Fortunately the alterations have not everywhere gone far enough to obhterate all the original phases of the iron formations. Grada- tions may be observed between original cherty iron carbonate or greenalite phases of the forma- tions and the dominant alteration products, ferruginous cherts and jaspers and iron ores. The 47517°— VOL 52—11 34 530 GEOLOGY OF THE LAKE SUPERIOR REGION. former are found in protected places beneath slate or other impervious cappings; the latter occur in portions of the formations exposed to percolating oxidizing waters. The former are ferrous compounds, unstable under conditions of surface weathering; tiu' latter are the stable oxides, end products of weathering. The ferruginous cherts, jaspers, and iron ores furthermore retain textures characteristic of carbonate and greenahte, thereby betraying their derivation from these substances. This is especially noticeable in the ores and cherts derived from green- ahte, the pecuhar granular shapes of the greenahte being conspicuous in its derivatives. The red, brown, and j'ellow colors of the altered phases of the formations, the ores and ferruginous cherts, contrast strongly with the gray and green of the original cherty carbonate and greenahte, making the alterations conspicuous to the eye, especially along fissures in the original rocks. The secondary alterations of iron carbonate and greenahte rocks to iron ore involve (1) oxidation and hydration of the iron minerals in place, (2) leaclaing of sihca, and (.3) introduc- tion of secondary iron oxide and iron carbonate from other parts of the formations. These changes may start simultaneously, but the first is usually far advanced or complete before the other two are conspicuous. The early products of alteration therefore are ferruginous cherts — • that is, rocks in wliich the iron is oxidized and hydrated and the sihca not removed. The later removal of sihca is necessary to produce the ore. The secondary introduction of iron oxide and iron carbonate in cavities left by the leaching of sihca is of httle importance in the alteration of the greenahte rocks to ore. In the alteration of the carbonates to ore it is fre- quently a conspicuous feature. The alteration of the original rocks of the iron formations to ore may therefore be treated under two main heads — (1) oxidation and hydration of greenahte and siderite, producing ferruginous chert; (2) alteration of ferruginous chert to ore by leaching of sihca, -with or without secondary introduction of iron. OXIDATION AND HYDRATION OF THE GREENALITE AND SIDERITE PRODUCING FERRUGINOUS CHERT. The oxidation of the cherty iron carbonates and greenahtes to hematite or hmonite pro- duces ferruginous cherts of varying richness. (See Pis. XLII, C, D; XLIII-XLY.) During these changes tlie iron minerals for the most part are altered in place, but iron ma}' also be transported and redeposited. Evidence of this is abundant in the stalactitic and botryoidal ores lining cavities or incrusting secondary quartz crystals and numerous veins of ore cutting across the beddmg of the formation. It will be sho\\^^ in the follo^v•ing discussion, however, that the principal enrichment of the ore takes place in connection with the removed silica, although in several districts the introduction of iron is very important. The oxidation and hydration of the original iron imnerals are expressed in the following reactions: 4FeC03 (siderite) + nHjO + 20 = 2Fe203.nH20 + 400^. 4Fe(Mg)Si03.nH20 (greenahte) + 20 = 2Fe203.nH,0 + 4SiOj. The alteration of the iron minerals is facUitated by smaU amounts of acids carried by per- colating waters. Carbonate of iron is soluble ^nth difficulty in pure water and not easily soluble with an excess of carbon dioxide. On the otlier hand, it is easil}' soluble in either of the stronger acids, sulphuric or hydrocliloric. Sulphuric acid results from the decomposition of the iron sulphide in the original carbonates and in the adjacent pyritiferous greenstones and slates. The reaction may be^ FeS^ + H,0 + 70 = FeSO, + H,SO, . , This is aided in turn by carbon dioxide in the water. Thus the iron sulphide is oxidized to ferrous sulphate^ ^\^th the simvdtaneous production of sulphuric acid, wliich attacks the iron carbonates and changes them to soluble ferrous sulphate. In the Micliipicoten ihstrict, where glacial erosion, has cut deep, sulphides are found abundantlv with the carbonates. Sulphate of iron is present in veins in the ores of the Iron River chstrict. Baj-ley " found the white efBorescence characteristic of Menominee ores to be essentially sochum sulphate with tlie for- mula of Glauber salt, Na^SO, + lOHjO, which he regards as the result of decomposition of pyrite oBayley, \V. S., The Menominee iron-bearing district ot Michigan: Mon. I'. S. Geol. Survey, vol. K\ 1904, pp. 390-391. PLATE XLIII. 531 PLATE XLIII. Photomicrographs of greenalite granules. A. Greenalite rock (specimen 45178, elide 15652) from 100 paces north 500 paces west of the southeast corner of sec. 22, T. 59 N., R. 15 W., Mesabi district, Minnesota. Without analyzer, X 50. The slide is selected to show both the fresh and the slightly altered granules. Note the peculiar greenish-yellow color of the granules, their irregular shape, and their curving tails, some of which seem to connect with adjacent granules. The homogeneous greenish yellow colors represent the unaltered parts. The bright-green and dark-green colors represent grunerite which has been developed from the alteration of the greenalite. The dark green is perhaps in small part iron oxide. Described on pages 165-168. B. The same with analyzer, X 50. The unaltered portions of the granules are nearly or quite dark under crossed nicols. ^\^lere the granules have altered to griinerite the polarization colors appear. The matrix consists of fine-.grained chert in which the individual particles are very irregular in shape and size. Described on pages 165-168. 532 U. S. GEOLOGICAL SURVEY MONOGRAPH Lll PLATE XLIII PHOTOMICROGRAPHS OF GRUNALITE GRANULES. PLATE XLIV. 533 PLATE XLIV. Photomicrographs of ferruginous chert showing later stages of the alteration of greenalite granules. A. Ferruginouschertwith granules(8pecimen 45063, slide 15563) from nearcenterof sec. 22, T. 60 N., R. 13\V. With- out analyzer, X 50. The granules are outlined and in part replaced by iron oxide. The matrix is chert. The complex nature of one of the granules is to be noted. Apparently one complete small granule is entirely inclosed in another large one. Described on pages 168-170. B. Griineritic ferruginous chert (specimen 45603, slide 15974) from Clark mine. With analyzer, X 50. The rock consists of chert and iron oxide and griinerite. The iron oxide is a yellowish-brown hydrated variety, which is with difficulty distinguished from the griinerite. The granules have been entirely obliterated. Described on pages 168-170. 634 U. S. GEOLOGICAL SURVEY MONOGRAPH Lll PL. XL!V PHOTOMICROGRAPHS. PLATE XLV. 535 PLATE XLV. Photomicrographs of granules and concretionary structures in Clinton iron ores. A. Granules in Clinton iron ore, from lower bed, Sand Mountain, New England City, Ga. Loaned by C. H. Smyth, jr. Without analyzer, X 40. Granules of black and dark-brown hydrated hematite stand in a matrix of calcite. The latter areas within the granules are also calcite. Traces of organic shells in these slides are abundant. The granule a little to the right of the center shows this especially well. There can be no doubt as to the fact that the granules are for the most part replacements and accretions about shells and particles of shells. It is apparent also that there is a marked tendency for the granules to take on rounded and oval forms regardless of the shape of the original particles of shell. Note the remarkable similarity of these granules in shape to the greenalite granules illustrated in Plate XLII, A, B. B. Green oolites in Clinton ore, from Clinton, N. Y. Loaned by C. H. Smyth, jr. With analyzer, X 40. Concen- tric layers of chloritic and siliceous substance, of various shades of green and yellow, surround angular, subangular, and rounded grains of quartz . The concentric greenish and yellowish bands under crossed nicols show black crosses characteristic of concretionary structures. The matrix is mainlycalcite. but there are present also small particles of quartz. 536 U. S. GEOLOGICAL SURVEY MONOGRAPH Lll PL. XLV PHOTOMICROGRAPHS. THE IRON ORES. 537 and muscovite. Iron sulpliides and chalcopyrite are also common as vein fillings. Sul])hates are found in mine waters. (See pp. 54.3-544.) Humus acids are also well known to aid in the solution of the iron. Precipitation of the iron from ferrous solutions would be caused (1) by direct oxidation and precipitation as limonite; or (2) by reaction with alkaline carbonate, producing iron car- bonate, which in this form in the presence of oxygen alters almost immediately to hydrated iron oxide; or (3) by loss of carbon dioxide. A small amount of secondary iron carbonate, where iron is carried in solution as bicarbonate, observed locally in each of the districts, is incidental to the mam process of oxidation producmg ferruginous cherts. The oxidation of the iron in the carbonate and greenalite goes on much more easily and rapidly than the removal of the silica and may afl'ect most or all of the carbonate or greenaUte, producing ferruginous cherts, before the removal of the silica has gone fai' enough to be appre- ciable. , An epitome of the storey for the formation is presented by almost any hand specimen of iron carbonate or greenalite. The ferruginous cherts are, therefore, intermediate phases between the original greenalite or siderite and the ore, and the principal removal of the silica is subse- quent to the formation of the ferruginous cherts. Given sufficient time and the other necessary favorable conditions and any part of them may become ore. In districts where greenalite is the dominant origmal iron compound, so far as can be determined, the layers of chert in the ferrugmous cherts prior to their alteration to ore are not veiy different m number, iron content, and degree of hj'dration from those in the greenalite rocks, indicating but little transfer of iron, though localh^ the segregation of silica and iron oxide into bands is more accentuated. In districts where carbonate is an important original iron salt, the rearrange- ment, transportation, and introduction of iron salts are quantitatively important. This is probal)ly due to the structural conditions described on page 538. Slight rearrangements of the iron ore are to be seen in the concretions composed of alternate concentric laj'ers of chert and iron oxide developed during the alteration. These develop both from the iron carbonate and from the greenalite. Not uncommonly oxidizetl greenahte cherts are found alongside of unoxicUzed iron car- bonate cherts. At first thought this would seem to indicate the readier oxidation of the greenalite than the carbonate, but it is not certam that this is the case, for it is sometimes found that the carbonate in these relations is secondary, and another possibility is that the greenalite was oxidized at the time of its precipitation rather than secondarily. ALTEKATION OF FERRUGINOUS CHERT TO ORE BY THE LEACHING OF SILICA, WITH OR WITHOUT SECONDARY INTRODUCTION OF IRON. PROCESSES INVOLVED. Ore may be formed (1) by taking awa}' silica from the ferrugmous cherts, leavmg tlie iron oxide; (2) by taking out silica and introducing iron in its place; or (3) by adding iron to an extent sufficient to make the percentage of sdica a small one. In the last case there would necessarily be a large increase in volume. Quantitative tests show that (1) is of greatest importance, that (2) is effective only in some of the ores derived from carbonates, and that (3) is practically negligible. Measurements of pore space of the ores derived from the alteration of ferruginous cherts of greenalitic origin brmg out the facts that pore space approximates the volume of silica which has been removed (see pp. 184-185), when there has been little slump; in other words, the filling of the pore space in the ores by sUica would nearly reproduce the composition of the fer- ruginous cherts. It wHl be shown also that the leaching of silica from the ferruginous cherts derivetl from greenalite alterations does not materially affect the character of the iron oxides, especially their degree of hydration, and that therefore the nature of the ore of the deposit is primarily determined by the changes which the greenalite undergoes when it alters to the oxide bands of the ferruginous cherts. Measurements of pore space in ores derived from ferruginous cherts, which in turn have been derived from the alteration of iron carbonate, show that the pore space is less than the 538 GEOLOGY OF THE LAKE SUPERIOR REGION. volume of the silica wliieh has beeu removed. (See p. 24L) This is due i)artly to shimp, but mamly to the fact that secondary iron oxide partly fills the openings. The ran<^o in wliich there is conspicuous absence of evidence that iron has been transported to any considerable extent is the Mesabi, where the flat dip exposes a larye portion of tlie for- mation directly to oxidizing; waters, and oxidation works down more or less uniformly from the surface, leaving; few imoxidizcd portions to contribute soluble iron salts to be earned down and mixed witli deeper oxidizing solutions following cluuincls from the surface. In the other districts, where the evidence of the carrying of iron is ])Iiiin, the formations are so tilted that the underground courses of oxidizing waters from the surface pitch deeply, lea^-ing unoxi- dized iron formation above as a source for soluble iron salts, which may be taken into solution and carried down, and, by reaction wdth oxidizing waters, precipitate the iron oxide. This deej) circulation of oxidizing waters afforded by steeply tilted formations permits the leaching of silica at tlepth, thus providing openings in wliich the iron carried m solution from the upper unoxidized portions of the formation may be deposited. It is m tliis essential that the ilesabi conditions differ from those of the other ranges. Silica dissolved from the iron formations has been in small part redeposited in veins, both in ore and rock and in the crystallized quartz linings of uiany cavities in the ore, and in part has joined the run-off. The process is going on to-day, for mine and svirface waters carry' siUca (see pp. 540-544), anerior, in deep wells in the Paleozoic of the upper Mississippi Valley, in the granites of the Piedmont area of Georgia, and elsewhere. Their characteristics seem not to be related to certain kinds of rocks or ore deposits, but to depth and stagnant conditions. Chlorine is present in minute quantities in original igneous rocks and in nearl}' all surface waters. Its salts tend to remain in solution, while the salts of other acids are more largely precipitated. With a given amount of water, there seems likely to be, therefore, a progressive relative accumu- lation of chlorine salts. Such is the case in salt waters at the earth's surface, where a large factor in the accumulation is the lack of sufficient circulation to carry off and dilute the salt waters that are developing by evaporation. In deep underground Maters there is essentially the same condition of stagnancy, and therefore we suggest jirogressive accumulation of soluble chlorine salts. In the shallower mine waters the rapid circulation and accession of fresh waters from the surface prevent such accumulation of salt. The proportion of sodium chloride to calcium cliloride in deep mine waters in the Lake Superior region becomes relatively less with increase in depth, indicating that the increasing content of chlorine is able to hold not only all sodium present but larger amounts of calcium. The materials in solution under any conditions must be regarded as representing the residual solutions from which all possible insoluble minerals have already crystaUized out. All the Lake Superior mines, both iron and copper, are associated with basic rocks in which calcium greatly predominates over the sodium, so that whenever the sodium is taken care of by the clilorine present there should always be a considerable excess of calcium available. Lane, who has given special attention to deep mine w* aters and who has brought together the analyses above quoted, otl'ers quite "another explanation for the characteristics of these deep waters. He beUeves them to be connate or fossil sea waters, included in the rocks, both igneouf and sedimentary^ during submarine deposition. The fact that they differ from present sea water in having so large a proportion of calcium chloride he ascribes to a possible change in composition of the sea water during geologic time in the direction of increasing the proportion of sodium chloride as compared with calcium chloride to the present known proportion of sea water. We do not follow him in this conclusion because of the fact, already cited, that these pecuhar salt waters seem to be characteristic not only of marine sediments but of sediments of subaerial origin, of surface eruptives, and of plutonic igneous rocks. They are related to depth and stagnancy rather than to kind of rock or geologic horizon. There seems to be no adequate reason for regarding these waters as fossil sea waters, for all the essential kinds of conditions which produce the salt water of the ocean are present. LOCALIZATION OF THE ORES CONTROLLED BY SPECIAL STRUCTURAL AND TOPOGRAPHIC FEATURES. From the foregoing discussion it appears that the iron ores constitute concentrations in the exposed parts of the iron-bearing formations accomplished on the average mainly b}^ the removal of associated silica, leaving the iron oxidized and in larger percentage, but to an important extent accomplished also by solution, transportation, and redeposition of the iron when it was still in its soluble ferrous condition. The agents of alteration are surface waters carrying oxygen and carbon dioxide from the atmosphere. The accessibihty to the iron-bearing formations of these agents therefore determines the location, shape, and size of the deposits. The structural conditions favoring such accessibihtj' have been summarizeil in the earlier part of this chapter (see pp. 474-475), and are discussed in some detail in connectit)n with the ores of the individual districts. They may be merely mentioned here. The most favorable condition is afforded by wide area of exposure of the formation, which in turn is a function of the dip. Fractures, » iDgall, E. D., Report on mines and mining on Lake Superior: Ann. Rept. Oeol. Survey Canada for 18$7-SS, vol. 3, pt. 2, 1889, p. 2SB. THE IRON ORES. 545 impervious basements, and varying porosity also serve to concentrate the circulation. Ores are not found, however, in some places where area, fractures, and impervious basements seem to be favorable for ore concentration. This is beheved to be due in some part to the denseness of the cherts in these places, preventing access of water. Wherever the rocks are dense the silica is not removed. The amphibole-magnetite cherts, the unaltered greenalite and siderite rocks, and the quartzites associated with the iron-bearing formations all have very Uttle pore space, as shown by a considerable number of determinations. Silica is not removed directly from these rocks. On the other hand, the ferruginous cherts, resulting from the alteration of cherty iron carbonates and greenalites, contain pore space averaging about 5 per cent, developed by the lessening of the volume of the iron minerals during their alteration from the ferrous to the ferric form. Tliis pore space is so distributed as to give the water access to all parts of the rock mass. The size of grain is so small that for each grain there is a large surface in proportion to volume. But even the ferruginous cherts are locally so dense that they do not allow ready access of water. Several possible reasons may be suggested for this unusual density. (1) The ferruginous cherts at these places may not have been derived by alteration from cherty car- bonates or greenahtes but may have been deposited directly in their present form as chemical sediments mth small pore space. It has been shown that this could easily go on with the deposition of greenahte and carbonate. This explanation would seem to be especially Ukely to hold for certain of the amphibohtic cherts of the Keewatin, wliich are intimately associated with basalt flows both above and below and wliich it is entirely conceivable might have been originally deposited in a condition different from those of the cherty carbonates and greenalites of the later iron-bearing formations. (2) Metamorpliism of the cherts under pressure after pore space had been developed by oxidation of the iron minerals may have closed the openings before the siUca had been taken out. Cherts wliich have been much folded and contorted at so great depth as to be deformed without fractures are almost invariably dense. The Keewatin iron- bearing formations are the oldest and have naturally suffered more from such metamorpliism than the later formations, and this may be a factor in the barrenness of the Keewatin. On the other hand, larger areas of the upper Huronian are comparatively Uttle deformed and pore spaces formed by the oxidation of the iron minerals have remained substantially open since upper Huronian time. (3) The openings may have been closed by infiltrated sihca and iron. In the Marquette jasper, secondary materials completely heal the rock. The relative importance of these conditions affecting pore space varies from place to place and between the different iron-bearing formations, and this variation is beheved to account in large measure for the marked differences in enrichment of different formations and different parts of formations. Undoubtedly the processes of secondary concentration above described tend to affect to a greater or less degree all the exposed surface of the iron-bearing formations. It is not unlikely that in long periods of slow denudation ores may have actually covered all of this surface. It is equalljr obvious, however, that the covering had various depths, depemlmg on a consider- able variety of structural conditions. The glacial denudation has scraped off ore which may once have developed at the surface, and little has developed since. There remam only the lower parts of the deposits left by denudation. A discussion of the structural conditions governing the ore deposits is therefore really a discussion of the conditions determining their lower limit and configuration. The structural and topographic conditions of each of the dis- tricts are summarized in other chapters. QUANTITATIVE STUDY OF SECONDARY CONCENTRATION. The nature of the secondary concentration of Lake Superior iron ores has been in the past inferred almost entirely from qualitative evidence. The extensive commercial develop- ment of the ores of this region during recent years now makes available data for quantitative study of the origin and concentration of the ores. Although there is a great similarity in the secondary concentration of all the iron ores of the Lake Superior region, certain local difler- 47517°— VOL 52—11 35 546 GEOLOGY OF THE LAKE SUPERIOR REGION. ences require that each of the several districts be discussed independently. This is done in the chapters on the several districts. The average change in secondary concentration, based on all available analyses (seep. 181), is graphically expressed in figures 20 (p. 189) and 31 (p. 245). ALTERATIONS OF IRON-BEARING FORMATIONS BY IGNEOUS INTRUSIONS. ORES AFFECTED. The changes described in the foregoing sections have completed the development of the ore deposits of the Mesabi, Gogebic, Menominee, part of the Marquette, Crystal Falls, Iron River, Florence, and Cuyuna districts, wliich yield roughly 93 per cent of the total ore mined annually in the region. Other ores, such as the hard ores of the Marquette and Vermihon districts and the magnetic rocks of the Mesabi and Gogebic, have suffered certain additional vicissitudes of anamorphic alterations by igneous intrusion, thus becoming the hard, dense, recrystalhzed, more or less magnetic, dehydrated, and silicated ores described below. (See Pis. XXXV, p. 470, and XLVII.) The development of some of these characteristics may have been synchronous with the deposition of the iron-bearing formation under the influence of contemporaneous igneous extrusives, discussed on page 527, but whatever the probabiMty of this there is no doubt that characteristics of this kind have been developed mamly bv later intrusives. The intrusion of small masses of igneous material, as the dikes in the Gogebic district and certain of the bosses in the Marquette district, has apparently but slightly' metamorphosed the iron-bearmg formation, ^\^lere great masses of igneous material have come into contact with the iron-bearing formation, however, marked results have followed, as near the Duluth gabbro, the gabbro of the western Gogebic district, and the intrusives of the western Marquette district. POSSIBLE CONTRIBUTIONS FROM IGNEOUS ROCKS. The characteristic features of the amphibole-magnetite rocks of the iron-bearing forma- tions described above become more accentuated in approach to the igneous rocks, leaving no doubt that they are the metamorphic result of the intrusion of the gabbro. The facts available indicate to some extent also the processes through which this result is accomplished. The question first to be answered is whether or not the iron-bearing formation owes its character- istics near the contact to direct contribution from the hot intrusives or to the recrystallization of substances already in the iron-bearing formation. The essential similarity of composition of the amphibole-magnetite rocks with that of the ferruginous cherts (see p. 204) argues against large introduction of materials from the gabbro. Had such materials been introduced on a large scale they would probably have considerably changed the proportions of the elements present, for otherwise it would be necessary to assume that the materials contributed from the gabbro had been in the same proportion as those originally present in the iron-bearing forma- tion. The magnetite in the gabbro is titanic, while that in the adjacent iron formation is not. The higher sulphur content in tlie amphibole-magnetite rocks may mdicate direct contribution of sulphur, though this may also be original in the iron-bearmg formation. (See pp. 550, 552.) Wliether or not there was some small introduction of materials from the gabbro, the bulk analyses of the amphibole-magnetite rocks are so similar to those of the other phases of the iron-bearing formation as not to require the assumption of delivery of hot solutions from the gabbro to the iron formation. Furthermore, there is no regular variation in the composition of metamorphic phases of the iron-bearing formation through the several hundred feet from the contact for which these phases are known in many places to extend. Finally, the very fact that the metamorphic phases of the iron formation extend so far and so uniformly from the gabbro contact argue against their development by accession of materials from the gabbro. It is conchuled, therefore, that the princijial efl'ect of the intrusion of the gabbro into the iron-bearing formation was that of recrystallization of substances already present and not by contribution of solutions. PLATE XL VII. 547 PLATE XLVII. Photomicrographs of ferruginous and amphibolitic chert of iron-bearing Biwabik formation near contact with duluth gabbro. A. Actinolitic, griineritic, and magnetitic chert (specimen 45141, slide 15621) from southeast of center of sec. 17, T. 60 N . , K. 12 W. , Mesabi district, Minnesota. Without analyzer, X 50. This rock is close to the contact with the Duluth gabbro and shows the tj-pical alterations characteristic of the contact. The chert is in much larger particles than in the western portion of the range away from the contact. The particles fit in somewhat regular polygonal blocks. The iron oxide is magnetite instead of hydrated hematite, and actinolite and griinerite are present. The amphiboles are in small quantity in the slide shown, but the short actinolite needles may be seen inclosed in the quartz. (See PI. XXXV.) B. Actinolitic slate (specimen 9555, slide 3190) from Penokee Gap, NW. \ sec. 11, T. 44 N., R. 3 W., Wisconsin. In polarized light, X 165. The section is a typical actinolitic slate. The quartz is completely crystallized. The magnetite has mostly well-defined crystal outlines and is manifestly the first mineral to crystallize, being scat- tered uniformly through the section without any regard to the actinolite and quartz and therefore included by both of them. The actinolite is in its characteristic blades and sheaf-like forms, hav-ing a radial arrangement of its fibers. It is as plainly the second mineral to crystallize, as needles of actinolite everjTvhere penetrate the quartz, but never the magnetite. The quartz constitutes a background for the magnetite and actinolite and includes them in such a manner as to make the. conclusion certain that it must in the main have crystallized subsequently to the formation of the magnetite and actinolite. (See PI. XXXV.) 548 U. S. GEOLOGICAL SURVEY MONOGRAPH LH PLATE XLVII PHOTOMICROGRAPHS OF FERRUGINOUS AND AMPHIBOLITIC CHERT OF IRON-BEARING FORMATION NEAR CONTACT WITH DULUTH GABBRO. THE IRON ORES. 549 TEMPERATURE UNDER WHICH CONTACT ALTERATIONS WERE EFFECTED. The significant discovery by Wright and Day," of the geophysical hiboratory of tlie Carnegie Institution of Wasliington, that quartz crystalhzed below 575° dift'ers in its properties from quartz crystallized above tliis temperature affords a satisfactory means of determining tlie temperatures at which the quartz of the iron-bearing formation has crystallized. Doctor Wright has kindly determined for us the properties of the quartzes in specimens from different parts of the Lake Superior iron-bearing formations, some of them clearly developed under katamorphic conditions, some of them near the contact with the gabbro. His observations are as follows: Properties of quartz crystals from iron-bearing formations. Speci- men No. Number of sec- tions cut. Average diameter (mm.). Circular polarization. Twinning, etch flgures.u R. L. H.-fL. Character of inter- growth. Number not twinned. Number twinned. Character of twinning. A B 29955 29450 5 6 4 6 7 5 1.5 2.0 4 1 6 Regular do 2 3 fi 4 4 Regular large patches. Do. Regular. Often irregular and small. 3 3 1 3 2 n Etched 1} hours in cold commercial hydrofluoric acid. .\. Crystalline quartz in ore from Vermilion district. B. Crystalline quartz in ore from Mesabi district. 29955. Coarsely recrystallized iron-bearing formation, 300 feet from gabbro contact, northwest of Paulson mine camps, Gnnflint district, north- eastern Minnesota. 29450. Coarsely recrystallized iron-bearing formation in actual contact with Duluth gabbro at east end of Fay Lake, Gunflint district, north- eastern Minnesota. The quartz of Nos. A and B occurs in clear crystals and free from fracture.s. The usual -f- and — unit rhomlio- hedrons are present; also the prism faces. On A crystals there is also present the rhombohedron (1121) and a trigonal trapezohedron form; this in itself is proof that the A quartz was formed below 575°. The aljove observations show conclusively that the A, B, and 29955 quartzes [distant from gabbro contacts] have not been heated above 575° ; that they were formed below that temperature. Specimen 29450 [at gabbro contact] is less regular in its behavior and resembles in that respect the quartz of .some pegmatites. It is not as shattered as granite quartzes usually are and yet is not so regular as the definitely lower temperature quartzes. I concluded that in the pegmatites such quartz was formed probably near the inversion temperature 575°, because pegmatite dense quartz is definitely the low a form while some pegmatite quartz is definitely high 6 quartz. This was proved on one and the same dike. It seems to me probable, therefore, that the temperature of formation of the quartz band in specimen 29450 was not far from 575°. It is obvious from these results that the iron-bearing formation as a whole has not been fused, for its fusion temperature is certainly higher than 575°. This conclusion, together with the one above referring to the lack of transfer of material from the gabbro to the iron-bearing formation, emphasizes strongly the probabihty that the metamorphism of the iron formation near the gabbro was primarily the result of recrystallization below fusion temperature, with the aid of heat from the gabbro. CHARACTER OF IRON-BEARING FORMATIONS AT THE TIME OF INTRUSIONS OF IGNEOUS ROCKS. Wliat were the constituents originally present in the iron-bearing formations at the time of the intrusion? Were they the ferruginous cherts earlier developed from the alteration of cherty carbonates or gi'eenalite rocks, or were they the chcrty carbonates and greenalite rocks themselves? If prior to the intrusion of the igneous rock the iron existed as ferric hydrate, then the change to magnetite involved deoxidation. This, according to Moissan,'' will occur at .300° in 30 minutes in a hydrogen atmosphere. The presence of an actively reducmg agent of tliis type along igneous contacts, wliile perhaps locally probable, can not be proved on any a Wright, F. E., and Larsen, E. S., Quartz as a geologic thermometer: Am. Jour. Sci., 4th ser., vol. 27, 1909, pp. 421-427. 6 Moissan, H., Compt. Rend., vol. 84, p. 129C. 550 GEOLOGY OF THE LAKE SUPERIOR REGION. large scale. If prior to the intrusion of the igneous rock the iron was in the ferrous condition, either as greenahte or carbonate, then moderate heat was sufficient to produce magnetite by robbing the associated water of part of its oxygen. (See p. 526.) This alteration is thought in general to be a more common one than the reduction of iron to magnetite from the ferric state. In the Lake Superior region there is field evidence also that the development of the amphibole-magnetite rocks has been more largely accomplished by partial oxidation of the ferrous iron than by the reduction of ferric oxide. In places in the Lake Superior region, where there is good field evidence that the iron-bearing formation had been exposed and altered to ferruginous cherts before the introduction of igneous rocks — as, for instance, in the eastern part of the Marquette district or at Sunday Lake in the Gogebic district — it is found that the contact effect of the intrusives has been to produce the bright-red banded specular jaspers or black magnetitic jaspers rather than ampliibole-magnetite rocks. In the Marquette district it was long ago noted that the lower parts of the Negaunee formation in contact with intrusives devel- oped amphibole and magnetite, while the upper parts developed the banded specidar jaspers. The cement in these rocks is usually magnetite. Smyth" argued that tliis present ditference in the character of the rocks at upper and lower horizons, especially for the Republic trough, is so uniform as to indicate an original difference in the beds at these horizons. The magnesia content of the ampliibole-magnetite rocks for the most part seems to be like that of the original greenalites and carbonates rather than that of their altered derivatives, ferruginous cherts. In the alteration of carbonates or greenalites to cherts magnesia is lost. (See p. 528.) Had the ampliibole-magnetite rocks developed from the ferruginous cherts, it would be necessary to assume that magnesia had been introduced in just the percentage of the original siderite and greenalite rocks. Sulphur is also more abundant in the original phases of the iron-bearing fonnation than in its katamorphosed products, though no figures are available to show what the average sul- phur content is, because analyses have ordinarily been made of the greenalite and siderite where free from sulphur. Contact or deep-seated mctamorphism would not remove this sul- phur, and this is thought to be the probable explanation of the high sulphur in the amphibole- magnetite rocks. The alternative explanation is that sulphur had been introduced directly from the igneous rocks. CHEMISTRY OF ALTERATIONS. The chemistry of the alterations from original ferrous compounds, greenalite and siderite, to ampliibole-magnetite rocks presents less difficulty than that of the alteration of ferruginous cherts, or ferric compounds, to the amphibole-magnetite rocks. The former alteration requires partial oxidation of a ferrous compound ; the latter requires reduction of a ferric compound, which is thought to be much less common. On the assumption that the ampliibole-magnetite rocks had developed directly from the cherty iron carbonates and greenalites, the changes would be substantially as follows: *" Where the carbonate is nearly pure siderite, griinerite is produced, according to the following reaction: reC03+Si02=FeSi03+COj, with a decrease of volume of 32 per cent, provided the silica be a solid and the carbon dioxide escape. Where the original material was hydrous ferrous silicate, greenalite, simple dehydration only is necessary to form the griinerite. Where the iron-bearing carbonate bears calcium and magnesium in considerable quantity, instead of griinerite being produced sahlite or actinolite may be formed. Supposing the carbonate to be normal ankerite, sahlite is pro- duced, according to the following reaction: CaFeCACaMgCjOe-|-4Si02=Ca2MgFeSi40,j-l-4C02, with a decrease in volume of 37 per cent, provided the silica lie solid and the carbon dioxide escape. From ankerite actinolite may be produced, according to the following reaction: 3(CaFeC20„.CaMgC20e)-|-8Si02=Ca2Mg3Fe3Si024-t-SCOo+4CaC03, with a decrease in volume of 23 per cent, provided the silica be a solid, the CaCOs formed remain as a solid, and the carl)ou dioxide escape. o Mon. IT. S. Oeol. Survey, vol. 28, 1897, p. 530. » Vau Hlse, C. R., A treatise ou metaraorphism: Mon. U. S. Geol. Survey, vol. 47, 1904, pp. S34-S37. THE IRON ORES. 551 If a more ferriferous and less calcareous iron-bearing carbonate be taken, it would not be necessary to suppose any calcium carbonate to have separated. The iron-bearing carbonates may be very impure, just as limestones may be impure; and in this case there may develop various other minerals. In proportion as impurities are mingled with the carbonates, other amphiboles and the pjToxenes, micas, garnets, and other heavy minerals such as olivine may abundantly develop; and thus there may be produced a great variety of rocks, such as garnetiterous magnetite rocks, micaceous griinerite rocks, etc. As the impurities become abundant and the silicates other than griinerite, sahlite, and actinolite more prominent, the altera- tions become nearly those of the fragmental rocks. Between the two there are, of course, all gradations. But as a matter of fact, the two silicates which most extensively form by the alterations of the iron-bearing carbon- ates in the zone of anamorphism are actinolite and griinerite. Where these reactions are complete we may have, in place of the iron-bearing carbonate, actinolite rocks, griinerite rocks, and all gradations between them. Where the iron-bearing formation is originally greenalite, the alteration to the amphiboles would be simply one of dehydration. The development of magnetite directly from the iron carbonates is possible by the following reactions; 2FeC03 + FeSj -f 2H2O =Fe30, + 2H2S + 2C02,° 3FeC03 + H,0 =Fe30, + 3C0. + H, 3FeC03 =Fe30, + CO + 2C02,° BFeCOj -t- O =Fe30, + BCO^." Carbon dioxide is driven off at temperatures probably as low as 400°. At these and higher temperatures the ferrous iron remaining will rob the water of its oxygen, forming magnetite. Siderite at red heat passes into a magnetic oxide with the formation of both carbonic acid and carbonic oxide. According to Dobereiner this reaction takes place as follows :o 5FeC03=3FeO.Fe203+4C02+CO. Glasson.b however, says that 4FeO.Fe203 results, at first giving two parts of CO., and one of CO, but that later the proportion changes to five parts of COj and one of CO."^ Van Hise " says, again : Observation in the field show.s beyond question that the change fi'om iron carbonate to magnetite takes place on an extensive scale. Wliich of the above reactions is tlae more important may be an open question. The alteration of greenalite to magnetite is possible by the following reaction: 3FeSi03nH20 + O =Fe30, + SSiO, -f nH^O. ■^Tiich of the above rocks develops at a given place depends not only upon the original composition of the rocks, but upon the nature of the alteration. For instance, where in the original rock silica is subordinate and nearly pure siderite abundant, a quartzose magnetite may develop, as at various places in the Lake Superior region. WTiere the conditions are such that the silicates form, the development of the actinolite or gi-iinerite uses up both the iron carbonate and the silica, and an actinolite rock or a gi'iinerite rock may be produced. Wliere silica was originally an abundant constituent both magnetite and the silicates are likely to develop. Thus we have various proportions of all the min- erals, producing the magnetite-quartz rocks, the actinolite-magnetite-quartz rocks, the griinerite-magnetite-quartz rocks, the actinolite-quartz rocks, and the griinerite-quartz rocks. <2 BANDING OF AMPHIBOLE-MAGNETITE ROCKS. <■ Usually a given formation, or member, does not show a perfectly homogeneous arrangement of the mineral particles. The original sedimentary rock is banded, and the different bands have different compositions. Naturally the trans- formation of these bands produces different combinations of minerals. Moreover, during the recrystallization there is a tendency for minerals of the same kind to segregate. Hence, in any of the above cases, where as a whole a certain Bet of minerals are dominant within a rock, a single mineral, or two combined, may be largely segregated in bands; and in the alternate bands the other minerals be largely segregated. Thus a banded rock, consisting mainly of magnetite and quartz, may have a banded appearance as the result either of the segregation of the quartz and magnetite in sepa- rate bands or, more commonly, the segregation of more quartz and less magnetite in one band and less quartz and more magnetite in another band. In a similar manner alternate bands may be made up of actinolite or griinerite with quartz o Van Hise, C. R., op. cit,., p. 838. b Cited by Gmelin- Kraut, Anorganisehe Chemie, vol. 3, p. 319. c Chambcrlin, R. T., The gases iu rocks: Pub. Carnegie Inst. No. 106, 1908, p. 61. d Van Hise, C. R., op. eit., p. 8.39. 'Idem, pp. 839-840. 552 GEOLOGY OF THE LAKE SUPERIOR REGION. in variiius proportions, and of actinolite or griinerite with magnetite in various proportionp. In still other instances the bunding may be due to the combining of actinolite or griinerite, magnetite, and quartz in various proportions. In general, therefore, the alterations of the rock do not destroy the original sedimentary banding, but, on the contrary, emphasize it. The staking banded appearance of actinolitic and griineritic rocks is one of their most characteristic features. BECRYSTALLIZATION OF atTAKTZ. The recrystallization of quartz under these anamorphic reactions has multiplied the size of the prain maiw times, as mentioned in the discussion of the individual districts. The rccrj-s- taUization of quartz has largety followed the development of magnetite, for magnetite with crystal outlines is often observed to be completely inclosed in large clear quartz crystals with no strain effects. HIGH STTLPHXTB CONTENT OF AMPHIBOLE-MAGNETITE ROCKS. The amphibole-magnetite rocks usually carrj^ a higher percentage of iron sulphide than other phases of the iron-bearing formations. If iron sulphide plays the important part assigned to it in the early portion of this discussion (see pp. 518-519), iron sulphides may be supposed to have been locally deposited throughout the iron-bearing formations with the carbonates and greenalites. These would be the first substances to be altered by the surface waters and, going quickly into solution, would greatly accelerate the concentration of the ore, but during the alteration of iron carbonate or greenalite to amphibole-magnetite rocks there is no opportunity for oxidizing solutions to get at the sulphides and hence the}" remain. The refractoriness of the amphibole-magnetite rocks also prevents subsequent oxitlization. In the Gunflint Lake dis- trict of Minnesota the sulphide is in the form of pyrrhotite, which, according to Moissan " and Allen, * is developed through the application of heat to pyrite. An alternative explanation of the high sulphur is that it was secondarily contributed by the hot intrusives. For this there is no direct evidence. SECONDARY IRON CARBONATE LOCALLY DEVELOPED AT IGNEOUS CONTACTS. In a few localities, as at Gunflint Lake, Minnesota, in the Animikie district, and at Sunday Lake, in the Gogebic district, coarsel}^ crystallized iron carbonate is found close to the igneous rock, this material doubtless being produced by recrystallization of the original finer carbonate. CONTACT ALTERATIONS NOT FAVORABLE TO CONCENTBATION OF ORE DEPOSITS. The anamorphic changes above described do not fav'or the transfer and segregation of con- stituents of the u'on-bearLng formations. They tend rather to combine them. Localh* there is evidence that iron is carried in solution under these conditions, in the fact that cements in fractures are largely magnetite and the iron is usually in coarser bands. If the intrusions come before the original iron-bearing formation has become porous tlu'ough the loss of its silica, the rocks do not have the openings for the transfer of solutions. Even had openings existed in some places, the deep-seated pressures exerted by great batholiths, like the Dulutli gabbro, have been sulficient to make tlie rock undergo rock flowage, thereb}- closmg openmgs. If other conditions were favorable there would still be the lack of abundant surface waters to leach the silica. So far as the iron-bearing formation ]ia32. t> Allen, R. C, Iron formation of Woman River: Eighteenth Ann. Rept. Ontario Bm-. Mines, pt. 1, 1909, pp. 254-262. 556 GEOLOGY OF THE LAKE SUPERIOR REGION. Guy II. Cox lias assembled tlie vjirious experimental data on the su})ject and supplemented them b_y laboratory experiments of his own. From meteoric solutions under ordinary temperatures at the surface the precipitates of iron are ferric hydrates containing 29 per cent of water, which rapidly changes in contact with water into limonite. containing 14.44 per cent of water. The presence of alumina, lime, and magnesia to combine with the iron may prevent dehy- dration." If left for several years, the ore becomes dehydrated and crystalline.* Increase in temperature and pressure on the solutions at tlie time of precipitation will lower the hj'dration of the precipitated salt. At a temperature of 500° magnetite may be precipitated directly from solution. Slight variations in the degree of hydration in a precipi- tate are determined by tlie form in which the iron is held in solution, by the precipitating agents, and by the strength of the solutions, though so far as experimental data go the range of variation due to these causes is small. Secondary alterations have little eflFect on anliydrous ores, but liydrous ores may easily lose part of their water by moderate increase in temperature and by pressure such, for instance, as that involved in freezing, where the water is allowed to escape. It appears also that in an ore containing various hydrates, solution will dissolve the highest hydrates, leaving the residue in a lower state of hydration, but that the redeposition of tlie dissolved part as a higher hydrate may result in net increase of hydration for the residue and dissolved parts combined. It appears, therefore, that conditions of high temperature and pressure, either during the original deposition of the iron salts or during their secondary alterations, favor the development of anliydrous salts, thereby explaining the occurrence of crystalline hematite and magnetite in the iron-bearing formations near igneous contacts or where djmamically metamorphosed. It is shown elsewhere that [magnetite, perhaps even hematite, may have been precipitated directly from the hot solutions coming from some of the basic igneous rocks, or that the iron salts may first have been deposited as greenalite and iron carbonate which subsequently altered under conditions of high temperature and pressure to magnetite and hematite, or that the iron salts were first deposited as greenalite and hematite, subsequently altered to limonite, and then dehj'drated by the high temperature and pressure of anamorphic conditions to hematite and magnetite. In all these cases the heat from some adjacent igneous rock or the pressure developed from rock flowage seems, from field evidence, to be an essential factor. However, hematite and various hytlrates are found minutely interliedded in parts of the iron-bearing formations where there is no evidence of the effect of unusual heat or pressure. A hand specimen may show several layers of iron oxides with varymg degrees of hydration. These differences persist in the ferruginous cherts and jaspers and in the ores into which the ferruginous cherts and jaspers grade. Moreover, they seem to be independent of distance from rock surface and of dip of beds. In steeply inclined beds layers with different degrees of hydra- tion may be found to continue from the surface to great depth with no relative change in hydration. These remarkable and persistent variations in hydration in closely associated layers ma}' have been due to — 1. Differences in the original substances in different layers, whether carbonate or greena- lite. The iron-bearing formations were originally anhydrous iron carbonate and hytlrous silicate, both of which have altered when weathered to hydrous oxides. It has not been ascer- tained that there is any specific difference in degree of hydration of the alteration products of the greenalite and carbonate, though on the whole the beds in the Mesabi district, containing the most greenalite, are the most hydrous. 2. Difference in time of alteration of the greenahte and carbonate, vnth accompanying slight variations of temperature and pressure. The hydration of different layers has taken place at o Spring, W., Neues Jahrb., vol. 1, ISDO, pp. 47-ti2 (cited by Moore, E. 8., Eighteenth Ann. Rept. Ontario Bur. Mines, pt. 1, 1909, p. 194). t> Wittsteln. G. C, Vierteljahresschrltt fiir Pharmacie, vol. 1, 1852, p. 275 (cited by Moore, E. S., Eighteenth Ann. Rcpt. Ontario Bur. Mines, pt. 1, 1909, p. 194). THE IRON ORES. 557 different times when the temperature conditions anil jiressure conditions may have been sHghtly diflferent, although of these differences we have no knowletlge. 3. Selective secondary alterations of the iiydrates formed by the first alteration of the green- alite and carbonate. Freezing (seasonal and glacial) and moderate depth of cover may tend to dehydrate the ores and probably have contributed to the low average degree of hydration of the bedded hematites. So far as experimental evidence goes, these ores would have their highest degree of hydration at the time of precipitation, and all influences acting upon tliem subsequent^, even moderate seasonal variations in temperature and moderate depth of burial, would tend toward lowering the degree of hydration. It might be expected that the result of seasonal variations in temperature and the pres- sure of overlying rocks would result in a uniform variation in hyth'ation from the surface down- ward. No evidence of this sort has been found in the ore bodies. It should be noted, however, that the effect of freezing would be toward dehydration at the surface and the effect of pres- sure would be toward dehydration with depth. Instead of uniform change iia hydration one way or another from surface to depth, the most conspicuous change in hydration is between closely interbedded layers of the iron-bearing formations. I The selective effect of solution and redeposition might have influence; for instance, waters percolating rapidly along a certain bed or fissure might dissolve the more hydrated ores, carry them off, and redeposit them, leaving the residue with a lower degree of hydration. Slight original variations in hydration would thereby be emphasized. Other unknown causes may be operative. According to Stremme," hydration is favored by salt content and carbon dioxide content of the altering solutions. The salt and acid content apparently influence the degree of hydra- tion of the u'on oxide by lowering the vapor pressure of the solution. Each ii-on hydrate is supposed to have its own vapor pressure, which is the minimum pressure of water vapor with which the hydrate can remain in equilibrium at any given temperature. We may conclude in general that the hydrous ores of the Lake Superior region have devel- oped under ordinary concUtions of temperatiu'e and pressure near the surface, that the anhy- drous ores exhibit the effects of heat and pressure, and that the differences in hydration of closely intermingled layers of the iron-bearmg formations have requu'ed some influence of a selective sort, the nature of wliich may be suggested but not proved. SEQUENCE OF ORE CONCENTRATION. We have touched upon each of the factors going to determine the present character and structural relations of the ores. To complete the picture we have now to dwell upon the chron- ologic development of the ores. The beginning of the processes of secondary concentration must be placed for the Archean ores in early Huronian time and for the middle Huronian ores in the time between the middle and upper Huronian. Iron-formation fragments in the basal conglomerates of these divisions tell to some extent what had previously happened to the iron-bearing formations of the ohler land. At the base of the upper Huronian rich ferruginous detritus was formed at the beginning of upper Huronian time. In certain places the iron-bearing formation within the upper Huro- nian was exposed by erosion before Keweenawan time and went through a set of changes in the time interval between the Huronian aiid Keweenawan similar to those that affected the lower Huronian iron-bearing formation in inter-Huronian time. This is shown by the detritus of the Keweenawan basal conglomerate and by the development of red jaspers and hard ores from the soft varieties near the contact of Keweenawan and upper Huronian in eastern .Gogebic dis- trict. In those districts in which great masses of Keweenawan rocks were laid down upon the Huronian rocks before the iron-bearing formation had been exposed to weathering, the concen- tration of the ore could not have begun until the Keweenawan was cut through in the erosion a Strenime, H., Zur Kenntnis der wasserhaltigen und wasserfreien Eiseno.xydbilduDgen in den Sedimentgesteinen: Zeitschr. prakt. Geologic vol. IS, No. 1, 1910, pp. 18-23 (reviewed in Econ. Geology, vol. 5, 1910, p. 499). 558 GEOLOGY OF THE LAIvE SUPERIOR REGION. period precedincr Cambrian time, and it is rather probable that this hmitation also applies to other districts. Clearly the process in each district began when, as a result of the great oro- geiiic movements and the attenilant denudation, the iron-hearing formation was exposed to the weathering forces. In most of the districts this occurred in the great time gap represented by the unconformity between the Keweenawan and the Cambrian. At this time were concen- trated most of tlie great ore deposits of the upper lluronian of the region and the ores at the middle and lower horizons of the Negaunee formation of the micklle Huronian. Wherever the Cambrian remains in or near the iron districts it contains iron-ore frag- ments, jaspers, and clierts in its basal conglomerate. In the Menominee district these are rich enough to be mined. The process of ore concentration was therefore well advanced before Cambrian time. In the Alesabi district remnants of Cretaceous beds overlie some of the ore deposits, par- ticularly in the western parts of the range. At the basal horizons of these beds are detrital iron ores derived from the Biwabik formation. Here, then, the concentration was well advanced as early as Cretaceous time, and there is little doubt, from the similar relations of the ores to the Cambrian in other regions, that the ores of tiie Mesabi district were well concentrated even by Cambrian time. The process of enrichment has undoubtedly continued until the present time. It there- fore appears that the circulating waters have had eras in which to perform their work; indeed, a part of pre-Paleozoic time and all of the Paleozoic, Mesozoic, and Cenozoic. Frequently during pre-Cambrian time the ii-on-bearing formations were metamorphosed by igneous intrusions, the principal effect of which was to recrystalUze the original phases of the iron-bearing formations, yet unaltered, to refractory ampliibole-magnetite rocks able to resist the ordinary katamorphic ore-concentratmg agencies. The alteration to ores of portions of the iron-bearing formations so modified was practically stopped at the times of the intrusions. In all the districts since the beginning of final concentration many thousands of feet of strata have been removed by erosion. During the process of denudation the ore deposits in each district began to be secontlarilj- concentrated shortly after the iron-bearing formation was exposed at the surface and for a long time they continued to increase in size. It is probable that after a sufficiently long period the growth of the deposits practically ceased, for denudation would finally remove the ores at the surface as fast as they formetl below the surface. However, change would not stop. The ore deposits formed would continue to migrate downward pari passu with denudation. On account of the pitch, lateral migration would accompany downward migration. At any given time the masses of ore would extend from the surface to the depth at which descending waters were effective. We therefore must conceive of the secondarily concen- trated iron-ore deposits as slowly migrating downward through thousands of feet, being always just in advance of the plane of erosion. So far as the original iron-formation layers were rich enough to be ores without secondary concentration, these statements do not apply. The amount of ore existing at an}^ one period tlirough much of preglacial time may have been roughly constant, although there was doubtless considerable variation depending on topo- grapliic and climatic conditions. At times the processes of denudation would go on rapidly; at other times they would be stayed for long periods, depending on the post-Keweenawan history of the Lake Superior region. The important steps of tliis history are (1) the great pre-Cambrian mountain making and erosion, (2) subsidence and Paleozoic sedimentation, (.3) the post-Paleozoic uplift and denuda- tion, (4) the deposition of Cretaceous rocks upon parts of the region, (5) the post-Cretaceous uphft and succeeding denudation, and (6) the Pleistocene ice incursions. 1. In the pre-Cambrian period of mountam making and denudation the ore deposits probably reached their full development, and indeed they maj^ during the latter part of this ancient time have been of greater magnitude than they are at present, although possibly not so rich. In the Menominee district the Upper Cambrian sandstone and the Ordovician lime- stone cap the Huronian formations and even some of the ore deposits. The upward extension THE IRON ORES. 559 of the iron-bearing formation was removed before Upper Cambrian time. It is clear, therefore, that the main concentrations of iron oxide for these deposits must have taken place in pre- Cambrian time. The basal conglomerates of the Cambrian carry ore fragments from previ- ously altered formations. If, as is probable (see below), Cambrian and Ordovician or Silurian strata capped the beds in other iron-bearing districts of the Lake Superior region, it is all but certain that ore concentration was equally advanced in these other districts, although where erosion has extended farther below the Paleozoic than in the Menominee district later events have had a greater influence upon the present condition of the ore deposits. The later stages of this period of denudation were marked by the development of a great peneplain, over which, it may be assumed, the ore-concentrating processes acted slowly. 2. After this period of denudation the Paleozoic sea encroached upon the Lake Superior region. Where the iron-bearing formations were reached by the sea, detrital ores were formed at the base of the Cambrian. The entire region was deeply buried beneath the Paleozoic deposits. Probably so long as the region remained below the sea the processes of concentra- tion practically ceased antl the mass of the ore deposits remained nearly stationary. Sea water does not chemically affect the iron oxides. 3. Wlien after Paleozoic time the region was again raised above the sea and denudation began, little enrichment took place until the major portion of the Paleozoic rocks was stripped from the region. Over much of the region these Paleozoic rocks were entirely removed, and the pre-Cambrian Huroniau surface again emerged from below the Cambrian deposits. In the Menommee district and the southeastern part of the Crystal Falls district the Paleozoic deposits were not completely removed from the iron-bearing formations, and here considerable quantities of detrital ores are found at the base of the Cambrian. In most of the region erosion did not stop at the Paleozoic but extended downward for a greater or less depth into the Huronian rocks, and it is presumed that where this took place the ore deposits migrated downward precisely as durmg the pre-Cambrian period of denudation. 4. Erosion continued until the end of the Cretaceous period of base-leveling, when the area was again reduced nearly to an uneven plain and locally was overridden by the sea and capped by Cretaceous rocks, at least as far east as the Mesabi district. The basal strata of these beds carry detrital iron ore from the Biwabik formation. At the end of this period the processes of downwartl denudation and concentration were greatly diminished in speed. 5. Durmg the period of the post-Cretaceous uplift denudation and the migration of the ore deposits again went on, but to what extent is uncertain. It is highly probable that m the Menominee district the topography of the Huronian rocks is largely pre-Cambrian and the present depressions to a large extent are reexcavated pre-Cambrian valleys. The same is true of the Felch Mountain tongue of the Crystal F^lls district. On the borders of the Marquette district, also, Cambrian deposits are found. However, it is now a matter of conjecture as to how far the present topography is redeveloped pre-Cambrian topography and how far it is post-Cretaceous. 6. The last great event in the development of the ore deposits was the glacial incursion of Pleistocene time. So far as the ore deposits are concerned, the work was of two kinds, glacial denudation and glacial deposition. The quantity of ore which was removed during the first stage of Pleistocene time, that of glacial erosion, was enormous. Almost the entire zone of decom- posed rocks which must have been adjacent to the ores has been removed. The ore deposits were certainly truncated to at least an equal depth. Glacial erosion also in many places cut deeper into the soft ore bodies than into the adjacent hard rocks, and thus produced subordinate valleys, as is finely illustrated in the Mesabi district. The abundant fragments of hard iron ore in the glacial drift furnish evidence of the large amount of ore wliich has been removed b\' the glaciers. It is certain that still greater quantities of soft ore have been removed, although on account of its softness it has been broken into minute fragments and therefore furnishes little evidence of its removal. The foregoing considerations lead to the certain conclusion that the glacial truncation seriously reduced the amount of available iron ore in the Lake Superior region. WTiile the pi'ocess of concentration has continued since glacial time and has tended to 560 GEOLOGY OF THE LAKE SUPERIOR REGION. enrich and deepen the deposits, there is no doubt that the gain since the glacial incursion is insignificant as compared with the loss of rich material during the glacial period. Wlien the glaciers receded, the clean-cut ore bodies were covered to a greater or less depth by deposits of glacial drift. This relation may be seen to the best advantage in the great open pits of the Mesabi district, where the soft, clean ore extends directly to the drift, not derived from the ore but brought from the north. The contacts in many places are of almost knifeUke sharpness, there being practically no ore in the basal layers of the drift. It appears from the foregoing discussion that wliile the quantity of ore in the Lake Superior region has alwaj^s been large since Cambrian time, there have been numerous vicissitudes in its history during which the quantity of ore alternately increased and decreased. ORIGIN OF MANGANIFEROUS IRON ORES. Manganese exists in a series of minerals remarkably similar to and usually in association with those of iron. The origin and secondary concentration of the manganese minerals have been regarded in general as following very closely those of the iron. The subject has not been specifically studied for the Lake Superior region. It may be noted here merely that the man- ganese tends to be concentrated in the upper parts of the Lake Superior iron-ore deposits, and that as secondarily concentrated it consists piincipally of manganese dioxide (pj-rolusite) and subordinately of manganese carbonate. In the general study of the manganese deposits of the Appalachians and other parts of the United States it has been found that this is a common but not invariable relation of iron and manganese. In some deposits also the relation is reversed, the iron being above, the manganese below. Where they are associated with clay, not in the Lake Superior region, thei'e seems to be a tendency for the concentration of clay at the surface relative to the manganese. Iron and manganese oxides and cla}- are the most stable of the common constituents of the belt of weathering, and hence all of them tend to become residually concentrated as compared with other substances originally associated wnth them. The vertical distribution of these three substances is taken to be a function of their relative stabihty under various conditions of weathering, but the available information does not seem to warrant more specific statements. PART OF THE METAMORPHIC CYCLE ILLUSTRATED BY THE LAKE SUPERIOR IRON ORES OF SEDIMENTARY TYPE. Starting with the ferrous iron and dominance of silicates in the original igneous rocks, the development of the ore deposits is a process of continuous katamorphism. From the original igneous rocks and their included veins containing a small percentage of iron there is developed an iron-bearing formation — cherty iron carbonate or greenaUte — containing 25 or 30 per cent of iron, which, on further alteration at the surface, becomes concentrated to 50 or 60 per cent or more. The iron-bearing formation and included ores may themselves be broken up to yield materials for later sedimentary iron-bearing formations. The upper Iluronian iron-bearing formations, the greatest and most i^roductive of the Lake wSuperior region, ma}- be regarded as including materials not only from the chemical alterations of the older greenstones but from the destruction of the older iron-bearing formations of the middle Iluronian and Archean. These formations have undei-gone the extreme of katamorphism. Nature's great concentrating mill has developed a liigh-grade end product, both chemical and mechanical, through a series of concentrations. The changes have been those of simplification and segregation of mineral compounds, mai'ked increase in volume, when all substances entering into the reaction are taken into account, incoherency of substance, and net liberation of heat, all of them typical of the katamorphism or destructive processes affecting the earth's surface. No sooner have the ores reached their maximum incoherency through katamorphic changes than constructive agencies begin their work. It may be more correct to say that they begin before the destructive agencies have finished. The ores become cemented and strengthened; they tend also to become dehydrated and more or less magnetic. As they become buried THE IRON ORES. 561 beneath the surface, owing to the deposition of later sediments, and as they become folded, their volume is decreased by an elimination of pore space and moisture, they are recrystalHzed, are shghtly deoxidized to magnetite, in small part combine with siUceous and other impurities to produce sihcates, and are frequently rendered scliistose, producing the hard specular ores. The mineralogical change is one froin simple to less simple compounds. The net change in energy is loss, due to the energy given off in volume decrease. The process is a characteristic one of anamorphism, which affects all rocks under similar conditions. The anamorphic changes in the ores are best shown in the oldest or Aixhean iron-bearing formations. More marked anamorphic results are produced under the influence of igneous intrusions. The contrasting katamorphic and anamorphic changes affecting the ore deposits constitute a partial metamorphic cycle." Beginning with a coherent igneous rock, incoherent ore deposits are developed through kataniorphism and in turn a part are rendered coherent again through anamorphism. The mineralogical changes are at first from complex to simple and later from simple to complex. The changes at first are essentially those of simplification and segregation and later this process is arrested and on a smaller scale reversed in the development of the complex silicates. The ores are not essentially dispersed to again become constituents of igne- ous rocks, although certain of the amphibole-magnetite rocks associated with the ores are not easily distinguishable from igneous rocks. The cycle, therefore, so far as observation goes, is not complete. There is throughout a net loss of energy. TITAXIFEROUS MAGNETITES OF NORTHERN MINNESOTA. The great gabbro mass of Lake and Cook counties, i\Iinn., contains much magnetite, both disseminated and segregated into ore deposits. Complete gradation may be observed between gabbro carrying little magnetite and magnetite carrying little of the ferromagnesian con- stituents and feldspars. The knowai deposits are extremely irregular, with gradations between themselves and the gabbro and containing within themselves much gabbro material. They weather very much like the gabbro and might be easily unnoticed on the weathered surface. There has been little exploration for these ores. A few drill holes have been sunk in the region south of Gunfhnt Lake, some of them revealing depths of ore aggregating several hundred feet. The known deposits seem to be distributed in irregular zones roughly parallel to the north or basal margin of the Duluth gabbro. The composition of the ore averaged from 3,556 feet in 14 drill holes is 43.8 per cent of iron. The range is from 54 to 20 per cent. The high titanium content renders the ores of doubtful value for the present. Where the gabbro comes into contact ^vith the iron-bearing Gunflint formation both formations carr}' magnetite so similar in texture that it is difficult to tell one from the other. However, on analj'sis the gabbro magnetite is found to be titaniferous, while that of the Gunfhnt formation is not titaniferous. This fact seems to argue against any considerable transfer of material from the gabbro to the iron-bearing formation during its alteration. The titaniferous magnetites of northeastern Minnesota are direct magmatic segregations in the Duluth gabbro, according to all geologists who have studied them, including Irving, Merriam, Bayley, Grant, Winchell, Clements, Van Hise, Leith, and others. The complete gradation from gabbro with a small amount of original magnetite to a magnetite with small amounts of amphibole and other gabbro minerals can be seen in almost any part of the titanif- erous magnetite deposits. It is scarcely necessary to repeat the detailed petrologic evidence so fully given by the writers named. Evidence is given elsewhere for the intrusive character of the Duluth gabbro. It cooled far beneath the surface, where there was not easy escape for its solutions. This fact is taken to explain its retention of its iron oxides. It has been argued under an earlier heading that where basic rocks of similar composition reached the surface large quantities of iron escaped and became available for ordinary sedimentary' deposition. o Leith, C. K., The metamorphic cycle: Jour. Geology, vol. 15, 1907, pp. 303-313. 47517°— VOL 52—11 36 562 GEOLOGY OF THE LAKE SIJPEIIIOR REGION. IVLVGNETITES OF POSSIBLE PEGMATITIC ORIGIN. The ore in the Atikokan district is a magnetite, higiil^- iinprcgnatcd with amphibolos and sulphides and showing extremcl}' close and intricate relations to associated diorite. It difFers from the magnetite of the gabbro of Minnesota in being nontititniferous and in being separated by defmite boundaries — in many places plane surfaces — from the adjacent wall rock. The apparent absence of iron-bearing formation, the general lack of banding, the high content of amphibole corresponding to that in the associated diorite, the content of sulphides, and the extremely intricate structural association with the diorite are not easy to explain if the ore is sedimentary and owes its character to complex intrusion by the basic igneous masses. Nowhere in the Lake Superior region is intrusion known to completely destroy banding, nor does it develop so much coarsely crystalline amphibole and iron sulphide with lack of parallel texture. On the other hand, both character and relations suggest pegmatitic intrusion or igneous after- effects, similar to those described by Spencer " for the New Jersey magnetites or by Leith * for certain western magnetites. The evidence for pegmatitic origin of the ores of the Atikokan district is weak. This district lies outside of the principal area studied in connection with this report, but from our examination of it we suggest this origin as a plausible one from the facts available. Certainly this district seems to show marked variations from most of the districts of the Lake Superior region — variations which seem to call for another mode of derivation. Minute pegmatitic veins of quartz or iron oxide or both are common in the ellipsoidal basalts of the Vermilion district. In the coarser phases they may be seen to be intimately and irregularly mixed with the rock, and grading out toward the finer phases they tend to take on more definite vein outlines. In the Keewatin series as represented in tlie Vermilion district it is in many palces dilhcult to determine whether the iron-bearing formation is a magmatic segre- gation of greenstone, a vein material of a pegmatitic nature, or an ordinary iron-bearing sediment derived from them. In Plate XLVIII are shown gradations from the basalt through siliceous and jaspery phases to ordinarj' banded iron-bearing formation. These intermediate phases seem to be of a pegmatitic nature. BROWN ORES AND HEMATITES ASSOCIATED T^^TH PALEOZOIC AND PLEISTO- CENE DEPOSITS IN WISCONSIN. ORES IN THE POTSDAM. In the driftless portion of the Potsdam area north of Wisconsin River in western Wisconsin there are many small patches of hematite and brown ore, closely associated with upper horizons of the Cambrian (Potsdam) sandstone. Many of these patches lie on the tops and slopes of liills, but some of them follow the valleys. During the early days of mining in Wisconsin these ores were smelted locally at a furnace in Sauk County, but for 30 years they have not been mined, principally because of the small- amounts available. The origin of these ores is not clear. Occurring near the upper horizons of the Potsdam, some of them may represent residual accumulations due to erosion of the overh'ing Ordovician limestone. Samuel Weidman "^ believes that part of them at least are results of later valley filling by spring and bog solutions. BROWN ORES IN "LOWER MAGNESIAN " LIMESTONE. At Spring Valley, in Pierce County, Wis., are notlules and irregular masses of limonite in clays, resting upon the eroded surface of the "Lower Magnesian" limestone, particularly in old drainage courses on the surface of this lunestone. Quoting from .Mlcn:<* o Spencer, A. C., Franklin Furnace lolio (No. 161), Oeol. Atlas U. S., U. S. Geol. Survey, 1908, pp. 6, 7. t Loilh, C. K., Bull. V. S. Geol. Survey No. 338, 1908, pp. 75-89. « PersonaU'oiumunication. d Allen, U. C, statement prepared for this monograph. See also Allen. R. C, The occurrence and origin of the brown iron ores of Spring Valley, Wisconsin: Eleventh Keport. Michigan Acad. Sci., 1909, pp. 95-103. .' PLATE XL VIII. 663 PLATE XLVIII. FeREUGINOUS chert or JASrER, OF POSSIBLE PEGMATITIC ORIGIN, IN BASALT. A. Partly silicified basalt (specimen 2S564) from Vermilion district, Minnesota. In the ledge this is observed to grade imperceptibly into the little-altered basalt of the region. B. Chert, green silicate, and iron oxide (specimen 28565) from Vermilion district, Minnesota, more definitely seg- regated into bands, grading imperceptibly into the rock sho-mi in A on the one hand and into that shown in C on the other. C. Same (specimen 28566), with larger proportion of iron in bands. This is an amphibolitic ferruginous chert or jaspilite of a type often seen in the iron-bearing formations. In the ledge from which this series of specimens was collected, it was quite impossible to find any plane of separation between basalt and iron-bearing formations. 564 U. S GEOLOGICAL SURVEY MONOGRAPH Lll PLATE XLVIII FERRUGINOUS CHERT OR JASPER, OF POSSIBLE PEGMATITIC ORIGIN, IN BASALT. THE IRON ORES. 565 Spring Valley ia a small town on Eau Galle River reached by a spur from Woodville, on the Chicago, St. Paul, Minneapolis and Omaha Railway. Iron ores were discovered in the vicinity of Spring Valley about 20 years ago. Thorough prospecting developed a number of deposits, two of which, known as the Oilman and the Cady deposits, are being mined. The Oilman was opened about 1890 and has been in operation more or less continuously since that time. In 1893 a furnace was erected at Spring Valley tor utilizing the Oilman ores and numerous charcoal ovens were built in the vicinity for supplying fuel for the furnace. Wood soon became scarce and coke supplanted charcoal as a fuel. The original plant has been partly replaced by a more modern one. GEOLOGY AND TOPOGRAPHY. The Upper Cambrian sandstone underlies the valleys and lower hill slopes. The uplands are formed by limestone of Lower Ordovician age. The strata are conformable and flat-lying. The topography is that of the maturely dissected plateau, and is essentially of preglacial origin. Eau Galle River and its tributary creeks are flowing through partly filled valleys. If the valleys were to be filled to the average height of the ridges the resulting surface would be a plain. A plain probably once existed here as part of a greater one which extended over a surrounding broad area. The present topography may lie explained as resulting from the uplift of this -ancient plain, giving the streams new erosive power. Before glacial time the streams had sunk their valleys through the Ordo\dcian limestone and well into the underlying Cambrian sandstone. During the glacial epoch the valleys were partly filled by glacial wash. OILMAN BROWN-ORE DEPOSIT. The Oilman deposit rests upon an eroded surface of the Ordovician limestone, near its base, on the upper slopes of a ridge above the valley of a small creek tributary to Eau Oalle River. It is on the railroad and is li miles west of Spring Valley. The deposit covers several acres and in outline is very irregular, as shown liy the mine workinos which are open shallow excavations, the deejjest being not more than 30 feet. The ore is a brown hj'drated hematite and occurs as nodules and concretions mixed irregularly with ocherous clay, sand, chert fragments, and nodular con- cretions of sand and clay. Locally the deposit shows rough and irregular bedding, but the general absence of beddin" is conspicuous. The limestone presents an uneven surface to the bottom and sides of the deposit. In one place a wall of limestone some 6 or 8 feet high, showing undoubted e\ddence of having been eroded while exposed to the air, abuts directly against the ore. In places the ore comes quite to the surface, but as a rule it is covered by a foot to several feet of clay. All the mining is done by hand. The larger nodules of ore, called "rock " ore, are picked by hand from the clay and sand in which they are embedded. Some of them are very large and need to be broken up by blasting. But most of the ore in the Oilman mine is removed with the impurities in which it occvu's and put through barrel washers. The following is the analysis of a three months' sample of "rock" and "wash" ore: Analysis of ore from Gilman mine. Fe 43. 6 SiO. 24.00 A1263 2.3 CaO 58 MgO. P.... S Mn.. 0.30 . 14 .018 .80 CADY BROWN-ORE DEPOSIT. The Cady deposit is 2i miles northwest of Elmwood and about 5 miles southeast of Spring Valley. It covers several acres on the top and upper slopes of a hill that rises steeply some 200 feet above the valley of Cady Creek. As in the Gilman deposit the ore rests on the Ordovician limestone. At the time of visit in 1906 the deposit had not been opened, l)Ut the ore was exposed in numerous pits and trenches. According to W. H. Foote, a shaft went down through 80 feet of ore and struck a face of limestone at that depth which was at an angle of 60° with the horizontal. Ore was followed down this face for 40 feet more with no bottom. The following analyses indicate the character of the ore in this shaft: Analyses of Cady Creek ore. Thickness (feet). Fe. SiO.. Mn. P. 10 10 16 22 2S 34 40 45 50 55 60 65 59.12 49.96 47.79 32.96 46.56 52.02 37.91 55.11 53.66 52.02 52.22 54.18 9.0 14.33 20.5 45.25 22.17 U.82 35.34 2.03 .83 1.39 2.13 2.47 2.51 1.82 1.73 2.72 2! 25 1.91 1.33 Brown lump ore. . . . 073 Do Do ... 077 Do 054 Do . ■ . . . 068 Do Do ..... 062 Do Do.. . . . . .... 063 Do 566 GEOLOGY OF THE LAKE SUPERIOR REGION. The ore contains a Bomewhat higher percentage of iron, has a greater proportion of rock ore, and is associated with a less amoiint of impurities (sand, clay, etc.) than the Oilman ore, but is otherwise exactly similar to it. Mining has recently bc},'uii. The oro is delivered to the bins at the base of the hill by an aerial tram. The descending loaded buckets return the empties to the to]) of the hill, ORIGIN OF SPRING VALLEY BROWN-ORE DEPOSITS. The ores near Sprinj^ A'lilley are of superficial ori^jjiii, beiuf^ deposited upon the eroded surface of limestone and other rocks. Allen has shown, from a consideration of the thickness of the strata once overlying the present ores and their probable content of iron, that the now known dejjosits were probabl_y not the result of direct downward slump of residual materials but are rather sediments transported laterally along drainage channels after the country hud been cut down to the elevation of the ores. Allen shows further that since the formation of these deposits erosion has cut through them and around them, with the result that the adja- cent territory has been lowered, leaving the deposits on the tops or slopes of hills. He con- cludes that the ore deposits of Spring Valley were laid down in lakes or marshes that existed along the drainage courses on the old post-Devonian peneplain, or on the valley bottoms, as may have been the case in the Giiman and Cadj- deposits, where the ore abuts directly against eroded limestone faces. The marshes and lakes were finally drained as a result of uplift of the land which enabled the streams to erode vertically at a greatly increased rate. Narrow valle3^s were formtd in the older, broader ones. The outer margins of the old valleys correspond with the upper slo]>es of the hills forming the present valley sides. It is on tliese upper slopes that the ores characteristically occur. As erosion progressed ore-covered areas would naturally come to occupy liigher and higher relative elevations, owing to the resistant nature of the ore beds. In this way would result ore-covered hilltops, as illustrated by the Cady deposit. Weidman," who has made a survey of the region surrounding Spring Valley, while accepting the general view that tiie Spring Valley ores are of superficial origin and were deposited upon the eroded surface of the limestone and associated rocks, is inclined to place the date of their origin long after the period of peneplanation of tlie region. This alternate hj^iiothesis supposes the ore to have been formed in these valleys after they were eroded to a considerable depth — 200 to 300 feet — in the peneplain and perhaps even at the still later stage when the valleys were in the process of being filled again with alluvial material. The deposits lie in secondary and tertiary valleys and on slopes opening outward toward larger valleys, and the massive, lumpy character of the ore indicates that it may very well have originated in the manner of iron-spring deposits, accompanied by more or less slope wash and slumping of clay and sand wliile the valleys were bemg filled. Since the valley's were partly filled and the ores formed, erosion has removed 30 to 40 feet of alluvial material from the valleys and a variable amount of the ore. This explanation as to the date of origin of the ore — namelj', at the time when the valleys were well developed — seems to apply very well to the Giiman ore deposit, where most of the ore has been removed and where the relation of the ore deposit to the topography can be clearly observed, • and it jirobably also applies equally well to the Cady deposit, where mining is not sufficiently advanced to show the actual conditions. POSTGLACIAL BROWN ORES. Postglacial iron ores are known in many parts of the Lake Superior region. They are ordinary bog deposits to which iron is being contributed in solution under the influence of organic material and deposited by oxidation. Nowhere is their thickness known to be over a few feet. Lj'ing, however, directly at the surface, they frequentlj' attract attention and for man}' years have been subject to intermittent exploration. a Weidman, Samuel, Geology of northwestern Wisconsin: Bull. Wisconsin Geol. and Nat. Hist. Survey. (In pi«paration.) THE IRON ORES. 567 CLINTON IRON ORES OF DODGE COUNTY, WIS. OCCURRENCE AND CHARACTER. Iron ores of Clinton age, similac to ores of the same horizon in the Appalacliian region, appear in Dodge County, in soutlieastern Wisconsin. The shipments to the end of 1909, wliich figure in the total for the Lake Superior region, have aggregated 570,886 long tons."^ The ores outcrop in a narrow belt extending for about a mile north and south on a westward-facing scarp caused by the overlying Niagara limestone. The underlymg rock is Ordovician shale. The dip is eastward at the rate of about 100 feet to the mile. The beds are lens shaped along the outcrop and range in tliickness up to a maximum of 37 feet. Mining operations have followed them 400 feet down the dip, and they are known by drilUng to extend farther. Wells have shown the occurrence of ore m the southeast corner of the county and near Hartford in tluck- nesses ranging from 4 to 20 feet, and a diamond-drill hole near Kenosha, 60 miles to the south- east, cuts 18 feet of ore. The iron beds, if continuous eastward to Lake Michigan, a distance of 35 miles, are nowhere more than 800 feet below the surface, for the Niagara limestone which overlies them has this thickness and it outcrops all the way to the lake. If we assume an average tliickness of 10 feet, an extension down the dip of 2,000 feet, and contuiuous extension southward to Kenosha (which is doubtful), the amount of ore in these deposits would be 600,000,000 tons. The ore is a slightly hydrated hematite, running from 29 to 54 per cent m iron and aver- aging perhaps 45 per cent, high in phosphorus, with the typical granular, oolitic, or flaxseed forms so characteristic of the ores of the Appalacliian area. The matrix is calcite. Bedding is distinct and false bedding is common. The granules he vnth their flat sides parallel to the bedding. The individual granules have been worn shiny by water action and aggregates of them have been rounded into pebbles. Under the microscope the iron-oxide granules are found in part to be amorphous and in part to have the concentric structure of ooUtes. Clastic grains of quartz or of iron oxide commonly form the nucleus, surrounded by alternate layers of iron and siUca. On treatment with hydro- chloric acid the iron is dissolved, leaving little globular particles of amorphous siUca, forming at first casts of oolites but on drying falling in, giving a basin-shaped indentation on one side. In the Clinton formation of the East some of the granules have the structures of replaced marine shells, but these have not been noted in the iron-bearing formation of this horizon in Wisconsin. The origin of some of the amorphous granules is observed in experimental precipitation of ferric hydroxide in laboratory solutions where the precipitate is allowed to settle slowly. There is then observed a marked tendency fc^r the aggregation of iron oxide into granules identical in shape and size with the granules observed in the Chnton ores. These granules are of the type regarded by Lehmann'' as liquid crystals, a globular form precedmg develop- ment of crystal structure and indefinitely grading into it. In materials that have- strong crystalHzmg power this globular stage is soon passed or is not even observed. In substances weak in crystallizing power, such as iron oxide, the tendency inherent in the substance itself to group or crystallize does not go beyond this stage of globular aggregation. Along the top of the ore body, at the contact with the overlying limestone, is a thin layer, rangmg from less than an mch to 6 inches m tliickness, of a hard, compact bluish hematite, heavier than the oolitic ore and ruiming about 10 per cent higher in metallic iron than the main body of oolitic ore. In this hard bed there is no trace of the oolitic structure. However, there is an apparent gradation from one to the other. The contact between the ore and the Niagara limestone might be termed a "knife-edge," as it is perfectly well defined, showing no gradation whatever from the iron into the limestone. The lower contact of the ore body mth the underlying calcareous shale is similar to the upper contact. Under the microscope some of the calcite grains in the limestone near the contact are a Lake Superior iron ore shipments for 1909 and previous years, compiled bj' Iron Trade Review, Cleveland. ' Lehmann, O. , Fliissige Kristalle sowie Plastizitat vpn Kristallen im AUgemeinen, molekulare Umlagerungen und Aggregatzustandsander, angen, Leipzig, 1904. 568 GEOLOGY OF THE LAKE SUPERIOR REGION. observed to be partly replaced by iron oxide, while other large calcite grains are the result of recrystallizntion. However, those are not common. In the lower contact the calcito Ls dis- colored by tlie iron oxide, but where tliis iron stain occurs we do not always find any evidence of replacement. For the most part the surface of contact of ores and overlying limestone is even, but locally the beds finger into one another. ORIGIN OF THE CLINTON IRON ORES. For a fuller discussion of the origin of the Clinton iron ores the reader is referred to the pubUcations of the geologists who have studied the Clinton ores of the Appalachians, especially to the recent work of Burchard in Alabama." The ores are not minetl on a large scale in the Lake Superior region and have not been studied in the same detail as those of the Algonkian and Archcan. However, a comparison with the ores of the Algonkian and Archean in the Lake Superior region iliscloses certain contrasts, wliich are probably significant of origin. The Clinton ores constitute beds uniform in lithology, with no evidence of local concentration or replacement or residual masses of unaltered material, and the adjacent beds are not altered or iron stained, as they are where secondary concentration has occurred. The hematite is therefore probably not the residual result of the alteration of preexisting rocks. On the other hand the granules and aggregates of granules making up the ore are distinctly weatherworn and he with their flat sides parallel to the strongly marked bedding and current bedding, pomting strongly to the deposition of the iron in essentially its present mineralogical condition in shalhnv waters. The Clinton ores therefore differ from the Lake Superior Algonkian and Archean ores in being deposited as ferric hydroxide under shallow-water or shore conditions rather than as some ferrous compounds in quiet water, as is characteristic of the pre-Cambrian iron tleposition. That the waters were marine is indicated by the character of the beds both above and below, carrying marine fossils, and also by the similarity of these ores to Clinton ores of the eastern United States, in which marine fossils are plentiful. It is also clear from the waterworn granules, current bedding, and oohtic structure that the waters were movuig, suggesting shore conditions. The discontinuity of the beds aiid their variation in thickness also suggest locally varying shore conditions. But many features of the history of the deposition of these ores are yet obscure. No satisfactory answer has yet been made to the question why these ores have developed at this particular horizon in the Paleozoic and not at other horizons. The final answer to this problem must involve the study of the Clinton ores of all of North America. SmiMARY STATEMENT OF THEORY OF ORIGIN OF THE LAKE SUPERIOR IRON ORES. The Lake Superior iron ores include the genetic types described in the follo-n-ing paragraphs. 1. Lake Superior sedimentary type: Iron brought to the surface by igneous rocks and con- tributed either directly by hot magmatic waters to the ocean or later brought by surface waters under weathering to the ocean or other bodj' of water, or by both; from the ocean deposited as a chemical sediment in ordinary succession of sedimentary rocks; later, under conditions of weath- ering, locally enriched to ore by percolating surface waters. To tins class belong most of the producing iron ores of the Lake Superior region, those of the Michijucoten district of Canada, and most of the nonproducing banded iron-bearing formation belts of Ontario and eastern Canada. 2. Magmatic segregation type: Ores brought to the outer part of the earth in molten magmas but were retained in them during crystallization, with the result that the ores form part of the rock itself, just as do the feldspar and other minerals. Such are the titaniferous magnetites, which contain refractory silicates and in places sid])hur and ])hos]diorus in dele- terious quantities. Although these ores are known in enormous quantities in the Duluth gabbro of northern Minnesota they are not mined. o Burchard, E. i'., The Clinton or red ores of the Binningham district, .Mabaina: Bull. U. S. Gcol. Surrey Xo. 315, 1907, pp. 130-151. THE IRON ORES. 569 3. Pegmatite tyjDe: Ores which are carried to or near the surface in magmas and are extruded from them in the manner of pegmatite dikes, after the remainder of the magma has been partly cooled and crystallized. They are deposited fi'om essentially aqueous solutions mixed in varying proportions with solutions of quartz and the silicates and have had no second concentration. To the pegmatite tyjje are doubtfully assigned the ores of the Atikokan dis- trict of Ontario, and possibly also certain magnetites of the Vermilion district. (See p. 562.) No detailed study of the Atikokan ores has been made. Ores of tliis type have been mined in small cjuantity in the Atikokan district. 4. Clinton setlimentary type: Sedimentary "flaxseed" ores deposited in shaQow waters, presumably from weathering of the land areas in wliich the iron is either disseminated in igneous rocks or has undergone some of the concentrations outlined in the three preceding paragraphs. They have suffered no essential second alteration. These are the ores in the vicinity of Iron Ridge, Wis. 5. Brown or hydrated ores, associated with Paleozoic and Pleistocene deposits: Residual or bog deposits in limestone as at Spring Valley, Wisconsin, or in glacial drift. Also abundant in ores of the Lake Superior sedimentary type. The associated substance is largely clay and they are therefore not susceptible of second concentration. Each of these classes of ores has counterparts in ores mined elsewhere in the country, except the Lake Superior sedimentary ores, the only ones which have undergone a second con- centration. From this class have been produced 99 per cent of the iron ores sliipped from the Lake Superior region and annually 80 per cent of the iron ores mined in the United States, a fact that indicates the great importance of a second concentration. All the ores have been derived ultimately from the interior of the earth, whence they were delivered by igneous eruptions to ])oints near or at the surface, there to undergo various dis- tributions and concentrations under the influence of meteoric waters acd gases. The varia- tions in composition, shape, and commercial availability of an ore have been controlled by variations of conditions untler which the ores have reached the surface and have been dis- tributed. The titaniferous magnetites rejiresent ores brought nearly to the surface but not aUowed to escape. The pegmatites rejiresent ores which have been crystallized in the act of escape. The pre-Cambrian sedimentary formations of the Lake Superior region were derived largely from basic rocks of not dissimilar composition that reached the rock surface, though usuaUy under water, in wluch case they crystallized as ellipsoidal basalts. The eruptions to which is due primarily the introduction of most of the known ores have come up along the zone of the present Lake Superior basin. The copper ores of Keweenaw Point and the silver ores of Silver Islet have been brought up by similar igneous rocks at a httle later date along the same zone. Along the strike of the Lake Superior zone during Kewee- nawan time igneous rocks also brought up the cobalt, nickel, and silver ores of Sudbury and Cobalt. The minerals and petrographic relations of the Keweenawan, cobalt, and nickel ores bear many similarities, suggesting possible differentiation from essentially the same magma. It is suggested that the entire Lake Superior and Lake Huron region is a great metallographic province from which the early extrusions brought up iron salts and the later extrusions were differentiated into the copper, silver, cobalt, and nickel ores. OTHER THEORIES OF THE ORIGIN OF THE LAKE SUPERIOR PRE-CA]\ffiRIAN IRON ORES. Whitney," Wadsworth,* Winchell,"^ HiUe/ and others have held the Lake Superior pre- Cambrian ores of the sedimentary type to be of igneous origm. Winchell's arguments are nearly aU based on the similarity of the textures of the iron-bearing formations to those of a Foster, J. W., and Whitney, J. D., Report on the geology and topography of the Lake Superior land district, pt. 2, The iron region: Sen- ate Docs., 32d Cong., special sess., 1851, vol. 3, No. 4, 406 pp. 6 Wadsworth, M. E., Proc. Boston Soc. Nat. Hist., vol. 20, 1881, pp. 470-479; Bull. Mus. Comp. Zool., Geol. ser., vol. 1, 1880, p. 75. <; Winchell, N. H., Structures of the Mesabi iron ore: Proc. Lake Superior Min. Inst., vol. 13, 190S, p. 203. d Hille, F., Genesis of the Animikie iron range, Ontario: Jour. Canadian Min. Inst., vol. 6, 1904, pp. 245-287. 570 GEOLOGY OF THE LAKE SUPERIOR REGION. igneous rocks. For instance, the concretions are compared with bombs, the spaces left by the leaching of silica are regarded as amygdahiidal cavities, tlie breccias are regarded as volcanic breccias, the bedding is regarded as flow structure, the slump of the ores in contact with wall rocks is regarded as the result of flow of lava over the bluff represented by the wall rock. In this view Winchell practically reaches a conclusion similar to that of Wadsworth, who believed that theores and jaspers are cliiefl\' eruptive and described the jasper and ore as intruded into the country rocks in wedge-shaped masses, sheets, and dikes. These resemblances between iron ores and igneous rocks are so superficial that they would scarcely be taken seriously by most ol)servers, and conclusions as to igneous origin ignore so many fundamental facts of composition, texture, and structural relations described in these reports that it is not believed necessary to attempt to refute them. In earlier reports Winchell " presents a different view of the origin of the ores, as follows: A chain of active volcanoes, having explosive emissions, extended across northeastern Minnesota about where the Mesabi iron range is found. This was near the shore line of the Taconic ocean, and was accompanied by land- locked bays and perhaps by fresh-water lakes. Such marginal volcanoes had a chemical effect on the oceanic water, causing the precipitation of silica and probably of iron. Its basic lavas and obsidians were attacked by the hot waters and were converted by encroaching silica into jaspilite. Near the shore such glassy lavas were eroded by wave action and distributed so as to form conglomerates and sandstones. Such action would have distributed lavas wholly silici- fied as well as those which were yet glassy, and the detritus of both would necessarily mingle with detritus from the Archean. Such lavas would exhibit great contortion and in places great brecciation, the same as later lavas, and these breccias must have been mingled sometimes with the products of detrital action. After prolonged activity of the volcanoes most of the deposits and of the lavas which were submarine would be permeated by secondary silica, but carbonate of iron would permeate the mass where carbonic acid had freer access, as in the lagoons into which streams drained from the land surface to the north. This view Winchell also applies to the Vermilion range. He argues that the iron of the iron-bearing formation was first deposited as a ferric oxide and that the ferruginous cherts making up the greater part of the formation to-day are origmal oceanic deposits laid down essentially in the present form. In volume 6 of tlie "Geology of Minnesota" he argued that the solutions formed from the igneous rocks acciunulated in the rocks to the point of saturation and that precipitation came later as a result of cooling. This discarded view of Winchell obviously has more points in common vnth the theory of origin outlined in the present monograph than his more recent views, although important differences are still to be noted. In a report on the Baraboo range Weidman * reached the conclusion that the iron ores of that district were originally precipitated in bogs and shallow waters as limonite and hematite associated ^nth slate, that they were then covered by the dolomite, tilted up, and ernded, and that the deposits to-day are essentially the same in lithology as they were when depositeil with the exception of certain minor vicissitudes in the way of dehydration, recrystallization, etc. The deposits might under this theory extend to indefinite depths — indeed, as far as anj' of the sedimentary formations of the district — and in this way the}' would contrast with the distribution of the ores determined primarily by a secondary concentration from the surface. In view of the evidence of secondary concentration found in other parts of the Lake Superior region the burden of proof must rest with one who attempts to exclude secondary concentra- tion of the Baraboo ores. Deep drilling in the Baraboo district has seemed to show a diminu- tion in thickness and grade of ore beds and a relative increase of iron carbonate with increase in depth, pointing to secondary concentration from the surface as the agency which has been largely responsible in developing the ore bodies. The dilVcrence in opinion as to tlie origin of Baraboo ores here indicated is really one primarilj' of emphasis. Weidman emphasizes the primary deposition in rich beds; we believe that the primary deposition, wliile a large factor in localizing ores, has been supplemented by considerable secondary concentration to develop the commercial ore deposits. o The geology of Minnesota, vol. 5, 1900, pp. 997-99S. 6 Weidman, Samuel, Bull. Wisconsin Geol. and Nat. Uist. Siu-vey No. 13, 1904, pp. 142-14(3. THE IRON ORES. 571 GENETIC CLASSIFICATION OF THE PRINCIPAL IRON ORES OF THE WORLD. Iron ores are known to have been developed by a great variety of igneous and metamor- phic processes. In almost any genetic classification of ore deposits iron ores will be repre- sented in each of tlie divisions, contrasting thereby with the less abundant precious metals. Moreover, it is likely that certain iron-ore deposits would fall outside of any such classifica- tion and others would require assignment to two or more of the divisions. The following classificatioa of the iron ores of the world has been constructed with the idea of showing the correlatives of the Lake Superior pre-Cambrian ores and the wide range of conditions under which the larger and better-known deposits have developed. 1. Macmatic segregations, usually in basic rocks. Titaniferous and silicated magnetites, weathering to limonites, epidotic and chloritic magnetites. On disintegration yielding mag- netic sands. Titaniferous magnetites of northeastern Minnesota and Adirondacks. Magnetite of Vysokaya Gora and Gorolilagadot of the Uralo, Russia. Silicated magnetites and specular hematites of pre-Cambrian of Kiirunavaara, Gellivare, etc., Sweden. Silicated magnetites of Kiirunavaara, Loussavaara, and TuoUavaara, Sweden. Titaniferous magnetites in Taberg, Sweden. 2. Igneous after-effects, usually from acidic rocks (pneumatolytic, pegmatitic, etc.), usually deposited ^^^thin or near parent igneous mass. Certain silicated magnetites of Vermilion and Atikokan districts of Adirondacks .and New Jersey, of Iron Mountain. Missouri, and of Iron Springs, Utah. Contact-silicated magnetites of Christiania, suggested by Backstrom and DeLaunay to be aqueous sediments contriljuted by associated jjorphyries. 3. Residual limonites resulting from weathering of igneous rocks. In this class are most of the laterite deposits resulting from the weathering of basic igneous rocks in tropical regions. The limonites of northeastern Cuba, constituting the weathered mantle of serpentine rock, are in enormous tonnage. 4. Sedimentary. A. Iron oxides, mainly syngenetic. Crystalline hematites of Minas Geraes, Brazil, the largest and richest known deposits of this type in the world. C'ambro-Silurian micaceous hematite and magnetite of Norway. Oolitic limonites, containing subordinate quantities of iron-silicate granules of various descriptions and iron carbonates, in Silurian Clinton rocks of Wisconsin and Appa- lachians and Newfoundland; in Jtirassic of Luxemlnirg, Lorraine, and elsewhere in Germany and in Cleveland district of England; in Tertiary of Louisiana, Texas, and Bavaria. Bog and lake limonites, sometimes in granules. In glacial lakes and bogs of Lake Superior j-egion. Small and nonproductive. Represented by Scandinavian lake ores, Finnish lake ores, lake and bog ores of eastern Canada, Massachusetts, and elsewhere. A 1. Iron oxides, developed mainly by secondary surface alterations of sedigenetic carbonates and silicates. Pre-Cambrian hematites and limonites of Lake Superior region. Paleozoic limonites of Spring Valley, Wisconsin. Brown ores of southern Appalachians, etc. A 2. Iron o.ridcs. resultinr/ from anamorphic alterations of sedimentary iron-bearing formations. Specular hematites and silicated magnetites derived from deep-seated anamorphism of oxides, especially of carbonates and silicates by deep burial, intrusion, or both. Marquette specular hematites. Hard lilue hematites of Vermilion. Silicated magnetites of Gunflint district of Minnesota, eastern Mesabi, western Gogebic, western Marquette, etc. B. Iron carbonates. Usually associated with coal or carbonaceous slates. Also various inter- mixtures of calcium and magnesium carijonates, with minor amounts of oxides and silicates. Huronian original iron carbonates of Gogebic, Marquette, Menominee, and other districts of Lake Superior region, altering at surface to limonites and hematites, and nt depth or by igneous intrusion to silicated magnetites and hemitites. Carlioniferous 1ilack-band ores of Pennsylvania, Ohio, and Kentucky, altering at surface to brown ores or pot ores in clay. 572 GEOLOGY OF THE LAKE SUPERIOR REGION. Tertiary black-band ores of Marj-land. Carboniferous lilack-hand ores of Germany. Carboniferous blatk-band ores of Wales and Scotland. Permian lilack-liand ores of district of Erzberg, in the northern Alps. C. Iron silicates. Greeiialite, glauconite, chamosite, thuringite, etc., with minor mixtures of iron oxides and carlionates. Hiironian original greenalite rocks of Mesabi district of Minnesota, derived largely from direct igneous contributions, as indicated under 2. .Mtering to hematites and limonites at sur- face and to silicated magnetites at depth or at igneous contacts. Lower Silurian chamosite ores of central Bohemia and chamosite and thuringite ores of Thuringerwald and vicinity, in Germany. D. Various combinations of above. It will be noted that the Lake Superior ores are represented in most of the princi})al classes here given. They also constitute an important subclass, the greenalite ores, developed by ac[ueo-igneous processes, not yet certainly idcntilicd elsewhere. Much the largest part of the world's production of iron ore has come in recent years from the sedimentary ores. The largest reserves are in that class. Also important for the future are the resiilual weathering ores of the laterite type, such as are found hi northeastern Cuba. The highest grades are reached in the sedimentary ores which, in addition to some 'purification by weathering in place in a parent rock, have been sorted and segregated during transporta- tion and deposition as sediments, and in the Lake Superior type, when again exposetl to the surface, have undergone further purification through katamorphjsm. These successive concen- trations have removed deleterious constituents, broken up complex silicates, and left the ores with a porous texture better adapted for furnace reduction than the ores of classes 1 and 2. The iron ores therefore illustrate both a wide range of ore-depositing agencies and the great increase of values effected by the reaction with meteoric waters and the atmosphere in the zone of katamorphism. One of the most striking features of the ore deposits of the sedimentaiy class is the preva- lence m them of granular textures, both oolitic and amorphous. The principal types of gran- ules are as follows : Green ferrous silicates: Greenalite, Fe(Mg)Si0^.nH20, amorphous. Glauconite, hydrous silicate or iron and potassium, amorphous, resembling earthy chlorite, in granules. Thuringite, 8Fe0.4(Al,Fe)203.6Si02.9H20, related to prochlorite, massive and fresh, oolitic when altered. Chamosite, SiOj 29 per cent, AljOj 13 per cent, FeaOj 6 per cent, FeO 42 per cent, H2O 10 per cent. Related to prochlorite. Oolitic. Oolites with concentric rings of quartz and some green silicate, of chloritic nature, undetermined. Found in Clinton and other ores. Hematite and limonite: Oolites consisting of concentric rings of silica and iron oxide. Amorphous granules .representing oxidation of scjme of the ferrous silicate granules mentioned above or replacing sliells. All the above granules lie in various cements of silica, iron oxide, and calcium carbonate. The correlation and origin of these various granular forms present an interesting field for monographic study. It is known that some are organic, as, for instance, tite glauconite and certain of the amorphous iron-oxide granules replacing shells. It is known further that proba- bly the larger part are inorganic, including the oolites and amorphous greenalite and iron oxide. As shown in another place (p. 525), both the greenalite ami iron-oxide granides form in ordinary chemical j)recipitates, and it is further suggested that they are perhaps related to Lehmaim's liquid crj'stals. It may be of interest to note that of the three common iron com- pounds, oxides, silicates, and carbonates, the two former appear in granules, while the last does not. The oxides and silicates have weak crystallizing power, which, according to Lehmann, is usually a.s.sociated with tlic (Ipvel()|)ment of granular or amorphous forms; the carbonates have strong crystallizing power, tending to give the surface definite and angidar outlines. CHAPTER XVIII. THE COPPER ORES OF THE LAKE SUPERIOR REGION. By the authors, assisted by Edward Steidtmann. THE COPPER DEPOSITS OF KEWEENAW POINT. GENERAL ACCOUNT. Although the authors have studied the copper of the Keweenawan series in many parts of tlie Lake Superior region and have visited the copper deposits frequently, they have made no systematic investigation of the ore deposits themselves. Since the publication of Irving's monograph" on the district by the United States Geological Sui'vey, the detailed mapping done by the Survey in this region has been confined to the iron deposits. It is nevertheless thought desirable to include in this monograph a general account of the copper deposits in order to summarize, as fully as possible, the present state of knowledge of the geology' of the Lake Superior region. The portion of tliis chapter dealing with the origin of the ores con- tains certain new features. The fallowing description of the ores is based jiartly on our own observations and largely on the published descriptions of Irving,"^ Rickard, '' Lane/' Graton/ and others. The copper-producing district of Keweenaw Point follows the axis of the point in a general northeasterly direction for 70 miles and has a width of 3 to 6 miles. The richest portion of the belt is the central portion, in Houghton County, adjacent to Portage Lake (see PI. XLIX), in association with the upper lava flows. The copper is metallic. With the exception of the comparatively small amount of coarse copper — "mass" and "barrel work" — sorted out at ^he mines, all the ores are subjected to crushing by steam stamps, followed by concentration. The principal gangue minerals of the copper of this district are calcite, quartz, prelmite, and laumontite, with smaller but still considerable quantities of analcite, apophyllite, natro- lite and other zeolites, orthoclase, datolite, epidote, chlorite (delessite), and native copper. Rarer associates are, according to Prof. A. E. Seaman, of the Michigan College of Mines,* ailularia, agate, anliydrite, algotlonite, azurite, aragonite, argentite, amethyst, annabergite, ampliibole, ankerite, barite, braunite, biotite, bornite, cerargyrite, chalcocite, chloanthite, clu^'socolla, chalcopyrite, clilorastrolite, cuprite, covellite, clinochlore ( ?), dolomite, domeykite, fluorite, gypsum, hematite, iddingsite, jasper, kaolinite, keweenawite, limonite, magnetite, martite, marcasite, malachite, melaconite, muscovite, mohawkite, niccolite, pyrite, pyrrhotite, phillipsite, powellite, saponite, selenite, stibiodomeykite, semiwhitneyite, serj)entine, silver, siderite, talc, whitneyite, thomsonite, wad, and wollastonite. Though this group of minerals a Irving, R, D., The copper-bearing rocks of Lake Superior: Mon. U. S. Geol. Survey, vol. 5, 1883. b Rickard, T. A., Tile copper mines of tlie Lake Superior region, New York, 1905. c Lane, A. C, Tlie geology of Keweenaw Point — a brief description: Proc. Lake Superior Min. Inst., vol. 12, 1907, pp. 81-104: The geology of copper deposition: Am. Geologist, vol. 34, 1904, pp. 297-309. ti Graton, L. C, Silver, copper, lead, and zinc in the Western States: Mineral Resources U. S. for 1907, pt. 1, U. S. Geol. Survey, 1908 (Michigan, pp. 496-523; Copper, pp. 571-644). ' Personal communication, 1910. 573 574 GEOLOGY OF THE LAIvE SUPERIOR REGION. is cliaracteristic of the deposits in general, they may vary in importance in the difTeront tyi)es as well as in tiie difTerent parts of the distrirt. Calcite is the most abundant associated mineral in the transverse vems and conglomerates; ejjidote is the most abundant in the dipping veins. The genetic sequence of these minerals is discussed uniler the origin of the ores. The copper constitutes (1 ) veins intersecting the northwestward dipping beds of the Keweenawan series described in Chapter XV and (2) beddeil deposits formed by infiltration or replacement of both the conglomerate and amygdaloidal beds of the Keweenawan series, chiefly in the beds below the "Great" conglomerate, which is the dividing line between the lower part of the Keweenawan, where traps predominate, and the upper part, where sediments predominate. (See fig. 75.) Copper deposits have not been found in felsitic beds and compact traps, except in minute quantities in the latter, where they are closely associated with amygda- loid beds. Rich cores of native copper are reported to have been drilled ,on the Indiana property, in 0nt6nagon County, from a verj^ dense felsite, which appears to be intrusive. Development, however, has not reached the productive stage. Only one bed above the "Great" conglomerate contains copper, and this is the Nonesuch shale, which carries a little disseminated copper throughout its extent and has been worked m the Porcupine Mountain district. _o*^o° Basicflows with . Level of Lake Superior ,o*,e»interbedded conglomerate '{< __!;gl,^,;^^!^^tar?— ,■ y j„y^^ yy/At a^^ Cambrian sandstone Copper-bearing lodes KEWEENAWAN SERIES Figure 75.— Cross section of Keweenaw Point near Calumet, sliowing copper lodes in conglomerates and amygdaioids. The deposits earliest exploited were the veins transverse to the strike of the beds in the Eagle River area at the northeastern extremity of the district ; the next were the veins parallel to the strike, though not uniformly to the dip, in the Ontonagon area at the southwest end of the district. The vein deposits, especially those in the Ontonagon district, are characterized by masses of copper, being in this respect distinguished from the amygdaloidal and conglomerate copper deposits, in which the copper is, as a rule, much more minutely disseminated. The amygdaloidal deposits were the. next to be opened, principally in the central portion of the district, but also in the Ontonagon area. The conglomerate deposits occurring only in a small area in the vicinity of Calumet, in the central portion of the district, were the last to be opened. (For summary of history see pp. 35-37.) In 1907 73.1 per cent of the ore mined came from amygdaloidal lodes and 26.9 per cent from conglomerate lodes, the vein deposits at present being practically nonproducing, although of the total production from the district approximately 3 per cent is sorted out at the mines as coarser mass material. The grade of the ores is low and is becoming lower. In the early days of mining much ore above 3 per cent was mined. In 1906 the average grade for the district was 1.26 per cent, and in 1907 it dropped to 1.1 per cent, and to 1.05 per cent in 1908. Onlj' four mines in 1908 worked ore yielding an average of 1 per cent or more in metallic copper. In 190S the richest iodes mined carried less than 2 per cent metallic copper, while the poorest yielded but little over 0.5 per cent. The grades and amounts mined from the principal mines in 1907 are as follows:" o Mineral Resources U. S. for 1907, pt. 1, V. S. Geol. Survey, 1908, p. 500. U. S. GEOLOGICAL SURVEY MONOGRAPH Lll PL. XLIX Veins See list below for explanation of numbers LIST OF VEINS Lake lode (amygdaloid) Nonesuch lode (conglomerate and sandstone) Arnold lode (ash bed amygdaloid) (Equivalent to No. 1 1?) Forest lode (amygdaloid) Branch lode (amygdaloid) Calico lode (amygdaloid) Evergreen lode (amygdaloid) Butler lode (amygdaloid) Knowlton lode (amygdaloid) Winona lode (amygdaloid) Atlantic lode (amygdaloid) Pewabic lode (amygdaloid) Allouez or Boston and Albany lode (conglomerate) Calumet and Hecia lode (conglomerate) Osceola lode (amygdaloid) Kearsarge lode (amygdaloid) Isle Royale lode (amygdaloid) Baltic lode (amygdaloid) R27W R26W MAP SHOWING LOCATION OF COPPER-BEARING LODES AND MINES ON KEWEENAW POINT. See page 573. THE COPPER ORES. Ore output and grade of the principal Michigan lodes in 1907. 575 Lode. Ore (tons). r.rade (per cent). Calumet .■ 2,400.000 1,900.000 2,350.000 1.250.000 750.000 1.835 1.06 .87 Baltic Kearsarge Pewabic a Osceola '. 895 Actual total and average 8.041,361 1,250,853 1 67 All other lodes .62 a Partly estimated. A little native silver occurs with the copper in some lodes. Averaojed on the total tonnage in 1908, the silver yield was 0.023 ounce to the ton. Native silver is present in all the deposits, but is particularly characteristic in the veins of the Eagle River and Ontonagon areas, where also mass copper is abundant. The amygdaloidal and conglomerate deposits have great extent along the strike, the Kearsarge lode, for example, being actively mined almost without break for a distance of 12 miles and other lodes being mined for 2 miles along their strike. They have been followed down the dip to a maximum distance of more than 1^ miles and a vertical depth of about a mile, making these mines among the deepest in the world, and are still found to be productive, although of somewhat lower grade. The depth to which mining may be carried is not yet known. That it should be possible to mine at a profit ores as low as 0.5 per cent at a depth of a mile is due to the remarkable uniformity and continuity of the deposits along both strike and dip. Shoots of richer ore pitching parallel to the strike of the beds — as, for mstance, the northward-pitchmg shoot of the Calumet and Hecla conglomerate — are known in a few places, but these are themselves so extensive that their existence and alternation with leaner portions of the beds have been ascertained only after years of extensive mining. TRANSVERSE VEINS OF EAGLE RIVER DISTRICT. The veins of the Eagle River district, in the northern part of Keweenaw Peninsula, cut vertically across the strike of the betls of sediments, traps, and amygdaloids. The veins are not commonly formed by the filling of a simple fissure, but by a large number of subparallel, anastomosing fissures with blocks of small rock inclosed between, forming rather a fracture zone. The productive zone is i-n the amygdaloid beds immediately below the Allouez con- glomerate and above the greenstone. The veins vary fi-om mere seams to those 20 or 30 feet wide, being widest where they cut across loose-textured amygdaloidal beds and not exceeding a width of 3 feet where they are in contact with compact traps. The greatest depth reached La the minmg of transverse veins is 1,600 feet, in the Cliff mine. The texture of the rock traversed by the veins also controls the ore content, the veins being rich where they cut porous amyg- daloidal layers and poor where they cut compact layers. Many of the amygdaloid beds them- selves are rich enough to be productive adjacent to transverse veins. The gangue materials associated with the copper of the Eagle River veins are mainly calcite, quartz, prehnite, and laumontite, but analcite, apophyllite and other zeolites, orthoclase, datolite, epidote, natrolite, and other minerals are found. Native silver is present. Veins containing only calcite are generally bare of copper. The copper is scattered through the gangue in thin films penetrating other minerals or in coarser fragments filling interstices between other minerals, or occurs in lenses, in this occurrence usually with a crystalline form. Mass copper also is found here, the masses ranging up to many tons in weight and many of them containing fragments of wall rock. Irving " believes that these veins are replacements along fissured zones rather than fillings of open fissures. As evidence he cites the gradation between vein and wall rock, the replacement of wall rock by copper masses, the occurrence of fragments of wall rock in the vein and in the 1 1rving, R. D., Mon. U. S. Geol. Survey, vol. 5, 1883, pp. 422-426. 576 GEOLOGY OF THE LAKE SUPERIOR REGION. copper masses, and the greater width of the veins adjacent to amydgaloidal beds than of those in contact with dense traps. The origin of tlie copper ores is discussed on pages 580 et seq. Transverse lissure veins are not restricted to the Eagle River district, but are present in nearly every mine on Keweenaw Peninsula. In the southern districts, however, these veins, as a rule, contain no copper, or at least not enougli to make them productive. Many of them are barren even where they cross productive beils of amygdaloids. No mines are now operating in the Eagle River district. Explorations have recently been conducted there with a view to further mining. The mines which have produced ore in this district are the ^tna. Empire, Delaware, Amygdaloid, Copper Falls, Central, Phoenix, and Cliff. DIPPINO VEINS OF ONTONAGON DISTRICT. The dipping veins of the Ontonagon district are noted chiefly for the great amount of mass copper that has been removetl from them. Tiie principal "mass " deposits are in tiie group of amygdaloids, traps, and conglomerates corresponding roughly to the strata between Portage Lake and the area covered by the u})per sediments. The veins are fillings of fractures following the strike of the beds. Many of those within weaker portions of the bed — for instance, along amj'gdaloidal layers — have a dip steeper than the bedding. Those that lie between two different beds are likel}' to dip at the same angle as the beds. The veins vary in width from a few inches to many feet. The veins between different beds are more lO'cely to be narrow; those cutting amygdaloidal beds may consist of a wide fracture zone, with fi-agments of rock interspersed with vein minerals. Slickensided walls locally bound the veins, but on the whole the contact is irregular. Irving'^ believed these veins, as well as the transverse veins of the Eagle River district, to be largely replacements of wall rock. Transverse veins (crossing the strike) are present also throughout the mines of the Ontonagon district, but they are unproductive except where they cross dipping veins. The chief vein materials associated with the copper are epidote and calcite, but the other minerals above named as generally associated with copper are present. The copper occurs in irregular hackly masses, some of wliich are many tons in weight. One mass found in the Minnesota mine in 1857 weighed 420 tons. The large proportion of mass copper originally mined in this district gradually decreased and the production of amygdaloidal copper increased. In 1908 the production was derived wholly from amj'gdaloid lodes. The principal producing mines are the Adventure, Mass, Michigan, and Victoria. Recent explorations have shown additional copper deposits. AMYGDALOID DEPOSITS. The copper deposits in am3'gdaloids are by far the most numerous and most productive in the Keweenaw Point region. The amygdaloids are the uj^jier, and in some places the lower, vesicular portions of the many lava flows, vnth here and there an interbedded detrital la^'er. The thickness of the productive portion of the amygdaloids varies from a few feet to 35 or 40 feet. The depth to which amygdaloid beds are productive has not been determined : the greatest depth yet reached is shown in the Quincy mine — 5,280 feet along the incline, or 4,008 feet vertically. The copper deposits in the amygdaloids, though lean in places, are much more continuous along the strike than those in the conglomerates, several mines miles apart working the same bed. There are very unusual variations in strike in the vicinity of the Baltic, Trimountain, and Champion mines. The dip of the amygdaloids flattens out below and also to the northeast along the strike. The Quincy lode has a dip of 55° at the surface and 37° at a depth of about 5,000 feet along the inchne. The Atlantic lode dips 54°, the Wolverine 40°, and the "Baltic" 70°, the dip thus showing a considerable variation even in a small area, tliough in general being stec])er in the southern part of the region. 11 Irving, R. D., Mon. U. S. Oeol. Survey, vol. 5, 1883, pp. 422^36. THE COPPEK ORES. 577 In amygdaloidal beds the copper occurs in cavities in amygdules partly filled by other minerals, alon"; cleavaounds in weiglit; in e])idote, quartz, and ])rehnite bodies it occurs as thread and flakelike impregnations; in fohaceous, lenticular chloritic bodies it forms flakes between cleavage planes and oblique jomts, or here and there — this is more particularly true of fissure veins — it replaces the chloritic selvage-like substance till it forms literally pseudomorphs, some of which are several hundred tons in weight. In the Baltic and adjacent mines are considerable quantities of black sulphides near the surface, but even here they are not in sufficient amount to have economic value. The amount of these sulphides decreases greatly with increasmg depth. The amygdaloids are productive only where broken. Usually they have both strike and dip fractures in addition to very irregular fracturing. Commonly the strike fractures are not exactly parallel to the beds but cut across them at acute angles. Many of these fractures show shckensiding, ]iroving considerable differential movements. At the Quincy and Baltic mines the amygdaloid is lean where there are cross fractures, but a little distance away from the cross fractures it is rich. In some places the copper goes down into the compact rock beneath the amygdaloid, following zones of Assuring, alteration, and replacement. In a number of places productive amygdaloid occurs below a heavy trap bed, as at the Winona, Quincy, Atlantic, Wolveiine, and Baltic mines. The mines operating in the amygdaloidal deposits and ]iroducing 60 per cent of the total output of the Keweenaw Point district in 1908 were the Calumet and Hecla, Tamarack, Osceola, Quincy, Centennial, Wolverine, Tecumseh, Franklin, Isle Roy ale, Atlantic, Baltic, Trimountain, Champion, Winona, Allouez, Ahmeek, Mohawk, Adventure, Mass, Michigan, and Victoria. The distribution of these mines and the lodes upon which they are operating are shown on Plate XLIX (p. 574). The Calumet and Hecla, Osceola, Ahmeek, Wolverine, and Mohawk are on the so-called Kearsarge lode, wliich has been developed for an extent of about 14 miles, the largest deposit in the district. The Wolverine has the richest deposit, running about 1.35 per cent of refined copper. The ore runs as low as 0.7 per cent in other mines. Below 0.7 per cent it has not been found profitable to mine. South of Portage Lake the only lode which has a large production is the Baltic. Its surface extent is about 4 miles and the yield averages about 1.1 per cent. COPPER IN CONGLOMERATES. Only two workable beds of conglomerate have been found among thirty or more beds distributed through the Keweenawan series untlerneath the "Great" conglomerate — the Allouez ("Boston and Albany") conglomerate and the Calumet and Hecla conglomerate — and even these are workable only in a small area. A number of other conglomerate beds have been found to contain small impregnations of copijer but not enough to be proiluctive. The Allouez conglomerate is being worked by the Franklm Junior mine and the Calumet and Hecla con- glomerate by the Calumet and Hecla and Tamarack mines. (See PI. XLIX.) The Calumet and Hecla conglomerate is the richest and largest copper lode in the district and ranks among the fii'st two or three lai-ge copper deposits of the world. It is famous as the principal source of copper of the Calumet and Hecla Company, which has been the greatest ttOp. cit.,pp. 421-422. 47.517°— VOL 5-2—11 37 578 GEOLOGY OF THE LAKE SUPERIOR REGION. ilividend payer in the history of mining. This conglomerate thins both to the north and south. At the North ITeclti nime it is not more than 8 feet wide in one place. It tliuis so ra])i(lly \o the soutli that on tlie Osceola property it has not been discovered. Thus the Calumet and Hecla conglomerate is essentially a lens. The Calumet and Ilecla conglomerate bed is prochictive only in the 2 rrules covered by the Calumet and ILecla and Tamarack pioperties. Nortli of this area the bed was mined by the Centennial mine without success, and to the south it was mined by the owners of the Osceola before they sunk down to the Osceola amygdaloid. The conglomerate dips 39° W. at the surface, flattening to 36° with depth. It is followed down the dip from the outcrop to a maximum depth of 8,100 feet by the Calumet and llecla Company, representing a vertical depth of 4,748 feet. A vertical shaft belonging to the same company about a mile from the outcrop on the hangmg-wall side passes through the lode at a depth of 3,287 feet and goes to a depth of 4,900 feet. One of the Tamarack shafts reaches a depth of 5,229 feet, being the deepest shaft in the world. The conglomerate lode has increased in tldckness from 13 feet at the surface to 20 feet in the deepest workings. The upper half (stratigraphically) is richer than the lower half of the bed. The copper content of the Calumet and Ilecla conglomerate was formerly about 4 per cent near the surface, and now a mile verticall}' below the siuface is 1 to H per cent and averages for the mine shipment 1.83 per cent. The copper is of lower grade and less regularly dis- tributed in the Tamarack part of the same bed. This decrease in grade of the ore worked with increase in depth is partly a real one and jiartly due to improvements anil lower costs, enabling lower-grade ores to be worked. The richer ores of the Calumet and Ilecla conglom- erate constitute a shoot pitching to the north and extending to the Centemiial ground. The upper half of the conglomerate bed is finer grauied than the lower half. It contains more interstratified sandstone layers, called sandstone bars, which are usually barren but in places are verj' rich. In some places they separate the conglomerate into two parts ami in such places the values may be either above or below the sandstone. The conglomerate is well cemented to both foot and hanging walls. Below the conglomerate are several amygdaloidal beds. Immediately over the conglom- erate is a trap, 300 or 400 feet tliick, which separates it from the fu-st amj'gdaloid. The cross- section maps of the formations, made from the drifts of the deep shafts intersecting the beds for thousands of feet above and below the conglomerate, divide the lavas into two classes, traps and amygdaloids. The traps form the greater part of the sections and man}- of them are hundreds of feet in thickness. The amygdaloids compose a much smaller portion of the sections and are usually thin. The copper values are very small in the trap and amygdaloids, both above and below the conglomerate. The rich portions of. the conglomerate are usuall.y light colored; the poor portions are dark. Tliis is a practical distinction by mining men, who speak of the lean conglomerate as "black" and mean by tliis that wherever it is m tliis condition the values are low or lacking. Tliis difference in color is due to the fact that the alterations, a jiart of wliich resulted in the deposition of the copper, have bleached the conglomerate. In many jilaces in the rich conglomerate aureoles of hghter-colored material ma}^ be seen at the outer parts of pebbles and bowldei-s. The Allouez conglomerate, worked by the Franklin Junior mine, varies between 8 and 25 feet in thickness, with 3 to 4 feet of sandstone at the base. It is of lower grade than the Calumet and Hecla conglomerate, averagmg about 0.5 per cent in copper. The pebbles in the conglomerates are mainly porphyritic felsite with diabase and amvgda- loids m subordinate amounts. Locallv, as m the Calumet and Hecla conglomerate, granitic and quartz porphjny pebbles are abundant. The original cementing materials were siliceous and feldspatlvic particles, l)ut these have been replaced largely by secondary calcite and cpidote, with chlorite, and where the conglomeiates are proe great, on closer inspection for the most part disappear. That much of the copper was introduced as filling and replacement of wall rocks admits of no doubt. Several hypotheses are still open as to the source of the copper and the manner in which it was transferred and redeposited. PREVIOUS VIEWS OF NATURE OF COPPER-DEPOSITING SOLUTIONS AND SOURCE OF COPPER. Irving,'' Wadsworth,'' and nearly all other geologists who have studied the copper-bearing rocks believe that the source of the copper was in the basic igneous rocks, and that so far as it was derived from the sediments, its ultimate source was still the basic igneous roclcs, because tlie sediments came from those rocks. This belief is founded principally' on the uniform and close association of copper with the basic igneous rocks and the known existence of copper sulphides minutely disseminated through some of the coarser igneous rocks. The source of the copper was believed by Pumpelly '^ to be in tiie overlying sediments. Smyth « believed that the ores did not come from the adjacent wall rocks but from a deep- seated source, the nature of wliich does not appear from his report. The conditions and agents under which the copper has been supposed to have been taken from the adjacent rocks and concentrated have been variously mterpreted. Irvmg,/ Pumpelly, " <" Irving, R. D., The copper-bearing roelis of Lake Superior: Men. U. S. Geol. Survey, vol. 5, 1883, pp. 424-428. I> Idem, pp. 425-420. <■ Wadsworth, M. E., The origin and mode of occurrence of the Lake Superior copper deposits: Trans. .\ni. Inst. Mln. Eng., vol. 27, 1S98, pp. 694-090. See also Miiller, Albert, Verhandl. Naturf. (lesell. Basel, 1857, pp. 4U-4.'i.S; Hauermann, Hilary, (Jiiart. Jour. Geol. Soc., vol. 22, 1886, pp. 448-403; Wadsworth, M. E., Notes on the iron and copper districts of Lake Superior: Bull. Mus. Comp. Zool. Harvard Coll. Geol. scr., vol. 1, 1880, p. 126. d Pumpelly, Raphael, The paragenesis and derivation of copper and its as-sociates on Lake Superior: Am. Jour. Sci., 3d scr., vol. 2, 1871, pp. 188-198; 24.3-258; 347-355. 'Smyth, n. L., Theory of origin of the copper ores of the Lake Superior district: Science, new ser., vol. 3, 1S90, p. 251. / Irving, R. D., op. cit., pp. 419-420. f Pumpelly, Raphael, op. cit., pp. 353-355. THE COPPER ORES. 581 Wadsworth," Lane,'' and others have been inclined more or less strongly to the theory of concentration under the direct downward movement of meteoric waters. Pumpelly has also implied that concentration may have occurred when sediments were still below sea level. Lane "^ has suggested that the waters were salt waters of the type now found in the deep copper mines, and that they represent fossil sea waters or fossil desert waters, which in the tilting of the series have migrated downward. Van Hise'' has argued that while meteoric waters have done the work, it has been during their upward escape after a long underground course. Smj'th^ assigned the first concentration of the ores to ascending solutions from a deep-seated source not specified. OUTLINE OF HYPOTHESIS OF ORIGIN OF COPPER ORES PRESENTED IN THE FOLLOWING PAGES. The copper ores are characteristically associated with basic igneous rocks. The source of the copper-bearing solutions lies in these igneous rocks. The original copper-bearing solu- tions were hot. These solutions may be partly direct contributions of juvenile water from the magma, partly the result of the action of meteoric waters on crystallized hot rocks. ASSOCIATION OF ORES AND IGNEOUS ROCKS. From 60 to 70 per cent of the copper produced in this region comes from the amygdaloids. The veins of mass copper also are all in igneous rocks and these veins are richest where they lie parallel to or intersect amygdaloidal beds. The ore-bearing rocks are characteristically near thick rather than thin flows. Barren conglomerates are interbedded with productive flows. The only productive conglomerates, the Calumet and Hecla and the Allouez, are associated with thick flows. Especially is the overlying flow tlaick. Not only is the association of the ores and the igneous rocks cons])icuous in the producing district, but throughout the Keweenawan area of Lake Superior traces of copper are widely distributed in the igneous rocks. Copper is associated principally with basic igneous flows, but it is now reported in drilling in felsite, supposedly intrusive, at the Indiana mine. Copper sulphide is also reported by Wright ^ in association with intrusive gabbros and ophites of Mount Bohemia. ORB DEPOSITION LIMITED MAINLY TO MIDDLE KEWEENAWAN TIME. It is beheved that the original deposition of the copper was limited mainly to middle Keweenawan time, or, if not, at least to the cooling period of the igneous rocks of that time. As shown below, the wall-rock alterations associated with the ores seem to be characteristic of hot water. Some of the gangue minerals are hot-water deposits. Bowlders of some barren conglomerate beds show mineralization wliich was developed before they were broken from the parent underlying ledge. The deposition of the copper was an episode in the work of cementation of both sedimentary and igneous rocks, which certainly began as soon as the beds were deposited but which continued to the end of the volcanic period of the middle Keweenawan and even longer. Pumpelly's work, mentioned below, shows that the copper was relatively late among the minerals introduced. The same thing is shown in some places by the absence of deformation effects upon the copper. The late introduction of the copper is argued by Smyth B from the contrast of minerals first deposited in the copper-bearing series with those coming later and carrymg the copper, the first, accortlmg to him, bemg developed under condi- tions of weathering before the series was folded, and the second being developed after the series was folded. n Wadsworth, M. E., The originand mode of occurrence of the Lake Superior copper deposits: Trans. Am. Inst. Min. Eng., vol. 27. 1S9S, p. 695. 6 Lane, A. C, The theory ot copper deposition; .\ni. Geologi-st, vol. 34, 1904, pp. 297-.'!09. (■Lane, A. C, The chemical evolution of the ocean: Jour. Geology, vol. 14, 1906, pp. 221-225. d Van Hise, C. R., .\ treatise on metamorphism: Mon. U. S. Geol. Survey, vol. 47, 1904, p. 11.3G. t Op. cit., p. 251. / Wright, F. E., The mtrusivc rocks of Mount Bohemia, Michigan: .\nn. Kept. Michigan Geol. Survey for 190S, 1909, pp. 301-.TO7. ffSmytii, H. L., Theory of the origin of the copper ores of the Lake Superior district: Science, new ser., vol. 3, 1896, p. 251. 582 GEOLOGY OF THE LAKE SUPERIOR REGION. Wadsworth " cites tlie extension of copper in a continuous mass from one flow to another as cvidonco of introduction "after tlie copppr-l)cann Adams, F. D., Jour. Geology, vol. 17, 190!), pp. 1-18. 597 598 GEOLOGY OF THE LAKE SUPERIOR REGION. is the Iluronian series. The Kewecnawan and Huronian series together make up the Algon- kian system. In certam districts the Iluronian is separable into three divisions, marked by uucotd'ormities; in other districts it is separable into two divisions, and in still other districts it is not yet divisible. The most serious questions therefore arise in the correlation of the Huronian formations of the several districts. In correlating the Huronian rocks the following principles are used: 1. Relations to series or groups of known age; that is, to recognizetl horizons. In using this criterion the relations of the Huronian to the Arcliean and to the Keweenawan are especially helpful, for these rocks are readity recognizable and afford datum planes from which to work up and down. The upper Huronian (Animikie group) adjacent to Lake Superior, being con- tinuous through so much of tlus region, is also very helpful as a recognizable datum plane. 2. Unconformities. The unconformities between the divisions of the Iluronian are of great assistance in correlation. Wliere all of the Huronian is pi'esent, separated by two uncon- formities, there is naturally no difficulty in separating it into lower, middle, and upper. T\niere, however, the Huronian has only one unconformity or where in a disconnected district only one division of the Huronian is present, the unconformities fail to be a determinmg factor. 3. Lithologic likeness of the formations. Tliis criterion is of assistance, but it clearly has severe Imiitations, because again and again the geologic conditions have been the same, pro- ducing like formations at different times. This is illustrated by the remarkable similarity of the iron-bearing formations of the upper Huronian, middle Huronian, and Archean. The natural behef that they were of the same age long acted as a bar to progress. 4. Like sequence of formations. Similar sets of formations in the same order are of much greater unportance as a criterion for correlation than the likeness of single foimations. But conditions producmg similar sets of formations have frequently recurred tluring geologic time. For instance, when a sea transgresses over a land area, there are normally formed in ortler a psephite, a psammite, a pelite, and a nonclastic formation. 5. Subaerial or subaqueous origm. Closely connected with the third and fourth criteria is the question whether the deposits were formed under air or under water. It is clear that the conditions of the formation of these two classes of deposits are so different, and therefore the nature of the formations wliicli maybe contemporaneous so variable, that there is difficulty m correlating the two. Also it is plain that the difficulties in correlating disconnected conti- nental deposits are scarcely less great. On the other hand, the correlatmg of subaqueous deposits with one another is relatively easy. 6. Relations with intrusive rocks. The older the series the more intricately it is likely to be cut by intrusive rocks, and this relation is of assistance in connection with the other criteria. However, as there have been igneous intrusives, both acidic and basic, in great quantities up to middle Keweenawan time, this criterion has relativeh' small utility in the correlation o^ the Lake Superior Huronian. 7. Deformation. The amount and nature of deformation are of assistance in correlation. On the whole the older the series the greater and more intricate the deformation. Thus in this respect the Archean rocks exceed all later series. The Keweenawan is much less deformed than the other pre-C'ambrian series. But the differences in deformation of the Huronian divi- sions may not be so marked in a single district as to give unportant assistance in the discrimi- nation of these divisions from one another. AJso a particular division of the Huronian may be much deformed in one district and not in another. 8. Degree of mctamorphism. The degree of mctamorphism is of some assistance in cor- relation. On the whole the older rocks are more metamorphoseii than the younger rocks, but this criterion has limitations, since witliin comparativeh' short distances the closeness of folding and the quantity of igneous intrusions may greatly vary, and tliese are very important factors in ])roducing metamorphism. The criterion relied on more than all others in the correlatioii of the Cambrian and post- Caml)rian formations — that of fossils, showing similaiity of the life on the earth at the time the equated formations were laid down — is not available for the pre-Cambrian rocks of the Camlation of ]irt-CDd Dlack Klvuraiou «„^».». Waterloo ana, WIkod- »1Q, ¥vx River valley. Utnbldliulct. QiinnirK Lake db- trkt. PlEton Point. Anlmlkleor Loon Lake dUirtct. Cnyuoa dUlrfel. VmnlUoodlilttei 's^^r^SS UlehlplcwtendltirtBi North gaw^arUta Uppw, Not MHilinnJ. tiul gtnulvM In upper Nol Idmlinw] Uoulilfullr pnuDi boubltully isoenl A twill OnaSiaOl. it). Abunt. AbKOt. AbMM. Ab«nt AbMnt. AbMnf, Ab«n>. Abral. Oabbro*. dlaba«j, »lc. EmtMrru* gnniu (In- mislvf). Dululli EBbbro. Unoootmrnlly Acldlfi and butc In- vStsr BIwjblk rormalloti (Iran beaiUic and Pokeg^^iarUlt*. ilisntinioiEPCnuille, Imnitlvc Into rocks Sutltomli (slalp. KluiDcroU-i! whJrh or IhF Kulrt Lakr slfllp and OgUhlic caiiKlunieniti' ol llie Vermilion dUttlot. C'aiiriL™ (Diii'iiFbj. *° Onbluo and "red Conglomerole, saad- Bailo and aeldic la- trutlva and oitru- Ulddlft. AbK'Dl. "tS-^u'^Tt.^ 1 a Lo-». 1 onelamaraloi. troup). and uiruilta. MlcUcuom.- .J. 10 To 0)* toulh pully npLuwl b> IhB vot nuranlu CUiki- 0«^<* qiurtill* rlnvoilonglalnulra and nlnirira iHvlnf mtmbor. VuiuantonMlloo UklUfunmtilaK. FakhK^lil IninuKai ond ot- MIcftlBuniiiB (-nan- V'ulrw) fontia Una .ni ti- dlvldsl loiu Ida J^bar'.Tii'iirijI.W m»nb«, and Tiad- Uncntonnlty (julnueicT viii>i, ;,',ria!.m,'r- MlflilncnniBilBla. Ill- du.lliwinwulWioI dOUllthll BfT. utd IIIB r.t.d OIlIW Mlu'lilcuniiig alalB. Inrludlnx \nleui Iron-bwrlniniBin- Uncnitone Inlnj. Htm and oxtru- Tjl^ilate. North Moond con- ElomcialB and Arpln roDflODionto andqua^ulte. Uarahaliumconclam- Uualhan conttotnef- alo. - — rnwrnfonnlly^ niironUin <|iiarltlli4. Paalbly U.H„dloi( KeAOHiiiwoniiiiniU jlbly (Ubdlvldwl Huronlan qunrliltm. Buronlan qUBrti- KB.™-'- <)iiulill>< (uprar Oninlli-, Inlnulvo Fnwloni dolomlto, maliilv ituroRilIu, iDcliiJlne lion- liMrlnsiiuinibiirla Its lowr liorlion, Swloy slnie. DantlMu i]iiar1>1lu. IVAterlw ijuartilu; piadblymlddlallii- (Iran bautni). lultrbnldcdquani- Itnamltlates. Black Ilale Vlrtlnl«("BvU)uli-l 9law.lacludlncl>»r- aunntui ransailds (Iron bearlog. but nan prod uctlvB). AbBDl. \bwll. SodlMiU uat Sod- UMdlalluroDlui, NaiBuno' totmiiJon lion) BJuaa sJsts. AJIblk qunritlt.. NfcaiiDH a) (utma- llaii(lmnl>wrtn(j. AJlblk quuldlf ivokMilo) •lllilrun- Wrtu (falo nivrn- bm aw lop, RaDdvfltv daloiiillA — I'mmnfonnliyr — lUndvlIlF ilolomlK. einrunn quaruJi*. ^lotonlonnliy^ Abwnt. — Dnmnronullyt- RandrtUfdolamll* SluiKOoa quaniil* ^Itneonlofmlljr^ AUanl. BEurtnfi quartili*. m til* «i|iilva1«nl 'If ooilla and fliur- Eton quutilU). .^u lAurooUui Hirln ilo- tniiltt lolo K*- wlllDj. OnnliF, •vMillr. tut- Idoill* FBimM snelM. dUbuaSlkai. (iranHa and psnl- a».,.-^. rsjs; '■■"'■ Aelillrfolcanlcrwin. probably l^uita- Granitej uid pocphy- UranllsindcnrUin. inlnulTo into K^ waltn. OrBnK.aDdp-1-s. "r^-^ne^KS: 1 hatVDiln ivliii »lil>l. tht JBllir bnaati ofiii In a tn n^JToar haiuU of bot^lonnailaD- anwn rUIil*. "assss/ffl" UrMutDoa and Goclo and kIiHW. fiitlidTwilito. SSd OrMnichliu.FMO- glnncHndnuuIied Umn Khtita. tnta- porpli>rl»* Soadan lormatlaD >li0D btarlne and J^SddSr'pStrt boilD Icneoos and iMjalyvolewIorwik OrwD KhEiu and Op. clt., 1S8.S, p. i:u F. cDaly, R. A., The mechanics of igneous intrusion: .\ni. Jour. Sci., 4th ser., vol. 26, 1908, p. 30. . GENERAL GEOLOGY. 601 The Laurentian of the Lake Superior region as a whole is characterized by both massive and schistose phases. It is perhaps surprising that so large a proportion should be massive. I It is topographically rough in detail, the massive parts usually standing somewhat higher than the schistose parts, but altogether it forms a part of the Archean peneplain. GENERAL STATEMENTS CONCERNING THE ARCHEAN SYSTEM. Both Laurentian and Keewatin rocks appear in each of the important districts that have been considered in the detailed chapters. Manifestly the wide and irregular distribution of the Archean is a natural consequence of the fact that these rocks constitute the basement com- plex upon which later formations were laid down. Whether or not they are now at the sur- face at any particular locahty depends on subsequent deposition, folding, and denudation — that is, it depends on whether geologic agencies have brought them to the surface. If, in the future, erosion should cut the Lake Superior region to a depth of several thousand feet below the present surface, it would probably be seen that much the larger part of the area would be occupied by the Archean, and it is believed that the Archean everywhere underhes all later rocks. It appears from the foregoing characterization of the Keewatin and Laurentian that the Archean as a whole was a period of regional igneous activity. All succeeding series contain sedimentary rocks in large or dominant proportions; they are treated essentially as sedimen- tary series and the igneous rocks are considered with reference to the sediments. In the Archean, on the other hand, the igneous rocks, which make up more than 90 per cent and probably more than 95 per cent of the area, are primarily considered, and the subordinate masses of sedimentary rocks are discussed in reference to the igneous rocks. The igneous activity of Archean time was both plutonic and volcanic on a tremendous scale. Probably at present the plutonic igneous rocks of the Archean occupy a much larger area at the surface than the volcanic rocks, but this is doubtless due in large measure to the very profound erosion which has taken place since Archean time, and which has in consider- able measure removed the volcanic rocks and exposed the plutonic rocks. . A very characteristic feature of the Archean of the Lake Superior region is its likeness from one district to another, and this is so whether the lithologic types of rocks or their relations are considered. The foregoing description of the intrusive relations between the Laurentian and Keewatin is applicable with scarcely a change to each of the several districts. If a set of specimens from the Laurentian or Keewatin south of Lake Superior were unlabeled, they could not be disci'iminated from a set of specimens from the Archean northwest or east of the lake. There are, of course, some exceptional types of rocks which occur only locally, but these are extremely subordinate in their mass. This extraordinary paralleUsm of phenomena of the Archean of one part of the Lake Superior region with that of another part — and, for that matter, with the Ai'chean of other parts of the world — has led to the phrase that the Archean is "homo- geneous in its heterogeneity" — that is, while it is heterogeneous for any one district, it shows the same kind of heterogeneity in each of the other districts. Topographically also the Archean is a unit. Though rough in detail it is a great peneplain, the UTCgularities of wliich do not constitute regular lineaments, and it is thus m contrast to the Algonkian rocks, part of which usually stand above the peneplain surfaces with conspicuous linear features. Whether or not it is generally accepted that the Archean, as the term is here used, can be safely correlated with similar rocks of other geologic provinces, it can hardly be doubted that the Archean rocks of the different districts of the Lake Superior region form parts of a single great system. This conclusion is supported by substantially all the criteria in reference to cor- relation given on pages 597-599. The system wherever it occurs is in a basal position. It rests unconformably below all the series with which it comes into contact. The general lithologic hkeness of the heterogeneous mass is remarkable. The Keewatin rocks are largely submarine. The complexity of intrusives is greater than that m any other series. The deformation is 602 GEOLOGY OF THE LAKE SUPERIOR REGION. greater than in other pre-Camhrian series. The metamorphism is profound. Similarity of sequence of formations in difTerent areas of the Kcewatin is lacking, but in place of this are the prevalent intrusive relations which exist between the Kecwatin and Laurcntian. It is of interest to note that the oldest recognized Archean rocks are basalts, with tex- tures indicating both subaqueous and subaerial extrusion. The basement upon which they rest has not been identified. It is natural to turn to the Laurentian granites and gneisses, but wherever these are found in contact with the Kecwatin they are intrusive into it. "VMietlier some parts of the Laurentian represent the original basement or whether the Laurentian as a whole has formed the basement and has been subsequently fused, there is no evidence to show. GENERAL STATEMENTS CONCERNING THE ALGONKIAN SYSTEM. CHARACTER AND SUBDIVISIONS. The Algonkian system on the whole contrasts with the Archean in being dominantly sedi- mentary rather than dominantly igneous, in being less metamorphosed, in having distinctly recognizable stratigraphic sequence, and in topography. The sediments are largely water assorted and deposited but in part are probably subaerial. The iron-bearing formations are regarded as having an exceptional character, being derived partly from submarine volcanic rocks either in magmatic solutions or by the reaction of hot volcanic material with sea water, or both. (See p. 516.) The Algonkian system comprises in its fullest development in the Lake Superior region four unconformable divisions — lower Huronian, middle Huronian, upper Huronian, and Keweenawan. The Keweenawan series is essentially a unit geographically and lithologically and is considered as such in the following discussion. The Huronian series, especially the lower and middle Huronian, presents such variation in lithology and succession as to require its consideration under two main geographic subprovinces — (1) the northern subprovince, including the north shore of Lake Superior and westward extension into Minnesota, and (2) the southern subprov- ince, including the Gogebic and Marquette districts of the south shore of Lake Superior and the continuation of this belt eastward to the north shore of Lake Huron, and the Menommee, Crystal Falls, and Iron River districts of Michigan. NORTHERN HURONIAN SUBPROVINCE. LOWER-MIDDLE HTJRONIAN. LITHOLOGY AND SUCCESSION. The Huronian rocks unconformably above the Archean and unconformably below the upper Huronian (Animikie group) of the north shore of Lake Superior are extensive and thick. The unconformities above and below are great. At many places the comparatively flat-lying Animikie group may be seen resting upon the steeply inclined or vertical truncated edges of the middle or lower Huronian. The latter rocks consist mainly of conglomerates, graywackes, slates, and mica schists. In some places it is possible to divide them into two formations, the lower consisting dominantly of conglomerates and the upper dominantly of graywackes and slates and their metamorphosed equivalents. The most characteristic and widespread of these rocks are the conglomerates of the lower formation. Those which lie near the subjacent rocks from which they are derived are commonly coarse bowlder conglomerates. Their fragments vary in lithology, depending on the under- lying formation. They may be dominantly from granite, from greenstone, or from gneiss, or mixtures of these three in viirious proportions and also with other materials. Many of the conglomerates at .higher horizons have a fine-grained matrLx. Some of them have a slate matrix through which very nunu>rous isolated i)ol)l)los and bowlders are scattered in an irregular manner. These have bt'en called slate conglomerates. In certain localities the GENERAL GEOLOGY. G03 slate conglomerates are the only rocks found. Associated with the slate conglomerates in many places are beds of well-laminated slate and schist. As has been intimated, the upper formation consists commonly of pelites. The most extensive areas of pelite are those of the Vermilion, Rainy Lake, and Hunters Island districts.. At the west end of the Vermilion district, between the conglomerate (there called the Ogishke conglomerate) and the slate (known as the Knife Lake slate) is a thin iron-bearing formation (called the Agawa formation) which appears to grade toward the southwest into a calcareous slate. The latter is the only known representative of a limestone in the lower or middle Huro- nian of the northern subprovince. At this particular locality the succession is in certain respects similar to that of the middle Iluronian of the Marquette district, but by far the greater areas and masses of these, rocks in the northern subprovince exhibit no close analogy in succession or lithology with either the lower Iluronian or the middle Huronian of the south shore. IGNEOUS ROCKS. During the time of the deposition of the rocks under discussion there were very great outbreaks of igneous rocks, basic and acidic, plutonic and volcanic. Contemporaneous volcanic detritus is mingleil in varying proportions with ordinary sedimentary material, from a subor- dinate to a dominant amount, as at Kekekabic Lake. The contributions of volcanic material were so great as to make them quantitatively very important. Some of the larger of the plutonic masses are the intrusive granites in the Mesabi and Vermilion districts. The slates that have been intruded by great masses of granites and have been deformed have become pelite schists (mica schists). This phase is extensively illustrated in the Rainy Lake and Namakan Lake areas. The conglomerates under similar circumstances are metamorphosed to psephite schists or gneisses, as illustrated by the schistose conglomerates adjacent to the Snowbank granite in the Vermilion district. CONDITIONS OF DEPOSITION. Coleman" holds that the lower Huronian slate conglomerate at one locality in the Cobalt district of Ontario is a glacial till. He points out the likeness of the great masses of the slate conglomerate to modern glacial till and to the Dwyka glacial deposits of South Africa, and con- cludes that they are all till. However, even if the glacial origin of the conglomerate-bearing striated and grooved bowlders at Cobalt is accepted — geologists are not all agreed as to tliis — it does not follow that the Huronian conglomerate of the northern subprovince as a whole is of this origin, because, among other reason's, the Cobalt area is a long way east of the Lake Superior region. Wliether or not Coleman's conclusion as to origin applies to the lower-middle Huronian in tliis subprovince, it is regarded as likely that these rocks are essentially of terrestrial tleposi- tion because of their unassorted character, being made up principally of conglomerate and graywacke, lacking quartzite and limestone; because of the recurrence of conglomerates at many horizons through several hundred feet; because the extensive conglomerate beds, like the Ogishke, have a thickness and extent which are more easily explained by terrestrial than by subaqueous sedimentation, which, according to Barrell,'' is not likely to produce conglom- erates over 100 feet tlnck; and finally because the part of the lower-middle Huronian nearest the granite or greenstone of the Archoan is locally a recomposed rock, which has not been sorted CORRELATION. The criteria under which the formations under discussion are classed as middle or lower Huronian are the following: They rest upon the Archean and are below the Animikie group, or upper Huronian; they are separated from these rocks by unconformities; they are exten- sively cut by both basic and acidic igneous rocks; they are similar in their deformation and oColeman, A. P., The lower Huronian ice age: Jour. Geology, vol. 16, 1908. p. 154. 6 Barren. Joseph, Relative geological importance of continental, littoral, and marine sedimentation: Jour. Geology, vol. 14. 1900, pp. 433-446: also personal communication. GO-t GEOLOGY OF THE LAKE SUPERIOR REGION. degree of metamorphisiii. It thus appears that the assignment of the rocks under discussion to the general place of lower Iluronian and middle Iluronian is unquestioned. But as large portions of these rocks may be land formations, they can not be exactly correlated with the aqueous de])osits of the middle and lower Huronian to the south. The deposition of land sedi- ments may well have begun earlier than that of the aqueous deposits or it may have continued later. On earlier maps jjublished by the United States Geological Survey the rocks here named lower-middle Iluronian appear as lower Huronian. As earlier continental deposits are likely to be removed by later erosion, however, it is probable that part, probably the larger j)art, of these rocks are of middle-Huronian age. It has already been noted that in northeastern Minnesota there is a similarity in succession to the middle Huronian of the Marquette district. UPPER HTJBONIAN (ANIMIKIE GROUP). LITHOLOGY AND SUCCESSION. The upper Iluronian of the northern subprovince extends from a point some distance east of Nipigon Bay, on the north shore of Lake Superior, westward through Thunder Bay to the Mesabi cUstrict of Minnesota, thence southwest and south to the Cuyuna, Little Falls, ("arlton, Cloquet, and St. Louis River cUstricts of Minnesota. The belt extending from Nipigon Bay t^) the Mesabi district consists from the base up of the following rocks: L Conglomerate, quartz slate, and quartzite. These reach a tliickness of 200 feet on the Mesabi range. Farther east, in the vicinity of Gunflint Lake antl Thunder Bay, the tluckness becomes only a few inches or a few feet. 2. Iron-bearing formation, 700 to 1,000 feet thick in the Mesabi district and thinning somewhat toward the east and west. 3. Slate, best exposed in the Thunder Bay district. Thickness unknown, but large. Throughout the northern part of this belt the sediments are gently inclined to the south at angles ranging from 5° to 20° and locally even up to 4.5°, with pitches of gentle minor folds in the same direction. In general the upper Huronian is not schistose but has' suffered contact metamorphism where it is in contact with the Keweenawan gabbro and granite and other large intrusive masses. It rests unconformably against the older rocks to the north, the unconformity being marked by areal relations, differences in steepness of dip, amount of schistosity, kinds of metamorphism, relations to intrusive rocks, basal conglomerates, and topogra])hy. The unconformity is one of the most conspicuous in the Lake Superior region. The line of contact is easily recognized by casual field observation. That the essential continuity of the upper Huronian is obvious is indicated by the early use of the term Animikie not only for the upper Huronian rocks on Thunder Bay, but for those in the Mesabi chstrict. In the area southwest of the Mesabi district, in the St. Louis River and Cuyuna districts and the country to the west, the upper Iluronian consists principally of slate, carrying lenses of iron- bearing formation, with many intrusive and possibly extrusive rocks and certain rare quartz- ites, the horizon of which is not satisfactorily determined but wliich are probably basal to the division. The upper Huronian in this area contrasts markedly with that along the Mesabi range and farther east in being closely folded, in the abundance of its intrusive rocks, and in possession of cleavage, as well as in the differences in lithologic character just noted. It is suggested i])]). 214, 528, 611) that the structural differences may be related in some way to proximity to the axis of the Lake .Superior syncline, or that the Mesabi and eastward belt of the upper Huronian may represent a shore phase of deposition, while the upper Huronian of the Cuyuna area to the south may be an offshore phase. IGNEOUS ROCKS. Intrusive into the upper Huronian are the great Duluth gabbro of northern Minnesota, the basic siUs of the Gunflint and Animikie Bay liistricts (Logan sills), a few basic dikes and possibly sills in the Mesabi district, a granite mass on the east end of the Mesabi range, and GENERAL GEOLOGY. 605 more abundant basic and intrusive masses in the Cuyuna district. Most of the intrusives are of Keweenawan age. Contemporaneous volcanic rocks have not been recognized. Extrusive rocks rest on the Animikie in the Cuyuna district. It has been shown that many of the capping chabases of the Nipigon area may be extrusive. These are doubtless middle Keweenawan, but some of them may be late Animikie. CONDITIONS OF DEPOSITION. The upper lluronian is a unit for the region, hence the conditions of deposition are dis- cussed on pages 612-614, after the southern subprovince has been treated. CORRELATION. The correlation of the upper Huronian of the northern subprovince with that of the southern subprovince is discussed on page 610. SOUTHERN HURONIAN SUBPROVINCE. LOWER HXrnONIAN. LITHOLOGT AND SUCCESSION. The lower Huronian of the southern subprovince reaches its fullest development in the Marquette district, where it consists, from the base up, of the Mesnard quartzite, Kona dolomite, and Wewe slate. In the Gogebic district the lower Huronian includes similar quartzite and doloinite named respectively the Sunday quartzite and the Bad River limestone, but the slate overlying the limestone is absent. Although the north shore of Lake Huron does not fall within the area covered by this report, it is desirable to consider the position of the series there because that is the district to which the term Huronian was first applied. The lower Huronian of the north shore of Lake Huron includes a great clastic formation above wliich is a limestone. In most places the clastic forma- tion comprises a conglomerate at the base, above this a quartzite, and above this a slate. In other places the conglomerate is almost immediately overlain by the Umestone. The succession is very similar in its essential features to that of the lower Huronian of the Marquette district. The lower Huronian is represented in the Menominee, Iron River, and adjacent districts of Micliigan and Wisconsin. It consists of a quartzite (the Sturgeon quartzite) followed by a dolomite (the Randville dolomite); but in the Iron River district the quartzite and dolomite are interbedded and for them the new name Saunders formation has been introduced. The lower Huronian partakes of the major structure described for each of the districts. As a whole the folding is not as intense as in the Archean. Cleavage is usually lacking, jointing is abundant, and bedtling is easily discerned. The quartzite of the lower Huronian of tliis subprovince represents a cleanly assorted sand, now strongly indurated, more or less iron stained, and locally showing fracturing and rock flowage, but retaining its original bedding structure as a conspicuous feature. It therefore contrasts in many respects with the lower Huronian of the northern subprovince. The dolomite overlying the quartzite is very cherty and shows more evidence of deforn^ation than the quartzite. The weathering of this dolomite emphasizes the folded and brecciated chert layers and serves to make the formation easily identifiable. IGNEOUS ROCKS. In the areas which are certainly known to be lower Huronian, contemporaneous igneous activity was not important. This applies to all the districts south of Lake Superior, as well as to the area north of Lake Huron. In this respect the lower Huronian contrasts with the miildle and upper Huronian and to a more marked degree with the Archean. The contrasts between 606 GEOLOGY OF THE hAKE SUPERIOR REGION. the Archean and tlie li)\ver Huronian in tliis respect are contributory evidence of the uncon- onuity l)et\veeii tlie two. (See pp. 617-018.) The volcanic activity of Archeun time appar- ently liad died out comi)letely in this Huronian subprovince before tlie dej)osition of the rocks unquestionably belonging to the lower Huronian. Later intrusive rocks cut the lower Huronian in small dikes. The [)ost^Huronian or Keweenawan granites of the Florence district of Wisconsin doubtless also cut the lower Huronian, but exposed contacts are only those of the granite and upper Huronian. CONDITIONS OF DEPOSITION. It has appeared that the lower Huronian south of Lake Superior and on the north shore of Lake Huron comprises first a great clastic formation consisting from the base up of conglomerate, rpiartzite, and slate. Over this is a largely nonclastic formation now represented by a dolomite, and localy above tliis in the Marquette district is another clastic slate formation. The essential subaqueous origin of the lower Huronian is believed to be showTi by the cleanly assorted nature of the sediments, the ripple marlcs of a shore rather than a stream type, and extensive beds of hmestone. It remains to be proved that such thick and continuous Umestone formations may be produced as terrestrial formations. Finally the conglomerate at the base of the group contrasts stronglj^ with the arkose and thick conglomerate masses at the base of the middle-lower Huronian of the north shore, and is beUeved to be more character- istic of aqueous sedimentation. It therefore appears that at the beginning of lower Huronian time the conditions in the southern subprovince had become those of normal sedimentation in which the material destroyed by the epigene agents was sorted and hiid down in beds one upon another, the lithologic character varying from time to time. This is evidence that the erosive forces of air and water were working as at ])resent. Moreover, as emphasized by Chamberlin and Salisbur}'," it is evidence that the weathering processes possessed their full efficiency, and this would favor abundant vegetation.'' With the beginning of Huronian time at the latest commences the part of the history of the world to which Lyell's principles of uniformity'^ are apj)licable. These ancient Huronian rocks have no lithologic peculiarity which can discriminate them from the rocks of much later age; indeed, there is notliing to indicate that when they were laid down the con- ditions were in any respect different from those which prevail to-day, ^\-ith the sole negative point that fossils have not been found. CORRELATION. The Gogebic, Marquette, and original Huronian districts are approximately in an east-west line and the prevailing strikes of the lower Huronian in all but the Crystal Falls and Iron River districts are in the same general direction, favoring the correlation of the rocks of the different districts. In each district the lower Huronian rests with profound unconformity upon the underlying Archean or basement complex, the unconformity being marked where ex])osed by differences in lithology, by metamoqahism, and by the presence of a basal conglomerate, and being shown also by the areal relations and relations to intrusive rocks. The lower Huronian is overlain unconformably by the upper Huronian (Animikie group) in all the districts, and by the middle anil upper Huronian in the Marquette, Crystal Falls, original Huronian, and Menominee districts. The lower Huronian of the southern sub])rovince has no counteipart in the northern sub- province, though it occu]Hes the same general position in the succession as the lower-middle Huronian of the northern subprovince. a Chamberlin, T. C, and Salisbury, R. D., Geology, vol. 2, 1900, pp. KS-ltB. b Van Hise. C. R.. A treatise on metamorphism: Mon. U. S. Geo!. Survey, vol. 47, 1904, p. 477. <: Lycll, Charles, Principles of geology, vol. 1. 10th ed., lSii7, pp. 305-326. GENERAL GEOLOGY. 607 MIDDLE HURONIAN. LITHOLOGY AND SUCCESSION. The middle Huronian is represented in the ilarquette, original Huronian, Crystal Falls, and ilenominee districts. In the Marquette district, where it was iirst discriminated and is best developed, it consists from the base up of the Ajibik quartzite, Siamo slate, and iron-bearing Xegaunee formation (nonclastic). On the north shore of Lake Huron the broader features of the middle Huronian are analogous with those of the Marquette district — that is to say, the rocks comprise a clastic formation below, consisting of a conglomerate at the base and over this a quartzite, both so thick and extensive that they have been mapped separately, antl above these clastic formations a clierty limestone. In the Crystal Falls district the middle Huronian is represented principally by the volcanic Hemlock formation, containing iron-bearing slate near the top. The iron-bearing Xegaunee formation is doubtfully, present; the Ajibik quartzite is present near the northeast corner of the district, near the Marquette district. Volcanism seems to have intervened between the deposition of the lower Huronian antl the u])per Huronian, making lithologic correlation diffi- cult. It is to be noted, however, that the Clarksburg volcanic rocks of the Marquette district began to be extruded in middle Huronian time, and tljese are therefore to be partly correlated with the Hemlock volcanic rocks of Crystal Falls. In the Menominee tlistrict the miildle Huronian is taken to be represented by cherty quartz- ite, heretofore not separated from the Randville dolomite of the lower Huronian. There is evidence also in the jasper and iron pebbles in tlie conglomerate at the base of the upper Huronian that an iron-bearing formation corresponding in position and character to the Negaunfie was present in the district before upper Huronian time, but no remnants of this are now known. IGNEOUS ROCKS. In the Marquette district the middle Huronian is associated with part of the Clarkslaurg formation of basic intrusive and extrusive rocks. In the original Huronian district igneous rocks are lacking in the middle Huronian. The presence of igneous rocks in the middle Huronian of the Marquette district and their absence in the middle Huronian of the original Huronian district may perhaps be correlated with the presence in the former, and the absence in the latter, of an iron-bearing formation. (See pp. 506-507.) Hemlock volcanic rocks form the princijjal part of the middle Huronian in the Crystal Falls district. In the IMenominee district volcanic rocks are absent from the division. The Keweenawan ( ?) granites of Florence County doubtless also cut the midtlle Huronian, though they nowhere come into contact with it at the surface. CONDITIONS OF DEPOSITION. The extensive formations of cleanly assorted, well-rounded, ripple-marked sands, now quartzites, of the middle Huronian, both south of Lake Superior and north of Lake Huron, point toward subaqueous deposition. The pure nonclastic iron-bearing formation south of Lake Superior and the cherty limestone formation north of Lake Superior point in the same direction. Still further is this shown by the association of these rocks with partly subaqueous volcanic rocks of the Clarksburg formation. The iron-bearing formation and possibly some of the associated slates have a close genetic connection with some of the associated volcanic rocks. In the Crystal Falls district the middle Huronian was principally a time of extrusive volcanism, partly subaqueous. The volcanic rocks are interbedded with the slates and iron- bearing rocks, subaqueously deposited. In the Menominee district the middle Huronian is represented only by shreds of quartzite and perhaps by the iron-bearing Negaunee formation. 608 GEOLOGY OF THE LAKE SLTERIOR REGION. The quartzite is very cherty, as if derived from decomposition of the Randville dolomite, against which it rests. It is well betlded and well assorted. At one locality there seems to l)e a conglomerate with well-rovmdcil bowlders near its base. On the whole the evidence favors subaqueous deposition of the middle Huronian. CORRELATION. The middle Huronian rocks in the Marquette and original Huronian districts are correlated on the basis of similar succession of clastic and nonclastic rocks, similar relations to the lower Huronian, similar east-west trend, similar metamori^hism, and the fact that they are subaqueous in both districts. They diiVer in that the nonclastic formation of the Marfiuette di.strict is an iron-bearing formation and that of the original Huronian district a limestone, that associated igneous rocks are present in the Marcjuette district and not in the original Huronian district, and that in the Marquette district the overlying rocks are upper Huronian and in the original Huronian district no upper Huronian is present, although to the northeast in the Sudbury basin rocks probably to be correlated with the mitldle Huronian are overlain unconformably by rocks with upper Huronian characteristics. The middle Huronian of the Crystal Falls district, being largejy volcanic, may be correlated lithologically with the lower part of the Clarksburg formation of the Marquette district. So far as the Ajibik and Negaunee formations are present in this district they are correlated directly with formations of the same names in the Marquette district. They occur, however, in the northeast corner of the Crystal Falls district, the area nearest to the Marquette •district, and the correlation is of little aid in correlating the middle Huronian as a whole. The middle Huronian of the Crystal Falls .district is principally a great assemblage of volcanic rocks Ij'ing between the lower Huronian and upper Huronian and differing from the dominantly sedi- mentar}^ middle Huronian of other districts. Its correlation is therefore based principally on its position in the geologic column. The middle Huronian of the Menominee district is correlated with the middle Huronian of other areas almost entirely on the basis of its stratigraphic position, unconformably above the lower Huronian and unconformably below the upper Huronian. As it consists only of a rem- nant of quartzite, lithologic comparison with the middle Huronian of other districts is of no value. The equivalents of the middle Huronian have not been identified in the other districts of the Lake Superior region, though it is possible that future work may result in its identification in the Florence and Iron River districts. trPPER HURONIAN (ANIMIKIE GROUP). LITHOLOGY AND SUCCESSION. The upper Huronian of the southern subprovince consist mainly of a thick slate foimation carrying two or more iron-bearing beds or lenses near its base ami possibly othei-s higher in the group. In the Gogebic district it consists from the base up of the Palms formation, the iron-bearing Ironwood formation, and the Tyler slate. In the Marquette district it consists from the base up of the Goodrich quartzite, the iron- bearing Bijiki schist, and the Mchigamme slate. In the Menominee district the lower iron-bearing part of the upper Huronian is called the Vulcan formation and the upper slate the Michigamme ("Hanbury") slate. The Vulcan formation is subdivided, from the base up, into the Tradere iron-bearing member, the Brier slate member, and tlie Curry iron-bearing member. In the Crystal Falls district a similar subdivision into Vulcan and ^lichigamme is made, but there not only are the members of the Vulcan formation not discriminated, but the forma- tion is iuterbcdded near the base of the slate and is treated as a member of the .Micliigamme GENERAL GEOLOGY. 609 and not as a distinct formation, although it is mapped separately. On former maps of the Crys- tal Falls district" the iron-bearing rocks were not given a separate name, but were mapped with the slate as upper Huronian. In this report they are correlated with the Vulcan formation and called the Vulcan iron-bearing member. In the Calumet district the upper Huronian is divided into the Michigamme slate, the Vulcan formation, and a third formation at the base, the Felch schist. The Vulcan formation is subdivided into three iron-bearing beds and two slate beds. In the Felch Mountam district the slate is absent except where the district opens out to the west; the Vulcan formation is not subdivided and the Felch schist forms the base of the upper Huronian. The Vulcan and Felch formations of this district correspond respectively- with the "Groveland" and "Mansfield" formations of the earlier mapping of the district. The reasons for the change of names are given on ])agcs 303-305. In the Iron River district the upper Huronian is represented by the Michigamme slate, inter- bedded near the base of which is an iron-bearmg member that has been correlated with the Vulcan formation, although the evidence is not conclusive that certain iron-formation bands classed as Vulcan may not belong stratigraphically higher than the Vulcan formation as typically developed in the Menominee district. The same remarks may be made concerning the Florence district in Wisconsin. Throughout the southern subprovince the Michigamme slate is closely folded and in much of the area, especially in the vicinity of the intrusive rocks it has a strongly developed cleavage. Bedding is usually to be observed except in places where there has been exceptionally good development of cleavage. The iron-bearing formations and quartzites also have been closely folded, but lack cleavage. IGNEOUS ROCKS. Basic intrusive and extrusive rocks in the upper Huronian are represented in this subprov- ince by the Clarksbuj'g formation of the Marquette district; by the Prescjue Lsle area of the Penokee-Gogebic district, where volcanic rocks, lavas, and tuffs were built up during the larger part of uppei' Huronian time, and by basaltic schists of the Menominee^ Crystal Falls, Iron River, and Florence districts. In individual occurrences it has not been found possible to determine whether these basic igneous rocks are intrusive or extrusive or even to exclude the possibility of the rocks being pre-Huronian. Some of the intrusive rocks are probably of Keweenawan age. Granites of probable Keweenawan age intrude the upper Huronian and associated basaltic extrusives in the Florence district. CONDITIONS OF DEPOSITION. The conditions of dei^osition of the upper Huronian ui this subprovince are discussed on pages 612-614. CORRELATION. There can be little doubt about the correlation of the upper Huronian in the several districts of the southern subprovince. The rocks as a whole are easily eroded and heavily drift covered and therefore have few outcrops, with the residt that areal connections have not been every- where traced, although they probably exist. The upper Huronian of the Marquette district opens on the west and southwest into a gi-eat slate area, which, so far as Icnown, is the same slate area as that surroundmg the Crystal Falls district, and thence extends south and south- west into the Menominee and Iron River districts. Tlu-oughout the subprovince the greater part of the upper Huronian is slate and the u'on-bearing formation is characteristically near the base of the gi'oup. In metamorphism, folding, amount of intrusive rocks, and relations to intrusive rocks the upper Huronian within the province is a unit. a Mon. U. S. Geol. Survey, vol. 36, 1899. 47517°— VOL 52—11 39 610 GEOLOCIY OF THE LAKE SUPERIOR KEGTOX. From a study of the structural facts alone it may not be aflirincd tliat the unconformity at the base of the upper Iluronian of the southern subprovince represents a considerable time interval. However, when this unconformity is considered in connection with the deep erosion and local absence of the middle Iluronian between two divisions, which are identified on satis- factory evidence, as upper Iluronian and lower Iluronian, it is evident that the time break represented may be a large one. Great time intervals are known to be represented in other parts of the geologic column, as, for mstauce, between the Paleozoic and Mesozoic in parts of the West, where structural evidence is slight. The correlation of the upi)er Huronian of the southern and northern subprovinces is scarcely less clear. In each subprovmce the basal member is quartzite and slate, followed by an iron- bearing formation and then by thick slate. The differential metamorphism Ls similar m the two subprovmces. In both the upper Huronian rests with strong unconformity upon Archean or middle or lower Iluronian. In both it is unconfonnably beneath the Keweenawan. On the north shore it dips gently to the south under the Lake Superior syncLine ; in the northern part of the southern subprovince the upper Huronian of the Gogebic district dips steeply to the north under the same syncline. The identity in the succession of formations in these two subprovinces, their position immediately below the Keweenawan, and their general structural alliances with that series give such strong evidence of equivalence that no one can seriously doubt that the upper Iluronian of the two regions is essentially contemporaneous. If one saw the flat-lymg, little-altered upper Huronian at one locahty and the most folded and metamor])hosed phases at another far distant and had not proved their continuity, he might think that the rocks of the different localities belonged to different divisions, but in many places the various metamorphosed and unmetamorphosed phases have been found to connect. GENERAL REMARKS CONCERNING THE UPPER HURONIAN (ANIMIKIE GROUP) OF THE LAKE SUPERIOR REGION. CHARACTER. The Animikie is the only group that is continuous throughout the Huronian subprovinces. It is the principal iron-bearing group. Although it has been described m connection with each of the subprovinces, a further general description is here presented to emphasize its unity over the Lake Superior region. The upper Huronian was deposited on a remarkably uniform peneplain. Remnants of this peneplain appear from beneath the upper Huronian hi the Mesabi, Anhiiikie, and Gogebic districts. The post-Animikie and post-Keweenawan folding have resulted in the tdting of this plam to various angles and it is truncated by later peneplains. In each of the districts in which a full succession is found there is a clastic formation at the bottom, a niiddle iron-bearing formation, and an upper slate formation. The bottom chxstic formation consists of a con- glomerate at the base, which m the northern subprovince and the northern part of the southern subprovince passes up into a shale or slate and in most places linally into a quartzite. In the different districts, and in the same district, the relative proportions of conglomerate, quartzite, and slate vary, as does also the particular phase which Ls adjacent to the iron-bearing formation. For instance, m the Marquette district conglomerate and quartzite are domuiant hi the Goodrich quartzite and there is comparatively little slate. In the Penokee-Gogebic district conglomerate and slate are dominant in the Palms formation and the quartzite is a thin formation at the top. In the Mesabi district the Pokegama quartzite is somewhat similar. In the Animikie, Menominee, and Crystal Falls districts the clastic formation is very tlihi hideed. Over the clastic formation is the iron-bearmg formation, which hi the Marquette district is known as the Bijiki schist, in the Menommee district as the ^■ulcan formation, in the Crystal Falls, Iron River, ami Florence districts as the Vulcan iron-bearing member, in the Gogebic district as the Ironwood formation, in the Mesabi district as the Biwabik formation, and in the Cuyuna district as the Deerwood iron-bearmg member. This u'on-bearuig formation is by far GENERAL GEOLOGY. 611 the most persistent and important oi' those of the Lake Superior region. In it are probaljly 9.5 per cent of the known ore reserves. It is not a pure nonchistic formation, but has interstratified slaty hi vers of variable thickness. A number of these layers have been recognized in the Mesabi and Gogebic districts. In the Menominee district one of them is of sufficient thickness to constitute a distmct member of the formation and is known as the Brier slate member; it sepa- rates the two ore-bearing momljers of the Vulcan, the Curry and Traders. The maximum tliickness of the u'on-bearing formation for the region is 1,000 feet. In parts of the region the iron-bearing formation does not lie at a definite hoi-izon l^etween the coarse clastic sediments at the base and the shales aljove, but appears as more or less isolated and overlapping lenses entirely withm the slate which forms the upper part of the upper Huronian. This is the characteristic occurrence of the iron-bearing formation of the upper Huronian m the great area extending south and west from the Mesabi and St. Louis River districts, including the new Cuyuna range, and of the triangular area of Michigan between the Marquette, Menominee, and Gogebic districts, mcluding the Florence, Iron River, and Crj^stal Falls districts. The iron-bearing formation in this relation to the slate appears also in the western part of the Marquette district. Iron-bearing lenses of this kind seem on the whole to be more numerous near the base of the slate than elsewhere, but in many places it is not known what their stratigraphic position really is, the rocks both above and below them being slate. It will be noted that the sharply delimited, extensive ii-on-bearmg formations, occurrmg at a defuiite horizon above the lower clastic formations of the upper Huronian, border the okler formations on the northwest and southeast sides of the Lake Superior syncline, and that the discontinuous lens-shaped parts of the formations in the slate are located far from the contacts with the older formations. The suggestion is made that this ilifl'erence may be due to original difference of conditions of deposition near the old shore against which the upper Huronian sea washed, as compared with the conditions off shore. Above or associated with tiie iron-l^earing formation is the upper slate formation known as the Michigamme slate in the Marquette, Crystal Falls, Calumet, Menominee, Iron River, and Florence districts, the Tyler slate in the Penokee-Gogebic district, the Virginia slate in the Mesabi, Cuyuna, and adjacent districts, and the Rove slate in the Vermilion district. It occupies a large area in Michigan south of Lake Superior, an immense area west of Lake Supe- rior extending far into central Minnesota, and a very large area about Thunder Bay and vicinity. It probably extends westward beyond the western boundary of Minnesota and widens out in this direction. It is entirely possible that this formation wiU ultimately be found to connect beneath the later formations with the slates of the Black Hills of South Dakota and even with the Belt series of Montana. Indeed, the areal extent of this formation is far greater than that of all the other Huronian sediments of the Lake Superior region. The formation being for the most part a slate and so soft as to be extensively covered by the drift, exposed sections in which to measure its thickness are rare. Also cleavage in these sections has so obscured bedding that estimates are worth little. In the Penokee-Gogebic district, where such a section is exposed, the possible maximum thickness has been estimated at about 12,000 feet, but this is probably too large. Seaman and Lane "^ suggest 4,000 feet. The rocks of this formation in the Mesabi and Animikie districts are principally shales. Elsewhere they are principally slates. At Carlton and Cloquet, on St. Louis River, tlie forma- tion is niuch folded and has a slaty cleavage, and farther to the southwest, where intruded liy masses of granite and diorite, it locally becomes so metamorphosed as to pass into a schist. A like change is noted in the character of the upper formation, the Tyler slate, at the west end of the Penokee district, where it is intruded by igneous rocks. Conspicuous in the slate at many horizons are seams and lenses of pyritiferous and gra- phitic slates. These are so characteristically associated with some of the discontinuous non- bearing lenses, originally iron carbonate, as to serve as guides in exploration. u Lane, A. C, and Seaman, A. E., ^ioles on the geological section of Michigan: Jour. Geology, vol. 15, 1907, p. 686. 612 GEOLOGY OF THE LAKE SUPERIOR REGION. The slate as a wliole gives evidence by its composition of being less leached of its bases than average slates or residual clays. The cojnposition also sliows that it must have been derived from rocks on an average more basic than granites. In figure 76, prepared Ijy S. H. Davis, the mineralogical composition of the upper Huronian slates, calculated from chemical composition, is compared graphically with that of a variety of other clays and soils. The upper Huronian slate and ii'on-bearing formations arc interbe(hled locally with abundant basaltic extrusive rocks, partly subaqueous, and tuffs in the southern subprovince. In the northern subprovince these are yet known definitely only in the Cuyuna district of Minnesota. QUARTZ CLAY AND FERRIC OXIDE SILICATES Figurl; 7ti. — Triangular diagram comparing the amounts of undecomposed silicates, quartz, and residual weathered products, siit-li as clay and ferric oxide, in dilTerent kinds of muds, shales, and weathered rocks. For description of method of platting see page 182. The mineral com- positions are calculated from chemical analyses. Dotted lines with arrows indicate the progressive change in proportions of constituents between the unaltered and altered rocks. The diagram brings out clearly the fact that the upper Uuroniau shale represents the little- decomposed ddbris of a basic igneous rock. CONDITIONS OF DEPOSITION OF THE UPPER HXTRONIAN (ANIMIKIE GROTTP). Any hypothesis of the conditions of deposition of the upper Huronian must be built around the following salient facts: The succession of a thin fragmental base, an iron-bearing formation, and a thick mud deposit, and the tiiinness, evenness, and wide extent of the basal conglomerate and quartzite. The fact that the upper Huronian rests upon a Hat plane beveling alike hard and soft, resistant and nom'esistant rocks, without residual or terrestrial deposits at the base. GENERAL GEOLOGY. 6L3 The association of discontinuous iron carbonate lenses with graphitic slates at different horizons, pointing strongly to bog or lagoon conditions. The lack of sorting or decomposition in the slates as shown by analyses. Contemporaneous volcanism, partly submarine, probably relatetl to the deposition of the ore, so associated with the upper Huronian as to indicate subaqueous origin for at least a part of it. The hypothesis which seems to fit this group of facts better than others which have suggested themselves to us is this: 1 . The first upper Huronian event was the advance of the upper Huronian sea to a shore line somewhere north of the present northern boundary of Lake Superior. In the area of Michigan and Wisconsin it passed over middle and lower Huronian rocks which were nearly flat-lying and perhaps not much eroded. On the north shore it passed over middle and lower Huronian rocks which had been closely folded and deeply eroded. This advance was perhaps accompanied by some planation or scouring of the land area, as suggested by the evenness of this plane and the manner in which it bevels alike soft and hard formations and by the absence of residual or terrestrial deposits beneath the cleanly assorted fragmental base of the upper Huronian. Had the land been base-leveled by terrestrial erosion prior to the advance of the sea, that advance would seem likely to have flooded the river mouths and required them to build' up to grade, resulting in the development of terrestrial deposits, including much mud, in advance of the encroaching sea, to be ultimately covered by it, and not removed. It is entirely conceivable that farther to the south the upper Huronian sea may actually have advanced over this zone of terrestrial deposition, but that in the Lake Superior region the sea had encroached upon the upper portions of the rivers and was cutting into the rock. There seems to be an absence of sea cliffs to the north of the present upper Huronian beds, but this may be explained by later erosion. The advance of the sea over a gently sloping surface was accompanied by deposition of a thin conglomerate and sand formation spread evenly over a large area. Barrell " has shown that \\'ith tlie low gratlient characteristic of such advance the conglomerate at the base is likely to be very tliin, if not altogether lacking, being worn out by littoral abrasion, and in modern instances being observed to disappear a short distance from the shore. Conglomerates of this sort may be thick and coarse only around monadnocks standing above the plane of transgression. The deposit of the upper Huronian sea seems to be similar to the thin fragmental base of the Cambrian, which was laid down by the Paleozoic sea advancing also from the south over a flat surface. The absence of conglomerate in the Cambrian except around monadnocks is well known. 2. Then came the deposition of the iron-bearing material. Tliis is a chemical precipitate requu'ing either quiet conditions of deposition or extreme rapidity of deposition to account for the lack of interbedded coarse fragmental sediments. It has been argued in another place (pp. 506 et seq.) that the thick iron-bearing formations near the base of the upper Huronian, such as those of the Mesabi, Gogebic, and Menominee districts, find their essential explanation in their genetic relation with basic volcanism, wluch furnishes sources for unusually abundant deposition of iron salts. The abruptness of the change from quartz sand to iron-bearing forma- tion and the usual lack of any fragmental material in the iron-formation layers seem to imply some unusual change of conditions, probably not related to topographic or climatic changes. 3. The advance of the upper Huronian sea overlapped the Lake Superior region but may not have progressed much farther north. We fuid no record of it farther north, though allow- ance must be made for much erosion. The flatness of the plane would require that planation or scouring should l)e weakened diu-ing the northward transgression. The rivers would then be able to hold their own against the sea, and deposition of river alluvium in the form of great deltas may be supposed to have predominated over marine fragmental deposition. Then were o Barren, Joseph, Relative geological importance of continental, littoral, and marine sedimentation: Jour. Geology, vol. 14, 1906, pp. 433-446; also personal communication. 614 GEOLOGY OF THE LAKE SLTPERTOIl REGION. built up the thick masses of mud deposits characterized by discontinuous, pyritiferous, {):raphitic seams, and iron-car))()iia(o lenses at different horizons, wliich seem better explained by delta and lagoon contlitions than by any other hypothesis tliat has been suggested. A.ssociation with subaqueous extrusions is thus explained. So far as deltas are terrestrial the upper Huron- ian muds are terrestrial. The lack of tlecomposition of the muds and the graphitic material associated with iron carbonate, indicating the probable existence of peaty material associated with bog deposits, favor the view tliat the climate may have been contimiously cool and wot, for nowhere are the conditions for hick of decomposition, bog formation, and absence of oxidation of carbon so well developed as in a district where a continuous covering of water prevents the access of oxj'gen. In warmer regions or in those in whicii a part of the year .is hot and dry the organic material is likely to be oxidized, giving an abundance of carbon dioxide for attack of the rocks. Contemporaneous basic igneous extrusions, so abundant in the upper Hiu-onian, doubtless furnishetl an unusual source for mud, by their decomposition when hot," through the agencies of acid solutions, through the agencies of the atmosphere acting upon sulpliides and thereby freeing sulphuric acid for attack on the adjacent rock, and finally perhaps by reaction of the hot lavas with sea water. In figure 76 (p. 612) is indicated the direction of alteration of basalt by hot sulphuric-acid solutions of the Hawaiian Islands. The most altered pliase represents rock which has not been transported. It is to be noted that the direction of alteration is some- what dilTerent from that of weathering. It is entirely possible, if not probable, from the posi- tion of upper Huronian slates in the diagram, that they have been derived from the katamorphism of basic igneous rocks, both by ordinary weathering and by the unusual alteration of hot acid solutions associated with the igneous rocks themselves. The upper Huronian sediments are therefore regarded as the combined result of an advanc- ing sea scouring, perhaps cutting the old surface, of a source in which basic volcanic rocks form a distinctive part, and of the final deposition of a great mud delta. The building up of the upper Huronian, developing terrestrial conditions toward the close, fm-nishes an appropriate setting for the inauguration of the great Keweenawan period of terres- trial sedimentation which followed after an interval of erosion. KEWEENAWAN SERIES. As the Keweenawan is a unit to a greater extent than the Huronian or the Archean, being located along the border of Lake Superior with large inland extensions, and as the general out- line of the history of the Keweenawan has been given in Chapter XV, we give here only the briefest summary of the salient features of the series. LITHOI.OGT AND SUCCESSION. It has been seen tliat the Keweenawan is separable uito three divisions, a lower, middle, and upper. The lower Keweenawan was formed during a period of sedimentation and con- sists of conglomerates, sandstones, shales, and limestones. This division of the Keweenawan is not very thick, but it is widespread. The maximum measurement is 1,400 feet. The michlic Keweenawan represents a time of combined sedimentary and igneous action, containing many alternations of sedimentary and igneous deposits. In general the igneous activity greatly domi- nated in the early part of middle Keweenawan time, but was less dominant in the later part. Upper Keweenawan time was again a period of normal sedimentation. At the base of the upper Keweenawan are thick conglomerates, which are overlain by shales and these bj- a very thick sandstone formation. As contrasted with the Huronian the Keweenawan sediments are dominantly clastic. Xon- clastic sediments are found only in one locality, in the Nipigon-Black Bay district. Moreover, the clastic formations are coarse, being dominantly either psepiiitic or psammitic. Only sub- a Maxwell, Walter, Lavas and soils of the Hawaiian Islands: BiUl. A, Exper. Sta. Hawaiian Sugar Planters Assoc., 1905, pp. 8-22. GENERAL GEOLOGY. 615 ordinately are pelites present, the single important representative being the shale of the upper Keweenawan. Another feature in which the Keweenawan sediments contrast with the Huronian is tliat they are largely derived from the igneous rocks of the series itself. IGNEOUS ROCKS. Tlie igneous rocks of the middle Keweenawan are both plutonic and volcanic. They include basic, acidic, and intermediate varieties, the basic rocks being dominant. As the detritus of the middle and upper Keweenawan is derived largely from the igneous rocks of the period itself, in arriving at an estimate of the mass of igneous intrusions and extrusions of this time we must consider not only the original igneous rocks, but the sediments wliich are derived fi-om them. The mass of the Keweenawan volcanic and ])lutonic facies is enormous. CONDITIONS OF DEPOSITION. It is probable that the sediments of the Keweenawan were largely land deposits. (See pp. 416-418.) The principal arguments for this conclusion are their prevailing red color, their little-assorted, feldspatliic nature, and their rapid alternation with abundant extrusive rocks having textures that are ordinarily associated \vith subaerial cooling, in contrast with the tex- tures of subaqueous cooling so common in the volcanic rocks of the lower Huronian and the Keewatin. But it is also probable that a portion of them were deposited under water. In the discussion of orogeny (pp. 622-623) it is shown that the Lake Superior basin was formed largely in Keweenawan time, and it is highly probable that this basin contained water. CORRELATION. The correlation of the different areas of the Keweenawan is a simple problem. The great area of Keweenawan, extending from Keweenaw Point through northern ^Michigan into Wiscon- sin ami Minnesota and thence northeastward to the Thunder Bay district and Lake Nipigon, is almost continuous. Therefore the only problem of correlation so far as the general series is con- cerned is that of the rocks of Isle Roj'al, Michipicoten, and the areas on the east coast of Lake Superior. The placing of these rocks in the Keweenawan is based on their position at the top of the pre-Cambrian, the unconformity at their base, and their remarkable likeness in litliology, succession, deformation, and metamorphism to the rocks of the main Keweenawan area. Though all these points bear on the question, it was the likeness of the lavas of these areas to those of the main area and their interstratification with red sandstones and conglomerates which led the earlier geologists who worked in the Lake Superior region to recognize the identity of the separated areas of Keweenawan rocks. The problem of fixing the exact relations of the Keweenawan and Cambrian is not so simple The evidence as given in Chapter XV is in favor of the Algonkian age of the main part of the Keweenawan. PALEOZOIC ROCKS. The Keweenawan is the latest period which this monograph treats in detail. On the gen- eral geologic map (PI. I, in pocket) the Paleozoic and later rocks are shown as covering a large part of the area south of Lake Superior, but they are all represented l)y one color, for it is not our purpose to consider the post-Algonkian formations separately. The Paleozoic rocks are mentioned only in so far as they are related to the Proterozoic — that is, the Algonldan and Archean. For the most part the formation which overlies the Proterozoic rocks is a sandstone, which is generally recognized as of Cambrian age. Its basal portion where in contact mth iron-bearing formations consists of detrital ferruginous rocks. This formation is everywhere in a substan- tially horizontal attitude, thus conti-asting strongly with the Proterozoic rocks. In general the relations between this sandstone and the Proterozoic rocks are those of most profound uncon- 616 GEOLOGY OF THE LAKE SUPERIOR REGION. formity, luid tliis is true whichever of the more ancient series undeilics the sandstone. The manner in which tiu^ ('ambrian sandstone cuts unconformahly across tiie several series of the pro-Cumbrian is well illustrated on tlie east side of the ])re-Cumbrian area of the Upper Penin- sula of Michigan and northern Wisconsin. Here the Cambrian is fossiliferous. The uncon- tormnblo relation to the Arcliean is splendidly illustrated alono; the Lake Superior shore north of Mary jjeriods of epeirogenic movement, orogenic movement, and erosion, each of these intervals being markctl by an unconformity. The first of these unconformities, tliat at the top of the Archean, is the most conspicuous, represents a strong lithologic contrast, and lias been by all geologists taken as an 47517°— vol52— 11 40 626 GEOLOGY OF THE LAKE SUPERIOR REGION. essential datum plane in mapping and working out the geologic history. The unconformities separating the divisions of the Iluronian and the Iluronian from the Keweenawan are of differ- ing value, but all represent important structural and time breaks. The unconformity at the base of the Cambrian is one of the first magnitude and is coextensive with the great uncon- formity at this horizon outside of the Lake Superior region. Of the five periods of deformation, three stand out consi)icuously — that at the close of the Archean tliroughout the region, that at the close of the lower-middle Huronian principally on the north shore, and that at the close of the Keweenawan priiicipally along the axis of the Lake Superior basin and on the south shore. These areas of folding had been shore zones of heavy Huronian and Keweenawan deposition. As is common, the shore zone was a place of recurrent upheaval and subsidence, marked orogenic movement, igneous activity, and sedimentation. To these many causes combinetl is due the complexity of the geolog}' of the region. These shore conditions may bear some relation to the tact that part of the Lake Superior region south of the international boundary is one of the great iron-producing areas of the world. It has been a source of surprise to many that the adjacent Canadian region, in which the geology seems to be in a general way similar, has not been found to bear iron ore in anytlung hke the abundance of the States to the south. But by far the larger ]>art of the iron-ore deposits in the States occur in the middle Huronian and dominantly in the upper Huronian formations. The middle Huronian is known in a general belt fringing the main pre-Cambrian area of Canada along the north shore of Lake Superior and extending northeastward tlirough Lake Tiniiskaming. It may exist also farther north in the interior of the Canadian pre-Cambrian, but, to judge prin- cipally from the facts observed on the north shore of I^ake Superior, the interior jjre-Cambrian region of Canada was probably above the sea during middle and upper Huronian and Kewee- nawan time and only continental deposits were formed in it. The up])er Iluronian, the {)rin- cipal iron producer, is but scantily represented along the soiithem margin of the Canadian pre- Cambrian. The only iron-bearing formation which has an extensive occurrence in the great pre-Cambrian shield of Canada is tliat of the Keewatin series. The Keewatin iron-bearing formation has not been largely productive. If the apparent scarcity of middle and upper Huronian rocl1 Cloquet district, geology of 213. 214,215 375 Cobalt district, geology of. ' ' 'g^j ^ °"'''"-; :!:;;!:::!:::::::::;:'592,626 Coke, use of ., ,„ ^ ,, . . 4/-4S Colby mme, geology at 03^ Cole, G. A. J,, and Gregory. J. W., on ehipsoidal structure 511 Coleman, A. P., on gold ores -nj on lake basihs on northwestern Ontario 1 jo. Coleman, A. P., and Willmott, A. B Coleraine, building of Collins, W. H., on region north of Lake Superior Commonwealth, geology near Commonwealth district. See Florence district. Competition in mining, effects of Concentration, mechanical, process of 539-540 Concentration, secondary. See Secondary concentration; Weath. ering. Conglomerates, copper in Copper ores, area of, extent of V-Za area of, map showing ' c- . association of, with igneous rocks 5gj chapter oa See also Keweenaw Point, copper ores of. character 01 -„„ 0/3 deposition of, chemistry of 589-590 "<'^'^°' ■..'.'-.'.'..'.... 582-586 , t™"^"' : 5S1-5S2 deposits of, types of jjp depth and, relations of -„, e.xtentof „. 0/5 grade of ^^^ mine waters from c-q mineralogy of '"!-'"!!!;!;!;!!!;;"573^574,582 mining of, history of 3j_3g occurrence of 3„, modeof ' „, origin of ! . 569 production of 431 -150, 603 95,160 44 95 . 321 41 577-57! 573-593 Cretaceous rocks, deposition of . ^fjs detritaiorcsm '.y^'.'.'.'.'. ::.'::::.:.:::: i9^m analysis of distribution and character of '.'.■.■.■ 'i59,'l'78^i79,'2U,'21S, 616 "°°-°': ■. 178-179, 196-197, 460*503 phosphorus m jg ' Crj'stal Falls, geology near 295 323 Crystal Falls district, correlation in exploration in . .TOS, 606, 610-011,617 484 eO^^OSyot 291-300, 441, i;07-609, 619 map showing ,^, iron ores of, alteration of character of 546 323,325 composition of 324-32' magnetic phases of ^g plate showing .„ proportion of relations of reserves of 462 501 , 507, .i59 489 secondary concentration of 326 5.39 volumetric diagram of location and area of ... ; map of 580-592 575 relations of, to other ores 691-S9'> wall rocks of, alteration of 582-585 Copper Lake, hon at jj„ Copper River, geologj- near 378-379 Cordierite homstone, distribution and cliaraoter of. ■. 173-174, 200 Corey, G. W., and Bowen. C. F., on Menominee district '345 Correlation, chart showing 59g details of. See particular systems, etc. principles of .„- Courtis. W. M., on silver minerals Coutchiching schists, correlation of occurrence and character of •j.jg 599 593 598 147 Cox, G. H., on hj'dration of ores 556^557 Creosote, recovery of 48 353 - 32,291 mining in, history of 3„ ' physiography of !!!!!!!"!! 90^97 ,07 Cuba, iron ores of 495 CuB mine, geology near „„g Current River, geology near "" 205 206 Curry member, correlation of 'ggg distribution and character of. 335^340 347 ,-, "•™°'-<'<" ;::;::;;:::;;: '345 (. iirry mine, geology near 339 Cuyuna district, correlation in ; 698 610-611 description of " ' ' 32 211 exploration in 484^485 S™l°Sy<" !!!!!!. '211-216,' 375. 604-605 iron ores of, alteration of ,,„ , „ 646 analyses of 220 character of ' 2I9_2')o composition of 2''0-223 figure showing 221 222 distribution of 'jig magnetic phase of 217-219 486 phosphorus in ' oon ■"'^■''"^sof !!!!!!!!!!!!!!;!;;;;'5oi,507 reserves of ,„„ J 489 secondary concentration of 223-224 639 section of, figure showing '210 structure of 216 223 topography on '^^g manganese in " ^^ maps of 212 mining in, history of ^^ ^g phosphorus in " 224 physiography oi jn 213-214 623 D. Daly, R. A., on intrusion on lavas Uam Lake, geology near Xil Darling, J. n., on earth movement Davis, C. A., on Marquette district Davis, W. M., on physiography Dead River area, description of geology of ^y_; , '"'^Po' ■■ -.'^''''^''^^:'... 286 Deception Lake, geology near 208 370 iron near ..!...!... 207 Deerhunt mine, geology at Deer Lake, geology near Deerwood, iron near nig Deerwood member, alteration of 223-224 correlation of ..!!.'.'.'.'.'.'.'."598,610-611 distribution and character of 212-216 '■■™.i° -!!!!!!!'!!!;!;;;;;;;;.. "460 magnetic rocks in 216-219 structure of ,^7 rocks of, correlation of. . structure of. 600 510-511 451 96 110 287 301 254 630 INDEX. Page. Deformation, changes due to ^^ description of 620-621, B2Wafi Deltas, formation of 45^459 Doming, A C, aid of ^" Density, relation of , to cubic feet of ore, diagram showing 480 Dialjase, composition of • ■"' Diamonds, discovery of in drift *^^ Diemer, M. E., work of ^^^ Dilces, distribution and character of 136,236,411 Dip needle, use of, in prospecting ■'■' Disappointment Lake, geology at 129, 131, 133, 134, 199 Disappointment Mountain, geology at H^ Districts, ore-bearing, list of 31-32 Docks, ore, list of ■•* Mewol *^ Dog River, geology on \^ Dori5 conglomerate, correlation of °^^ distribution and character of 150, 151, 154-155 petrography of 154-lo5 Douglas County, Wis., copper of ^5*0 Drag folds, description of 347-348 figure showing 350 Drainage, character of 8' description of 33-.J4 Drainage, ancient, character of 8G-87 modification of 113. 435 figures showing 1'3 Drift, ages of 435-436 areas of 4o4-45o deposition of ^'■- 435-437 distribution of 439-441 drainage of areas of ^5 obscilration by ■*30 stratification of *27 topography of 454-455 modification of 455-459 See aim Pleistocene; Glaciers: Moraines; Kamcs: Outwash plains, etc. Driftless Area, lakes in 438 location of ^^ map showing ^8 physiography of 454 view of ^"i Drilling, exploration by 484-485 Drumlin.s, formation of 433-434 Dubois, II. W., and Mixer, C. T., analysis by 281 Duluth. geolog>- near 452-453. 458-459 physiography near 458 Duluth, Lake, formation of 444-445 map of : 445 Duluth escarpment, description of 112-115 glaciation of 431 view of If 2 Duluth gabbro, correlation of 598 distribution and character of 137, 159, 177, 198. 201-202, 372-373, 410, 414-415 dikes of 137 intrusion by 1.31,137,372-373,426,561 metamorphism by - 546 plate showing 548 relations of 202-203,372-373 segregations in 561 view of .- 112 Dumortierite. occurrence of 515 E. Eagle TI arbor, geology near 413 mining near 36 Eagle U iver district, copper veins of 575-576 geology of 425 section in 380 silver In 575 Eames, II. H., on Mcsabi district 42 Eastern sandstone, relations of 388-389 Eeliel, E.C., on production of manganiferous iron ores 4.S8 Eleanor. Lake, geologj- near 153, 154 Eleanor slate, correlr tion of 598 distribution and character of 160, 154 Page. Elevations, height of , 33,86,94 EHtman, ,\. TI., on Minnesota geology 371,:i7.')-376 Ellipsoidal structure, occurrence and character of 120, 148, 151 origin of 502,511-512 plate showing 120 significance of. In ore genesis 510-512 Ely, geologj- near 119,122,123,124.126 iron ores near 1.37-138 character of 140 composition of 139, 140 secondary concentration of 142-143 changes in 142-143 figure showing 143 Ely greenstone, acidic flows interliedded with 121 age of 127-128 clastic rocks associated with 121 correlation of 598 distribution and character of 119-122 intrusions in 121-122, 128-129 mineral composition of 120-121 relation of, to Soudan formation 124-128 topography on 93, 119 Ely Lake, geology near 122 Embarra,ss granite, correlation of 598 distribution and character of 159, 178, 415 intrusions l)y 178 Embarrass Lake, geology near 160, 176 Emerald Lake, geology near 122,123,126 England, iron ores of 495 English Bay, geologj' at 368 Epsilon Lake, geology near 136 Erosion, amount of 89-90, 109,558-559 relation of, to iron-bearing sediments 50S-506, 5.58-559 topography due to 86,98-99,109 Eruptive rocks, iron in 512-513, 569 mineralization of 569 relations of, to iron ores 506-516 solutions from 587-588 See also Lavas; Igneous rocks. Escarpments, age of 116 description of -• 112-116 distribution and character of 110-111 origin of 111-112, 117 structural relations of 116-117 Eskers, formation of 434 Exploration, cost of 47 methods of 484-1S6 Fall Lake, geology at 119 Faulting, description of 620 evidence of .' 87, 104, 114, 117 Influence of, on physiography 87, 98-99, 101 , 104, 112-115, 1 17 figure showing ! 112 Fault scarps. Sec Escarpments. Fay Lake, geology near 131 Felch, iron ore near 327 Felch.Mountain district, correlation in 598,609 geology of 302-305,386,609 iron ores of 326-327 analysis of 327 relations of 501 secondary concentration of 328 ■ volumetric diagram of 353 location and area of 32, 302 physiography of 107 stnut lire in 623 Felch schist, correlation of 327, 609 distribution and character of 302, 303,306,307, 609 ' relations of 305 Felsite, copper in 574 Fence River district, geology of 293,295,296-298 Fence River, physiography of 107 Fenner, C. N.,on igneous rocks 511 Fernckes, G . , work of SS9 Field work, correlation of laboratory e.xperiments and 527-529 Fish Creek, geology on 379, 415 Flambeau River, geology on 357 INDEX. 631 Page. Florence district, correlation in 598,610-611,617 exploration in 484 geology of 320-323, 376, 379-380, 606 iron ores of, alteration of 546 character of 323- 325 composition of 324-325 production of 461 relations of ,' 601 , 507 reserves of 492 secondary concentration of 326 structure of 475 volumetric diagram of 353 location and area of 32, 320 map of ! Pocket. mining in, history of 39-40 Flowage, rock, alteration by 554-555 Fluor I.'iland, geology on 368 Folding, extent of 123, C20-622 , figure showing 123 Fond du Lac, physiography near 452 Foster. J. AV., and Whitney, J. D.. on Marquetle district 96 Fourfoot Falls, geology at 345 Fox River valley, correlation in 598 geology of 365 map of . . .^ 359 Freda sandstone, deposition of 426 distribution and character of 384, 414, 417, 426 Freedom dolomite, correlation of 598 distribution and character of 360-361 iron in 460 Freight rules, relation of. to grade of ores 494 Fuel, nature of 47-48 Fumee. Lake, geology near 336 G. Gabbro. metamorphism by 546 Gabbro plateau, character of 91-92 monadnocks on 92 Galiimichigami, Lake, geology near 131, 136, 199, 202 Gary, E. n., on iron ore and freight rates 494 Gate Harbor, geology near 413 Geikie, A., on igneous rocks 511 on iron ores 508-509 Geography, maps showing 31, 32 outline of 30-32 Geologic history, rgsumfi of 023-026 Geologic map of Lake .Superior region Pocket. Geography, physical, account of ■ 85-117 Geologic knowledge, growth of 72-73 Geologic work, history of 70-84 Georgian Bay. copper ores at 626 Geology, general 597-1)26 Germany, iron ores of 495 Giants Range, definition of 41 description of 169 geology of .' 160, 172-173, 176, 177, 178, 179 Iron of 165 mining on, history of 42 physiogiaphy of 103-105 structure of 175 Giants Range granite, correlation of J9S distribution and character of 135-136, 162 intrusions of 170 phosphorus in 194 relations of 162-163 topography on 94 Gilbert, G. K., on glaciation 450 Gilman deposits, iron ores of 565 iron ores of, analysis of 565 Glacial deposits, distribution and character of 179, 216, 308-309, 355, 559-560 See also Pleistocene deposits; Drift; etc. Glacial epoch, description of 427-453 See also Pleistocene. Glacial lakes, beaches of 449-452 deposits in 452-453, 455 distribution and character of 441-448 Page. lilacial lakes, formation of 427 4,3^ tilting of, etiect of 448^449 Glaciation, effects of 33,91,92,98,106,114-116,427-453 erosion by ■_ 427 period of 427 Glaciers, advance of 427-429 advance of, map showing 428 effects of 427 contrasts in 43Q constructive work of 4.33-441 See also Moraines, etc. erosion by 43(^32 melt ing of 435 retreatot !'! 429-435 soiu-ce of 427 transportation by 432-433 See also Drifts. Glauconite, relation of, to iron 503 Goetz Lake, geology near 153 Gogebic district, correlation of (joe 617 geology of 214, 380, 423! 600 iron ores of, alteration of. 54(5 analyses of 4gi magnetic phase of ^gQ production of 49-51,69,461 relations of 601 .507 reserves of 489^92 secondary concentration in 475 539 structure of 475 4gg manganese in 4gg mine waters in, analysis of ,-. 543 mining in, history of 40 See also Penokee-Gogebic. Gogebic Lake, geology- near 385, .388 Gold ores, distribution and character of 595-596 Gold, mining of, history of 46 Goldthwait, J. W., on glaciation 451 Goodrich mine, geology near 265 Goodrich quartzite, correlation of. 598, 608 distribution and character of 251, 265, 283, 285, 288, 289, 608 iron ores in 270-272 relations of 264, 265, 266-267, 268 topography of 106 Goose Lake, geology near 258 Gordon, A. T,, analysis by 179, 191, 193 Gordon, W. C on Black River geology 384-385,388 Graben faulting, description of 112 figure showing 112 Grace mine, gold in 595 Grand Portage Bay, geology at 370 Grand Rapids, geology near _ . 164 iron near 164 Grand Traverse, physiography near 433 Granite Bluff, geology at 306 Granite Island, geology at 414 Grant, U. R., map by 355 on Gunflint Lake district 201 on Keweenawan 398-399. 400-401 on physiography 91, 92, 9,S-99, 100 on Wisconsin geology 378 Grant^burg, glaciation at 4.37 t treat conglomerate, deposition of 425 distribution and character of 381-387, 390, 413, 416, 418 relations of 574 Great Lakes, drainage to 33-34 history of 448-449, 456-459 maps showing 457, 458 structural relations of m transportation on 490-497 Great Palisades, geology near 371-372 section at, figure showing 371 Greenalite, alkaline solutions producing, source of 525 alteration of 187, 197, 210, 530, 5.37 chemistry of 550-551 plate showing 532, 534 analyses of 1G7 carbonate ores and, relations of 526 632 INDEX. Page. Greenalite, carbon dioxide producing, source of 527 conii»os:tion of 528 dei)osition of 503. 521-522 distribution and character of 165-108, •102, 572 iron in 10.^108 nature of 522. 525 oxidation ancl hydration of 530,537 phosphorus in ^^^ photomicrofiraphs of 524. .'532 plates showint; 120.474 Green Bay lobe of Labrador glacier, extent of 428 Greens'one conglomerate, correlation of 59S distribution and character of '-^^ iron in ^' - Oreenwater Lake, iron of '50 Gregory, J. W., and Cole, G. .\. J., on ellipsoidal structure 511 Gros Cap. geology at 151, 152, 153, 154. 393 Gros Cap greenstone, correlal ion of 598 distribution and character of 150-151 intrusions of '54 petrography of 151 Grout, W. F., on the Kewecnawan 376 Groveland, iron ores near 327 Gi-ovcland formation, distribution and character of 296, 304, 305 topography of - '^^ Gunflint formation, analysis of 204 correlation of 598 distribution and character of 198, 199-200 iron ores in 200, 203-204, 460. 480 magnetite in 501 metamerphism of 200 phosphorus In 1^5 relations of 203 section of, figure showing 199 structure of 199-200 topography of 102 GunHint Lake district, correlation in 598 description of 198 geology of 136, 172, 177, 198-203. 209. 604 iron ore from 203-204 photomicrograph of 524 physiography of 101, 1U2 H. Hall, C. W., on Cuyuna district 213 on glaeiation "155 on Kewecnawan series .- 377 Hall, R. D., analysis by 158, 173, 518 Hamburg slate, distribution and character of 356 topography of 108 Hanbury n ill, geology at 335 Hanbury slate, correlation of 267, 307. 330. 340. 598 topography of 1"^ Hancock mine, geology at 307 Hanging valleys, submerged, occurrence of 114 Hartford mine, ore from, washing tests on 281 Hawkins mine, folding at, view of 180 Helen fonnation, correlation of 598 distri bution and character of 150, 152-153, 155 intrusions in 154 iron in 152,155.460 petrography of ■ 153 relations of 153-154 Helen mine, description of 150. 157 geology near 152, 154 hanging valley near, view of 432 history of 45-16 Hematite, deposition of 527 mining of, history of 45 occurrence and character of 479, 534-566. 572 Hematite Mountain, glaeiation near 431 Hematitic chert, plate showing 406 Hemlock formation, correlation of 598, 007,617 distribution and character of 291,294-296,323,607 relations of 297,300,507 topography of ' 107 Hcnnansvllle limestone, distribution and character of 306, 307,330,345-346 Heron Bay, topography near 95 Page. nibbing, Minn., geology near 105,10 iron ores near 476 section near, figure showing 180 Highlands, elevations in 86 rocks oJ 85 subdivisions«f 85 topographic development of 85-89 Sec a^so Uplands; Monadntxjks; Peneplain. History, geologic, rfoumfi of 623-026 History of mining 35-. 2.';2, 254-255, 287, 309-310, 322, 330-331, 597, .59»-li(XI extension ol 623 gold in 595 iron ores of 46,149,460-461,501,507,517-518,626 alteral ion of. 554-555 proiiuction of 517 relations of 504. 506 relations of 227,257,260.504.506 Sc( also particular formalions. Kekekaliic. L.ike, geology near 133, 134. 130, 603 Kettle Kiver, geology near 378-379 Kettles, format ion of -• 438 Keweenavvan series, age of 415-416, 420 area of 419 copper ores in 574, 580, 581-582 correlation of. 305.598. B14-G15 deposition of 416-418,424-125, 426, 615. 025 distribution and character of 137, 159.177-178.198.201-209.212.215.224-220,2(4- 235,250,251.300-305,366-420, 597, 602, 614-(il5 faulting of. 420-421 , 620, 622, 6M folding in 622 grain of 407-408 Ustory of, r&iunfi of 424-426. 615 igneous rocks in .395-412, 425. 615 nomenclature of 395-407 source of 411-412 intrusions of 171-172, 197, 215, 278-279, 377-378, 418, 424, 425-426 iron horizons in 460 jointing in 420 metamorphism of 423-424 relations of 202-203, 234-235, 378-379, 384-386, 388-389, 414-416, 420, 619 section of 384 figure showing 99 sediments in 412-413 structure of 376,383,620-625 figure showing 419 subdivisions of 366-367. 614 thickness of 418-419 topography on 98, 99-102 unconformity at 619 volume of 419-120 See aho particular fonnafions. Keweenaw district, location and area of 31 mining in, history of 35-37 physiography of. 97, 100, 116 production of 36 topography of 91-92, 94, 100 Keweenaw escarpment, description of 115 Keweenaw lobe of Labrador glacier, extent of 428 Keweenaw Point, copper ores of 573-593 copper ores of, character of 573 extent of 573 grade of 574 minerals of 673-574 occurrence of, mode of 574 production of 575 geology of and near 380-385. 408, 409, 412-413, 415, 418, 425 maps of 380,574 section of, figure showing 99,574 silver at 575 structure on 383 veins on 574.575-576 Kcyes Lake, geology near 321,322 Kimball Lake, geology near I53 Kin g, F. II., on glaeiation 439 Kitchi schist, correlation of 598 distribution and character of 262, 254-255, 260, 287 Kloos, J. II., on Kcwecnawan series * 397 Knife Lake, geology near X19 Knife Lake slate, correlation of 593 distribution and character of 132-135 intrusion of 133-134 lithologyof 133-134 mlner.il character of I34 Page. Knife Lake slate, relations of 135 structure of 133 thickness of 135 topography of 94, 119 Kona dolomite, correlation of ^ 598.605 distribution and character of 252-25.3.258.605 relations of 258, 200,305 topography of 105 L. Laboratory sTOthesis, correlation of field work and 527-529 Labrador glacier, advance of 428 Lac la Belle conglomerate, distribution and character of 381 Lake Angeline mine, washing tests on ores from 281 Lake basins, formation of 431-432 section of, figure showing '. 432 Lake Huron shore, correlation on 598 Lake Michigan lobe, extent of 428 Lake of the "Woods, geology at 122 Lake of the Woods district, correlation of 69,s8-590 on glaeiation 439-440 on Isle Royal 389-390 on Keweenawan series 382.383,385,398.400-403.405.407 414,421 on mine waters 644 Lane, A. C, and Seaman, A. E., on Lake Superior sandstone 616 Larsen, E. S., and \Vright, F. E., on quartz crystallization 549 Laterite, association of, ^vjth iron ores 503 Laurentian liighlands. location of 355 Laurontian peneplain, extent of 88 Laurentian scries, batholiths of 145 correlation of 304. 000-601 deposition of 023--517 variation in. relation of. to iron 510-518 See also Eruptive rocks: Igneous rocks. Lawson. .\. C. on glaeiation 450 on iron ores 509-510 on Lake of the Woods and Rainy Lake districts 144-145, 147 on Minnesota 371,374 on physiography 94, 98, 101, 1 12, 450 on subcrustal fusion, theory of 146 Leaching, process of 537-539 Lee Hill, geology near 119, 122, 123, 127. 131 Lehmann. O.. on liquid crystals .'V2.5, 572 Leith. 0. K.. on physiography 103-104,432 work of 30,44-45 Lerch Brothers, analyses by 193, 238 INDEX. 635 Page. Life, pre-Cambrian, existence of 617 Lighthouso I'uint, Kcology near 254 Lime, relation of, to phosphorus 196, 249, 282, 281 figure showing 190. 249, 282 Limonite, formation of 519-520, 571 natiH-e of 620-521 Linear monadnocks. See Monadnocks, linear. Literature, list of 73-84 Little Falls district, geology of 213, 214, 375 Little Presque Isle River, geology near 227 Loess, distribution and character of 438 Logaa, W. E., on Keweenawan series » 392, 393 on Michipicoten Island 391 section by 367 Logan sills, correlation of 598 distribution and character of 198, 202, 208, 374, 410 intrusion of 208, 420 relations of 202-203, 593 silver in 593 topography on 101, 102 Long Lake, geology near 119, 132, 153 Loon Lake, Mich., geology near 201, 370 iron at 4G, 209 Sfc fl/so Anlmikie district. Loretto mine, geology near 332, 330 Low, A. P., on iron ores 508 Lower Magnesian limestone, distribution and character of . . 300,361-3(12 iron ores in 504-565 Lowlands, description of 108-110 geologj^ of : 108-110 M. McFarlane, Thomas, on Keweenawan series 397 on Maniainse Peninsula 392 on Nipigon Bay 308 on silver 593 Mclunes, William, on Tlunters Island and Thunder Bay region. 94-95, 101 McKays Mountain, geology of 209 McKenzie, geology near 206 Magma, iron from 513-514, 568, 571 solutions from. S(c Solutions. Magnetic phases of ore, occurrence of 185. 216-219 5fc aZso Amphibole-magnetite rocks. Magnetic survey, prospecting by 44, 48tV-488 Magnetite, deposition of 527 occurrence and character of 479, 480, 486 origin ol : 562 Magnetite, titaniferous, character of 561 origin of 501 , 568 Magpie Valley, geology near 151 Mahoning mine, concretions in 192-193 concretions in, analyses of 193 geology near 165 iron ore from, plate showing 468 Maniainse Peninsula, geology of 391-393, 418, 425 section on 392-393 Manganiferous ores, occurrence and character of 488, 560 Mansfield, geology near -• 295 Mansfield slate, iron in 295-296, 324 occurrence of 294, 295-296, 303 relations of 296 topography of 107 Map of Lake Superior region 86 showing topographic development 87 showing topographic provinces 88 Maps, geologic, accuracy of 73 of Lake Superior region Pocket Map, index, of region 31 Marathon conglomerate, correlation of 598 distribution and character of 356,357 Mareniscan series, name of 226 . 254 Marenisco, geology near 226 Mariska, geology near 162 Marquette district, acknowledgments concerning 251 correlation in 598, 599, 606, 610-611, 617 exploration in 485 geology of 251-269,429. 441,605-610,618,620-621 gold of 595 Page. Marquette district, iron ores of, alteration of 546, 554, 610-611 iron ores of, analyses of 273, 491 character of , 274-275. 503 classification of 271-272 composition of 273-274 distribution of 270 magnetic phases of 486 occurrence of 270-271 figure showing 270 phosphorus in 279-283 concentration of 281 . 283 figures showing 280. 282 plates showing 464. 468. 470 production of 461 proportions of 402 relations of 507 reserves of 489-492 secondary concentration of .' 275-279, 539 conditions of 275 sequence of 278, 279 volume changes in 276-277 figures showing 276, 277 section of, figure showing 270 structure of 475 topograph y of 476 location and area of 31-32. 251 map of Pocket mining in, history of 38-39 physiography of 96, 10.'i*-106. 252 production from 39. 51-00.69 structure of* 252-253. 623 figure showing 253 Marshall Hill graywacke, correlation of 598 distribution and character of 356,357 topography of 108 Martin, L., on Keweenawan series 420-421 on physical geography 85-117 on Pleistocene -127— i59 Mass mine, geology at 271 history of 36 Mastodon mine, geologj' at 295 Matawin district, geology of 149-150 iron of 150 Mead, W. J., diagram method devised by 182-183 on iron ores 137-143, 179-197, 235-250. 270-283. 286, 323-328, 346-3.54. 3(;2-365. 460-571 work of 518 Meadow mine, phosphorus in 194 Meeds, A. D.. analysis by 191 Menominee district, correlation of 59S, 599, 610-611 cross sections in, plate shovilng 346 geology of 329-346, 605. 007-609, 616 iron ores of. alteration of 546 analyses of 3.50-351. 491 character of 352. .503 composition of 350-352 magnetic phases of 486 position of ^59.610-Gll production of 461 proportion of 462 relations of 501 . 507 reserves of '. 489-492 secondary concentration of 3.53-354. 539 structure of 346-350. 475-476, 623 figures showing 346,347,348,349 volumetric diagram of 353 location and area of 32. 329 map of Pocket mining in, history of 39 physiography of 96. 105-106. 329. 433 production of 39,61-65,(19 Menominee River, geology on 321,322,344-.345 Merriam, W. N.. map by 320 on Steep Rock Lake district 147-148 Merritts, discovery by 43 Mesaba, iron at and near 44, 172, 185 meaning of name 159 Mesabi district, correlation in 598, 610-611 definition of 41, 159 636 INDEX, Page. Mesabi district, description of 159 exploration in 47, 485 geology of 159-179, 604-605 glaciation in, map showing 443 history of 41-44 iron ores of, alteration of 545 alteration of, plate showing 548 analyses of 181, 183, 185, 193, 197, 491 characteristics of 183-185,503 composition of 180-183, 193, 197, 555 figure showing 182 distribution of 179-180 niagnetio phases of 185,480 phosi)horus in 192-19G concentration of 194-195 figures showing 192, 190 plates showing 468, 474, 532 production of 401 proportion of 462, 477 relations of 180, 501 reserves of 489-492 rocks associated with, alteration of 191-192 secondary concentration of 186-191, 475, 538, 539, 558 figu7 P showing ISO sequence of 197 structure of 180, 475-476, 4S(J figure showing 180 volume changes of 188-191 figures showing ' 188. 189, 190 location and area of 32 map of '. Pocket mining in, history of 42-44 magnetic portion of, character of 43 development of 43-44 mine waters in, analysis of 543 mining in 497-498 physiography in : 105 plate showing 532 production of 05-68, 09 structjre in 623 view of 180 Mesas, distril)ution and character of 100-102 structure of, figure showing 101 Mesnard quartzite, analj'ses of 257 correlation of 598, 605 distribution and character of 252-253, 250-258, 605 relations of .• 257-258, 260, 305 topograph y of 105 Metamorphism, cause of 545-554, 559, 582 cycle of 5a)-561 effects of 545-554, 559,582-586 relation of, to secondary concentration 552-553 temperature of 549 See also Igneous rocks. Michigamme Lake, geology near 262.267 Michigamme mine, geology at 261 history of 38 Michigamme Motmtaui district, geology of 293, 295-298 physiography of 107 Michigamme River, geology on 295 Michigamrn/? slate, correlation of 267. 323, 598, 608-609, 611 distribution and character of.... 251, 2G7-268, 283, 285, 288, 289, 291 298-299, 306, 307, 309. 31 1-318. 321-323. 330. 340-342. 608-609 intrusions in 345 relations of 265, 206-267,268,313-314, 338-339, 343 structure of 312-313.341-342 topography of 100. 329 Michigsm, bibliography for 74-77 geology of 380-414 investigations in 35. 38. 71-73 iron ores of 461.507 production of 401 physiogniphy of 100. 433 Sec also pnrticular districts. Michigan mine, gold of 596 See aho Minnesota mine. Michipicoten district, copper ores of 580 correlat ion in 598-599, 615 description of 150 Page. Michipicoten d istricl, extensions of 155-156 geology of 150-156,390-391,423,425 gold of 505 iron ore of, analyses of 156 distribution and character of 156-157 reserves of 492 secondary concentration of 157-158 location and area of ' 32 map of 88 mining in, history of 45-40 physiography of. 95,150,431,456 section in . . 391 Michipi(roten Harbor, geology near 151, 154, 155 Middle cont^lorncrate, distribution and character of 381-387 Middle River, geology on 379.415 Mikado mine, geolog>' at 230.238 Minerals, source of 509 Mine waters, analyses of 543.579 composition of 540, 543-544. 579 Mining, history of 35-09 See also Copper; Iron: Silver; Gold. Minnesota, bibliography for 78-^ copper ores of 580 geology of 307-379, 425, 429 investigations in 72 iron ores of * 401 production of 461 lowlands of 110 map of 212 physiography of 91-92, 99, 110-117 titaniferous ores of 501 See also particular districts. Minnesota lobe of Labrador glacier. See Red River lobe. Minnesota mine, history of 36 ores of 36, 576 See also Michigan mine. Minnesota River valley, geology of 224 Mirmpsota tax commission, aid of 30 Misquah 11 ills, elevations in 92 Mississippi River, drainage to 34 pond ing of 438 Mixer. C. T.. and Dubois, H. W., analysis by 281 Mohawk mine, geology at 178 Moissan. H., on metamorphism 549-550 Moisture, relation of, to cubic contents of ore 483-484 relation of. to cubic contents of ore. figure showing 480 Mokoman. Ont.. geology near 149 Monadnocks, character and cause of 90 Monadnocks, Unear, descriptions of 98-106 structure of. figure showing 101 Mona schist. coiTelation of 598 distribution and character of 252. 254-255. 287 Monoclinal ridges, distribution and character of 99-100, 102-108 structure of. figure showing 101 Monopoly in mining, effects of 41 Monroe mine, folding at, view of 180 Montreal River, geology near 413,414 Moose Lake, geology near 119.127.129,131 Moraine, ground, formation of 433 topography of 433 view of 436 Moraine, recessional and interlobate, formation of 435 Moraines, terminal, formation of. 434 view of 434 Morrison Creek, geology on 313. 316-317 Morton, geology near 224 Mosinee conglomerate, correlation of. 598 distribution and character of 356, 357 Motley, geology near 215 Mountain Iron mine, geology near 160,162,164 history of 43 views in 180. 432 Mount Houghton, geology at 382 Mount f Tough ton felsi tc. correlation of '. 382 distributiim and character of 381 Mud Lake, geologj' near 254, 257 Musse.v. IT. R..on Marquette district 38 on Vermilion district 41 INDEX. 637 N. Page. Namakon Lake, geology at 146 Nashwauk, geology near IfiO National mine, history of 3li Necedah, correlation at 598 geology near 358 maps showing 358, 35' of 209, 426 Thwaites, F. T.,on Wisconsin geology 376,379 Tilden mine, geology at 236 Titanium, occurrence of 477, 533, 561 Topographic development, history of 85, 90 map showing 87 Topographic provinces, types of 85 map showing 88 See filto Pleistocene, provinces of. Topography, character of 33-34 modification of 455-459 relation of, to iron ores. ..; 544-545 Tower, gcologj- near 119, 122, 123, 124, 131, 133, 134 Tower Hill, geology near 122, 123 iron ores near 137 Traders mcrabor, correlation of 598 distri\)ution and character of 335-3^0, 346-347 iron ore of 346 relation of 342 Page. Traders mine, geology near 33(^337 Transportation, methods and cost of 495-197 Trap Range. gcoIog>- of 226-226. .385 Trenton limestone, distribution and character of 360.361-362 Two ltarl)ors Bay. geology near 373 Tyler slat^-, correlation of 598, 608, 611 distribution and character of 225-226,227,232-233.(308 relations of 233, 385 U. Ulrich, E. O. , fossils determined by 320 on Iron Mountain 319-320 Unconformities, description of 617-620, 62Mi26 United States, iron reserves of 492 United States Gcologicjil Survey, work of 72-73 l^pham. Lake, description of 443. 453 Upham, W., on glaciation 440.442 Uplands, development of 89-90 districts of, description of 91-98 glaciation of 91 monadnocks on 90 position of ; 89 relief in 89 soil of 91 valleys in 90-91 Upson, geology near 230 V. Valleys, character of 90-91 Van Ilise. C. R., on copper ores 581 on formation of ILmonite 519 on Keweenawan series 402 on metamorphism 551 on phj'siography 116 work of 29. 92-93 Van llise, C. R.. and Bayley, W. S., on Marquette district 96, 10.V106 Van Ilise, C. R., and Irving, R. D., on physiography 103 Veins, copper. Sec Eagle River; Ontonagon; Keweenaw Point. Vermilion district, correlation in 598.599-600.611 exploration in 47. 485 geology of 118-137, 603.61 1. 618 intrusive rocks of 135-136 iron-bearing rocks of 118 iron ores of 137-143. 4Sfi alteration of 554-555 characteristics of 140-141 composition of 139-140 origin of '. 570 plates showing 466.564 production of ; 461 proportion of 462 relations of 506. 507 reserves of 1 4S9 secondary concentration of 141-143 changes in 142-143 figure showing 142 structiue of 137-139 topography of 476 location and area of 32.118 map of 118 mines in, sections of, figures showing 138 mining in, history of 40-41 phosphorus in 143 production of 68, 69 physiography of 92-94. 431 structure of 123-124 figure showing 123 topography of 119 Vermilion Lake, formation of 442 geology near 119.122,129,131,132 gold near 40 Vermilion range, mining on. history of 42 Virginia, greenalite from vicinity of, plate showing 474 Virginia slat e, analyses of 1 73 correlation of 598.611 distribution and character of 159, 172-174. 212-213 phosphorus in 194 relations of 174 INDEX. 641 Volcanic vents, distribution and character of • 411-412 Volcanism, occurrence of *''17 Volume changes. Sec Secondary concentration. Volunteer mine, geology at 259,2fj5 Vulcan formation, alteration of. ■ 354 correlation of 598, 60S, OlO-Gll distribution and character of 302-304,.10G-307, 327, 330, 008 iron ore in 303-304, 335-340, 346, 351 , 460, 503 relationsof. . 305,334,338-339,342-343 structure of 314,338 topographyon . .■. 107,329 Vulcan member, correlation of 59S,G10-G11 distribution and character of 291 , 297-299, 312-318. 321-322. 323 iron ore in 299, 314-315, 323 magnetic phase of 317-318 relations of 313-314 structure of 315-317 W. Wadsworth, M. E., on copper ores 580-582 on iron-bearing rocks 500 on iron ores 570 on Iveweenawan series 397-398, 400-403, 404 Walcott, C. D., on fossils of Lake LSuperior sandstone 340 Wall rock, alteration of 582-585 alteration of. analyses showing 583 "Warping, elTects of. ._ 44s_449 evidence of _ 87,450 Water, circulation of, in rock 186 circulation of, figure showing 186 Waterloo district, correlation in * 598 geology of 3G4 map of 359 Waterloo quartzite, correlation of 598 distribution and character of 364 Waucedah. geology near 333^ 340 Wausau district, geology of 355-357 Wausaugraywacke. correlation of 593 distribution and character of 356 topography on 108 Waushara district , map of _ 359 Wawa tuff, coiTelation of 598 distribution and character of 150, 151-152, 153 petrography of 151-152 Wawa, Lake, geology near 151,152 Weathering, absence of 502 effect of, on concentration _ '. 500 processes of 585-586 relation of, to iron ores 503-505, 516, 539-540 Weidman. S., on age of peneplain , . 88 on drainage 91 Pago. Weidman, S., onglaciation 4.37.439 on iron ores : 564,566,570-571 on physiography . . 98, 107-108, 116 on Wisconsin geology 35S-365, 377 West I*ond, geology near 382, 3a3, 385, 410 West Seagull Lake, geology at 131 West Vulcan mine, section of, figure showing 349 Wewe slate, correlation of 598,605 distribution and character of 252-253,258-259,605 relations of 259, 2C0 topography on 105 Weyerhauser. geology near 357 While Iron Lake, geology near 119, 122 Whitney, J. D.. and Foster. J. W., on Marquette district 96 Whittlesey, Charles, on Mesabi district 42 Willmott, A. B., and Coleman, A. P., on Michipicoten ranges 95 Wilson, A. W. G., on Minnesota geology 367-369,374 Winchell, A. N., on ICeweenawan series 360, 395-410 on nomenclature 395-407 Winchell, N. H., pnglaciation 438 on iron ores 56^-570 on Iveweenawan series 397, 399 on Mesabi district 42 on Minnesota geology 370 on physiography 92. 98, 104 Wirmebago, Lake, formation of 443 Wisconsin, bibliography for „ 77-78 correlation in 598 Clinton ores of 45 copper ores of 580 geology of 355-365, 376-380, 413-414, 429 hematite in'. 45 investigations in 72, 73 iron ores of 461 production of 461 physiography of 97-98, 100, 107-108 figure showing 116 See also particular disfiicts. Wisconsin stage, glaciers of 427, 454 Wolverine mine, history of 36 ores of 576, 577 Woman River, geology near 126,507,555 Wood alcohol, recovery of 48 Woodward, R. W., analyses by.' 404 Wright, F. E., on igneous rocks 410-411,424,582 Wright, F. E , and Larsen, E. S., on quartz crystallization 549 Wurtz, H., on silver minerals 593 Z. Zenith mine, sections of, figures showing 138 Zapffe, Carl, on Cuyuna district 216-224 47517°— VOL 52— 11- -41 O ]■■ ■•v«|i>iij*lu|1 c I Zaurvntfttfi Suronian wflTTj* ginn l j [ u lm» ^ l[v ^IIB is? I, liiiiiiiiQ 9 0. ALCONKIAN U. S. GEOLOGICAL SURVEY MESABI DISTRICT KEWEENAWAN SERIES MINNESOTA ByC.Iv-Leith (Corrected toJamiaryl. 1911) Scale esTfiro Conlour interval 20 fcei Dolum is mean eea le*^„ . BlevBlionorLala-SiirprioriBoOSftjet 190S LEGEN D ALGONKIAN HURONIAN SERIES ARCMEAN LAURENTIAN SERIES KEEWATIN SERIES UPPER HURONIAN (ANIMIKIE GROUP) j KmlHrmss^mlc Plabaac andDuluUi gBbliro Vli-giiUa nlme Bivmblk^^i^ui" MONOGRAPH Lll PLATE ' U. S- GEOLOGICAL SURVEY GEORGE OTiS SMITH. DIRECTOR GEOLOGIC MAP AND SECTIONS OF THE MARQUETTE IRON-r Oi-ij^inallyprepHi-'Ml b\(".W A:im IIihc iiiul W.S.n;iylev.JH!Kl.U.'viscil LnUcccm urClfvelH-nd CUVls lion rmn(>,in,v. Oiivni-lron Mii)inii Compniiy.aiul j»p"l«if*'f Si-al*' snhr.o QUATERNARY CAMBRIAN ALGONKIAN Contour interval 60 feel LEGEND HURONIAN SERIES ERUPTIVE UPPER HURONIAN (ANIMIKlE GROUPJ MIDDLE HURONIAN W/jfJ/////A \ \ ZT nSM liiiiki srhist Coodrith quail/.iii- Negimnt-f fnimaiini. W'rwe slule MONOGRAPH Lll PLATE XVII ; OF THE MARQUEITE IRON-BEARING DISTRICT, MICHIGAN p Mild W..S,naylev.I»M(> Ur•^^se(i in Di^<(jinl)tM- 1.I.910. In iiulude <*xplnrn|]<»ns )livpi'Ii'oil MiijJnjS ) ■oiiipmiy.njid gpologicaJ vi-ork orA.K.SpujjiH ii uikI nUioi-s Scale RrJ^O Contoiii inti'i-\'al !,0 ft-el LEGEND ARCHEAN LAURENTIAN SERIES Jii» HchiHl '%' .^'- LEGEM D ALGONKIAN unONIAN SERIES KEWELNAWAN I?) SERIES , „^„ ^„ „, UinliigiuiiiUK dliUi-' ViOi-iuiii-rm-li"un.ii;5 [..tlilT tfr^Mi s<.-lii«lH iiu-l>uii£g ^HVurackcH, ,, membri- of [v] ! I* UONOGIUPH Lll fUTE ' OUTCROPS mi MIc* tilt* irit£ Mignatlte S Quango qe QusrU con£lomerMo q"c Qustti jchm qm*c Quart; fnici tchist 1 Slile IC2 .Schistote gieenitons «>.. ActinolltB tchiit • ic ActmolitB JChiit ind Conglomerii c ..Conglomerile ct Cirbon>c«ou( ttate CSC Carbantceout ichut c(» Cherty fetrugmouj slate Cg» .Chorty grlinente Schid mo Iron sUte and non ore 'e Fragmont. g»i )c Garnotifp'ous ithiat 6" Gfaywactio gd G'eenatonedior.te v^ji msC.Gatnetiferoui rnit ■/■■; \ ii..,G'o«fi ichiit and iton slal« .■ GreenitonB ,;r Granite g & a Gteenstons and ilate hie Hoinblenda ichitt hmii:. ... Harnblardai Air lES'ooo Contour uiUi-*b1 40(hot ItKK) MONOGRAPH Lll PLATE XXVI SEDIMENTARY ROCKS 1 ?l «s s i f $ * 1 1 » 5 "■ ••MMMiMMa^Mfvm««ai^>ViqMa ii!i 'nnuiiit ;iii ,! 11 ail {ill t ; . » ; iiii iliiilWililiii' iiiiiiliiiiiiliiil^ :>:ii>iti.'^'i'i'i iWl iii w Fi!J ilpilWiiiiliiiiiiii mm M Mmy ItUUIIiiitlilili mm