^^niTTil.l.l.t.l.i L'J'I'I'l^lTflllU!. Mn AaTiciiltiira[»p ;-:. rn:^.o:¥: r; i Du.'^S.e^ iiXj...!^./S.°l.'>sM_ I'RESKSTED BV Digitized by the Internet Archive in 2009 with funding from Boston Library Consortium IVIember Libraries http://www.archive.org/details/crystalfallsironOOclem DEPARTMENT OE THE INTElilOR MONOGRAPHS OP THE United States Geological Survey VOLUME XXXVI WASHINGTON GOVERNMENT PRINTING OFFICE 1899 UNITED STATES GEOLOGICAL SURVEY CHARLES I). WALCOTT, DIRECTOR THE CMSTAL FALLS IRON-IARING DISTRICT OF MICHIGAN liY J. MORGAN CLEMENTS and HENRY LLOYD SMYTH WITH A CHAPTER ON THE STURGEON RIVER TONGUE WILLIAM SHIRLEY BAYLEY AND AN INTRODUCTION BY CHARLES RICHARD VAN HISE WASHINGTON GOVERNMENT PRINTING OFFICE 1899 "A^o»). CONTENTS. Page. Letter of trans.mittal ,. xv Introduction, iiY C'h.\rle8 Riciiaku Van Hise xvii Outline of monograph xxix Pakt I. — The western part tiF the Crystal Falls district, isy J. Morgan Clements. 11 Chapter I. — Introdiction 11 Previous work in the district 13 Mode of work 22 Magnetic observations 24 Chapter II.— Geographical limits, structure and stisatigraphy, and physiography 25 Geographical limits 25 Structure and stratigraphy 25 Physiography 29 Topography 29 Drainage 31 Timber and soil 36 Chapter III. — The Archean 38 Distribution, exposures, and topography - 38 ReLations to overlying formations 39 Petrographical characters 10 Biotite-granite (granitite) 40 Gneissoid biotite-granite, border facies of granite 43 Acid dikes in Archean 45 Basic dikes in Archean 46 Schistose dikes 46 Massive dikes 48 R^sum6 49 Chapter IV. — The Lower Hdronian series 50 Section 1.— The Randville dolomite 50 Distribution, exposures, and topography 50 Petrographical characters 51 Relations to underlying and overlying formations 53 Thickness 53 Section 2.— The Mansfield slate 54 Distribution, exposures, and topography 54 Possible continuation of the Mansfield slate 55 Petrographical characters 56 Gray wacke 56 Clay slate .and phyllite ■■ 57 Origin of clay slate .and phyllite 58 Present composition necessarily dift'erent from that of rock from which derived 58 Analysis of Mausfield slate 59 Comments on analysis 59 Comparison of analysis of Mansfield clay slate with analyses of clays 60 Comparison of analysis of Mansfield clay slate with analyses of other clay slates. .. 61 Siderite-slate, chert, ferruginous chert, and iron ores 62 Rjlations of siderite-slate, ferruginous chert, and ore bodies to clay slate 63 Relations of Mansfield slate to adjacent formations 63 Relations to intrusives 63 Relations to volcanics 64 V VI CONTENTS. Chapter IV. — The Lower Huroniax series — Continued. Section 2. — The Mansfielil slate — Continued. Page. Structure of the Mansfield area 64 Thickness 64 Ore deposits 65 General description of Mansfield mine deposit 66 Relations to surrounding beds 68 Comijosition of ore 68 Microscoiiical character of the ores and associated chert bands 69 Origin of the ore deposits 70 Conditions favorable for ore concentration 72 Exploration 73 Section 3. — The Hemlock formation 73 Distribution, exposures, and topography 73 Thickness 74 Eelatious to adjacent formations 75 Relations to intrusi ves 77 Volcanic origin 78 Classification 79 Acid Tolcanica 80 Acid lavas 80 Khyolite-porphy ry 81 Texture 83 Aporhy olite-porphy ry 87 Schistose acid lavas 87 Acid pyroclastics 94 Basic volcanics 95 Basic lavas 9.5 General characters 95 Nomenclature 93 Metabasalts 98 Nonporphyritic metabasalt 98 Petrographical characters 98 Chemical composition 103 Porphyritic metabasalt 103 Petrographical characters 104 Chemical composition 105 . ' Variolitic metabasalts 108 EUip.soidal structure 112 Origin of ellipsoidal structure 118 Amygdaloidal structure 124 Flattening of amygdaloidal cavities 12l) Alterati on of the basalts 126 Description of some phases of alteration 127 Pyroclastics 135 Eruptive breccia 135 Volcanic sedimentary rocks 136 Coarse tuffs 137 Fine tuffs or ash (dust) beds 142 Relations of tuffs and ash (dust) beds 143 Volcanic conglomerates 143 Schistose pyroclastics 145 The Bone Lake crystalline schists , 148 Distribution 148 Field evidence of connection with the volcanics 149 Petrographical characters 150 CONTENTS. VII CnAPTKR IV. — Thk I.owkk Hukonian skriks — Continuod. Section 3. — The llciulock t'onnation — Continued. I'agn. Normal scdiinontaiit^s (if the Hemlock formation 152 Economic products 1,")3 Building and ornamental stones 153 Road materials 154 CHAPTEU V. — TllK UlTKI! HURHNIAX SICHIISS 155 Distril)Ution, oxposurea, and topograpbj- 1.55 Magnetic lines 1.56 Thickness 1.57 Folding 158 Crystal Falls syncline 158 Time of folding of tbo Upper Huronian 161 Relations to other series 162 Relations to intrusives 16i Correlation 164 Petrograpliical characters 165 Sedimentary rocks 165 Microscopical description of certain of the sedimentaries 169 Igneons rocks 174 Ore deposits 175 History of opening of the district 175 Distribution 175 Western half of sec. 34, T. 46 N., R. 33 W 176 Sec. 20, T. 45 N., R. 33 AV , 176 The Amasa area 177 The Crystal Falls area 178 Character of the ore 180 Relations to adjacent rocks 182 Origin 183 Size of the ore bodies 184 Methods of mining 184 Prospecting 185 Production of ore from the Crystal Falls area 186 Chapter VI. — The Intrusives 187 Order of treatment 188 Age of the iutrusives 188 Relations of folding and the distribution of the intrusives 189 Section 1. — Unrelated intrusives 190 Classification 190 Acid iutrusives 190 Geographical distribution and exposures of granites 190 Biotite-granite 191 Micropegmatites 192 Muscovite-biotite-granite 193 Relations of granites to other intrusives 194 Dynamic action iu granites 194 Contact of granites and sedimentaries 194 Evidence of intrusion 195 Basic intrusives 198 Metadolerite 199 Geograiihical distribution 199 Petrographical characters 199 Maeroscoiiical 199 Microscopical 200 VIII CONTENTS. Chapter VI. — The Intrusives— Continued. .Section I. — Unrelated intrusives — Continued. Basic intrusives — Continued. Metadolerite — Continual. Page. Relations to adjacent rocks 20S Relations to Lower Huronian Mansfield slates 203 Relations to Lower Huronian Hemlock volcanics 204 Relations to Upper Huronian 204 Relations to other intrusives 204 Contact metaniorpbism of Mansfield slates by the dolerite 204 Spilosites 206 Analyses of spilosites 207 Desiuosites 207 Adinoles 208 Analyses of adinoles 208 Comparison of analyses of rtormal Mansfield clay slates and the contact prod- ucts 209 No endomoriiliic effects of dolerite intrusion 211 Metabasalt 211 Ultrabasic intrusives .' 212 Picrite-porphyry (porphyritic limburgite) 212 Geographical distribution and exposures 212 Petrographical characters 212 Gray tremolitized picrite-porphyry 213 Dark serpentinized picrite-porphyry 217 Classification 220 Section II. — A study of a rock series ranging from rocks of intermediate acidity through those of basic composition to ultrabasic kinds 221 Diorite 222 Nomenclature 222 Distribution and exposures 223 Petrographical characters 223 Description of interesting variations 226 Sec. 15, T. 42 N., R.31 W 226 Across river from Crystal Falls 227 Southeiistof Crystal Falls 227 Analysis of diorite 231 Gabbro and norite 233 Petrographiciil characters 233 Description of interesting kinds of gabbro 240 Hornblende-gabbro in sec. 15, T. 42 N., R. 31 W 240 Sees. 15, 22, 28, and 29, T. 42 N., R. 31 W 241 Hornblende-gabbro dikes 243 Bronzite-norite dike 244 Sec.29,T.42N., R.31 W., 1200 N., 200 W 245 Dynamically altered gabbro 247 Relative ages of gabbros 249 Peridotites ^249 Distribution, exposures, and relations 249 Petrographical characters 249 Peridotite varieties 252 Wehrlite 2.53 Amphibole-peridotite 253 Gradations of amphibole-peridotite to wehrlite and olivine-gabbro 254 Process of crystallization 257 Analysis of peridotite 259 CONTENTS. IX CiiAriKK \'I. — TiiK iNTUfsivES — Coutiiiued. Si'ction 11. — A study of a rock series, etc. — Continued. Peridotites — Continued. I'eridotito varieties — Continued. Page. Pcridotito from sec. 22, T. 42N., R.31 W., I'JSION., loOW 26() Relations of peridotites to other rocks 261 Ago of peridotites 262 General observations ou the above series 262 Textural characters of the series 262 Chemical composition of the series 263 Relative ages of rocks of the series 265 Part II. — Thk kastern part of the Crystal Fall.s district, including the P'elch Mountain range, by Henry Lloyd Smyth. Chapter I. — CiEOORAPHicAL limits and physiography 32& Introduction 329 Preliminary sketch of geology 331 Character of surface 331 Drainage - 334 Chapter II. — Magnetic observations in geological mapping 336 Section I. — Introduction 336 Section II. — Description of the magnetic rocks 338 Section III. — Distribution of magnetism in the magnetic rocks 339 Section IV. — Instruments and methods of work .' 341 Section V. — Facts of oliservatiou and general principles 344 (1) Observed deticctions when the strike is north and south and the dip vertical 344 (2) Detleetious of the horizontal needle 345 (3) Deflections of the dip needle. , 347 (4) Horizontal and vertical components when the magnetic rock dips vertically 349 (5) Horizontal and vertical components when the magnetic rock dips at any angle 350 (6) Determination of depth 854 (7) Summary 356 Section VI. — Applications to special cases 3.56 (1) The magnetic rock strikes east or west of north and dips vertically. 357 (2) .The magnetic rock strikes east and west 359 (3) Two parallel magnetic formations 361 Section VII. — The interpretation of more complex structures 366 (1) Pitching synclines 367 (2) Pitching anticlines 370 (3) Formations split by intrusives 371 (4) Summary 372 ■ Chapter III. — The Felch Mountain range 374 Section I. — Position, extent, and previous work 374 Section II. — General sketch of the geology 383 Section III. — The Archean 385 Topography 386 Petrographical characters 387 Section IV. — The Sturgeon quartzite 398 Distribution, exjiosures, and topography 398 Folding and thickness 399 Petrographical characters 401 Section V. — The Randville dolomite 406 Distribution, exposures, and topography 406 Petrographical characters 408 X CONTENTS. Chapter III. — The Felcu Mountain range — Continued. Page. Section VI. — The Manisfield scliists 411 Distribution, exposures, and topography 411 Petrographical characters 412 Section A'll. — The Groveland formation 415 Distribution, exposures, and topography 415 Petrographical characters 417 Section VIII. — The mica-schists and quartzites of the Upper Huronian series 423 Petrographical characters 425 Section IX. — The iutrusives 426 Chapter IV. — The Michigamme Mountain and Fence River areas 427 Section I. — The Archean 428 Section II. — The Sturgeon formation .' 430 Section III.— The Raudville dolomite 431 Distribution and exposures 431 Folding and thicliness 432 Petrograpliical characters 434 Section IV. — The Mansliold formation 437 Distribution, exposures, and topography 438 Folding and thickness 438 Petrographical characters 439 Section V. — The Hemlock formation 440 Distribution, exposures, and topography 440 Folding and thickness 441 Petrographical characters 442 Section VI. — The Groveland formation 446 Distribution, exposures, and topography 446 Folding and thicliness 448 Petrographical characters 448 Chapter V. — The northea.stkrn area and the relations between the Lower Mar- (jtette and the Lower Menominee series 451 Chapter VI. — The Sturgeon River tongue, by William Shirley Bayley 458 Description and boundary of area 4.58 Literature 459 Relations between the sedimentary rocks and the granite-schist complex 461 The IJasement Complex 463 The gneissoid granites 463 The amphibole-sihists 465 Origin of the amphibole-schists 466 The biotitc-schists 467 The intrusive rocks 469 Comparison of the Sturgeon River and the Marquette crystalline series 470 The Algoukian trough 471 Relations between the conglomerate and the dolomite series 472 Relations between the dolomites and conglomerates and the overlying sandstones 473 The conglomerate formation 473 Important exposures 474 Petrographical descriptions 477 The dolomite formation 479 Important exposures 480 Petrographical descriptions 481 Slates and sandstones on the Sturgeon River 481 The igneous rocks 482 The intrusive greenstones 482 Petrographical description 482 The banded greenstones 485 Petrographical description 486 ILLUSTRATIONS. Pago. Plate I. Colored map ahowiug the distribution of pre-C';imbriau and other rocks in the Lake Superior region, and tlie goographical relations of the Crystal Falls district of Michigan to the adjoining Marquette and Menominee districts of Michigan 11 II. Topographical map of the Crj-stal Falls district of Michigan, including a portion of the Marquette district of Michigan In pocket. III. Geological map of the Crystal Falls district of Michigan, including a portion of the Mar([uette district of Michigan In pocket. IV. Portion of geological map of the Menominee iron region, by T. B. Brooks and C. E. Wright 18 V. Generalized sections to illustrate the stratigraphy and structure of the northwestern part of the Crystal Falls district of Michigan " 28 VI. Generalized sections to illustrate the stratigraphy and structure of the southern part of the Crystal Falls di,striet of Jlichigan 28 VII. Generalized columnar section 30 VIII. Map of a portion of the Crystal Falls district, showing iu detail the glacial topog- raphy and illustrating the development of the Deer Kiver 32 IX. Sketch of the Mansfield mine as it was before it caved iu, iu 1893 66 X. A, Reproduction of the weathered surface of a variolite; B, Reproduction of the polished surface of a variolite 110 XI. Colored reproduction of an ellipsoid, with matrix, from ', Quartz phenocryst with rhombo- hedral parting 268 XX. yl, Micropoikilitic rhyolite-porphyry; /?, Micropoikilitic quartz-porphyry 270 XXI. J, Very fine-grained micropoikilitic rhyolite-porphyry viewed without analyzer; B, Very fine-grained micropoikilitic rhyolite-porphyry viewed with analyzer 272 XXII. A, Perlitic parting in aporhyolite; B, Perlitic parting in aporhyolite 274 XXIII. A, Schistose rhyolite-porphyry; B, Aporhyolite breccia 276 XXIV. .4, Schistose rhyolite-porphyry ; /?, The same viewed between crossed nicols 278 XXV. ^,Amygdaloidal texture of basalt; i>, Amygdaloidal vitreous basalt 280 XXVI. A, Amygdaloidal vitreous basalt ; B, Amygdaloidal vitreous basalt showing sheaf-like aggregates of feldspar 282 XXVII. ^, Reproduction in colors of amygdaloidal basalt; i>, Pseudo-amygdaloidal matrix of ellipsoidal basalt ; C, Water-deposited pyroclastic 284 XXVIII. A, Fine-grained basalt with well-developed igneous texture; B, Illustration of the obliteration of the igneous texture of a basalt by secondary products when viewed between crossed nicols 286 XI XII ILLUSTRATIONS. XXIX. J, Basalt showing characteristic texture in ordinary light; B, Basalt showing obliteration of texture between crossed nicols 288 XXX. A, Basalt showing iu ordinary light a distinctly amygdaloidal texture; B, The same basalt with its amygdaloidal texture obliterated when viewed between crossed nicols 290 XXXI. A, Basalt affected by calcification process; B, Basalt affected by calcification process viewed between crossed nicols 292 XXXII. A, Illustration of perlitic parting in a fragment from a basaltic tuff; B, Sickle- shaped bodies in volcanic tuff 294 XXXIII. A, Water-deposited sand; B, Gradation iu water- deposited volcanic sediment 296 XXXIV. A, Contact product of granite ; B, Brecciated matrix between ellipsoids 298 XXXV. .4, Contact between granite and a metamorphosed sedimentary ; /?, Contact between granite and a metamorphosed sedimentary viewed between crossed nicols 300 XXXVI. A, A variety of spilosite with white spots ; B, A variety of spilosite with white spots viewed between crossed nicols 302 XXXVII. -J, Normal spilosite or spotted contact product; .B, Normal spilosite of somewhat different character 304 XXXVIII. J, Passage of spilosite into demosite; i?, Occurrence and alteration of bronzite in bronzite-norite 306 XXXIX. A, Biotite-granite viewed between crossed nicols ; B, Mica-diorite viewed between crossed nicols 308 XL. A, Quartz-mica-diorite-porphyry ; B, Quartz-mica-diorite-porphyry viewed between crossed nicols 310 XLI. J, Porjihyritic poikilitic hornblende gabbro; i?, Poikilitic hornblende gabbro 312 XLII. ll, Moderately fine-grained hornblende gabbro showing parallel texture; £, Mod- erately fine-grained hornblende gabbro showing parallel texture viewed between crossed nicols 314 XLIII. A, Normal granular hornblende gabbro; B, Schistose hornblende gabbro viewed between crossed nicols 316 XLIV. J, Moderately fine-grained hornblende gabbro; iJ, bronzite-norite 318 XLV. J, Bronzite-norite-porphyry ; i>, Feldspathic wehrlite 320 XLVI. J, Feldspathic wehrlite viewed between crossed nicols; Bj Feldspathic wehrlite 322 XLVII. Relations of magnetic beds to variation and dip 352 XLVIII. Relations of magnetic beds to variation and dip 362 XLIX. Geological map of the Felch Mountain Range 374 L. Geological map of a portion of the Crystal Falls district 450 LI. Geological map of the Sturgeon River tongue 458 LII. Map of exposures in sec. 7 and portions of sees. 8, 17, and 18, T. 42 N., R. 28 W 474 LIII. Schist conglomerate from dam of Sturgeon River 476 Fig. 1. Reproduction of a portion of the geological map of the Upper Peninsula of Michigan, by William A. Burt. 1846 15 2. Enlarged reproduction of a portion of a map of the Lake Superior land district, by Foster and Whitney 17 3. Enlarged reproduction of a portion of a geological map of the Upper Peninsula of Michigan, by Rominger, Brooks, and Pumpelly, 1873 18 4. Granite-porphyry with inclusions of gneissoid granite 45 5. Illustration of the effect on the topography of the differential erosion of basic dikes and granite 46 6. Concentric cracks formed by the caving in of the Mansfield mine 65 7. Sketch of the surface of the outcrop of an ellipsoidal basalt, showing the general char- acter of the ellipsoids and matrix 112 8. Sketch showing the concentration of the amygdaloidal cavities on one side of an ellipsoid, this side prob.ably representing the side nearest the surface of the flow 113 9. Ellipsoids with sets of parallel lines cutting each other at an angle 114 10. Reproduction of illustration of aa lava, after Dana 120 11. Profile section illustrating results of diamond-drill work 177 ILLUSTRATIONS. XIII Page. Fi(i.l2. Sketch illustrating CDiitortion of Uppor Huron I an strata 179 13. Skotili show i mi; chaiigo of strike of Upper lliironiaii beds, T.DI>r<; OF TIIK ARCHKAN AND IIURONIAN SERIES. The Crystal F.iUs district liad now been an area of deposition for a very lonj;- time, and a great tliickness of sediments had accumnlated. A profound pliysical revohition next occurred, the greatest since Archean time. The region was raised above the sea and was folded in a most com- plex manner. As a consecpience, the more conspicuous folds vary from a north-south to an east-west direction. The closer folds in the northeastern piu-t of the area are nearly north-south. In the centi-al part of the area the closer folds strike northwest-southeast. In the eastern and southeastern parts of the district the closer folds are nearly east-west. All of these folds, however, have steep pitches. It therefore follows that the region was subjected to great compressive stresses in all directions tangential to the surface of the earth, and that the yielding was mainly in one direction * here and in another there, although on every fold there is evidence of yield- ing in two directions at right angles to each other. Some of the folds are very close, as in the case of the Huronian area between the two Archean ovals in the northeastern part of the district, and in the Felcli Mountain range. In other areas — as, for instance, in the Crystal Falls syncline — - the major fold is somewhat open. However, upon the open folds are super- imposed folds of a higher order, so that the detail structure is very compli- cated. So far as known, the district has nowhere been faulted. Subsequent to or during the late stage of this time of folding there was a period of great igneous activity, probably contemporaneous with the Keweenawan. At this tin:ie there were introduced into both the Lower and the Upper Huronian rocks vast bosses and numerous dikes. The intrusives vary from those of an ultrabasic character, such as peridotites, tlii'ough those of a basic character, such as gabbros and dolerites, to those of an acid char- acter, such as granites. These intrusives, while altered by metasomatic changes, do not show marked evidence of dynamic metamorphism — there- fore the conclusion that they were introduced later than the period of intense folding, already described. A few illustrations are mentioned. The Archean and other great massifs are less profoundly altered than are the softer and weaker deposits of the Huronian. In these more rigid formations, such as the granites and quartzites, all phases of alteration by granulation and recrystallization are XXIV INTRODUCTION. beautifully exliibited. The Sturgeon River area affords one of the best- known illustrations of a schistose conglomerate the matrix of which has com-- pletely recrystallized and, therefore, can not be discriminated from a gneiss of igneous origin, but contains numerous pebbles and bowlders flattened in the plane of schistosity. The great Hemlock volcanic formation varies from rocks which ai"e altered chiefly by metasomatic change to those which have become com- plete crystalline schists containing no vestige, either macroscopically or microscopically, of a texture or structure which may be interpreted as Igneous. SUBSEQUENT HISTORY. After the introduction of the intrusives the region was subjected to vast denudation, which reduced it approximately to its present configuration This period of erosion continued until late Cambrian time, when the sea again overrode the district and deposited upon the older rocks Upper Cam- brian sediments. Long after the deposition of the Cambrian, and perhaps later Paleozoic rocks, the district was again raised above the sea, and the major part of the Cambrian deposits have been removed, although they are found in patches throughout much of the district, and occur as a continu- ous sheet just east of the area discussed. The district niay have again been submerged in Cretaceous time ; but if so, the deposits formed were removed after the area finall}'' emerged from the sea. Since Cretaceous time the region seems to have been one of erosion. During the Pleistocene period a thick mantle of glacial deposits was spread over the entire district. Since Pleistocene time erosion has advanced far enough to uncover the rocks here and there. METAMORPHISM. The folding varied in its closeness in different parts of the district. Moreover, the formations are of very variable character, including a great variety of sediments and of igneous rocks. The formations, therefore, xarj greatly in their capacity to resist stresses. It thus follows that during the folding process certain formations yielded to a much greater degree than others. The amount of contained water and other conditions were also variable. As a result of these many variable factors, it is one of the most characteristic features of the district that there are to be found neai'ly all INTRODUCTION. XXV varieties <^f niotamorphism in various stages of advanceineiit. The working out of the details of the transformations of the different kinds of rocks during their processes of inetamorphism is one of the chief scientific results which has come from a study of the district. CORRELATIOX. In order to compare the succession in the Crystal Falls district with that in the adjacent Marquette and Menominee districts, the descending pre- Cambrian succession in each of the tlii'ee districts is here given in parallel columns, the formations which are thought to be equivalent being placed opposite one another : Descending succession of formations in the Marquettef Crystal Falls, and Menominee districts. MARQUETTE DISTRICT. Upper Marquette. (1) Micbigamme formation, bearing a sbort distanre above its base an iron-bearing horizon, and being replaced in much of the districtbj the Clarksburg vol- canic formation. (2) Ishpeming formation, being composed of the Goodrich quartzite in the eastern part of the district and of the Goodrich quartzite and the Bijiki schists in the western part of the district. ITjicon/ormity. Lower Marquette. <1) Negaunee iron formation. 1,000 to 1,500 feet. (2) Siamo slate, in places including inter- stratified amygdaloida, 200 to 625 feet thick. (3) Ajibik (luartzite, 7l)0 to 900 feet. (4) "Wewe slate, 550 to 1,050 feet. (5) Kona dolomite, 550 to 1,375 feet. (6) Meanard quartzite, 100 to 670 feet. CRYSTAL FALLS DISTRICT. Upper Huronian. (1) Michigamme formation, bearing^ a short distance above its base an iron-bearing horizon. (2) Quartzite in eastern part of dis- trict. Unconformity. Lower Huronian. (1) The Groveland formation, about 500 feet thick. (2) Hemlock volcanic formation, 1,000 to 10,000 feet thick. In western part of district also occupies the place of (1) and f3). (3) Mansfield formation, 100 to 1,900 feet thick. (4) Kandville dolomite, 500 to 1,500 feet thick. (5) Sturgeon quartzite, 100 to l.UOO feet thick. MENOMINEE DISTRICT. Upper Menominee. (1) Great Slate formation Unconformity . Lower Menominee. (1) Vulcan iron formation con- taining slates. (2) Antoine dolomite. (3) Sturgeon quartzite. Unconformity. Archeaii. Unconformity. Archean. Unconformity. A rchean. From the tliree columns it appears that the equivalents in the different districts can be made out with a considerable degree of certainty. There are, however, various differences, due to several causes. For Upper Hui'onian time, omitting the Clarksburg formation, the sue- XXVI INTEODUCTION. cession in the Marquette, Crystal Falls, and Menominee districts is substan- tially the same. The Clarsksburg formation in the Marquette district may be omitted from consideration, because it is volcanic and replaces in jjart the Michigamme and Ishpeming formations. The Upper Huroniau was a great period of slate and grit deposition. The chief difference which appears between the Menominee district and the other two districts is that in the former no iron ores have been found in the Ujjper Huronian within the district proper, although such rocks occur a short distance to the west, at Commonwealth and Florence, in Wisconsin. The succession for the Lower Huronian in the three districts can be paralleled with a high degree of probability. The chief differences are due to the disturbance of the great volcanic outburst in the western part of the Crystal Falls district and to the uneven surface of the Archean land at the beginning of Lower Huronian time. As a consequence of the latter, the waters did not reach the western part of the Marquette district and the northeastern part of the Crystal Falls district as early as the eastern part of the Marquette district, the central part of the Crystal Falls district, and the Menominee district. The transgression of the Lower Huronian sea for the region covered in these three districts was therefore from the southeast toward the northwest. The Negaunee iron formation of the Marquette district is equivalent to the Grdveland iron formation of the Crystal Falls district and the Vulcan iron formation of the Menominee district. The Siamo slate and the Ajibik quartzite of the Marquette district are approximately equivalent to the Hemlock volcanic formation in much of the Crystal Falls district, but in places where the latter formation displaces the j\Iansfield formation they are equivalent to only a part of the Hemlock volcanic formation. The Wewe slate of the Marquette district is equivalent in the western part of the Crystal Falls distinct to a part of the Hemlock volcanic formation, and in the southeastern part of the district is probably equivalent to a part of the Randville dolomite. It appears that the Siamo slate, Ajibik quartzite, and Wewe slate of the Marquette district, and the Mansfield and Hemlock formation of the Crystal Falls district, are equiv- alent to a part of the Antoine dolomite of the Menominee district. The great dolomite formation occurring in all of the districts is sup- posed to be equivalent, except that, as just explained, the de^iosition of INTKODUGTION. XXVII liiiu'stone contiimod longer in the southejistern part of the Crystal Falls district and in the Menominee district than in the remainder of the region. The absence of the limestone and lower formations in the* western two- thirds of the Marquette district and the northeastern part of the Crystal Falls district is explained by the fact that during early Algonkian time this part of the region was not submerged. The Mesnard quartzite of the Mar- quette district and the Sturgeon quartzite of the Crystal Falls and Menom- inee districts stand opposite each other. From the foregoing it is apparent that the tlu-ee districts together pre- sent a most interesting and complex structural problem. While there is sufficient similarity in the formations for one to feel considerable assurance of their general equivalence in the different districts, it is certain that the formations of similar kind did not begin and end at the same time. Slore- over, there are remarkable lateral transitions in sedimentation, as a result of the uneven surface of the Archean at the beginning of Algonkian time and because of volcanic outbursts. As a resitlt of the first of these conditions, it is necessary to equate fragmental formations which occur in the central and western parts of the Marquette district and the northeastern part of the Crystal Falls district, with nonfragmental limestones in the area to the east and south. Consequent upon the Upper Huronian volcanic outbursts in the Marquette district, the Michigamme and Ishpeming formations are largely replaced by the Clarksburg volcanics. Similar outbursts in the western part of the Crystal Falls district in Lower Huronian time placed volcanic rocks for this part of the district opposite the Mansfield slate and the Groveland iron formation. The foregoing relations, combined with the great variety and complexity of the sediments of the district, the presence of many forms of contempora- neous volcanic deposits, the intrusion of the widest ^•ariety of igneous rocks of various ages from Archean to later Algonkian time, and the complicated folding and metamorphism to which the district has been subjected, will readily convince one that the working c)ut of the detail structure of the district by Messrs. Clements, Sn^yth, Bayle}', ]\Ierriam, and others has not been accomplished without most painstaking and laborious work, especially as the region is covered by timber or brush and is overspread by a mantle of glacial deposits. OUTLINE OF THIS MONOGRAPH, Part I. Chapter I. The Crystal Falls district is situated on the Upper Peninsula of Michigan, and foniis a connecting link between the Marquette and Menominee districts of Michigan. A history of the previous work in the district, accompanied by a summary of the literature, is given, and there is reproduced a series of maps which indicate the development of knowledge concerning the distribution . of the rocks and their structural relations. As explanatory of the locations given, the mode of work is described and the object and method of taking magnetic observations is briefly outlined. Chapter II treats of the geographical limits, structure, stratigraphy, and physiography. The portion of the district here described includes approximately 540 square miles. Structurally it is closely related to the Maniuette district; the essential features being a northwest-southeast set of folds, with a superimposed series trending northeast-southwest. The oldest rocks belong to the Archean. They cover an oval area which is surrounded by the Algonkian rocks represented by the Lower and Upper Huroniau series. The Archean and Algonkian are overlain with strong uncon- formity by rocks of the Cambrian division of the Paleozoic. The drift deposits of the Quaternary are everywhere present. Only the pre-Paleozoic rocks, however, are discussed. The most noticeable topography is that of the drift, which in places is seen to be superimposed upon pre-Pleistocene topography. The maximum elevation is 1,900 feet above sea level, and the minimum 1,250 feet. The conclusion is reached that this portion of Michigan before Glacial times had been reduced to the condition of an approximate peneplain. The drainage is chiefly by a few large streams which flow into Green Bay of Lake Michigan. A small part is drained by streams flowing into Lake Superior. In portions of the area the drainage has reached an advanced stage, in other portions it is very youthful. The development of the drainage is illustrated in the ease of the Deer River. The timber and soil vary much in character. Chapter III treats of the Archean. The rocks of this nge form an elliptical core, following the axis of a northwest-southeast trending anticline. Exposures are few because of the superimposed drift. The Archean is overlain uncouformably by Algonkian sediments derived from the granite, and there is absence of contact action. These facts indicate that it was the floor upon which the over- lying sediments were deposited. Petrographically it consists chiefly of biotite-granite. On the periphery of the area a biotite gneissoid granite is very well developed. Some of this at least is of dynamic origin. The Archean is cut by acid and basic dikes which are now both schistose and massive. Chapter IV treats of the Lower Huroniau series. This series is subdivided into the following formations, from the base upward: The Raudville dolomite, the MausHeld slate, and the Hemlock volcanics. Section I. The Randville dolomite is poorly exposed. It consists of quartzose dolomite, grading down into quartz-schist and recomposed granite. It has evidently been derived partly from the granite, and is cousiderably younger than that. Its relations to the overlying formations were not observed. The thickness could not here be determined, but, in the area studied by Smyth, its maxi- mum is about 1,500 feet. XXX OUTLmE OF THIS MONOGRAPH. Section II. The Mansfield slate is best exposed in the vicinity of the town of the same name. It here occupies a valley, through which flows the Michigamme River. Petrographically this formation includes graywatkes, clay slate, phyllite, siderite-slate, chert, ferruginous chert, and iron ores, with various metamorphic products derived from them. The slate predominates. The Maualield slates are intruded by basic igneous rocks, which underlie them. They are overlain by volcanics, which contain fragments of the slates, and are hence younger than they. The Mans6eld slates strike north and south and dip on an average 80^ to the west. They represent the limb of a westward- dipping monocline. The maximum thickness of the belt is 1,900 feet. Followed south away from the point of maximum thickness it rapidly thins out and disappears. In the slate but a single ore body of commercial importance has been found. This is exploited by the one Bessemer ore-producing mine of the district. The ore body is presumed to have resulted from the alteration of a siderite, and the concentration in a favorable position of the iron from the portions of the ferruginous beds removed by erosion. A possible continuation of the Mansfield slate is suggested by the occurrence of small outcrops of somewhat similar slates about 5 miles slightly to the west of north. Section III. The Hemlock formation consists almost exclusively of volcanic rocks, both b.asic and acid, with crystalline schists derived from them. Sedimentary rocks play a very unimportant rule. Exposures are numerous west of the belts of previously described rocks, and where erosion has removed the drift the formation has a marked influence ou the topography. The thickness is estimated from the dip to reach 23,000 feet, but this is probably illusory because of reduplication due to folding. In the northern portion of the district the formation overlies the Kona dolomite. In the southern portion it overlies conformably the Mansfield slate. It is probable that volcanic activity began in the north and moved south, and that some of the volcanics to the north are contemporaneous with the Mansfield slates. The volcanics are cut by a few acid dikes. Basic dikes forming enormous bosses of basic rock are of freiiuent occurrence. The volcanic origin of the major portion of this formation is perfectly clear. Some of the volcanics 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. The Hemlock volcanics are divided into igneous and sedimentary rocks. Under the igneous rocks there are described both acid and basic lavas and pyroclastics. Under the sedimen- tary rocks there are described volcanic sediments, both of eolian and water-deposited character. By extreme metamorphism crystalline schists have been produced from l)Oth igneous and sedimentary rocks. The acid volcanics include rhyoliteporphyries and aporhyolite-porphyry. The rhyolite- porphyry shows interesting micropoikilitic textural characters. Some of the porphyries have been rendered schistose by pressure. Acid pyroclastics are scarce and were derived from the aporhyolite. The basic lavas correspond to the modern basalts. They are much altered. To indicate these facts and at the same time show their correspondence to the Tertiary aud recent basalts, they are called " metabasalts." The basic lavas include noni^orphyritic, porphyritic, aud variolitie types. A columnar structure was not observed, but an ellipsoidal structure is very common. This structure is described ill detail and the couclusiou reached that basalts possessing this structure were originally very viscous and correspond to the modern aa lavas. The amygdaloidal structure, which is almost univer- sally jiresent in the volcanics, is described aud illustrated. The alteration of the basalts is discussed aud speci.al cases described. As result of this alteration the textural as well as the mineralogical characters may be completely changed, and the volcanic origin of the resulting rocks could not be determined but for their association. In the zone of weathering, calcification is the controlling alteration jirocess. In rocks more deeply buried, silicification is the process which predominates. The pyroclastics comprise eruptive breccia, including thereunder friction breccias and flow breccias, and volcanic sedimentary rocks. The eolian deposits, which arc described as tuft's, grade from fine dust deposits up into tliose in which the fragments are bowlders. The water-deposited volcanic fragmentals are known as volcanic conglomerates, and likewise grade from those of which the particles are of minute size into those of which the fragments are of very large size. At various places clastic rocks occur which are now schistose, and whose exact mode of origin — that is, whether eolian or water deposited — could not be determined. Normal sediments consisting of slate with limestone OUTLINE OF THIS MONOGRAPH. XXXI liMises form u loiitiiulai' deposit in the volcuiiics uiiar tlie top of tin; formation. 'I'licy are inHijjiiiH- cant in quantity. Under the Uouo Lalvu crystalline schists there' arc included rocks of completely crystalline character, but which by field and microscopical study have beeu connected with the volcanics and are considered to have been derived from rocks similar in nature to them. The rocks composiun' the Hemlock formation arc little likely, owing to their somber color, to be much used for building or ornamental purposes. They offer, however, an inexhaustible supply of the best cjuality of road-Unildiug material. Chapter V treats of the I'pper Hurouian series. This series is connected in the northern part of the area described with the Upper Maniuette series of the ad.joiniug Marquette district, and is considered to correspond stratigraphically to the Upper Maniuette. Owing lo lack of exposures and to the intricacy of folding, the series could not be subdivided. It covers a great area surrounding the Hemlock formation, and extends beyond the limits of the map. The exposures are scanty. It inlluences the topography only in a very general way, being for the most part heavily drift covered. Its thickness could not be estimated. The rocks of this series wrap around the subjacent Lower Huronian series. The line between them is uudulatory. The indentations in the Lower Huronian represent minor cross eyuclines, and the jirotuberauees represent minor cross anticlines. The most prominent fold of the series is known as the Crystal Falls syncline. The strike of the axis of this syncline is in general to the south of west, and pitches in the same direction. The syncline is not simple, but has minor rolls, as shown in various exposures. This folding has beeu productive of extensive reibungsbreccias. The folding occurred immediately preceding the deposit of the Keweenawau series iu other parts of the Lake Superior region. The Upper Huroniau is penetrated by iutrusive rocks of acid, iutermediate, and basic composition. The rocks constituting the series may be divided into those of sedimentary and those of igneous origin. The sedimentary rocks are graywackes, ferruginous graywackes, micaceous, carbonaceous, and ferruginous clay slates and their crystalline derivatives, and thinly laminated cherty sideriite-slate, ferruginous chert, aud iron ores. In two places rocks of conglomeratic nature occur. The extensive folding which the series has undergone, coupled with the intrusions of the igneous rocks, has produced crystalline schists from the muds and grits. These are extensively developed iu the southern portion of the district in the vicinity of the Paint and Jlichigamme rivers. The igneous rocks which have penetrated the Upper Hurouian subsequent to the folding which affected it are not described under the series. Interlaminated with the crystalline schists there are, however, certain rocks, now perfectly crystalline hornblende- gneisses, which are presumed to have resulted from the metamorphism of either basic intrusive sheets or interljedded flows. The economic development of the district followed that of the adjoining Menominee district. The exploited ore deposits occur near Amasa, and iu the vicinity of Crystal Falls. The ore is hematite and limonite. The grade is non-Bessemer. The ore is associated with white and reddish chert, and this formation lies between carbonaceous slates. The ore bodies in general pitch to the west at varying angles, which correspond to the pitch of the axes of the syncliues iu which they occur. These minor folds in turn correspond to the western pitch of the main Crystal Falls synclinorium. The ore bodies are concentrates in synclinal troughs, as described by Van Hise for other Huronian districts. The mining is now for the most part undergouud, and is carried on iu open stopes. The greatest shipment of ore from the area, including the Lower Hurouian Mansfield mine, was .586,970 tons in 1892. The total shipment for 1898 reached 325,814 long tons. Chapter VI treats of the intrusives. There is here included a varied assortment of rocks, exhibiting iu common intrusive relations to the sedimentary and igneous rocks. The term " intrusive" is not to be interpreted as synonymous with the "dike rocks" of some authors. These rocks are never found to penetrate the Cambrian rocks, and have not been affected by the folding which meta- morphosed the Upper Huroniau sediments. They are presumed to be of Keweenawau age. The intru- sives have in some cases beeu injected along the axes of the folds, these representing the lines of greatest shattering, and hence least resistance. XXXII OUTLINE OF THIS MONOGRAPH. In Section I there are described a number of intrusive roclis which can not be connected genet- ically with one another. These comprise ordinary biotite-granites, with micropegmatitic varieties, muscovite-biotite-granite, metadolerite, metabasalt, and picrite-porphyry. AVhere the granites have intruded the Upper Huronian series they have contorted the strata, and include and metamorphose the rocks, producing muscovite-biotite-gneiss, and staurolitiferous and garnetiferous mica-schists. The metadolerites possess no special points of interest in themselves, but where they have intruded the Mansfield slates they have caused interesting exomorphism. The slates are converted into adinoles, spilosite8,and desraosites. Chemical analyses indicate the chief change which has taken place in the production of these rocks from the clay slate to have been in the increase of silica and soda as the contact is approached. There seems thus to have been a direct transference of sodiuip and silicon from the igneous rock to the sedimentary. The metabasalt dikes are of little interest. The ultrabasic picrite-porphyries are extremely altered. This alteration has produced in one case chiefly tremolite and in another serpentine. One of these serpentine picrite-porphyries is polar-mag- netic. The picrite-porphyries .ire presumed to have contained a vitreous base, and correspond to the modern Tertiary limburgite. In Section II there is given a study of a series of rocks varying from those of intermediate acidity through those of basic composition to ultrabasic kinds. The exposures of these rocks are found iu an area underlain by the Upper Huronian series, extending from Crystal Falls southeast to and a short distance beyond the Michigamme River. The prevailing rocks are, on the one hand, diorites of intermediate acidity, ranging to more acid rocks, tonalites, quartz-mica diorites, and granite. On the other hand, we have hornblende-gabbros, gabbros, norites, and, lastly, peridotites of varying luineralogical character. Only those kinds of rocks of which analyses have been obtained — mica-diorite, hornblende-gabbro, norite, and wehrlite — are discussed. The diorites are holocry- stalline rocks of medium to coarse grain. In texture they show some variation from those which are hypidiomorphic granular to those in which the texture is imi)erfectly ophitic. As facies of the dioritic magma there are described diorite, mica-diorite, quartz-niica-diorite, tonalite, and plagioclase-bearing granite. A quartz-mica-diorite-porphyry occurs in narrow dikes cutting the mica-diorite. Analysis of the mica-diorite shows it to stand upon the border between the lime-soda feldspar rocks and the orthoclase rocks. Thegabbros and norites arehulocrystalline rocks of moderately iine to coarse grain. They show considerable v.ariation in texture. The hypidiomorphic granular texture predominates, but some few show a good parallel texture. Others are noticeably porphyritic, a few have poikilitic textures, and less commonly there is an approach to the ophitic texture. Hornblende and labradorite is the must common mineral association, giving typical hornblende-gabbro. A monoclinic pyroxene at times becomes .abundant, giving a transition to the normal gabbro. Bronzite at times is the promi- nent bisilicate constituent of these rocks, giving bronzite-norite. A bronzite-norite-porphyry also occurs. Locally the hornblende-gabbro has been crushed, and there is produced therefrom a schistose rock which represents a transition to a hornblendo-gueiss. Of these rocks the hornblende-gabbro was first formed. It was intruded by normal gabbro, and both of these types were then cut by dikes of bronzite-norite and bronzite-norite-porphyry. The rocks included under the peridotites show con- siderable mineralogical variation. There is produced an amphibole-peridotite, which, when augite becomes predominant, grades to wehrlite. This in its turn becomes feldspathic, and indicates a tran- sition to olivine-gabbro. The amphibole-peridotite also becomes feldspathic and quartzitic, indicat- ing a transition toward diorite. When the order of crystallization of the minerals composing the granular rocks of the entire series described above is considered, it is seen to have been as follows: Bronzite is apparently the oldest. The olivine and monoclinic pyroxene come next and are of essen- tially the same age. Mica and hornblende follow, and are contemporaneous. Then comes plagio- clase, orthoclase, and qu.artz. A consideration of the chemical analyses of the rocks above described shows them to belong to a series ranging from a diorite, on the one hand, to hornblende-gabbro and norite and to peridotite on the other. On the acid side of the series variations are shown microscopic- ally, but of these rocks chemical analyses have not been obtained. It is not possible to state which of these rocks most nearly resembles in its composition the original magma of which the di£ferent OUTLINE OF THIS MONOGRAPH. XXXIII types ropreseut tlm differentiation products. Tlie lioruMende-jjiilibro i« that typn whicli apparently first loaclied its present geological position. It was followed in the acid part of the series liy the diorite, which in its turn was succeeded by the diorite-porphyry. Along the basic scries hornbleude- gabbro was succeeded by gabbro, followed by the bronzite-norite and the peridotite. In general the forces of differentiation have been toward increasing acidity and increasing basicity. Pakt II. Chapter I treats of geographical limits and physiography. The geographical limits of the area described arc given and a brief statement made concerning the eouditious under which the work was done. In the preliminary sketch of the geology the rocks represented are stated to range in ago from Archeau to early Paleozoic. In that part of the district north and west of the Michigamme River the Archean is exposed in several regularly outlined oval areas from 10 to 12 miles long and from 2 to 6 miles wide. The intervjils Ijetweeu these ovals are occupied by highly tilted metamorphosed sedimen- tary and igneous rocks of Algonkian age. In the southern and eastern portions of the district the edges of the tilted rocks are covered by the gently dipping Cambrian sandstone. Xevertheless, field work shows that the distribution of the Archeau in ovals, which is so characteristic for the areas north, also holds here. The chief surface feature is a rolling plain, which slopes gently to the southeast, and upon which is superimposed the glacial drift with its characteristic topograjihical features, multitudinous in variety and detail, but insignificant in relief. While the details of the topography are mainly glacial, the broader features have often clearly been determined by the presence of the more resistant Archeau and Algonkian rocks. The drainage is to the southeast, mainly into Lake Michigan, through the Michigamme aud the Sturgeon rivers. Details of the drainage have been determined by the distribution of the rocks. It is interesting to note that the Michigamme flows along the eastern edge of its drainage basin, having no eastern tributaries. Chapter II treats of magnetic observations in geological mapping. Certain of the rocks occurring in the Crystal Falls district contain magnetite in such quantity that they have a marked influence on the magnetic needle. Advantage is taken of this fact in the mapping of the rocks where exposures are w.auting. The instruments and methods of work used in making magnetic oI)Servatious are described. Facts of observation are mentioned, aud general principles are laid down. Applica- tion of these principles to special cases is then considered, and finally a description of the method used in the interpretation of complex structures follows. Chapter III. In Section I the position, extent, aud previous work done in the Feloh Mountain range is described. An abstract of the literature covering the area is given. Section II contains a general sketch of the geology of this range. The rocks range from Archean to early Paleozoic. These last are not considered for the jiresent. The Archeau is distributed in areas which represent the cores of large arches formed over the whole region by mountain building energy, and subsequently truncated by deep Cambrian denudation. The rocks, chiefly of sedimentary origin intermediate between the Archean and Paleozoic, to which the name Algonkian is applied, occupy a nari-ow strip ranging from a mile aud a half to less than a mile in width, and extending east and west for a distance of over 13 miles. This strip constitutes the Felch Mountain range. It is bordered on the north and south by the Archean. The lowest part of the Algonkian occupies parallel zones nest to the Archean lioth on the north aud on the south, aud is succeeded toward the interior of the strip by the younger members. The general structure therefore is synclinal, Iiut is not simple. The strip contains two or more synclines separated by anticlines. They have likewise been affected by cross folds, which give a different pitch to the axes of the east and west folds. The structure is also compli- cated by faulting. The Algonkian is divided into two series separated by an unconformity. In the first occur, from the base upward, the .Sturgeon quartzite, the Eaudville dolomite, the Mansfield schist, and the Groveland iron formations. Above these follows a youuger series which is undivided. Section III treats of the Archean. This limits the Algonkian rocks on the north and south, and is very well exposed. The topography is very rough. Usually, but not always, a topographical MON XXXVI III XXXIV OUTLINE OF THIS MONOGllAPH, (lepressiou, occupied by a swamp or by a stream, exists along the coutact betweeu the Archeau and the Algonkian. Petrographically the Archean consists of (1) granites or granitic gneisses, (2) gneisses, (3) mica-schists, (4) hornblende-gneisses, or amphibolites. The granites possess the usual characters of such rooks. The gneissoid members of this division are merely crushed granites, and are connected with the massive rocks by indistinguishable gradations. The gneisses are banded laminated rocks, the minerals of which have crystallized in parallel elongated forms. Subsequent to crystallization they have been acted on by great stresses. The mica-schists are even, medium- grained rocks, with generally well-developed schistosity. The original character of these schists is wholly indeterminable. Their relationship with the granites and gneisses is perhaps a reason for regarding them as derived from originally massive granites by dynamic metamorphism. The ' hornblende-gneisses, or amphibolites, are black or dark-green rocks, which are universally foliated. They occur in narrow bands in the granites and gneisses. Their boundaries are sharp and frequently cut the foliation of the amphibolites and of the gneisses. The field relation as well as the composi- tion of the amphibolites leads to the conclusion that they are old dikes of basic rocks which have been metamorphosed and recrystallized. Section IV treats of the Sturgeon quartzite. The Sturgeon formation is the most widespread member of the Algonkian series in the Felch Mountain range. It occurs in two parallel zones of varying width, immediately adjoining the Archeau to the north and to the south, except when displaced from thi.s position by faults. It is fairly well exposed. It frequently forms distinct lineal ridges, which, with but few exceptions, seldom rise to the mean altitude of the adjoining Archean. Owing to the comjileteness of recrystallizatiou, the original sedimentary features have almost been obliterated, so that it is difiScnlt to find places suitable for dip observations. SufiScient dips have been found to show that subordinate folds occur within the quartzite. The average thickness is probably not less than 450 feet, and may bo more. Petrographically the formation includes massive quartzites and mashed quartzites or micaceous quartz-schists, in some of which the relations of the quartz present unusual features. Section V. The Randville dolomite, consisting of crystalline dolomitio rocks, overlies the Sturgeon quartzite. The Randville dolomite covers a larger share of the surface in the Felch Moun- tain range than any other member of the Algonkian. Natural exposures are fairly numerous and very evenly distributed. Moreover, test pits and diamond-drill borings have shown the pi-esence of the formation in the covered areas. Relatively the dolomite is a weak rock, and occupies relatively low ground. An average thickness of 700 feet is estimated for the Randville dolomite within the Felch Mountain range. Petrographically the formation consists of a rather coarse-grained, thoroughly crystalline dolomite, with more or less abundant crystals of tremolite and a number of other minerals of minor importance. Section VI. The Mansfield schist is only exposed in certain test pits. Its presence has also been determined by diamond-drill borings. The thickness is so small — not more than 200 feet — that, though it weathers readily, it produces no noticeable eft'ects on the general topography. Petro- graphically it consi.sts of fine-grained mica-muscovite or mica-biotite-schists, probably derived from the metamorphism of a clasfc. It shows nothing of especial interest. Section VII. The Groveland formation is magnetic and has been traced by means of compass and dip needle. Excellent natural as well as numerous artificial exposures render the data concerning the distribution of the formation very satisfactory. The most prominent hills in the Algonkian belt owe their relief to the fact that they are underlain by the Groveland formation. Petrographically we may recognize two main kinds of rock. The usual kind consists of quartz and the anhydrous oxides of iron, while the other and much rarer consists essentially of an iron amphibole with quartz and the iron oxides as associates. Both of these kinds are clearly of detrital origin. The conclusion is reached, based on certain microscopical structures, that iron and silica were originally present largely in the form of glauconite. Section VIII. Mica-schist aud ferruginous quartzites of the Upper Huronian series occur in the eastern part of the Felch Mountain range. The rocks constituting the series are soft iron- OUTLINE OF THIS MONOGRAPH. XXXV stained mica-schists, with thin, interbandod beds of ferrugiuous aud miciiccous (juartziti!. Neither kind shows traces of chistic origin. V'rom their structures and general relations they are believed to have been derived from sedimentary rooks by metamorphism. Section IX. The Algonkian rocks are cut by iutrusives, among which both acid and basic rocks are reiiresented. The acid rocks are granites occurring in narrow dikes. No dikes of granite .ire known to cut the Kaiidville or Manstield formations. Chapter IV treats of the Miehigamme Mountain and Fence River areas. These areas occur in the central part of the district. In the Fence River area the structure! is very simple. In the Miehi- gamme Mountain area the structure is comple.x. Section I treats of the Archean. The prev.alent rock is granite, cut by acid and basic dikes. Section II treats of the Sturgeon form.-itiou. This is scarcely known as a distinct Algonkian member in this area .apart from the Randville formation. In one section purely clastic sediments were observed, for which it is convenient to retain the name. These exposures cousist of slates iind gray wackes, with some layers of a coarser texture. Section III treats of the Randville dolomite. In the Fence River area the dolomite lies on the east side of the Archean and occupies a belt about one-half mile in width, aud extending from the mouth of the Fence River about 10 miles to the north and west, where it leaves the portion of the district studied. In the Miehigamme Mountain area the dolomite tops the low arcli in a broad crumpled sheet. The formest exposures occur on the sections made by the Fence River. No folds have been observed within this formation. The thickness probably varies from to 2,300 feet as a maximum. The rocks of the formation are chiefly chloritic and ophitic schists, with which are associated schists bearing biotite, ilmenite, and ottrelite ; greenstone, conglomerates or agglomerates, .and amygdaloids. The general characters of the schists are (1) a groundmass composed of chlorite, quartz, m.agnetite, epidote, and in some cases plagioclase microlites, and (2) the presence in this groundmass of much larger porphyritic individuals of several secondary minerals. As evidence of the origin of these schists, first, there is the absence of rocks possessing any sedimentary characters; next, lavas and also greenstone-conglomerates or agglomerates are undoubtedly present in the series ; furthermore, the minerals which compose the schist are those which would result from the alteration in connection with dynamic metamorphism of XXXVI OUTLINE OF THIS MONOGRAPH. igneous rocks of basic or intermediate chemical composition ; and finally, the grain and character of the groundmass, and in some slides tlie jireaence of plagioclase microlites disposed iu oval lines point directly to an igneous origin and to consolidation 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 composition from acid to basic. Section VI treats of the Groveland formation. This is of wide extent throughout this part of the Crystal Falls district, but its outcrops are limited to three localities. Its distribution has been determined by means of its magnetic properties. It is not topographically prominent, except in the Michiganime Mountain area, where it forms part of Michigamme Mountain. In the Fence River area' it is probably not folded. It there dips to the east. At Michigamme Mountain it is found iu several well-marked folds. The thickness of the formation is estimated to be approximately 500 feet. The rocks are iuterbauded ierruginous quartzite and actiuolite and griinerite schists, which still contain evidence of detrital origin. Chapter V treats of the Northeastern area and the relations between the Lower Marquette and the Lower Menominee. The territory included in the Northeastern area extends from the northern- most outcrops of the Fence River area to the northern end of the Republic trough, a distance of about 11 miles. Outcrops are scarce throughout this area, and the main conclusions are drawn from the magnetic work. Through the structural and lithological results of the magnetic work the gap between the Marquette and the Crystal Falls district is bridged, and it is shown with a high degree of probability that the Negaunee iron formation of the Marquette range is identical with the Groveland iron formation of the Felch Mountain range. Chapter VI treats of the Sturgeon River tongue. In the southeastern part of the Crystal Falls district and just north of the Felch Mountain range a tongue of fragmental rocks extending eastward has been studied. The extreme leugth is 12 or 13 miles. Its width at its eastern end is li miles; to the west it widens rapidly. It is bounded both to the north and to the south by Archean granites and schists ; to the east it is overlain by Paleozoic sandstones and limestones ; and at its west end it is covered by glacial deposits. Within the tongue two small granite islands occur. The Archean or Basement Complex rocks comprise gneissoid granites, hornldeude-schists, and biotite- schists, which are cut by dikes of greenstone and granite, and veins of quartz. The sedimentary rocks comprise conglomerates, arkoscs, quartzites, sericite-schists, clay slates, rocks that are probably tufaceous, dolomitic limestones, and calcareous sandstones and slates. They may be divided into a conglomerate series and a dolomite series. From the distribution of the exposures of the two series it is concluded that the conglomerate series is the older, and that conformably above it follows the dolomite series. The two form a westward-pitching syncline. The conglomerate series and the dolomite series are correlated respectively with the Sturgeon quartzite and the Randville dolomite of the adjoining Felch Mountain range. THE CRYSTAL FALLS IRON-BEARING DISTRICT OF MICHIGAN Part I.— THE ^A^ESTERN PART OF THE DISTRICT MON XXXVI 1 CONTENTS. Page. ChAPTKR I. — INTKOIIUCTION ; ^^ Previous work in the district 13 Mode of work ^2 Maj^uetic observations 24 Chapteh II— Gkograpiucal limits, structure and stratigraphy, and physiography 25 Geograiihical limits 2o Structure and stratigraphy 25 Physiography -" Topography 29 Drainage 31 Timber and soil 36 Chapter III.— The Archean 38 Distribution, exposures, and topography 38 Relations to overlying formation,s - 39 Petrograpliical characters '10 Biotite-granite (grauitite) 10 Gneissoid biotite-granite, border facies of granite 13 Acid dikes iu Archuan • 15 Basic dikes in Archeau '1'^ Schistose dikes 1'^ Massive dikes ■1° Re8umi5 . 49 Chapter IV. — The Lower Huronian series •'^O Section 1.— The Rand ville dolomite 50 Distribution, exposures, and topography 50 Petrographical characters - - °1 Relatious to underlying and overlying formations 53 Thickness 53 Section 2. — The Mansfield slate 54 Distribution, exposures, and topography s^l Possible continuation of the Mansfield slate ^>3 Petrographical characters 51) Graywacke ^'^ Clay slate and phyllite 57 Origin of clay slate and phyllite 58 Present composition necessarily different from that of rock from which derived 58 Analysis of Mausfield slate 59 Comments on analysis ''" Comparison of analysis of Mansfield clay slate with analyses of clays 60 Comparison of analysis of Mansfield clay slate Tvith analyses of other clay slates.. 61 Siderite-slate, chert, ferruginous chert, and iron ores 62 Relations of siderite-slate, ferruginous chert, and ore bodies to clay slate 63 4 CONTENTS, Chapter IV.— The Lower Huronian series— Contiuned. Section 2. — The Mansfield slate— Continued. P?ge. Relations of Manstield slate to adjacent formations ., 63 Relations to intnisives 63 Relations to volcanics 6l Structure of the Mansfield area 64 Thickness 64 Ore deposits - 65 General description of Mansfield mine deposit 68 Relations to surrounding beds 68 ' Composition of ore 68 Microscopical character of the ores and associated chert bauds 69 Origin of the ore deposits 70 Conditions favorable for ore concentration 72 Exploration - 73 Section 3.— The Hemlock formation .- 73 Distribution, exposures, and topography 73 Thickness 74 Relations to adjacent formations 75 Relations to intrusives 77 Volcanic origin 78 Classi fication 79 Acid volcanics • 80 Acid lavas 80 Rhyolite-porphyry 81 Texture 83 Aporliyolite-porphyry 87 Schistose acid lavas 87 Acid py roclastics - ^ Basic volcanics 95 Basic lavas "5 General characters 95 Nomenclature 95 Metabasalts 98 Nonporphy ritic metabasalts 98 Petrographical characters 98 Chemical composition 103 Porphyritic metabasalt 103 Petrographi cal characters 104 Chemical composition 105 Variolitic metabasalts 108 Ellipsoidal structure 112 Origin of ellipsoidal structure 118 Amygdaloidal structure 124 Flattening of amygdaloidal cavities 126 Alteration of the basalts 126 Description of some phases of alteration 127 Pyroclastics 135 Eruptive breccia 135 Volcanic sedimentary rocks 136 Coarse tuffs 137 Fine tufi's or ash (dust) beds 142 Relations of tuffs and ash (dust) beds 143 Volcanic conglomerates 143 Schistose pyroclastics WS CONTENTS. 5 ClIArTKK IV. — TllK LoWKR HURONIA.N SKUiKS — C'outiiimjd. Section 3. — The Hemlock formation— Continued. Page. Tl)<« Bono Lake crystallini^ scliists 148 Uistriliution 148 Field evidence of connection witli tlie voleanics 149 I'etrogiiipliical iharacter.s 150 Normal 8edimentarie.s of the Hemlock formation 152 Economic products 153 Building and ornamental stones 153 Road materials 154 Chapter V. — The Upper Hukonian series 155 Distribution, e.xposures, and topography 155 M.agnetic lines 156 . Thickness 157 Folding w 1.58 Crystal Falls syncline 158 Time of folding of the Upper Huronian 161 Relations to other series 162 Relations to iutrusives 164 Correlation 164 Petrographical characters 165 Sedimentary rocks 165 Microscopical description of certain of the sedimeutaries 169 Igneous rocks I74 Ore deposits j75 History of opening of the district I75 Distribution I75 Western half of sec. 34, T. 46 N.,R. 33 W 176 Sec. 20, T. 45 N., R. 33 W 176 The Amasa area I77 The Crystal Falls area 178 Character of the ore 180 Relations to adj acent rocks 182 Origin I83 Size of the ore bodies 184 Methods of mining Igj. Prospecting I85 Production of ore from the Crystal Falls area 186 Chapter VI. — The Intrusives 187 Order of treatment 188 Age of the intrusives 188 Relations of folding and the distribution of the intrusives 189 Section I. — Unrelated intrusives _ igo Classification 190 Acid intrusives igO Geographical distribution and exposures of granites 190 Biotite-granite 191 Mieropegmatites 192 Muscovite-biotite-granite I93 Relations of granites to other intrusives I94 Dynamic action in granites I94 Contact of granites and sedimeutaries I94 Evidence of intrusion I95 (3 CONTENTS. Chapter VI.— The Intrusives — Continued. Section I. — Unrelated intrusives — Continued. "P^ge. Basic intrusives 198 Metadolerite 199 Geographical distribution 199 Petrographical characters 199 JIacroscopical 199 Microscopical 200 Relations to adjacent roclis 203 Relations to Lower Huroniau Mansfield slates 203 Relations to Lower Huroniau Hemlock rolcanics 204 Relations to Upper Huroniau 20+ Relations to other intrusives 204 Contact metamorphism of Mansfield slates by the dolerite 204 .Spilosites 206 Analyses of spilosites 207 Desmositcs 207 Adiuoles 208 Analyses of adinoles 208 Comparison of analyses of normal Mansfield clay slates and the contact prod- ucts - 209 No endomorphic effects of dolerite intrusion 211 Metabasalt 211 Ultrabasic intriisives 212 Picrite-porphyry (porphyritic limburgite) 212 Geographical distribution and exposures 212 Petrographical characters 212 Gray treniolitized picrite-porphyry 213 Dark serpentinizod picrite-porphyry 217 Classification 220 Section 11. — A study of a rock series ranging from rocks of intermediate acidity through those of basic composition to ultrabasic kinds 221 Diorite , 222 Nomenclature 222 Distribution and exposures 223 Petrographical characters 223 Description of interesting variations 226 Sec. l.^T. 42N.,R. 31W 226 Across river from Crystal Falls 227 Southeast of Crystal Falls 227 Analysis of diorite 231 Gabbro and norite 233 Petrographical characters 233 Description of interesting kinds of gabbro - 240 Hornblende-gabbro in sec. 1.5, T. 42 N., R. 31 W 240 Sees. 15,22,28, and 29, T. 42 N.,R. 31 W 241 Hornblende-gabbro dikes 243 Bronzite-norite dike 244 Sec. 29, T. 42 N., R.31 W., 1200 N., 200 W 245 Dynamically altered gabbro 247 Relative ages of gabbros 249 Peridotites 249 Distribution, exposures, and relations 249 Petrographical characters 249 CONTENTS. 7 CiiArTKK \I. — TiiK iNiiiUsivBS — Coiitinuefl. Sootioii II. — A study ot':i rock series, etc. — Continued. Peridotite — t'outiuucd. Paje. Periilotite varieties 252 Wehrlito 253 Amiiliiliole-puridotite 233 Gr.adatious of ampbiboK'-peridotite to wehrlite and olivine-gabbro 254 Process of crystallization 257 Analysis of peridotite 259 Peridotite from sec. 22, T. 42N., R.31 W., 1990 N., 150 W 260 Relations of peridotites to other rocks ., 261 Age of peridotites 262 General observations on the above series 262 Textiiral characters of the series 262 Chemicil composition of the series 263 Relative ages of rocks of the series 265 ILLUSTRATIONS Page. Pl-ATE I. Colored insii) showing the distribution of pre-Cambriiin and otber rocks in the Lake Superior region, and the geographical relations of the Crystal Falls district of Michigan to the adjoining Marquette and Meuominee districts of Michigan 11 II. Topographical map of the Crystal Falls district of Michigan, includiug a portion of the Marquette district of Michigan In pocket. III. Geological map of the Crystal Falls district cf Michigan, including a portion of the Marquette district of Michigan In pocket. IV. Portion of a geological map of the Menominee iron region, by T. B. Brooks and C. E. Wright 18 V. Generalized sections to illustrate the stratigraphy and structure of the northwestern part of the Crystal Falls district of Michigan 28 VI. Generalized sections to illustrate the stratigraphy and structure of the southern part of the Crystal Falls district of Michigan 28 VII. Generalized columnar section 30 VIII. Map of a portion of the Crystal Falls district, showing in detail the glacial topog- raphy and illustrating the development of the Deer River 32 IX. Sketch of the Mansfield mine as it was before it caved in, in 1893 66 X. A, Reproduction of the weathered surface of a variolite; B, Reproduction of the polished surface of a variolite 110 XI. Colored reproduction of an ellipsoid, with matrix, from an ellipsoidal basalt 116 XII. Mount Giorgios, viewed from its west flank, in April, 1866, illustrating the charac- teristic block lavas, from Fouque's Santorin et ses Eruptions, PI. VIII 120 XIII. Reproduction in colors of a basalt tuft' 140 XIV. Idealized structural map and detail geological map, with sections, to show the dis- tribution and structure of the Huronian rocks in the vicinity of Crystal Falls, Michigan 160 XV. Portion of Brooks's PI. IX, Vol. Ill, Wisconsin Geological Survey 172 XVI. Detail geological map of the vicinity of Amasa, Michigan 176 XVII. Detail geological map of the vicinity of Crystal Falls and Mansfield, Sheet 1 178 XVIII. Detail geological map of the vicinity of Crystal Falls and Mansfield, Sheet II 178 XIX. ^, Inclusions in a fractured (juartz phenocryst; 7i, Quartz phenocryst with rhombo- hedral parting 368 XX. ^, Micropoikilitic rhyolite-porphyry; 5, Micropoikilitic quartz-porphyry 270 XXI. ^, Very fine-grained micropoikilitic ryholite-porphyry viewed without analyzer; B, Very fine-grained micropoikilitic rhyolite-porphyry viewed with analyzer 272 XXII. j4, Perlitic parting in aporhyolite; £, Perlitic iiarting iu aporhyolite 274 XXIII. J, Schistose rhyolite-porphyry ; B, Aporhyolite breccia 276 XXIV. J, Schistose rhyolite-porphyry; i?, The same viewed between crossed nicols 278 XXV. ^, Amygdaloidal texture of basalt; i}, Amygdaloidal vitreous basalt 280 XXVI. J, Amygdaloidal vitreous basalt; B, Amygdaloidal vitreous basalt showing sheaf-like aggregates of feldspar 282 XXVII. A, Reproduction in colors of amygdaloidal basalt; B, Pseudo-amygdaloidal matrix of ellipsoidal basalt ; C, Water-deposited pyroclastic 284 XXVIII. J, Fine-grained basalt with well-developed igneous texture: i?, Illustration of the obliteration of the igneous texture of a basalt by secondary products when viewed between crossed nicols 286 9 10 ILLUSTKATIONS. Page. Plate XXIX. ^, Basalt showiDg characteristic texture in ordinary light; B, Basalt showing obliteration of texture between ci'ossed nicols 288 XXX. A, Basalt showing in ordinary light ii distinctly amygdaloidal texture; B, The same basalt with its amygdaloidal texture obliterated when viewed between crossed nicols 290 XXXI. A, Basalt aft'ected by calcification process; B, Basalt affected by calcification process viewed between crossed nicols 292 XXXII. A, Illustration of perlitic parting in a fragment from a basaltic tuff; B, Sickle- shaped bodies in volcanic tuft' 294 XXXIII. ./, Water-deposited sand ; B, Gradation in water-deposited volcanic sediment . .. 296 XXXIV. .1, Contact product of granite; /?, Brecciated matrix between ellipsoids 298 XXXV. A, Contact between granite and a metamorphosed sedimentary; B, Contact between granite and a metamorphosed sedimentary viewed between crossed nicols 300 XXXVI. A, A variety of spilosite with white spots; B, A variety of spilosite with white spots viewed between crossed nicols 302 XXXVII. A, Normal spilosite or spotted contact product; i?, Normal spilosite of some- what dift'erent character 304 XXXVTII. .1, Passage of spilosite into demosite; B, Occurrence and alteration of bronzite in bronzite-norite 306 XXXIX. A, Biotite-granite viewed between crossed nicols ; B, Mica-diorite viewed between crossed nicols 308 XL. A, Quartz-niioa-iliorite-porphyry ; B, Quartz-mica-diorite-porphyry viewed be- tween crossed nicols 310 XLI. J, Porphyritic poikilitic hornblende gabbro; /?, Poikilitic hornblende gabbro.. . 312 XLII. .-(, Moderately fine-grained hornblende gabbro showing parallel texture; B, Mod- erately fine-grained hornblende gabbro showing parallel texture viewed between crossed nicols. ..'- 314 XLIII. J, Normal granular hornblende gabbro; B, Schistose hornblende gabbro viewed between crossed nicols 316 XLIV. ,1, Moderately fine-grained hornblende gabbro; i?, bronzite-norite 318 XLV. J, Brouzite-norite-porpbyry ; /?, Feldspathic wehrlite 320 XLVI. J, Feldspathic wehrlite viewed between crossed nicols; /i, Feldspathic wehrlite. . 322 Fig. 1. Reproduction of a portion of the geological map of the Upper Peninsula of Michigan, by William A. Burt, 1846 , 15 2. Enlarged reproduction of a portion of a map of the Lake Superior land district, by Foster .nnd Whi tney 17 3. Enlarged reproduction of a portion of a geological map of the Upper Peninsjila of Michi- gan, by K'ominger, Brooks, and Pumpelly, 1873 18 4. Grauite-porpbyry with inclusions of gneissoid granite 45 5. Illustration of the effect on the topograjjhy of the differential erosion of basic dikes and granite 46 6. Concentric cracks formed by the caving in of the Mansfield mine 65 7. Sketch of the surface of the outcrop of an ellipsoidal basalt, showing the general char- acter of the ellipsoids and matrix 112 8. Sketch showing tlie concentration of the amygdaloidal cavities on one side of an ellip- soid, this side probablj' representing the side nearest the surface of the flow 113 9. Ellipsoids with sets of parallel lines cutting each other at an angle 114 10. Reproduction of illustration of aa lava, after Dana 120 11. Profile section illustrating results of diamond-drill work 177 12. Sketch illustrating contortion of Upper Huronian strata 179 13. Sketch showing change of strike of Upper Huronian beds, due to the folds 179 14. Sketch to illustrate the occurrence of ore bodies 182 us GEOLOGICAL SURVEY MONOGRAPH XXXVI PL GEOLOGIC MAP OF FART Ol'^ THl.; LAKE HUI'KRIOH REGKW Compilftd from Official maps of U. S.MiniiPsoUi.aiid Caiiadian Sui^'eys Scah.- Ah* The Afrnamuirr IrotvBrarirtfiSent^ . Ah5 Thf n'usconj>ui f'niliry Stntes Hi KONlA^ ■, Ah6 T/tf fmokeelran-Bttifinff Series Ah 7 Hi/- St Louis Sfrifj Ahe The Chippewa VitUey QuartMitei: Ah 9 The Blaek River TrowBearin^ .Srfiwfj Ah 10 The Aiumikie Series Ah 1 1 TTie Fhlded SrJUata afOinadfU urn STAT Ml THE CRYSTAL FALLS IRON-BEARING DISTRICT OF MICHIGAN. PART 1. THE WESTERN PART OF THE DISTRICT. By J. Morgan Clements. CHAPTER I. INTRODUCTION. The present report is an account of a portion of the Crystal Falls dis- trict of Michigan, so called from the most important town, Crystal Falls, the comity seat of Iron County. The iron-bearing district along the Paint River, near the site of the town of Crystal Falls, was first called in literature the Paint River district by Brooks.^ As soon as the town was begun, about 1880, the name of the town was applied to the district.^ It is situated on the Upper Peninsula of Michigan, adjoining the northeastern border of Wisconsin, and serves as a link connecting the two well-known iron-ore- producing districts of Michigan, the Marquette, and the Menominee. The Crystal Falls district is of itself of considerable economic importance, as will be seen, though not deserving to be ranked with either of the two above-mentioned iron districts. Since the geological relations of the rocks ' The irou-bearing rocks (economic), by T. B. Brooks: Geol. Survey of Michigan, Vol.1, Part I, 1873, p. 182. "Rept. Com. Miu. Statistics Mich, for 1881, p. 222. 11 12 THE CRYSTAL FALLS IRON- BEARING DISTRICT. of the Marquette district have now been ascertained/ it is hoped, by means of the determination of the succession in the intermediate Crystal Falls district, that the Menominee rocks may be closely correlated with those of the Marquette district. The accurate delimitation of the iron-bearing or coal-bearing forma- tions, or any other formations containing valuable mineral products, is of inestimable value to miners and investors. In the iron districts of Michigan alone innumerable test pits have been sunk in areas of solid granite, and at great distances outside of the possible iron formations, thus wasting large sums of money. Although the investigations carried on in the Crystal Falls district, the results of which are here recorded, do not enable us to point out definitely the places where the prospector will find iron deposits, they have enabled us to delimit in a broad way the various formations, and warrant the statement that iron deposits may occur in certain areas and that the prospector will assuredly not find iron dejjosits in certain others. The opportunity of studying the Crystal Falls district was given me tlu-ough Prof. C. R. Van Hise. In the prosecution of the field studies and in the preparation of the report I have availed myself of his advice and suggestions, which have been generously ofi"ered and which have been found of greatest value. To him I am most deeply indebted. The report is based not only on my own field work, but also on the field work done by a number of other geologists, whose notebooks have been placed at my disposal. The names of these geologists may be found on page 22. Among them, the notes of Mr. W. N. Merriam and Dr. W. S. Bayley have been found especially valuable. Mr. Merriam, assisted by Dr. Bayley, spent a season in doing very detailed work on the area shown on the sketch map at the bottom of PI. Ill, between Crystal Falls and Mansfield, and from this point northwest to some distance beyond Amasa. The mag- netic lines represented in this part were traced by Mr. Merriam, and the geology in general is the same that he outlined on his final field map I wish to thank Mr. C. K. Leith, who has been of the greatest clerical assistance, and Mr. E. C. Bebb, by whom the maps were drawn; also Mr. • The Marquette iron-bearing district of Michigan (preliminary), by C. R. Van Hise and W. S. Bayley; with a chapter on the Republic Trough, by H. L. Smyth: Fifteenth Ann. Rept. U. S. Geol. Survey, 1895, pp. 477-650. Ibid, (final), Mon. U. S. Geol. Survey, Vol. XXVllI, 1897. PKEVIOUS WORK. 13 J. L. Ividg-wiiy, by wlioin tlie colored plates of natural size specimens w(;re j)repure(l. PREVIOUS WORK IN THE DISTRICT. ( )u account of its comparatively slight economic importance, and also on account of its isolation, very little work of whicli the results have been published was done in this district prior to that on which this monograph is based. As a rule, the earlier observers began the season's work either in the Marquette or in the Menominee range, and working westward the Crystal Falls district was reached only as the season neared its close, or as the appropriation was nearly exhausted. The published work upon this district is given below in chronological order. X850. Burt, Wm. A. Report of linear surveys with reference to mines and minerals, in the Northern Peninsula of Michigan in the years 1845 and 184C. Dated March 20, 1847. Thirty-tirst Congress, first session, 1850; Senate documents, Vol. Ill, No. 1, pp. 842-882, with map. Dui'ing the year 1846 a linear survey was made of that part of the Upper Peninsula of Michigan described as being bounded on the north by the fifth correction line, on the south by the fourth correction line and the Brule River, on the east by ranges 23 and 26 W., and on the west l)y range 37 W. This includes in its limits the district under discussion. In the course of the survey, geological observations were made by William A. Burt, the deputy surveyor in charge of the work. The report and accom- panying geological map embodying the results of these observations are concealed among the Senate documents of the Thirty-first Congress. The following quotations from this report give all the observations on the part of the territory surveyed in which we are at jjresent interested: Topography. — Wcst of range 31 west, and north of the Brule River to the fifth correction line, is a tract of about 43 townships in which the rock is mostly greenstone and hornblende slates. This part of the surveyed district is less broken than that above described, and a large proportion of it may be denominated rolling lauds. There are, however, many ridges and conical hills of various heights upon this part of the survey, and also deep valleys of streams, many of which have ledges upon their sides. These general characteristics are often changed for cedar, spruce, or tamarack swamps, which are most numerous in townships 46, 47, and 48 N. [This includes the part of the district supposed to be the continuation of the northern Wisconsin peneplain (p. 31).] 14 THE CRYSTAL FALLS IRON-BEARING DISTRICT. Granite (and syenite). — These Tocks occupy au area of about 22 townships on tlie northeast part of the survey, between the fifth correction line and the south boundary of township 45 N., and east of range 32 W., in a series of irregular uplifts, frequently forming high cliffs and sloping ledges on the most elevated portion of this district. [This covers a part of the Archeau granite oval of the Crystal Falls district, as well as the large Archeau areas northeast of it.) Argillaceous slates. — The argillaccous slates alluded to in townships 42 and 43 N. are generally overlaid by deep drift; their boundaries, therefore, could not be satisfactorily defined. West of the Peshaknmine River these slates appeared to have undergone considerable chauge by igneous action, aud were often associated with an oxide of iron; but east of the Peshakumine no change by igneous action in the slates was observed, and on this part they have generally a reddish color They dip variously at a high angle, and are supposed to conform to the greenstone on the north and west, and to overlie or pass into the mica-slates on the south ; and in their middle portion they di|) about 90°, with strike nearly east and west. (These slates correspond to our least metamorphosed phases of the Upper Huronian.] Greenstone and hornblende slate. — Thcsc rocks occupy a larger area lu the district surveyed than any other class of rocks. They extend from the granitic and other rocks east of them westward beyond the survey. [See their outline on map, tig. 1.] The greenstone and hornblende slates form a less broken surface than the granitic range; and next to it is the most elevated range in this district, having an estimated altitude, in many places, of from 1,001) to 1,100 feet above Lake Superior. These rocks are frequently seen in the beds and bauks of streams aud in ridges aud conical hills of various heights, often forming precipitous ledges upon their sides The greenstone of this region is generally more or less granular and syenitic, with a dark green color when moist; its composition is hornblende, feldspar, and quartz — the former mineral greatly predominating. In some places the feldspar and quartz are nearly or quite wanting, leaving a granulated hornblende rock. Another variety of this rock was freciuently seen which was composed of the same ingredients but very tine grained and compact and having frequently a laminated or slaty structure, the cleavages of which generally dip from the granitic rocks at a very high angle Some of these hornblende slates have in their seams and cleavages a silky luster, from the presence of mica or talc in very fine grains All of these rocks are traversed by many quartz veins, from a line to 4 feet or more in width, and with still larger veins and dikes of more recent trap rock. This range is supposed to have become blended with the trap range of Keweenaw point as it passes under the red sandstone lying between them, and probably farther west the two are united in one range. [These are the altered and more or less schistose basalts and accompanying- fragmeiitals which are comprised in the Hemlock formation. [ Mica-siates. — Thcsc slatcs strctch along the southerly side of the argillaceous slates on the south part of the survey. They extend from the Brule River on a course east- uortheast for about 22 miles, in townshij)s 41 and 42 N., ranges 29, 31, and 32 VV.. aud have an average breadth of about 4 miles The mica-slates are supposed to dip northerly under the argillaceous slates at a high angle, varying at the surface from 45° to 80° PREVIOUS WORK. 15 This rock is conii)<)se(l of iniwi, ((uartz, and feldspar. Its lainiita' are uudiilatiiig or waved, but its cleavages, on a large scale, are even and regular Those luica-slates are best (levelo])ed on the south bouiidaiy of township 4L! N., ranges 31 and 32 W., in the beds and banks of the Peshakunime and Mesqua-cum-a- XXX III XXXII XXXI XXX XXIX x'Xvul Scale of miles 5 Fig. 1. — Reproduction of a portion of tbe geological map of the Upper Peninsnla of Michigan hy Wnj. A. Burt, 1846. cum-sepe, and at the falls near the junction of the latter stream with the Brula River. [These, according to our observations, are the most altered phases of the Upper Huronian.] That part of the geological map accompanying the above report which corresponds to the Crystal Falls district is reproduced iu fig. 1. 16 THE CRYSTAL FALLS IRON-BEARING DISTRICT. 1851. Foster, J. W., and Whitney, J. D. Report on the geology of the Lake Supe- rior land district. Part II. The iron region, togetlier with the general geology. Thirty-second Congress, special session, 1851; Senate documents, Vol. Ill, No. 4, pp. 406, with maps and section. In 1851 there was published a report by Foster and Whitney on the iron regions of the Lake Superior land district, together with the general geology. This gives the first connected account of the results obtained by the various surveyors who had been engaged on the Government survey of the Upper Peninsula of Michigan. Accompanying this report there are two colored maps and a section. The subdivisions of the rocks as made by Burt in the Crystal Falls district are not retained in this report by Foster and Whitney. The map is generalized, and the hornblende-slates, etc., of Burt are included under the general term "crystalline schists," and are placed by the authors in the Azoic system. There are represented here and there throughout this Azoic area a trajjpean knob and bed of marble. Tlie granite area shown on Burt's map is very much reduced in size, and no longer connected with the large granite areas to the east. The granite on the lower reaches of the Michigannne, in T, 42 N., R. 31 W., is here indicated for the first time. In these respects only does this portion of the map show a decided advance in knowledge of the distribution of the rocks. A copy of the map, showing the distribution of the rocks by symbols instead of colors, is reproditced as fig. 2. 1873. Brooks, T. B. The iron-bearing rocks (economic). Geol. Survey of Michigan, Vol. I, Part I, 1873, pp. 319. With Atlas Plate IV and general map, by Romiuger, Brooks, and Pumpelly. The next mention of the district that I have been able to find was made in 1873 by Maj. T. B. Brooks, in his report on the iron-bearing rocks of Michigan. However, this report seems to show a decided decrease in knowledge from that possessed by Burt concerning the geology of this district. It is trae that indications of iron had been seen, but the observations made were so meager that nothing could be done toward determining the relations of the rocks or unraveling the structure of the area. Upon the map accompanying the report (Geol. Survey of Micliigan, PKEVIOUS WORK. 17 1S73), a {jortion of wliicli is reproduced iu ti O <- t "> E i qo -i in PREVIOUS WORK. 19 II*' also <;-ives liis analysis (No. 68, p. 302 ot" Brooks's report) of au ore sample t'roin the district, and calls attention to its abnormally high water content, freedom from silica, and richness in iron as compared to those of the more eastern mines in tlie Menominee region. 1880. Brooks, T. B. Geology of the Menominee region. Geol. Survey of Wisconsin. Vol. Ill, Part V, 1880, pp. 430-0.35. Atlas folio. Pis. XXVIII ami XXIX, and PI. XXX, by C. E. Wright. In an article on the geology of the Menominee region, which wa.s written for the geological survey of Wisconsin, the same author brief!}' touched on that part of the adjoining Michigan territory which is included in the district under consideration. His observations were thus confined to a few exposures in a limited portion of the area. He attempts to correlate certain beds by means of their lithological character with those with which he was familiar in the Marquette district and refers them uniformly to the higher members of the Huronian. They are also so referred on the map which accompanies the report, thougli this is dated a year earlier than the date of publication of the report. That portion of the map covering a small part of the Crystal Falls region is reproduced on PL IV. The present survey enables us to add very little Jo this, aud these additions are chiefly of a petrological character. . 1881. RoMiNGER, Carl. Geology of the Menominee iron region. Geol. Survey of Michigan, Vol. IV, 1881. In 1880 Dr. Carl Rominger, at that time State geologist of Michigan, spent a season in the Menominee district, and in his report gives detailed descriptions of a few occurrences in the Crystal Falls district, to which I shall refer later on. He considers the rocks in general to belong to the Huronian, and distributes the beds among his diorite group, iron-ore group, and arenaceous-slate group, as given and defined in the previous report on the Marquette district No attempt at more definite coiTelation was made. isso. Van Hise, G. R. An attempt to harmonize some apparently coutlicUng views of Lake Superior stratigraphy. Am. Jour, of Sci., 3d series, vol. 41, 1891, p. 133. On December 30, 1890, Prof C. R. Van Hise read a paper on Lake Superior stratigraphy before the Wisconsin Academy of Sciences, Arts, and 20 THE CRYSTAL FALLS IROX-BEARING DISTRICT. Letters, the same article being published the following year in the American Journal. The iron-bearing series of this district was in this article referred to the Upper Marquette (Upiser Huronian). 1893. Wright, C. E. Report of State geologist from May 1, 1885, to June 1, 1888, hi Rapt, of the State Board of Geol. Survey of Michigau, 1893, pp. 33-44. State geologist, Charles E. Wright, in a report for the seasons from 1885 to 1888, inclusive, merely mentions the general strike of tlie rocks of the district, and makes no attempt to determine their age nor to unravel the structure. Wadsworth, M. E. Sketch of the geology of the iron, gold, and copper dis- tricts of Michigau. lu Rept. of State Board of Geol. Survey for years 1891-92, 1893, pp. 75-186. Dr. M. E. Wadsworth, who on Mr. Wright's dea,th succeeded him as State sreoloffist of Michig-an, mentions the occurrence of carbonaceous slates, of granite and melaphyre, and of conglomerate near Crystal Falls, but does not enter into a discussion of the relations of anv of these rocks. Dr. Wads- worth agrees with the correlation of Professor Van Hise, and places the Crystal Falls ore deposits in the Upper Marquette series (Wadsworth's Holyoke formation) (pp. 117, 132), but the evidence for so doing is not given in the report. He also is the first to recognize the volcanic nature of the rocks in the vicinity of Crystal Falls (p. 134). lS9o. RoMiNGER, C. Geol. rept. on the Upper Peniusula of Michigan. Geol. Survey of .Michigan, Vol. V, Part. 1, 1895, pp. 1-164. In liis report of work done on the Upper Peninsula of Michigan from 1881 to 1884, published in 1895, he follows the same plan, referring the various rocks exposed h\ mining operations to his different groups.^ Clements, J. Morgan. The volcanics of the Michigamme district of Michigan. Jour, of Geol., Vol. Ill, 1895, pp. 801-822. ^ In a preliminary article on this district, by the writer, published in 1895, the volcanic character of the rocks which cover a large area of the Crystal Falls district was emphasized, and in a sketch maj) in the same ' Geol. Survey of Micbigan, Vol. IV, 1881, p. 8. 'After the publication of this article, the name Michigamme haviug been applied to a formation, it was deemed advisable, in order to avoid confiisiou, to change the name of the district to the Crystal Falls district. PREVIOUS WORK. 21 article was oiveii an outline of the distribution of the various rocks for a portion of the ilistrict, with their stratig-ra[)hical succession (p. 803), the dis- cussion of the structure and correlation l)eino- left for the present report. The above-mentioned sketch map, with the maps by Burt, Foster and Whitney, Brooks, Brooks and Wright, the section by Foster and Whitney, and the section by Brooks, along the Paint and Michigannne rivers, are the oidy maps or sections which, so far as can be learned, have been published of that part of the Crystal Falls district under discussion. MISCELLANEOUS REFERENCES. JuLiEN, Alexis A. Apiit-ndi.x A. Lithology. Geol. of Michigan, Vol. II, 187.3, PI). 1-185. WiCHMANN, Arthur. Microscopical observations on the iron-bearing rocks from the region south of Lake Superior. Brooks's Geol. of the Menominee Iron Region, 1880, Chap. V, pp. 600-65."). Wright, Charles B. Geology of Menominee Iron Region. Geol. of Wis- consin, Vol. Ill, Part 8, 1880, pp. 665-741. Lane, A. O. In sketch of the geology of tlie iron, gold, and copper dejwsits of Michigan. Rept. of State Board of Geol. Survey for 1891-92, 1893, p. 182. Patton, H. S. Microscopic study of some Michigan rocks. Uept. of State Board Geol. Survey for 1891-92, 189.3, p. 186. During the progress of the Michigan and Wisconsin State surveys specimens from outcrops were collected, and descriptions of these discon- nected specimens are found in the State reports. References to the pages on which the individual descriptions may be found will be given under the petrographical discussion of similar rocks here described. UNPUBLISHED WORK. In 1891 a survey was organized by a private corporation, and put in charge of Prof C. R. Van Hise. He consented to take charge of this work on the conditions that all maps and notes should be available for this report and that no other compensation was to be made by the company. The object of this survey, known as the Lake Superior survey, was to study that part of Michigan of which Crystal Falls is the center, in order to determine the feasibility and advisability of opening up the mines of that district. This survey was vigorously prosecuted, and an excellent topographic map made of an area 32 miles north and south and 42 miles east and west, cover- ing a large part of four 1.5-minute atlas sheets of the United States Geological 22 THE CRYSTAL FALLS IKON-BEAKING DISTRICT. Survey. At the same time, in connection witli the topographic work, a reconnaissance geological survey was made. The following is a list of those who took geological notes for this survey: Andrews Allen, A. H. Brooks, W. S. Bayley, J. P. Channing, E. T. Eriksen, J. R. Finlay, F. J. Harriman, F. T. Kelly, E. B. Matthews, E. R. Maurer, J. A. McKim, F. W. McNair, W. N. Merriam, and H. F. Phillips. ' The following season was devoted to a detail study of the iron-bearing belts which had been outlined by the reconnaissance. This detail work in the western part of the district was prosecuted by parties in charge of W. N. Merriam, and in the eastern part of the district by parties in charge of H. L. Smyth. When they ceased work, the two areas mapped were sepa- rated in the north b}' about 12 miles, and a narrow belt separated the mapped areas to the south. During the season of 1894, under the direc- tion of Profes.sor Van Hise and assisted by Gr. E. Culver, and during part of the season by S. Weidman, I was engaged iu completing this unfinished work for the United States Geological Survey, preparatory to connecting this district with the Menominee iron-bearing district to the southeast. This work was carried on in 1895 by Dr. W. S. Bayley, S. Weidman, and myself, and the mapping of the district extended as far as the Menominee district. Mr. H. L. Smyth has written Part II of the present report, covering the portion of the district which was worked by his party. My description of the part of the district worked by me is based largely on my own obser- vations. Many of the facts of field occurrence, however, mentioned in the following paper were observed and recorded by the several men mentioned above, and were subsequent!)' verified by my own observations in portions of the area surveyed by myself, and by visits to localities in other portions. The topography of the greater portion of the district was taken by the members of the Lake Superior Survey. The remainder we owe to the topographical division of the United States Geological Survey. The areas covered l^v tlie respective organizations are shown on the sketch map below the topographical map (PI. 11). MODE OF WORK. As explanatory of the locations given in the paper, it is perhaps not out of place to give a brief description of the plan of work followed by the Lake Superior Division of the United States Geological Survey in this as MODE OF VVOKK. 23 ■well as in tlie otliiT Lako Superiiir iniu-buariuy (listricts wliicli lia\u been previously surveyed. The Upper Peninsula of Michigan affords an excellent example of the ■excellence wliicli can be obtained in the rectangular land survey, when properly carried oiit l)^■ the Government. The section corner posts originally established are in many cases still to be seen, and of course the bearing trees are even more common. Since the original survey the timber value has increased so much that in certain forested areas the section lines have been resurveyed. Not uncommonly trails follow the section lines for long distances. Moreover, the roads are frequently laid out along the section lines, thus giving permanent land boundaries. The section corners con- sequently offer the most reliable points from which to make locations. Ti'averses are made across each section, either frcm east to west or from north to south, and at varying intervals, according to the discretion of the geologist and the exigencies of the case. Each geologist is accompanied by a compassman, whose duty it is to determine the course of the ti-averses by means of a dial compass, and the distance traveled by pacing at the rate of 2,000 steps to the mile. Corrections are made at the corner and quarter posts. The compassmen employed are Michigan woodsmen, land lookers or cruisers as they are frequently called, and it is reniarkable with what accuracy they will pace mile after mile through swamp and over rough hills, windfalls, etc. The geologist explores the territory on both sides of the line followed by the compassman. Ledges are located by the geologist pacing to the compassman as he comes opposite him in a due east-west or north-south direction. With two coordinates thus determined, the ledges are located with reference to the starting point. For uniformity and to facilitate ref- erence and cataloguing, it is customary to give the location with reference to the southeast corner of the section. Thus, 1,000 N., 1,000 W., SE. cor. sec. 5, T. 42 N., R. 33 W., gives the location of the outcrop at the center of the section, and affords a means of finding that ledge which could not be so accurately and concisely stated by the use of any ordinary land- marks. Moreover, easily recognized landmarks, such as houses, quarries, etc., are few, and exceedingly great changes may occur very rapidly, such, for instance, as those caused by widespread forest fires, so that such a method of location is practically valueless. 24 THE CRYSTAL FALLS IRON-BEARING DISTRICT. MAGNETIC OBSERVATIONS. It has long been known that many rocks are possessed of decidedly magnetic properties, due to the presence in them of varying quantities of magnetic iron ore. By the mining engineers and prospectors this property has been turned to a practical use in aiding in the location of iron mines where the ore is of a magnetic kind. It is only in the past three decades that this property has been used to any extent by geologists as an aid in the interpretation of the structure of a region. So far as I can learn, the best published account of its use thus is in Brooks's report on the iron- bearing regions of Michigan.' Conclusive proof of its geological value was given in the mapping of the Penokee area, in 1876, by R. D. Irving of the Wisconsin survey.^ That area extends for about 60 miles northeast- southwest, and is on the average about 4 miles wide. For the eastern part of the Wisconsin area the outcrops are few, and Irving located the iron formation by magnetic work. Along that belt have been sunk shafts belonging to various mines which have raised quantities of ore, and in no case has a shaft sunk outside of the limit indicated by Irving come upon paying ore. By means of the dip needle and solar compass, observations were taken which enabled us to trace a curving magnetic formation and connect the outcrops, which were separated by about 16 miles. The same bed was further delimited, and the direction partly checked, by the occurrence, at varying distances along this course, of outcrops of rocks of the underlying formation. Since the second part of this report contains an exhaustive article on the methods and use of the magnetic needle,' the subject is not further treated here. The lines of maximum magnetic disturbance — or briefly, the magnetic lines — are represented on the accompanying general map, PI. Ill, by blue lines marked with letters I) and E. • Magnetism of rocks .and the use of the magnetic needle in exploring for ore, by T. B. Brooks. Geol. Survey of Michigiin, Vol. I, Part I, 1873, pp. 205-243. ■^Geol. of the eastern Lake .Superior district, by K. D. Irving. Geol, of Wisconsin, \o\. Ill, 1880, pp. 53-238. .\tlas sheets, XI-XXVI. 3 See Part 1 1, Chapter II, by H. L. Smyth, pp. 33(j-373. CHAPTEE, II GEOGRAPHICAL LIMITS, STRUCTURE AND STRATIGRAPHY,, AND PHYSIOGRAPHY. GEOGRAPHICAL lilMITS. The portion of the district here described extends from the north line of T. 47 N. to the south line of T. 42 N., and from the c.enter of R. 31 W. to the west line of R. 33 W., and contains approximately 540 square miles. Upon the small sketch majj at bottom of PI. Ill is outlined the por- tions of the district which have been studied and described by the different authors. The detail character of the formations is imknown for parts of the area under discussion. This is especially true of the north, west, and southwest parts, where, owing to the readily decomposable nature of the rocks, as determined by the few ledges observed, and to the drift mantle, very few outcrops are to be found. STRUCTURE AKD STRATIGRAPHY. The Crystal Falls district is not sharply defined petrographically, but is continuous with the Marquette district on the northeast and the Menomi- nee district on the southeast (PI. I). It is, however, remarkable for the vast accumulation of volcanic rocks, which, while by no means absent from the adjoining districts, do not there play so conspicuous a role. StriTCturally this district can hardly be better separated from the Menominee and IMarquette districts than it can be petrographically. The important sedimentary troughs of the two adjacent disti-icts are separated by an average width of 40 miles. The area between the districts on a. direct course is occupied chiefly by Archean rocks, with narrow infolded troughs of Huronian rocks playing a very subordinate role. At the east 25 26 THE CRYSTAL FALLS IRON BEARING DISTRICT. the Archean is overlain by the sedimentaries of the Paleozoic, the Cam- brian, and the Silurian. The connecting Crystal Falls rocks are west of this Archean dome. In the Marquette district the essential structural features have been shown ^ to be a g-reat east-west synclinorium, upon which more open north- south folds are superimposed. At the western end^ of the district the ■ superimposed north-south folds become close, and the Republic trough is a close fold with an axis in an intermediate position. In the adjoining Crystal Falls district there are also two sets of folds with their axes approxi- mately at right angles to each other. The closer folds are represented by the great anticline in the central part of the district. This anticline has its axial plane trending west of north and south of east, and the axis plunges down both at the north and south ends. The more open set of folds at right angles to the above set, is repre- sented by the Crystal Falls syncline, with its axis striking to the south of west, and plunging west. Farther south the axes of the folds become much closer and more nearl)^ east and west, thus nearly according in direction with the close folds of the Menominee district. Thus the structural features of the Crystal Falls district merge into those of the Menominee district, which joins the Crystal Falls district on the southeast, where the great structural feature is a synclinorium similar to that of the Marquette, but with its axis trending north of west and south of east. A glance at PI. Ill will show the presence in the eastern part of the northern half of the district of an oval-shaped mass of Archean, and, nearly surrounding this, a number of rock belts. The Archean ellipse is 11 miles long and 3 miles wide on the average. The rocks are mainly granite and gneiss. They are cut by rather infre- quent acid and basic dikes. Immediatel}^ surrounding the Archean is a quartzose magnesian lime- stone formation, to which the name Randville dolomite has been given.^ In the eastern half of the district described by Sm3^th, where more numerous exposures are found than occur in the western half, the formation has an estimated thickness of about lj500 feet.* Not only are the exposures ' Mon. XXVIII, cit., p. 566 et seq. 2Loc. cit., J). 570. 2 See Part II, Chapter IV, by H. L. Smyth, p. 431. \; n'"'-,M« «W '-'/v -'n- '-/v,^''-'-'A n'i''-'/n v' ''- '- ,\ ty-/ ' - V- - ^- I '- V- ,• '- >1 \ - ,' /-'\^y S- ^ //- I '1 \ AlaAIn Au AIn Ala /Rgr Au AInAla -^^r- -i%.\-,/.'-~s-,/-/'l>-'/-''^-/,./-,/,i--\-' - i^'-s-'/-/- .^'a>^-:v:v,;jaaKK-/ /--s-O/ri'- v-/^-i'L\-. /-'-'- \-/-'-\-,-'-i'^-\-//-i'l\-.^/- ^gr Ala AIn V / V , ' / \ , 1 - ~/"_,N v', -' x-N-,-,)-/^- s-,/-l/_N-/ /-/-■- I-,'/-/ -,N -<~ N.E- Au AIn Als Ala US BIENaCO LITH TJ v GENERM.1ZED SECTIONS THROUGH NORTH WESTERN PART OE CRYSTAL FALLS DISTRICT HORIZONTAL SCALE, 1 LMCH = 1 MILE. VERTICAL SCALE, I lNCH-1320 ?'EET. ELEVATION OF BASE LINES 1000 FEET. NOTE: Formations are brought to the surface only whel-e exposures have been obsenved ALGONKIAN ARCH E AN aJ UPPER HURONIAN Granile Sturricon am! Ajibik Koao amlRaiidvillt quartzile dolomite Hemlock formaUoii Growlaitd ^Xe^iiiinee UndixTded lonnation 6 us GEOLOGICAL SURVEY MONOGRAPH XXXVI PL VI sw ^'/i!/^ly thu entire C'lystal Falls district, was covered by Caiiil)riaii deposits. The thickness of tlie Cambrian deposits can not be determined. The next hi<.>lier })ortion of the geological time scale represented in the district is that part of the Pleistocene penod which in this part of the United States is characterized l)y the past existence of great ice-sheets. The evidences of the existence of the ice ai*e everywhere present, either in the rounding and polishing and scoring on the surfaces of the rocks exposed or in the character of the drift deposits. The direction of the ice movement was clearly from the northeast to the southwest, as is shown l^v the trend of the stria', which were observed upon the rounded rock out- crops in N'arious places. The thickness of the drift deposit varies very materially. In places it has been almost entirely removed by denudation, if in such places it ever formed anything more than a thin veneer upon the surface. In other places it reaches a very considerable thickness, as is shown by the glacial topography characteristically developed in T. 45 N., R. 32 W. As the present report is confined to the pre-Paleozoic rocks, no detail description will be given of these Cambrian and Glacial de2:)0sits, nor are they represented on the map, except in those places where it has been found impossible to map the underlying rocks. The generalized columnar section on PI. VII gives in condensed form our knowledge concerning the formations mentioned. PHYSIOGRAPHY. TOPOGRAPHY. The topography in its large features is pre-Glacial, and in some cases this older topography is rather distinct. For instance, in the case of the Deer River Valley, drift covers the gentle slopes and bottom, but is not sufficiently deep to completely hide the pre-Glacial Deer River Valley. In the southwestern part of the district west of Crystal Falls, or, more generally, west of the Paint River, pre-Glacial topography is seen in places. Here we find the drift as a veneer and only partly hiding the bed-rock topography, which depends mainly on the strikes, dips, and varying charac- ters of the rocks. It is so well known that this part of the country was at one time -covered by ice, that it is useless to cite such proof as the rounding and 30 THE CRYSTAL FALLS IRON-BEARING DISTRICT. scoring of the rocks and the character of the di'ift material, a good portion of which can be readily seen to have been brought from some other region, no such rocks as those forming it existing where the bowlders now lie. The ice-sheet left a deposit of di-ift, and we find the pre-Glacial topography essen- tially modified by it. As a result of this, the prevailing and most noticea- ble topography of the western half of the Crystal Falls district is that of the drift, and is characterized by short ridges and broken chains of hills, usually oval, though at times of very irregular outline, between which are lakes and swamps. The swamps are even occasionally found on rather steep slopes, where a thick spongy carpet of moss (sphagnum) retains sufficient moisture for cedars and other trees and shrubs characteristic of the Michigan swamps to grow. The swamps follow the carpet of moss up the hills to the spring line. The Glacial drift topography is especially marked where the drift was of considerable depth. These conditions are well exhibited in parts of T. 45 N., Rs. 31, 32 W., shown on the large-scale map, PI. VIII. Here, even though the ground is very heavily timbered, one may easily trace out the sinuous course of the eskers. When traversing the country, one is constantly descending into pot-holes or is climbing ridges, some of them 75 to 100 feet high, often with a crest only a few feet, in some places not more than 4 feet, wide. Where the drift mantle has been removed, the rounded character of the rock exposures is usually shown. This holds good especially for the more resistant rocks, such as the granites and massive greenstones. Slates and tuifs, weathering more readily, have in numerous cases had time since the ice retreated to be weathered into rough broken ledges, some of which show perpendicular cliffs. The elevations range usually from 1,400 to 1,600 feet above sea-level. The hills rarely rise more than 200 feet above the low ground at their bases. The extremes of height noted in the district are from 1,250 to 1,900 feet above sea-level, corresponding, respectively, to the valley of the Michigamme on the south and the watershed between Lake Superior and Lake Michigan on the north. Between these two extremes there is a strip of territory, 25 miles across from north to south, in which the variations in height are within the limits of 200 feet. A consideration of the slight difference of level which prevails over U. S. OEOLOGICAL SURVEY MONOGRAPH XXXVt PL. VII I Period. Formation name. _IO £LO Potsdam sandstone. 2-i l-O < CD o U. Colum- nar SEC- TION. ■Cp. Mansfield siate. LI_Li ^.^.^^^^^^ t^S I'll . r XJ, Randvilla dolomite. I ^gr. II" I I? Character cf rocks. Usual characters. Thickness unknnwn Yellowish to reddish brown sandstone, not thornughly cemented, therefore disintegrates readily, Found in patches in many places, and always lying either in beds wh'ch are horizontal or else possess slight dip to the south This may represent the initial dip with Vkhiqh the beds weie deposited. A series of very great bul unknown thickness. It consists of a'ternat- ing beds of slates, graywackes. siderite, and chert With Ihese, espe- cially associated wilh \he last two, are found hematite and limonite ore bodies of variable size and of great economic importance. From this series is derived nearly all the ore supplied by the Crystal ^a'ls distnct. In the southern part of the district, espec-ally well exposed in the vicinity of the Paint and Michigamme rivers, the slates and graywackes have bet^n metamorphosed into schists and gneisses. This series is cut by dikes of rock ranging from acid to ultrabasic, which have, in places, metamorphosed the sediments. The thickness of this vast pile o' volcanic ejectamenta can not be estimated with any degree of accuracy. It consists chiefly of in'.erbed- ded acid and basic lavas and associated tuff deposits, and the water- deposited materr Is derived from them. Near the top of the volcanics a lent'Cula: aiea of norma' sediments, slates witn lenses of limestone, is found. This formation is cut by acid and basic dikes 1500 Estimated to be about 1 ,500 feet thick. It consists of interbedded f rag- mentals, slates, and graywackes and. associated with these, fen uginous chert and carbonate From these last has been derived the ore found associated wiih them The Mansfield mine, by which is exploited the only ore body in the Mansfield formation, supplies the only Bessemerore of the Crystal Falls district These slates are cut and metamorphosed by basic dikes. The thickness 's that estimated for this formation in the eastern part of the district by Smyth, The prevailing rock is quartzose dolomite, of a veiy friable character. It shows the usual characters of granite. It is schistose on flanks of m.assif, and is cut by acid and basic dikes, which are mafsve ai-.d schistose. GENERALIZED COLUMNAR SECTION. PHYSIOGltAPHY. 31 the "Teatcr i)iirt of tlie Crystal Falls district has U-d Smyth to the conclusion that this portion of Michigan liad before Glacial times been reduced to the condition of an approximate peneplain. (See Part II, Chapter 1.) This peneplain is a continuation of tlie peneplain of northern Wisconsin, and lies between the northern Michigan base-level on the nortli and the central Wisconsin baselevel on the south, to both of which attention has recently been called by Van Hise.^ DRAINAGE. The greater heights in the Michigamme district are in the northern part, where some few of the hills rise to a height of 1,800 feet, and one to a maximum of 1,900 feet above sea-level; but tlie majority do not rise above 1,600 feet. The belt including these higher elevations extends about NE-SW. This belt represents the crest of the watershed, from which all streams on the northern side flow to Lake Superior, and on the southeastern side all flow to Green Bay of Lake Michigan. A part of this watershed is undivided, and it is not uncommon to find extensive swamps in which streams flowing to opposite sides of the watershed take their origin. The portion of the Crystal Falls district which is tributarj- to Lake Superior is so small that it will be totally neglected in the further discussion of the drainage. The topographical map, PI. II shows the general slope and drainage of the district to be SSE. The eastern part of the district is drained by the Michigamme^ River with its tributaries, the Fence (Mitchi- gan), and the Deer, while the Paint (Mequacumecum) River, with its main tributaries, the Hemlock and the Net, di-ains the west and northwestern por- tions. The Brule (Wesacota) flows along the southern part of the district, being for the most part just below the southern limits of the present map. It forms throughout its course the boundary line between Michigan and Wisconsin. The Paint flows into the Brule in sec. 12, T. 41 N., R. 32 W., and the Brule and the Michigamme unite in sec. 16, T. 41 N., R. 31 W., to ' A central Wisconsin base-level, by C. K. Van Hise : Science, new ser., Vol. IV, 1896, pp. 57-59, 219. A northern Michigan base-level : ibid., pp. 217-220. -The Indian names which the streams and lakes of this district formerly bore have either been dropped or else, in a few cases, have been replaced by translations, though most commonly they have been replaced by English names, which are altogether new. Those names which have been retained receive various spellings at the hands of dift'ereut authors, and even at the bauds of the same writer. The Michigamme Kiver, for example, is fre, p. li'6. 8 X ^ - =^M X - 175 feet, '{'he tew rock outcro])s are in all cases foiiiifl on the tops ami flanks of these hills, w here they have been exposed by denudation. At one jjoiut only has rock been found in situ near the river bed, and that is toward the mouth of the river. The conclusion is natural, since the river is 175 feet below these exposed rocks and has not reached rock, that it must be flowing' thr-Hiiih a i)reexisting depression or valle}- parti)- flUed by the drift of the Glacial epoch. The partial filHng of this valley at the time of the retreat of the ice to the northeast was accompanied by the filling of the depressions in the drift by the water flowing from the front of the melting glacier. After the depressions were filled, the overflowing water naturally followed the general southeastern slope, which exists throughout the area and is shown by the topographical maps and by the flow of the rivers. The immediate course of the water was determined by the former valley, which was not completely obliterated by the drift deposit. Drift barriers across the valley separating the ponded water, or lakes, from one another were cut through, the material eroded being spread over the bottoms of the lakes below. Thus was formed a chain of lakes, connected usually by narrow streams; the processes by which the channels were cut out and the lakes drained and filled up with the debris were going on at the same time. The result has been to obliterate the lakes to a great extent and to accentuate the char- acter of the stream. The final eflect of the processes, briefly outlined, would be to destroy the lakes entirely and produce a stream. By following on PI. VIII the Deer River from its mouth to its source, we may see the several stages in its development, which are also typical for other streams of the glaciated portions of the world. The river is about 20 miles long and has a width near where it enters into the Micliigannne of 20 to 30 yards. Near its mouth it is a slow-flowing, sluggish stream, which has nearly reached its base-level of erosion, and like many of the older streams of the Coastal Plain region of the United States is gradually filling portions of its channel with the silt and vegetable matter brought down from above. A short distance from its mouth it resembles such streams also in the HON xxxvi 3 34 THE CRYSTAL FALLS IRON-BEARING DISTRICT. meandering character of its channel. Tliis resemblance is still further enhanced by the presence of a remnant of a crescent-shaped cut-off, so characteristic of the old age of rivers. Just opposite this cut-off is a lake, which is of interest on account of its possessing two outlets, both leading into the river. Unfortunately this fact was observed on the topographical sheet too late to permit of a return to the field for the purpose of determining the cause of the presence of the two outlets. Passing up the stream we soon reach the lakes, which farther on become more numerous. The life history of these lakes is inseparably connected with that of the river. They reached maturity dui-ing or at the close of the Glacial epoch, and since that time their history is that of decline. This part of the history of these lakes may be brieflj' stated as follows: As the erosion continixes, the areas of water are reduced and the surrounding swanijD areas are correspondingly increased. If a lake were lai-ge and considerable inequalities existed in its bottom, two or more small lakes connected by the stream flowing through them may be formed. The final stage is a swamp, traversed by the slow-flowing river. The various stages in the history of the lakes are well illustrated on the accompanying map, P|. VIII, by the following series of lakes. In Nos. 1 and 2 the general character of such bodies of water, which may be con- sidered essentially as mere expansions of streams, is seen. No. 3, and Deer Lake, have long since reached maturity and are advancing rapidl}' to the point where they will each be separated into two bodies of water. No. 4 has already reached this stage, and in the swamp marked A we have the last stage, the swamp, with the stream flowing between peaty banks. On Light and Liver lakes, in the lower part of the Deer Ri-ver, we may see all but the last of these stages illustrated. The lakes are attached to the main river by very short streams. The main river after leaving the rapids above, where it accunmlates considerable detritus, enters a flat por- tion of its course partly occu})ied by the two lakes in question. Here, its rapidity being diminished, the stream deposits the detritus. Thus it has gradually built a delta, now for the most part covered by swamp growth. This tends to advance the shore line, and thus diminish the water area. The rapid cutting down of the barrier immediately below the lakes by the swiftly-flowing stream tends to lower the lakes and thus diminish their surface area still more. PHYSIOGRAPHY. 35 The couil)iiR'd effect of the draiiiiuy and tilHuf^ lias been to separate what was tormcrK- a long- narrow lake trending- NE-SW. into three rounded bodies of water, two of which are connected with each other, the larger of these two and the third bdcc being- connected with the main stream by very short necks. An artificial dam has been built across the narrow cliannel below the lakes, and the effect has been to flood the delta and unite the lakes into one large bod)' of water, occupying, approximately, the area covered h\ the glacial hike, thus restoring the conditions which existed before the natural barrier had been trenched. In the remainder of the course of the Deer River the tendency of other artificial dams to restore the river to its original condition, that of a series of connecting lakes, is well shown. These dams were built by lumbermen at the foot of the lakes or swamps when it was desired to retain a large body of water at these places. When, on the other hand, the desire was to enable the logs to pass rapids, a dam (marked B on the map) was built near the head of the rapids. The l)ack water would bring the logs to the dam, and on opening- the gates the flood would carry them over the rapids into the deeper water beyond. The Deer River thus, after having reached a somewhat advanced stage, has been rejuvenated by the Michigan lumbermen. A study of the small tributaries shows the same condition of things, although not on so large a scale nor so perfectly as in the main stream. The source of the Deer River is in the copious sjjrings which rise out of a spongy, marshy piece of ground less than 125 yards distant from Bone Lake, and about 20 feet below the usual water level of Bone Lake, and are really fed by the lake water percolating through the drift and appearing at this point. From the springs there is a depression which leads up to the lake. The highest point of this depression was about 3 feet above the normal water level of the lake. The outlet of Bone Lake is Ihe Fence River. The river leaves the lake at a point three-quarters of a mile distant from the head of the Deer River Valley. In order to obtain a supply of water for driving- the Fence River, Bone Lake has been converted into a reservoir. A dam was built at tbe outlet which raised the water about 4 feet, and the result was to turn some of the water of the lake into the Deer River, necessitating also a dam across this small valley near the lake shore. At present only a few strokes of the shovel would be necessary in order to turn the water of the flooded 36 THE CRYSTAL FALLS IROIST-BEARING DISTRICT. lake from tlie Fence into the Deer River, thus gaining for it a drainage area extending 7 miles farther north and including three large lakes, the main sources of tlie water suj)ply of the western branch of the Fence. I ha^-e no data which would enable me to show that the valley at the head of Deer River was ever a channel for the waters of Bone Lake. I am incUned to believe that such was not the case. For had it existed with the present slope, 20 feet in 375 feet, or even a much lower one, the water would have had a marked erosive power, and it would have cut back its channel mucli more rapidly than the Fence, which for a mile below the lake is a comparatively sluggish stream, and would have eventually captured Bone Lake and its feeders. The Deer River is still continuing the process of lengthening its chan- nel, and the springs which give it birth are gradually undermining the barrier at its head, so that it is possible that it will, unless artificially restrained, obtain much more water from Bone Lake than it does at present. A change in atmospheric and other conditions, which would insure a state of equilibrium Ijetween the incoming and outgoing waters, thus preserving the waters of Bone Lake at their present level, would be favorable for the final successful robbery of the upper Fence River system by the Deer River. This favorable condition, as may be readily seen, would be greatly increased in proportion as the increase of inflowing over outflowing water raised the level of the lake. TIMBER AND SOIL. The district was at one time very heavily timbered, with hard wood and ])ine, the former predominating on the whole. Along the flood j^lains of the large streams one finds sandy j)ine barrens where once there were heavy pine forests. On the headwaters the pine are found scattered through the hard wood. Individually these trees are very much larger and better tlian the thick and therefore smaller growth of the plains. Lumber- ing, which had been confined for years to the main drainage channels of the district, has of late been rapidly extended, following all the ramifica- tions of the tributary streams, until at present there remains in this district only a few years' cut of pine at the very headwaters of the rivers. Following the lumbermen comes the forest fire, which finds its most nourish- ing food in the dry resinous pine tops left by thein. The fires, once started. riJYSIOGliAPUY. 37 are not confined, however, to the cut pine, but .sjjread to tlie adjacent standing- ])ine and even into the hard-wood forests, carrying destruction with tliein, and leaving but the gaunt, l)are, and blackened truidcs to mark the sites of what were forinerh' thick forests. The i)ine-covered areas have a thin soil and are jtoorly adapted to agriculture. The areas covered with hard wood have, on the contrary, soil well adajtteil to the crops of the latitude. The advance of the lumberman has necessitated the damming and clearing of streams and the blasting of channels to permit the floating of the logs, and this has driven the fish> especially the sjjeckled trout, which formerly crowded all the streams, into the smallest and most inaccessible ones. Rutfed grouse, Bonasa umbellus, and deer are still rather plentiful in certain portions of the area, although the pot-hunter with set guns, spring nooses, and pitfalls is rapidly exterminating them. The deadh- character of such appliances is brought vividly to mind, when, as happened in my own case, one is suddenly arrested, while following a deer trail through the underbrush, by a hay wire noose around his neck, and he may be thankful if the bent sapling, having been bent so long- as to lose its elasticity, fails to spring up and render the device effective. CHAPTER III. THE ARCHEAN. DISTRIBUTIOlSr, EXPOSURES, AXD TOPOGRAPHY. The granite described in this chapter belongs to the oldest system in the district, and forms the western elliptical core designated on PL III as Archean. It is surronnded by sedimentary strata, which have a quaqua- versal dip away from the granite as a center. The portion of the Crystal Falls district, in which the granite outcrops, is about 13 miles long by 3 miles wide, its longest axis extending in a NW. and SE. direction and covering parts of Ts. 44, 45, and 46 N., Rs. 31 and 32 W. The exjjosures of granite are especially numerous in the southeast part of the o^•al area, where, owing to the proximity of large streams, the Fence and Deer rivers, and the consequent increased erosion, the drift has to some extent been removed. In the northwest part of the area, with rare excep- tions, all the rocks are deeply covered with drift. In general the topography of the area is that of the drift, but in the southern part it is seen to have been considerably influenced by the char- acter of the iinderlying rocks. The granite usually outcrops in small, rounded, and isolated knobs, whose relations to one another can onh' he conjectured. Where an occasional knob is composed of massive granite and more or less gneissoid granite, the exposed surface is so small as to prevent the observer from determining the relations between the two. Cutting the massive and schistose granite are certain long narrow masses of dark-colored rocks of rather fine gi-ain, and, with few exceptions, very schistose. From their geological occurrence it was concluded, in »\nte of their appearance, that they are dike rocks cutting the granite. The follow- ilig paragrajjli, quoted from the inanuscript notes of G. O. Smith, describes very cleai'ly their field occurrence: The gaps in this grauite ridge seem to indicate greenstone dikes, as here the granite usually has a facing of the greenstone more or less extensive, and often in the center of the gap there are several small areas of greenstone. In all cases the 38 RELATIONS OF THE ARCIIEAN. 39 greenstone is markedly more affected by weathering than is the granite. A stndy of the rehitions at the tew points of contact did not yield mncli more than negative resnlts, but these pointed to the intrusive character of the greenstones. UEIiATIONS TO OVERJOYING FORMATIONS. The relations of tlie granite to the sedhnentary rocks might be explained in two ways; the former may serve as the base of the latter rocks, or it may penetrate them. The occurrence of the granite in an elliptical sliape, with sediments surrounding it showing quacpiaversal dips, might be regarded as evidence of its intrusion in the Huronian sediments, and on this theory it would follow that the granite is of Huronian or post-Huronian age. If intrusive, it should be found to penetrate and metamorphose those sedi- ments. Against the intrusive character of the granite, and in favor of its pre-Huronian age, are the following facts: (1) There is a total absence in the surrounding sedimentary^ strata of any dikes which are lelated to the granite. (2) There is a total alisence of any metamorphic action, so far as observed, in the sedimentaries. (3) On the east flank of the granite core, on the west bank of the west branch of the Fence River in the SW. corner sec. 1, T. 45 N., R. 32 W., is a recomposed granite, which passes up into a tine sericitic quartzite, with false bedding. These rocks evidently derived their material from the granite, and hence mark the beginning of sedimenta- tion in this area. Thus the positive evidence confirms the negative, and since the granite underlies the oldest sedimentary rocks, whose age has been determined to be Huronian, the former is classified as Archean, that term being used here to designate those rocks of undoubted igneous character which form the foundation upon which rest the oldest determinable sedimentary rocks. It is not the province of this paper to enter into a speculative discussion of the origin of the Archean rocks of the district. For such a discussion the reader is referred to Professor Van Hise's exhaustive disquisition on the Principles of North American pre-Cambrian Geology,^ where the conclusion is reached that "the Archean is igneous and represents a part of the original criist of the earth, or its downward crystallization."'" The Archean has gradually reached the surface by the removal by erosion of the superjacent rocks. I Sixteenth Ann. Rept. U. S. «eol. Survey, Part I, 1896, pp. 571-874. - Loc. cit., p. 752. 40 THE CRYSTAL FALLS IRON-BEARING DISTRICT, PETROGBAPHICAIi CHARACTERS. The rocks of the Archean comprise biotite-granite, gueissoid biotite- grauite, and acid and basic dikes. BIOTITE-GRANITE (GRANITITE). The rock occupying the main and central ])art of the Archean area is a biotite-ffranite. This rock is also found to some extent on the border of the area. The rocks of this kind vary in color from light-graj' rocks to those having various tints of red, depending usually upon the degree of alteration. They vary also from medium to coarse grain. Some varieties show a decided porphyritic texture, and in some cases also an approach to a laminated structure. The porphyritic character is due to the presence of large crystals of feldspar, which stand out from the surrounding granitic groundmass, thus producing a typical granite-porphyry. The feldspar phe- nocrysts lie with their longer axes parallel, and thus help to produce an imperfect laminated structure. This parallel structure in the granite- porphyry is apparently analogous to the flow structure of the volcanic rocks, and probably was produced by movements in the magma before it had reached even a viscous state, as we find that the phenocrysts give no evidence of having undergone excessive mashing or torsion. The different textural varieties grade into one another in such a way as to indicate that the}^ are merely modifications of the same magma. In addition to these textural varieties, which are original, we find in certain places a passage from massive to schistose rocks, in which the schistosity is of dynamic origin, i. e., of secondary nature. In the thin sections these rocks show the normal granitic texture and the usual mineral constituents which characterize biotite-granites. The chief minerals are orthoclase, microcline, plagioclase, quartz, and biotite. Zircon and apatite are the accessory minerals present, and the secondary minerals include epidote-zoisite, chlorite, muscovite, rutile, and iron pyrites. Quartz occurs in grains forming the cement and molding around the other minerals. In one of the granites it has a peculiar saccharoidal char- acter macroscopically, and under the microscope such portions are resolved into very fine aggregates of quartz grains. The quartz is also frequently found in round blebs of varying size included in the best crystallized feldspar crystals. Thus the crystallization PETKOGKAPHIOAL CnAKACTEKS OF AKCHEAN.. 41 of the (|uartz, unless such qunrtz rej)reseuts the "([uartz tie corrosion" of the French autliors, conthiued through the entire time occupied by the crys- tallization of the feldspars, since it is included in the ohlest feldspar of the rocks, and also forms the matrix in which lie the youngest feldspars. Undu- lator\- extinction, so general in the quartzes, shows that the rocks have been subjected to pressure, and in some cases it has been sufficient to produce the extreme cataclastic structure of very greatly mashed rocks. The quartz includes numerous gas and fluid inclusions, the latter frequently with dancing bubbles and forming negative crystals, by means of which it is easy to orient the irregular grains. The quartz of one of the specimens was found to contain liquid inclusions, each of which, besides the usual bubble, held a small rectangular crystal. These crystals are trans- parent, with a light greenish tinge. A crystal similar in ajDpearance found in the same quartz individual is partly inclosed by a large (j -shaped bub- ble, and gave inclined extinction, though no further optical tests could be made upon it. Three kinds of feldspar are present : (1) A finely striated plagioclase ; (2) a feldspar, unstriated, or at most showing Carlsbad twins, and presumed to be orthoclase ; and (3) microcline, these last two being frequently inter- grown after the manner of pertliite. The plagioclase was the first feldspar to crystallize. It is invariably so altered that the twinning laminae are nearly obliterated, thus preventing accurate measurements. It is probably oligoclase; and if so, it is highly probable that much of the white mica produced by its alteration is pai'agonite instead of muscovite, a fact not determinable microscopically. The phenocrysts are orthoclase, usualh' in Carlsbad twins, and thus at first sight appear to have been the first feld- spar to crystallize; but I find that these ])henocrysts not uncommonly inclose small rectangular, more or less automorphic,^ crystals of plagioclase, which is in reality the oldest feldspar. Hence these orthoclases, notwith- standing their porphyritic character, are later than a part of the plagioclases. One phenocryst with Carlsbad twinning was observed in which one part of ' Automorph, Xenomorph; Uber die Eruptivgesteine iiu Gebiete tier Schlesisch-Maehrischen Kreideformatiou, by Carl E. 11. Kohrbach : Tsch. Jliu. Pet. Mit., Vol. VII, 188(3, p. 18. Idiomorph, AUotriomorph ; Rosenbusch : Mik. Phys., Vol. II, 1887, 2tl ed., p. 11. L. V. Pirssou has recently proposed in a paper, read before the Geological Society of America, on A Needed Terra in Petrology, the term anbedra for minerals which do not possess crystallographic outlines and are xeuomorpbic, in contradistinction to those which we properly call crystals and which are automorphio : Geol. Soc. Am. ,Vol. VII, 1896, p. 492, and Am. Jour. Sci., 4th series, Vol. II, 189G, p. 150. 42 THE CRYSTAL FALLS lEOX-BEARING DISTRICT. the individual shows microclinic striations. The other part was untwinned, and near the center of the phenocryst, bisected by the Carlsbad twinning- plane, was found a rectano-ular plagioclase crystal. The mierocline is usually the best crystaUized feldspar in the gTound- niass, and also by far the freshest. In the few cases in which it was observed in contact with plagioclase, the latter molded it, and is therefore older than the mierocline, which in its turn is older than the orthoclase. In one ease a mierocline individual showing the lattice structure over a portion of its surface possesses no twinning lamellaj in another portion, the twinning lamellpe fading until they totally disappear. Thus no sharp delimitation is apparent between the twinned and untwinned portions of the individual. In most slides all the feldspars are much altered, but even in those in which the mierocline is fresh the plagioclase and orthoclase alwaj's show alterations, the plagioclase altering most easily and usually being so changed that it is with difficulty that one can recognize the twinning lamellse. Hence some of them may have been taken for the nonstriated orthoclase. In an early stage of the alteration of the feldspars minute dark ferrite particles which impregnate them are hydrated, and this gives the feldspars a more or less distinctly red tinge. In a more advanced stage of alteration, muscovite and a httle epidote-zoisite are produced. Another alteration of the feldspar is alwavs associated with marked pressure phenomena, and hence is pre- sumed to be the result, partially at least, of dynamic action. This is the partial or complete granulation of the feldspar and the production from that mineral, with the addition from other sources of the iron and magnesia necessarv, of secondary white mica and quartz, and some biotite. It is highly possible that some of the small limpid grains considered to be secondary quartz are really an acid feldspar. Orogenic movements are also indicated by the bending of twinning lamellae, and were probably the partial cause of the twinning. Biotite occurs in plates, and as a rule shows better-developed crystals than does the feldspar, though it frequently occurs in decidedly ragged flakes. It is strongly pleochroic, showing absorption in the following colors: Pale straw yellow to yellowish brown, for rays vibrating perpendicular to cleavage, to very dark chocolate brown and greenish brown for those par- allel to cleavage. In the case of the biotite showing a greenish color this P1«:TK()GKA1'111CAL CHAUACTKRS of AKCIIEAN. 43 seems to be the result of ine'ipient alteration", since the edges of the flakes are ragu'ed, and in many cases almost the entire biotite of the section is altered to a chlorite, which shows ordinary white to li<>-ht o-reenish pleoch- roism, with the sinuiltaneous production of epidote and l)undles of needles with high single and double refraction, having yellowish or l)rownish color. These needles are taken for rutile. The biotite is found usually lying between the feldspar and quartz grains almost as though it had l)een tlie last product of crystallization. It has suffered crashing with the other minerals. Apatite and zircon were observed in a few crystals. No onginal iron ore was seen. As intimated above, by the use of the term "epidote-zoisite" the exact character of this secondary material is not always determinable. In some instances parts of an epidote crystal show^ the dee}) blue inter- ference color of zoisite, apparently indicating a mixture of the zoisite and epidote molecules, the latter predominating in the crystals.^ The remaining secondary minerals mentioned as occurring in the granite show their usual characters. GNEISSOID BIOTITE-GRANITE, BORDER FACIES OF GRANITE. About the central area of biotite-granite just described, and in part formina- the border of the Archean area, are rocks having a gneissic structure. With these are associated the biotite-granites. The gneissoid rocks in general are markedly darker in color than the granites, showing normally a rather dark gray. They vary little from one another in texture and are much fmer grained than the granites. The fine-grained condition of these schistose and banded rocks has perhaps a great deal to do with their dark color, though this is primarily owing to the amount of biotite present. In some of the specimens the bands can be readily distinguished under the microscope, and are seen to contain a white mica and a nxuch smaller amount of biotite. These two minerals are present in fine films between the crushed quartz and feldspar grains, gi\'ing to the rocks a very decided schistose character. These mica folia are much more numerous in certain areas than in others, thus producing a more or less perfect banding. The mica plates are not all regularly parallel, although ordinarily having a 'On some granites from British Columbia and the adjacent parts of Alaslca and the Ynkon district, by F. D. Adams : Canadian Record Sci., Sept., 1891, p. 3+6. 44 THE GEYSTAL FALLS lEON-BBAKING DISTKICT. tendency to this arrangement, and are usually parallel to the banding. The most perfect schistosity is thus developed parallel to the micaceous bands. The banding and the schistose structure are plainly of secondary oris'in, the result of dynamic action. Others of the gnessoid granites, however, when examined under the microscope, are decidedly massive, and it is only on a large scale that the banding shows distinctly. In such cases the cause of the banding could not be determined, and might by some be ascribed to differentiation, though, from the association of these gneissoid granites with those just described, it is assumed that the banded str icture is due to dynamic action. If this be the case, however, a complete recrystallizatiou has taken i)lace, and slight dynamic effects are now shown. The strike of the banding, wherever it was taken, was uniform, varying from N.-S. to nearly N. 45'^ W., agreeing, on the whole, with the trend of the Archean oval area. The microscope shows that the constituent minerals of the gneissoid granites are the same as those which compose the granites just described. These show also the same relations to one another and the same general char- acters as in the granites, except where mashing has completely obliterated the original texture, and hence no further description of them is necessary. The crushing to which the gneissoid granites have been subjected is very clearly shown in the present cataclastic condition of the quartz and feldspars. As stated above, both the gneissoid granite and the granite proper are found in the border area of the Archean. In those rocks in which the con- tact shows a gradual transition from the banded rock to the unbanded, the micaceous bands are clearly secondary, and are the result of the crushing of the original granite, these lines representing macroscopic and microscopic shearing planes along which the feldspar and quartz have been thoroughly granulated, and sericite and some biotite produced, as was found to be the case also in some of the granites. These rocks thus agree in their dynamic origin with a similar but apparently more extensive and better developed gneissoid border facies in the Morbihan (Brittany) granites, which have been described, and whose origin has been so cleai-ly demonstrated by BaiTois.^ Numerous other similar cases have been described recently from the Canadian granite massifs and from Sweden and other districts. ' Anu. Soc. G6ol. du Nord., 1887, p. 40. ACID DIKES IN ARCIIEAN. 45 ACID DIKES IN ARCHEAN. Observations upon diki's of acid rocks {•uttini^- the Arcliean <»-r;uiite are very i'ew, and we may suppose this to be partly due to tlieir occurrence in isolated knobs, which prevented the determination of the relations of adjacent exposures of rocks of sliji'litly different character. Some few dikes were, nevertiieless, ol)servetitueiits of tliu schistose eruptives, arranged according to tlieir rehitive importance, arc liiotite, liornlilende, clilorite, (juartz, feld- spar ('?), cah'ite, epidote, iron oxide, sphene, and ninscovite. Tile clear limpid grains whicli form the gronndmass are nndoubtedly for the most jiart (jnartz. No satisfactory results were obtained in the tests for feldspar, but it is highh' probable that some is associated with the quartz. Dark chocolate-brown to light-brown biotite is almost an invari- able constituent. In some cases it is aQCompanied by a little chlorite, which apj)ears not to have been derived from the biotite. In a few rare instances biotite is absent altogether, chlorite taking its place. The biotite and chlo- rite are usually found between the quartz grains. They have a parallel arrangement, and this gives the rock its schistosity. Biotite and epidote are found included in the grains of quartz of the groundmass. Muscovite is rarely present, but when found is in medium-sized automorphic plates. Ragged pieces of ore, either ilmenite or titaniferous magnetite, and sphene, secondarv to these, are found in almost all specimens, and in a few instances iron pyrites was observed. Calcite is invariably present in irregular, fairly large grains, almost equaling the quartz in quantity. Epidote is found in large qutmtity, both in crystals and in irregular grains, the crystals occurring among the bunches of biotite and included in the grains of quartz. The large amount of epidote in association with the calcite seems to point to the very basic character of the feldspar of the original rock. A bluish-green hornblende is rather frequently associated with the mica. In rocks in which the hornblende predominates mica is always pres- ent, but the reverse is not true, the most micaceous rocks being entirely free from the hornblendic component. The hornblende is found in lai'ge prismatic individuals without terminal faces. This mineral contains some of the other constituents of the rock in which it is found, such as quartz, epidote, and more rarely ircin oxides. The interspaces between the hornblende crystals are filled with irregular biotite flakes and with grains of quartz, epidote, and iron oxide. This hornblende is apparently one of the last, if not the last, mineral to develop. The hornblendic rocks are not nearly so schistose as the micaceous ones. The secondarv origin of the hornblende is clearlv shown in one of the sections which is traversed by a fissure; the hornblende can be seen extend- ing into, and in places crossing, this fissure. The other minerals are 48 THE CRYSTAL PALLS mON-BEARING DISTRICT. presumed to be secondary, but this can not be proved tor them. The schistose character of the rocks is evidence of dynamic action. The pres- ence of undulatory extinction was noticed in the quai'tz of some specimens, but its absence is the rule. However, from the absence of great pressure phenomena, and the remarkably fresh condition of the minerals composing the basic rocks, which contrasts strongly with the generallv altered condition of the minerals of the more refractory acid rocks including tliem, it would appear that complete recrystallizatiqn has occurred.^ The schistose structure can undoubtedly be referred to the dynamic action which resulted in the upturning of the sedimentaries and caused the de^■elo[)ment of schistosity in certain portions of the border of the granite. This dynamic action was in all probability also the chief force in the pro- duction of the secondary minerals. The schistosity of the dikes does not agree in direction with the gen- eral strike of the schistosity throughout the entire district, but is always nearly pai-allel to the long extension of the dikes. These dikes represent belts of weakness, and it is therefore natural that the movements should occur along these belts rather than across them. This schistosity of the dikes also furnishes a slight clue as to their age. Younger than the granites they cut, they must have occupied their present position at the time the dynamic revolution took place which resulted in the development of schistosity in the granite, as well as in the sedimentaries. It is impossible to bring the date of their intrusion within narrow limits. It seems very probable, however, that they were formed at the time of the extrusion of the basic Hemlock volcanics, though it is impossible to proye their connection with them. (2) MASSIVE DIKES. The only dike rock which retains to some extent its original texture is a much-altered medium-grained dolerite (diabase). The alterations it has undergone are those usual for such basic types of rock, and this one exhibits nothing peculiar or of special interest. An ophitic texture, while still recog- nizable, is more or less obscured by the uralite which has developed out of the pyroxene. The remnants of the original plagioclase feldspar ])resent show exceedingly slight pressure effects. The alteration processes would 'Principles of North American pre-Cambriau Geology, cit.,pp. 706-707. RfiSUMfi OF AIJCHEAN. 49 tlierefore seem to have been due to the action of percolating water, without special mechanical influence. Hence we may date the intrusion of tliia particular dike after the orogenic movements which affected the granite core, rendering portions of it schistose, and crushing all of it to a greater or less extent. These movements are presumed to have taken place just prior to or during Keweenawau time; and therefore the age of this dike is Keweeuawan or post-Keweenawan.' In the above- described granite massif we have a rock of pre-Huronian ao-e, as shown by its relations to the overlying sedimentaries. It possesses in general a coarse granular, and in places porphyritic texture. Along its border it contains portions which are much finer grained, darker than the rest of the mass, and very well banded. The boundaries between the banded rock and the granite at times are sharp, but frequently are very indefinite. This banded schistose portion is found to be due to pressure, causing the gradual passage from the granular granite to the gneissoid, schistose granite. One instance of undoubted inclusion of gneissoid granite by a true granite was observed. If the gneissoid granite was derived by pressure from the Archean granite, then the particular granite dike which includes the fragments must be of later age than the great mass of granite of the Archean area. The Ai'chean is cut by basic dikes of two ages. The earlier ones were rendered schistose, and the production of this secondary structm-e was accompanied by a total obliteration of the primary igneous texture and the production of a large amount of mica and hornblende. All the dikes were probably injected at the time of the volcanic activity when the vol- canics of the higher series were ejected, but no proof of their connection can be produced. They were, however, injected before the folding of the .area took place, as shown by their having been rendered schistose by it. A single dike belonging to the later series was studied. It is massive, and therefore was irrupted after the folding which produced the schistosity in the earlier series of dikes. It belongs probably to a Keweeuawan or post-Keweenawan period of eruption. ' For a discuasiou of the orogenic movements which aft'ected the Crystal Falls district, the reader is referred to p. 158 et seq. 3I0N XXXVI 4 CHAPTEE IV. THE LOWER HURONIAN SERIES. This series is represented in the Crystal Falls district by the following formations, given in order from the Dase upward: The Randville dolomite, the Mansfield slate, and the Hemlock formation. At the beginning of the deposition of the Lower Huronian series the entire district was covered by the pre-Cambrian sea, with the possible exception of a small island in the Archean area. SECTION I.— THE RANDVILLE DOLOMITE. The best exposures of this dolomite are found near the center of the district east of the western ellipse and in the extreme southeastei-n part of the district in the Felch j\lountain range. Both areas are described by Smyth, to whom we owe the name, and the reader is referred to his descrip- tion on p. 406 and p. 431 for the detail characterization of the formation. It will suffice for our purpose to state that it is a medium-grained crystallme dolomite. The few outcrops which I shall mention are important as showing the relations of the formation to the underlying rock, but are, petrographically considered, rather exceptional phases of the formation. Hence my descrip- tion will be brief. DISTRIBUTION, EXPOSURES, AND TOPOGRAPHY. The area in which the Randville dolomite immediately underlies the drift is a continuous zone adjacent to and surrounding the Archean core. The belt varies slightly in width along the sides of the ellipse. At the ends it is two or three times the width at the sides. This is due to the lower dip of the beds at the ends. Exposures are found m the area studied by me only on the northeast and southwest flanks of the granite core. The west branch of the Fence River follows the limestone area for a short distance in the northeastern part of the district, skirting the Archean 50 PETKOGHAPIIICAL CUAUACTERS OF KANDVILLE DOLOMITE. 51 • -•raiiito. On the whole, however, the Kiiudville (loh»inite has hud uo uuirked effect on the topoyrapliy or the drainage. PETROGRAPIIICAI. CHARACTERS. Ill general the Randville dolomite consists petrographicall}' of a fine- grained dolomite, with some quartz. This grades down through a calca- reous quartzite by increase of quartz into a true quartzite. The nearer the granite, the more quartzitic is the formation. At the southeast corner of sec. 2, T. 45 N., R. 32 W., on the west bank of the west branch of the Fence River, is a ver}- good exposure of the quartzite. Its derivation from the underlying granite is here shown. The rock is a very fiue-grained, almost novaculitic, quartzite. It shows current bedding in some places, though no true bedding was observed. Immediately below this quartzite is a very schistose rock, in which one can readily distinguish macroscopic- ally rounded to lenticular quartz areas, with masses of sericite flakes between them. The contact between tlie quartzite and the schistose rock seems very sharp when viewed from a short distance, but is found to l)e indefinite when closely examined. A close search was made along a con- tact for pebbles from the granite, but such were not found. However, small rounded pieces of vein quartz, most probably derived from the granite, were observed. The schistose rock in its turn grades down into a graj-ish granite, which is also more or less schistose. We have here evidently a transition from the granite, through the intermediate schistose recomposed granite, to the true sedimentary rock above. The meaning of this transi- tion is considered below. Under the microscope the cause of the schistosity of the rock inter- mediate between the granite and the quartzite is plain. Quartz and sericite, with some feldspar, are alone present in it. The quartz is grayish and granulated, and mashed out into oval areas representing original quartz grains. Various fragments constituting the areas are, however, angular and more or less equidimensional, and when not so never have a definite orientation of their longer axes. Between these large areas, but not between the individual small fragments constituting the areas, sericite is abundant. When the sericite is not predominant, the flakes lie in a fine mass of quartz grains, each of which agrees in long direction with the mica plates and large oval quartz areas. The sericite flakes are both included in this quartz. 52 THE CRYSTAL FALLS IRON-BEAfiING DISTRICT. and also lie between the grains. In one instance fragments of the original feldspar were found in the midst of such an area. These quartz-sericite areas are unquestionably of secondary origin, and the minerals have devel- oped in connection with pressure. They were probably produced from feldspar which existed in the original gi-anite. Whether this schistose rock was formed from a weathered but not transported granite, from an arkose or feldspathic sandstone, or from the solid granite, it is impossible to say. A similar sericite-schist which developed from recomposed granites has been described by Van Hise as occurring at several localities in the Marquette district.^ In these cases at places the fragmental characters are still sufficiently clear to admit of the statement that the rocks are sedimentary. In the Crystal Falls rock mashing has destroyed all original characters. The rock occupies an intermediate posi- tion between a metamorphosed sedimentary and a metamoiphosed eruptive, and grades on the one liand into the sedimentary and on the other into the eruptive. This makes it impossible to say whether it belongs exclusively to the one or to the other, or in part to both. Similar relations . in other parts of Michigan were explained by Rominger" as cases of progressive metamorphism of sediments, the granite being supposed to be the extreme stage of alteration of the sedimentary rock. Later the finding of basal conglomerates at or near these localities has shown conclusively that this explanation is incorrect, and it has been abandoned by Rominger. The quartzite, which immediately overlies the rock of doubtful char- acter, is composed of angular grains of quartz, between which are plates of sericite which have an imperfect parallelism, thus giving a certain degree of schistosity to the quartzite, possibly enough in places to warrant its being called a quartz-schist. The rock shows no conclusive microscojjical e\'idence of a sedimentary origin, but differs from the cherts, with which it might be confused, in the size of the grains and in the presence of sericite. This rock was originally probably chiefly composed of quartz sand, with some feldspathic material from the disintegrated granite. Coincident with the pressure which produced the striking schistosity in the underlying rock, this sand was also mashed, resulting in the production of sericite and quartz ' Mou. U. S. Geol. Survey Vol. XXVIII, 1897, p. 226. -The Marquette irou district, by Carl Rominger: Geol. of Michisau, Vol. IV, Parti, 1878-1880, pp. 15-52. lM:TROGKAPniCAL CHARACTERS OF RANDVILLE DOLOMITE. 53 trom the feldspiir and in the rrusliing- of the quartz grains, thus completely destro}'ing- the rounded chistic grains and oljliteratiug all the sedimentary character of the rock, except the macroscopic structure of cun-ent bedding. On the west side of the granite ellipse, at N. 1750, W. 1550, sec. 12, T. 44 N., R. 32 W., about 100 yards from the granite, to the nortli, and lower down on the slope of the same hill on which the granite is found, is found a carbonaceous quartzite or quartzose dolomite. The strike is N. 25°-35° W. The surfiice only is seen, so that the dip could not be taken. Microscopical examination shows the rock at the eastern side of the exposure to be made up of quartz grains held together by a fine-grained carbonate cement. This grades up to the west by increase of calcite and correspond- ing diminution of quartz to a quartzose dolomite. At N. 500, W. 1550, sec. 1, T. 44 N., R. 32 W., one-fourth mile distant from the granite, is seen another outcrop of a very dense quartzose dolo- mite, appearing macroscopically almost like a vitreous quartzite, but really with just enough quartz grains in it to enable the qualifying term "quartzose" to be appropriate!}- used. The brown ferruginous crust on the weathered surfaces point to a percentage of iron in the magnesium-calcium carbonate. The pure limestones are to be sought slightly farther away from the Archeau shore, where the conditions were more favorable for the production of a pure nonclastic sediment. REIiATIOlSrS TO UNDERLYING AND OVERLYING FORIMATIONS. At only the one place cited above has a contact between the granite and the Randville dolomite been found. It is probable that unconformable relations exist, even though no basal conglomerate has been discovered as evidence of wave action on the Archean coast. Relations between the Randville dolomite and the overlying- forma- tions have not been observed in the part of the district studied b}- me. THICKNESS. Reliable data for estimating the thickness of the Randville dolomite have only been obtained in that area surveyed by Smyth. (See p. 433.) According to his estimate, the formation possesses a maxinunn thickness of 1,500 feet. 54 THE CRYSTAL FALLS lEON-BEARmG DISTRICT. SECTION II.— THE MANSFIELD SLATE. The formation of the Lower Huronian, which is next higher than the Randville dolomite, is composed of sedimentary beds, in which a' slate pre- dominates. This formation is fonnd in its most t^q^ical development in a narrow valley through which the Michigamme River flows, and in which the village of j\Iansfield and a mine of the same name are situated. The valley and the slates are well known in the Crystal Falls district on account of their eco- nomic importance. For this reason the name "Mansfield slate" is here applied to this formation. DISTRIBUTIONS, EXPOSURES, AXD TOPOGRAPHY. The part of the valley occupied by the Mansfield slates begins at the northern section line of sees. 17 and 18, T. 43 N., R. 31 W., and extends due soiith for 3 miles to the southern section line of sec. 29 of the same township. The slate belt is widest at the north, being over one-fourth mile wide on the westren side of section 17. To the south it gradually dimini.shes in width, until it finally disappears in sec. 29. The strike of the sedimentary rocks is almost due north-south, except in a few places where the rocks have been gently flexed and the strike varies a few degrees. The dip is high to the west, ranging from 65° to 80°. The influence of the Mansfield slate belt upon the topography is strikingly shown by the depression in which the slates are found, and which contains the Michigamme River. The slates are surrounded on all sides by igneovis rocks which form fairly high hills, those to the west being composed of rocks of volcanic origin, those to the north, east, and south being intrusive, and later than either the sedimentaries or the vol- canics. The Michigamme River flows south through sec. 1, T. 43 N., R. 32 W., and meets the east and west ridge of intrusives in the northeastern part of sec. 12 of the same township and range. It cuts through this at an oblique angle, changing its course to the southeast In sec. 7, T. 43 N., R. 31 W., it leaves the intrusives and penetrates a short distance into the volcanic rocks, their contact not being able to cause a change in the course of the river, owing to the slight diff'erence in resisting power between the intrusives and the volcamcs. Still flowing to the southeast, it finds at THE MANSFIELD SLATE. 55 the Michigamnu' (lain, on the section line between sees. 7 and 18, near the southeastern and northeastern corners, respectively, the contact between the three kinds of rock, the sedimentaries, the volcanics, and the intrusives. Where the water leaves the eruptive and enters the sedimentary area the more easily erodible nature of the rocks of the latter is well shown by the ■falls which have been formed, the volcanics constituting the barrier over which the water plunges into a deep basin worn from the slates. Crossing the slates in the same direction, i. e., southeast, the river strikes squarely ao-ainst the intrusive dolerites and is deflected to the south, following the contact between the two rocks for a short distance, then gradually working to the west into the center of the sedimentary area, the river takes an almost directly southerly course, with only minor bends. In the slates the river has fairly low flat banks on both sides. In the southern portion of the area the valley is narrower, owing to the progressive narrowing of the sedimentary belt. As soon as the river leaves the Mansfield slate belt, it resumes the sinuous course it had before the Mansfield belt is entered, and flows between high banks through the intrusives, out through the sand plains near Lake Mary. POSSIBLE CONTINUATION OF THE MANSFIELD SLATE. In sec. 10, T. 44 N., R. 32 W., about 7 miles northwest of the extreme northern end of the Mansfield area of slate, there are one or two exposures of much crumpled interbedded brown and black slates. Their strike is about N. 16°-20° W., but owing to their plicated condition the dip varies from 55° southwest over to 85° northeast. The average dip, however, is presumed to be to the southwest, which is in accord with the general structure of the area. The slate exposures are surrounded by coarsergrained basic intrusives, dolerites, which outcrop within short distances on all sides. The nearest sedimentary beds are quartzose dolomite ledges which outcrop 1 J miles to the east, in sees. 1 and 12, T. 44 N., R. 32 W., rather close to the Archean granite. A section across the Lower Huronian rocks at this point shows the Archean granite overlain by quartzose dolomite, which is in its turn overlain b3' the slates. The relations which these rocks bear to one another are those which similar ones bear to one another near Michigamme Moun- 56 THE CRYSTAL FALLS IKON-BE ARING DISTRICT. tain/ and the slates of the two areas are consid red to he of the same age. Since the slates correspond stratigrai^hically to the slates of the Michigamnie Mountain and to those of the Mansfield area, they have been connected ou the map with the slates of Michigannne Mountain by a narrow belt included between dotted lines; but this belt is not based on any connecting exposures. These two ledges of slate are taken as the northernmost outcrops of the Mansiield slate formation, although a number of miles north and in direct continuation of them along the strike there was found a single doubtful out- crop of a graj'wacke, showing neither strike nor dip. Whether it represents a shallower water deposit contemporaneous with the slates it is impossible to say. However, on such slight evidence it was not deemed advisable to continue the slate belt to this point. PETROGRAPIIICAL CHARACTERS. A petrographical description of the Mansfield slate belt must neces- sarily be very brief, owing to the small area and to tlie scarcity of the exposures. The rocks of the Mansfield slate belt are graywackes, clay slates, phyllites, siderite-slates, cherts, ferruginous cherts, and iron ores, with the various rocks which have been derived from them l^y metamorphism. They vary from coarse-grained rocks to very fine grained slaty ones. The latter predominate, and for that reason this belt is called a "slate " belt. The color of the rocks varies from an olive green and purplish lilack to bright red for those which are very ferruginous and more or less altered. The ordinary detrital rocks may be divided into the coarser and the finer kinds. The first are the graywackes, and the second are the ordinary clay slates and ^shyllite. There is, however, a gradation from the one to the other. GRAYWACKE. The graywackes consist largely of grains of quartz and feldspar of unquestionably detrital origin. Associated with these is a large amount of mica, chlorite, and actinolite, with invariably more or less rutile. This last is in minute grains as well as in crystals. Many of the crystals show fine knee twins, triplets, and more rarely, heart-shaped twins. Tourmaline ' See Part II, Chapter IV, Sec. IV, by H. L. Smyth. I'ETROGRArillCAL COAHACTEKS OF MANSFIELD SLATE. 57 is sometimes pirst'Ut. The feiTO-niagnesiau minerals dc-velop cliiefly from the alteration of the feldspar, and from the finer detritns which is jjresmiied to have existed between the grains. As a consequence, the secondary minerals lie between the original grains. Many of the quartz grains are enlarged, and here the secondary minerals are included in the new areas of the enlarged grains. In numerous cases the new quartz occupies about as much space as the original grains themselves. This shows very clearly the porous character of the original sandstone. All original grains of the rocks show signs of extensive mashing. Some specimens contain a large amount of tourmaline in long slender crystals, which penetrate both the feldspar and the quartz grains. The presence of tourmaline is especially interesting as indicating that these sedimentaries may have been subjected to a certain amount of fumarole action. According to the proportion in which the vai-ious minerals have developed, we obtain sericite-, actinolite-, or chlorite- schists produced from the graywackes. CLAY SLATE AND PHYLLITE. The clay slates are dull and lusterless and are black, olive green, or red in color. They are usually impregnated with more or less iron pyrites in large macroscopical crystals. One can distinguish in them quartz, Avhite mica, a few needles of actinolite, rutile, hematite, with a small proportion of a dark ferruginous and carbonaceous interstitial material. The amount of iron which these clay slates contain varies considerably. In some, hematite is present in such quantity as to cause the slates to be appropriately called hematitic slates. Such, for instance, is the one forming the foot wall of the Mansfield ore body. The iron oxide gives to the slates a very bright red color where they are weathered. Tliese weathered hematitic slates are very commonly known in the district as red slates, or as "paint rock" or "soapstone," though rocks of very different character are also at times designated by these names. The phyllites have a silky luster and a bluish-black color. They are composed essentially of white mica quartz, some feldspar, innumerable minute crystals of rutile and dark ferruginous specks. These seem to differ from the rocks called here clay slates only in that they are more completely crystalline, the interstitial material of the slates having disappeared. 58 THE CRYSTAL FALLS IRON-BEARING DISTRICT. ORIGIN OF CLAY SLATE AND PHYLLITE. The origin of the clay slates of the Mansfield formation is probably to be looked for in the disintegration and decay of the Archean granite, and the subsequent metamorphisra of the resulting clay. For between the srranites and the slates no other rock masses are known to have existed from which the clay could have been derived. The phylUtes are presumed to have resulted from the metamorphism of the clay slates. PRESENT COMPOSITION NECESSARILY DIFFERENT FROM THAT OF ROCK FROM WHICH DERIVED. It is a well-recognized principle of rock weathering that in the altera- tion of rocks near the surface of the earth there is a relatively rapid diminution in the quantity of the more soluble constituents. Hence a clay shows a lower percentage of alkalies and alkaline earths than is found in the parent rock, with an increase in the percentage especially of alumina and water. This relation is made clear by Adams in a statement of the comparison of the composition of certain slates and granites:^ "On com- paring the analyses of a series of granites with those of a series of slates, as, for instance, those given in Roth's ' Gesteins Analyzen,' the latter are seen to be on an average considerably higher in alumina and much lower in alkalies, while at the same time they are lower in silica, which has been separated both as sand and in combination with the alkalies which have gone into solution, and in most cases contain more magnesia than lime instead of more lime than magnesia, as is usual- in granites." Adams con- cludes further, after a comparison of the alkalies in the slates and granites, that "The slates thus contain on an average about two-thirds of the amount of alkali present in the average granite."^ An examination of series of analyses of granites shows that while the percentages of soda and potassa vary considerably, now the one being predominant, now the other, on the whole in the typical granites the potassa is higher than the soda.' This is the relation which we would expect in the case of an ideally pure granite. 1 A further contribution to our knowledge of the Laurentiau, by F. D. Adams: Am. Jour. Soi., 3d aer., Vol. L, 1895, p. 65. "Loc. cit.jp. 65. ■^'Zirkel states that iu the weathering of granites the soda is much more readily removed than is the potassa; Lehrbuch der Petrographie, Vol. II, 1894, p. 32. PETROGRAPDICAL CnARACTERS OF MANSFIELD SLATE. 50 ill wliicli no aiiorthoclase replaces the orthoclase. As a consequence of tlie easier solubility of the soda, this relation between the two alkalies, soda and potassa, is maintained, and is often made more strildng in the clay slates. An average of 31 analyses of clay slates taken from various sources shows two and one-half times as much potassa as soda. In the case of the Mansiield slate this difiPerence has been increased, so that there is ten times as much potassa as soda present. ANALYSIS OF MANSFIELD SLATE. Mr. George Steiger, of the United States Geological Survey, has prepared a complete analysis (No. 1 in the following table) of a typical specimen of the Mansiield clay slate. Analyses Nos. 2 and 3 were pre- pared by W. Maynard Hutchings,' and numbered by him Nos. 2 and 5, respectively. Analysis of the Mansfield clay slate. Constituent. 1. 2. 3. SiO- 60.28 .69 22.61 2.53 .45 Trace. .13 .04 1.35 5.73 .54 .60 3.62 .03 None. .97 59.28 53.57 TiO. Al,03 21.85 1 5.80 24.53 6.51 FeaOa FeO MnO CaO .45 .76 BaO . . MeO 1.24 4.13 1.18 I 6.25 1.81 4.34 .97 7.65 K,0 NajO H.Q at 100 HiO above 100^' PjO, CO. . . . c Total 99.57 100. 18 100. 12 COMMENTS ON ANALYSIS. That which is the most striking about the analysis is the relative pro- portion of the alkaline earths, lime, and magnesia, the latter being present 'Notes on the composition of clays, slates, etc.. and nn sonip points in their contact metamor- phism: (ieol. M.ag., Vol. I, 1894, p. 38. 60 THE CRYSTAL FALLS IRON-BEARING DISTRICT. iu the greater quantity. As a rule, in all of the igneous rocks (and to the igneous rocks all clay slates owe their ultimate origin), except in the nonfeldspathic ultrabasic ones, the reverse condition exists, namely, the magnesia subordinate in quantity to the lime. The difiference in amount of soda and potassa is very striking and should be noticed, in view of cer- tain points to which attention will be called in subsequent pages. The percentage of alumina is higher than is usital in the clay slates. It will be noticed that considerable water is present, but in consideration of the char- acter of the rock this is to be expected. If anything, the value is rather lower than would be expected, indicating a possible loss of water due to the rock having already undergone some dynamic action. The carbon present is considered as offering trustworthy evidence of the presence of organic life at the time of the deposit of the slates, though no more satisfactory evidence of the existence of life has been found. COMPARISON OF ANALYSIS OF MANSFIELD CLAY SLATE WITH ANALYSES OF CLAYS. During the last few j^ears there have appeared in the Geological Magazine, from the pen of Mr. W. Maynard Hutchings, some very elaborate and suggestive articles upon the composition of clays, shales, and slates, and from one of these ^ I have taken two analyses of Carboniferous clays for comparison with the Mansfield clay slate. These two analyses, Nos. 2 and 3, p. 59, are from the very fine grained clays, in which the quartz was not distinguishable with the microscope, and are the analyses showing the highest and lowest percentages of silica. Mr. Hutchings says of his analyses that the samples were di-ied at 220° F., and that the titanic oxide was not determined but is contained in the silica and alumina. Concerning the clays, he writes: From these analyses it will be seeu that these clays would be capable, chemically considered, of transformatiou iuto very typical " clay-slates." Mineralogically they are clay-slates, having already undergone all, or nearly all, the mineral changes requisite to constitute the normal (unaltered) slates. Nothin.g more is needed but physical changes, such as compacting, arrangement of mica iu a plane, increase of size of mica, etc.^ The great similarity of these clays with the Mansfield clap slate is very evident. The only material difference which exists between them is in the 'Notes on the compositiou of clays, ."ilates, etc., autl on some points iu tlieir contact metamor- phism, by W. Maynard Hutchings : Geol, Mag., Vol. 1, 1894, p. 38. •Loc. cit., p. 38. rETllOGUAPUlOAL CHAKACTERS OF MANSFIELD SLATE. 61 liiyhor pLTcentiii^-e of water (•(tntaiued in tliu clay.s. This difference is natural, clays iisuall}' containing- about twice as much water as do the slates. COMPARISON OF ANALYSIS OF MANSFIELD CLAY SLATE WITH ANALYSES OF OTHER CLAY SLATES. In tlie following talkie there are given, for purposes of comparison with the Mansiield clay slate, analyses of typical clay slates, roofing slates from the Cambrian of Vermont and New York. Analyses of tyjncal clay slates. Constituent. SiO^ TiO, AUG, Fe.20, FeO MnO CaO BaO MgO KjO NasO H,0 at 100^ ... H2O above 100^ P2O5 CO; FeS, C Total 60.28 .69 22.61 2.53 .45 Trace. .13 .04 1.35 5.73 .54 .60 3.62 .03 None. .97 99.57 62.37 .74 15.43 1.31 5.34 .22 .77 .07 3.14 4.20 1.14 .34 3.71 .06 .87 .06 Trace. 3. 59.70 .79 16.98 .52 4.88 .16 1.27 .08 3.23 3.77 1.35 .30 3.82 .16 1.40 L18 .46 99.80 100. 05 67.61 .56 13.20 5.36 1.20 .10 .11 .04 3.20 4.45 .67 ((.45 ft 2. 97 .05 None. .03 100.00 67.89 .49 11.03 L47 3.81 .16 1.43 .04 4.57 2.82 .77 a. 36 6 3.21 .10 1.89 .04 100. 08 aHjO at 110-. b U2O above 110^. No. 1. Black slate, Bp. 32497, N. 450, W. 1620, sec. 17, T. 43 N., R. 31 W., Michigan. Analyzed by George Steiger. No. 2. Sea-green slate, Griffith & Nathaniel Quarry, South Poultney, Vermont. W. F. Hillebraud. No. 3. Black slate, American Black Slate Company, Benson, Vermont. W. F. Hillebraud. No. 4. Red slate, three-fourths mile south of Hampton Village, New York. W. F. HiUebrand. No. 5. Cireen slate, three-fourths mile northwest of Janesville, Washington County, New York. W. F. Hillebraud. Nos. 2, 3, 4, and 5 taken from Analyses of rocks and analytical methods, 1880-1896, Clark and HiUe- brand: Bull. U. S. Geol. Survey, No. 148. Nos. 2 and 3 are, respectively, C and F, p. 277, aud Nos. 4 and 5 are A aud D, p. 280. 62 THE CEYSTAL FALLS lEOK-BEAElNG DISTEICT. The strong similarity between the composition of these clay slates is at once apparent, and needs no further comment. The only marked differ- ence between the Huronian clay slate and the Cambrian ones is the higher percentage of alumina present in the former. . SIDERITE-SLATE, CHERT, FERRUGINOUS CHERT, AND IRON ORES. The two most interesting kinds of rock from the Mansfield slate belt are those known as the siderite- or sideritic slates and the cherts or ferrugin- ous cherts, according to the quantity of iron carbonate and iron oxide present. These alternate with each other, and are found also interstratified with the fragmental slates, and thus there can be no question as to their sedimei^tary character. The siderite-slates are of a light to dark gray color. They are well laminated, and in some places cleave rather readily along the laminae, though at other places they break with an almost conchoidal fracture. The weathered siderite slates are covered by a crust of reddish- brown hydrated iron sesquioxide. Microscopically the siderite slates are composed of siderite, or of sider- ite and exceedingly fine grained cherty silica. Roundish rhombohedra of siderite compose the purer sideritic portions. If one passes from the pure to the less pure slates, the siderite gradually diminishes in quantity, the silica grains increase correspondingly, and the rock grades into the chert which, in bands, is commonly associated with iron carbonate in the Lake Superior region. As the carbonate alters to the oxide or hydrated oxide ferruginous cherts are produced. The cherts are white to red, depending on the amount of iron oxide present. The manner in which the siderite alters to limonite and hematite, and the various steps of the process have been so well described and beautifully illustrated in Monograph XXVIII, that the reader is referred to that volume for further information. None of the brilliant red jasper or jaspihte, such as that found in the Marquette district, is associated with the Mansfield slates. Iron ores of economic importance, however, are found associated with these slates, and are described in detail farther on. None of the sideritic slates, ferruginous cherts, or ores, although interbedded with the fragmental slates, show any evidence of fragmental origin so far as the indi\'idual grains of the minerals composing them are concerned. PETKOGUAPIIICAL CUAHAGTEliS OF xMANSFIELD SLATE. 63 RELATIONS OF SIDERITE-SLATE, FERRUGINOUS CHERT, AND ORE BODIES TO CLAY SLATES. Owing- to the scarcity of tlie outcrops of the sedinientaries in thu i\Iaus- tic'ld Valley, it is practically impossible to decipher the relations of the individual lieds. Neither the study of the surface exposures nor the expo- sures in the mine workings liave given definite results. Tliat the heds repre- sent interbedded strata is well understood, but the sequence of the strata is indeterminable. It is of especial interest to determine, so far as possible, the relations of the ferruginous rocks, in order that the possible iron-ore deposits associated with theni may be found. A cross section through the Mansfield mine from east to west shows the following relations: The foot- wall of black heinatitic slate is overlain by 25 to 30 feet of fernaginous chert and iron ore. This stratum is succeeded by "red slate," so called by the miners, which is probably weathered greenstone impregnated with iron. This is followed b}' a conglomerate, and this by amygdaloidal greenstone, of the overlying volcanic formation. The ore body extends north and sovith, agreeing thus with the strike of the slates. All drifts end on the north in mixed ore, and on the south in mixed ore, with "quartz-rock" and "lime-rock" of the miners in some places. From these facts we may justly conclude that the ore-bearing ferruginous cherts exist in beds in the slates or as lenticular masses which agree in dip and strike with the surrounding slates. This conclusion is confirmed b}- test pits along the strike of the exposed beds, which have disclosed similar ferruginous cherts at various places for a distance of half a mile to the north. RKLATIONS OF MANSFIEIiD SLATE TO ADJACENT FORMATIONS. RELATIONS TO INTRUSIVES. The Mansfield slates are surrounded on three sides— east, north, and south — by coarse-grained basic eruptive rocks. The fact that the}' are so surrounded by these rocks, which cut them off in the direction of their strike, points to the later origin of these eruptives. Moreover, the quartzitic character of some of the sedinientaries shows that they could not have been derived from the eruptives which stratigraphically underlie them, for iu these no quartz is found. The quartzitic character would thus seem also to indicate that the slates are older than the intrusives. Wherever the 64 THE CEYSTAL FALLS lEON-BEAEING DISTEICT. igneous rocks and slates are in contact or in close association, the latter have been metainorpliosed, and adinoles, spilosites, and desmosites have been formed which are similar to those described as occurring in other areas along the contact zone of basic intrusives. Although no single instance of a dike penetrating the slates has been found, it can hardly be doubted from the relations which have been outlined that the slates are older than the intrusive dolerites. RELATIONS TO VOLCANICS. The sedimentaries are overlain by volcanics, both lava flows and tufa- ceous deposits. In these tuff's, at the northeast corner of sec. 7, T. 43 N., R. 31 W., angular black-slate fragments have been found similar in every respect to the slates of the Mansfield belt. From this it is clear that at least some of the volcanics are younger than part of the slate formation. In section 29 similar relations obtain, the <»nly difference being that the masses of slate and graywacke are inclosed in rather larger fragments in a volcanic conglomerate, and still retain very closely their normal strike. In the conglomerates near the Mansfield mine are found chert fragments and in some places fragments of iron oxide. These latter were evidently not included as oxide, but as fragments of cherty carbonate. Like the great mass forming the ore body, the fragments have since their deposition been altered, forming iron-oxide bodies of small size. Further discussion of the relations between the volcanics and slates will be found under the heading " Hemlock formation." STRUCTURE OF THE ]MA]SrSFIELD AREA. It has already been seen that the Mansfield rocks strike north and south and have a high westerly dip. The two possible explanations of this structure which are compatible with the facts in other portions of the area are (1) that they form a westward dipping monocline, and (2) that they are the western limb of an anticline. THICKNESS. As the sedimentaries forming the Mansfield belt now dip west at a very high angle, and as there is no evidence of duplication of strata due to fold- ing, I feel comparatively safe in giving an estimate of their thickness. The belt is widest at the north end, and there has a breadth of about 1,950 feet. THICKNESS OF MANSFIELD SLATF. 65 Till- i'.vi'i-aj^-o dip of tliu beds is 80^, and this gives a maxiumm thickuess (if l,!lO() feet. Toward the south the belt rapidly narrows, initil it is cut out l)y tile intruding (hilerites. A thickness of 1,500 feet is probably not tar from the average. To the east of the ^Mansfield slates is a belt, varying in width up to about 1,200 feet, in which are found large masses of metamorphosed slates, surrounded by intrusive dolerite. In this belt the slate masses still show a general north-south strike, with slight variations to the east or west, and a westward dip. One might, perhaps, consider this a slate area which has been ciHn])letely saturated with intrusives. If it should be so considered, this thickness should lie added to the estimated thickness of the slates as above given, but as intrusives predominate in it, the slate being, as it were, merely incidental, I have preferred not to include it in the belt with the slate. t)RK DEPOSITS. Although a great deal of exploring for iron ore has been done in the Mansfield slates, only one lai-ge body of ore has thus far been discovered, in which is the Mansfield mine. This mine is situated on the west liank of the Michi- gamme River, in sees. 17 and 20, T. 43 N., E. 31 W. The mine was apparently prospering wlien, on the nigdit of Septeni- ber 28, 1893, a cave-in occurred, letting in the waters of the jMichigamme River and drowning' 28 miners. N MAI,N SHAFT For two Lours after the caving occurred, the bed of the river below the mine was bare, the water tlowing into the mine worlviugs. The accompanying tigure, fig. C, prepared by J. Parke Channing, October 8, 1893, shows the relative position of the shaft and the river, and the couceutric- cracks caused by the caving of the mine. (Plan copied from address of presi- dent: Proc. Lake Superior Inst. Min. Eng., Vol. Ill, 1895, plate opposite p. 42. ) The timber shaft is near the center of tliese cracks. After the caving the mine remained idle until recently. At tlie present writing the DeSoto Mining Company has obtained control of tlie mine and, I understand, have freed it from water. MON XXXVI 5 Fig. 6. - Concentric cracks formed by the caving in of tlio 5I.iiiatield mine. 6(3 THE CRYSTAL FALLS IKOXBEARING DISTRICT. Ill May tbey began the task of diverting the cbauuel of the river to a point several hundred feet sonth of the old course. They have dredged out a cut 2,050 feet in length by 100 feet wide and 18 feet deep. At the upper end of the new channel a cofferdam containing 14,000 cubic yards of earth has been constructed, and where the waters join the old outlet several hundred feet below the mine another embankment has been constructed across the course of the old bed that has 8,000 cubic yards of earth. This task was a very expensive one, and it has been well completed, the old channel being perfectly dry. The turning of the river's course brings out with startling distinctness the criminal negligence or carelessness of those who were working the mine at the time of the accident. The upper tier of timbers in the mine are plainly seen, as also the ground that had been cut out to receive the set that was being gotten into place when the waters broke through. This shows the miners had worked up to within 12 feet of the water of the river. A great crack in the formation shows where the water first gained entrance. The ore made up the bed of the stream — was a portion of the bed in fact — and the walls of the mine were nearly vertical. The ore deposit had a width of about 20 feet. The water pressure must have been considerable, and the blasting of the ore (as it is hard, and explosives are needed to loosen it) shattered the thin piotection over the miners, permitting the water to find ready and unimpeded entrance Into the mine. An engineer could not have been employed and the wildest sort of guessing must have been done by those who had the work in charge. No sane man would have permitted the opening of the deposit so near the river's bottom." Owing to the long abandonment of the mine, the dii'ect sources of infor- mation have been closed. For a description of the ore body I am com- pelled to rely on such data as are available from existing notes and plats. I am especially indebted to a manuscript description of the mine by J. Parke Channing, and to Mr. C. T. Roberts, of Crystal Falls, for plats of the mine. The sketch of the mine here introduced, PI. IX, is compiled from an original drawing of J. Parke Channing, reproduced on the plate cited, and from data obtained from other sources. GENERAL DESCRIPTION OF MANSFIELD MINE DEPOSIT. The Mansfield mine has au ore body varying from 16 to 32 feet in width. It is in ahnost vertical position ; it has well-defined foot and hang- ing walls composed of impervious rock; it has a somewhat indefinite longi- tudinal extent. The ore is Bessemer and occurs in an iron-bearing formation, which corresponds in every particular to those of the other iron-bearing ' Report of Commissioner of Mineral Statistics of Michigan, George A. Newett, for 1896, p. 84. ::^^^g«ftas»«g?g»- ■=»;«<: r- feK^^>^.^, 5 il ORE DEPOSITS IX MANSFIELD SLATE. 67 districts ot" the Lake Superioi- region. 'I'lie ore was iirst toiuid in a test pit wliicli passed through feet of drift. The main working' shaft was then located about 100 feet west of this j)oiut. It was put down to a deptli of 4()0 feet before ore was struck. From this shaft crosscuts were driven east at average intervals of 70 feet, and the ore body was met at a distance vary- ingf from 74 feet at the first level to 10 feet at the sixth level. The cross- cuts, in every case after leaving the greenstone, pass through so-called red slate, at the maximum about 25 feet thick, before ore is reached, this rock constituting the hanging wall. From these data the dip of the ore body may be calculated to be about 80° W., agreeing well with the observed dip of the slates, which outcrop over the area. The thickness of the ore, as shown by the cross sections, averages about 25 feet. The extreme variation in thickness ranges from two sets, or 16 feet, to four sets, or 32 feet. The strike of the slates is north and south, and the trend of the ore body agrees with this. This brings its southern end under the original course of the Michigamme River as the stream bends slightly to the west, south of the shaft. An examination of the longitudinal (north-south) section through tlie ore body does not determine whether or not it has a pitch. The southern boundary is nearly vertical from top to bottom, while the northern boundary lengthens about 140 feet between the first and the fifth levels. In the northern end of the mine — that is, in line with the strike of the sedimentaries — the ore body terminates, in a more or less irregular Avay, in so-called mixed ore. This mixed ore continues to the north for over half a mile, as shown by the numerous test pits which have been bottomed in it. To the south of the mine shaft the ore body proper extends for 200 feet. It then changes its character, becoming a lean non-Bessemer ore. A long drift (335 feet) at the second level was run through this ore, and after leaving it penetrated a mixed ore, the so-called lime rock (sideritel) and quartz rock (chert?) of the miners. Three crosscuts along this drift show the ore body to vary from 20 to 30 feet in thickness, with the same foot and hanging wall as for the remainder of the mine. The same condition exists also lower down, as shown by a drift from the fourth level, 260 feet south. The figures on PL IX, giving longitudinal and cross sections of the mine, show clearly the dimensions of the ore body. 68 THE CRYSTAL FALLS lEON-BEARING DISTRICT, RELATIONS TO SURROUNDING BEDS. The foot wall of the ore is a black slate, described as being rich in hematite and liearing large crystals of iron pyrite. No crosscuts have been driven for any distance into the foot wall, so that it is impossible to say what thickness of the hematitic black slate there may be before the greenish pvritiferous slate begins. In places a gray "soapstone" takes the place of the black slate as the foot wall. The dump obtained by sinking the shaft in the material overlying the ore shows lai-ge masses of conglomerate, the pebbles of which are rounded and predominantly of volcanic rocks, with pebbles of chert and slate from the iron formation and slates below. These fragments are well rounded. The microscope also shows quartz grains with secondary enlargements, so that there can be no doubt that the rock is a true conglomerate. Similar conglomerates, except that the sedimentary fragments are w^anting, have been noticed farther north along the west side of the river. Just west of the Ijridge at IMansfield, near the mine, there is also a small exposure of con- glomerate, which shows an alternation of coarse and fine sediments, with a strike nearly north and south, and a dip of 80^ W. To the west, above this conglomerate, and not more than 15 to 20 feet distant, are found the lavas of the Hendock volcanics. According to the mine captain, the succession west from the ore body in the hanging wall is 20 to 25 feet of paint rock, or, as it is usually called, red slate, then conglomerate, then greenstone. It is difficult to diagnose the paint rock, as no specimens are to be had, but it is highly probable that it is a ferruginous and extremely altered lava sheet. Similar rocks are commonly found thus altered in association with the ores in the Penokee-Gogebic and Marquette districts. Lending weight to this conclusion is the fact that in some places an amygdaloidal green- stone has been exposed in test pits immediately above the iron-bearing formation. COMPOSITION OF ORE. The Mansfield mine up to the present time has raised only Bessemer ore, and is the only mine in the Crystal Falls district which has supplied any considerable quantity of ore of this character. An average of a num- ber of analyses gives the following composition for the Bessemer ore: OllE DEPOSITS IN MANSFIELD SLATE. 69 ]\lft;illir iron, 04.80; pliospliorus, 0.037; .silica, 3.70.' According- to Dr. N. P. Ilulst,- those ore deposits in tlie Menominee range which have ijoorlv defined walls carry a uiininiuni of phosphorus. This body, however, .shows that the same conditions do not exist at the Mansfield mine, since, while it has both sharply defined foot and hanging walls, it contains ])ut a low l)er cent of phosphorus. P'rom an e.xamination of the analyses from which the aliove average was obtained I find that the percentage of phosphorus shows a marked increase in the lower levels of the mines over that of the higher, and there is also a slight corresponding decrease in the content of metallic iron. Increase of phosphorus with depth is also found in the adjoining Menominee range, as noted by Messrs. E. F. Brown, ^ of '^he Pewabic mine, and Per Larsson,^ of the Aragon. It is impossible to state whether or not this distribution is due to the action of percolatino- water, as suggested by Hulst," Larsson,*' and other Michigan mining engineers. Onlv a large number of good analyses from carefully selected ores and asso- ciated rocks and a detailed study of conditions of occurrence could lead to any accurate determination of the reason for such distribution, and a dis- cussion of these reasons is by no means warranted by the few and imper- fect analyses of the Mansfield ores, which I have been able to obtain. The ore body changes in composition to the south of the shaft, as shown by the drifts in this direction. The ore in this part of the mine contains more phos- phorus, alumina, and calcium, and less iron. This low-grade lean ore then passes over into the banded chert and ore nnxed with the lime and quartz rock mentioned above. MICROSCOPICAL CHARACTER OF THE ORES AND ASSOCIATED CHERT BANDS. The ore varies from a soft limonitic hematite to a moderately hard hematite. It is for the most part opaque under the microscope, but in places shows bright-red to brownish-red color in transmitted li(i-ht In incident light the ore for the most part shows a dull-brown or reddish color, though in places it has a bright metallic reflection. In places in the ore ' An average of 62 per cant metallic irou and .030 per cent phosphorus is reported in Report of Commissioner of Mineral Statistics of Michigan (G. A. Newett) for 1896, p. 85. •The geology of that portion of the Menominee range east of the Menominee River, by X. P. Hulst: Proc. Lake Superior lust. Min. Eng., Vol. I, 1893, p. 28. 'Distribution of phosphorus and system of sampling at the Pewabie mine. Irou .Mountain, by E. F. Brown : Proc. Lake Superior Inst. Min. Eug., Vol. Ill, 1895, p. 49. *0p.cit.,p.52. ^0p.cit.,p.28. '0p.cit.,p.53. 70 THE CRYSTAL FALLS IRON-BEAEIXG DISTRICT. are spots, in wliicli is a large quantity of chert mixed with iron oxide. As such ferruo-inous-chert areas increase in quantity the ore grades into the ferruf>-inous chert and chert which is found associated with it in bands and lenticuLir areas. ORIGIN OF THE ORE DEPOSITS. The mode of occurrence and general characters of the ore body hav- ino- been described, we are now prepared to determine the cause of concen- tration of the iron at this particular point and the source. From the description it was seen that the appearance of the body of ore was that of a bedded deposit. The microscopical examination shows, however, that the ore presents no evidences of clastic origin. An examination of the cherts and rocks of the area which are interbedded with the ore, and also a studv of the southern contact of the ore body, shows that the ore is a chemical deposit, or the result of a replacement process, by which the original rock was largely removed, and its place taken by the present ore. It has been shown (p. 62) that the siderite bands pass into hematitic and limonitic chert bands. It has been seen that in the southern end of the mine the lean ore merges into a mass of ore bedded with chert and mixed with a rock called by the miners lime and quartz rock. I interpret this rock to be banded siderite and chert, possibly with some quartzite bands, all of which are found outcropping at the surface. The siderite evidently has been changed into iron oxide and the silica replaced by iron oxide, the banding of the original rock not having been destroyed thereby. Irving^ considered siderite to be the source of, similar ore and associated chert and jasper. Van Hise= has fully explained the process of the concentration of the ores of the Penokee-Gogebic and Marquette districts, and has applied the explanation to the other districts ' Origin of the ferruginous schists and iron ores of the Lake Superior region, by R. L). Irving : Am. Jour. Sci., 3d series, Vol. XXXII, 1886, pp. 255-272. -The iron ore of the Marquette district of Michigan, by C. R. Van Hise: Am. Jour. Sci., 3d series, Vol. XLIII, 1892, pp. 116-132. Iron ores of the Penokee-Gogebio series of Michigan and Wisconsin, by C. R. Van Hise: Am. Jour. Sci., 3d series, Vol. XXXVII, 1889, pp. 32-48. The Penokee iron-bearing series of Michigan and Wisconsin, by R. D. Irving and C R. Van Hise, Tenth Ann. Rept. U. S. Geol. Survey, Part I, 1890, pp. 341-507. The Penokee-Gogebic iron-bearing series of Michigan and Wisconsin, by R. D. Irving and C. R. Van Hise: Mou. U. S. Geol. Survey, Vol. XIX, 1892, pp. 245-290. The Marquette iron-bearing district of Michigan, by C. R. Van Hise and W. S. Bayley, with a chapter ou the Republic trough, by H. L. Smyth : Mon. U. S. Geol. Survey, Vol. XXVIII, 1897, pp. 400-405. ORK DEPOSITS IN MANSFIELD SLATE. 71 in tlio Liiko Suporior ivg-ion. I shall not do more, therefore, than to add that the investij^ations in this area have sliown the probable correctness of this explanation. It is very interesting- from an historical standpoint to note that as far back as 1868 Credner had made the suggestion, with reference especially to the ^Marquette district, tliat the ores were derived from an original iron carbonate. The following- quotation will show his idea of the processes of development of the ore:^ Spliaerosiderit wurde aus kohlensjiurereichen Gewiisseu abgesetzt, duioh eine theilweise Oxydatiou desselbeu eiitstaud Magneteisenstein, durch weitere Anfnalime von Sanerstoff' das Gemenge vou Magneteisenstein und liotbeisensteiu uiid endlich reiner Eotlieiseiisteiu; aus diesem sporadisch durch Zntritt von Wasser Brauneisen- stein. Credner's suggestion seems to have been lightly considered by other workers in that area. In 1886 Irving-^ suggested the theory of replacement of an original fen-uginous carbonate to explain the Penokee-Gogebic iron ores. This theory has since then lieen elaborated by Van Hise, and shown to have a wider application to the other Lake Superior ore districts. He has also traced the iron to its source in the rocks removed by denudation, and shows why it occurs in the positions in which the ore bodies are at present found to occur. Moreover, Van Hise has also explained the process of development in detail, and, what is perhaps far more important, the reason certain ores develop and not others. In its essentials, however, the process is the same as that suggested by Credner in the lines quoted above, though in them no suggestion of the replacement to which is due the em-ichment of the ore bodies is made. Much of the iron of the Mansfield ore is presumed to have resulted directly from the alteration of the feiTuginous carbonate in place, but a large amount was brought in from above by infiltrating waters. The ferruginous matter, which was taken into solution during the denudation of the area, has been carried down by percolating waters and deposited at places favorable for its accumulation. The beds are now on edge, off'ering the most favorable condition to percolation. The conclusion is obvious that these deposits ' Die vorsilurischen Gebikle der " Oberen Halbiusel von Michigan " in Nord-Amerilca, by H. Credner: Zeitschv. dentsch. geol. Gesell., Vol. XXI, 1869, p. 547. -On the origin of the ferruginous schists and iron ores of the Lalse Superior region, by R. D. Irving: Ani. Jour. Sci., Vol. XXXII, 1886, p. 263. 72 THE CRYSTAL FALLS IKON-BEAEING DISTRICT. were formed after the beds were tilted, and the iron derived from the upward extension of the rocks, which has been removed by erosion. CONDITIONS FAVORABLE FOR ORE CONCENTRATION. The conditions favorable for the accumulation of ore deposits have been ascertained by Van Hise from studies in the other iron-bearing dis- tricts of the Lake Superior region. He summarizes these results as follows:^ [1] The iron ore is contiued to certain definite horizons, known as tlie iron-bearing formations. . . . frtj All ore bodies have been found to be distributed very irregu- larly in these iron-bearing formations. Thi.s is due to the fact that they are secondary concentrations produced by downward percolating waters, and the ore bodies therefore occur at the places where water is concentrated, in accordance with the laws of the underground circulation of waters, [b] These places are just above an impervious formation, at the contact of the Upper Huroniau and Lower.Huroniau aud where the rocks are .shattered. \c\ The impervious basement formation m.ay be a surface volcanic, a subsequent intrusive, an argillaceous stratum, or any other impermeable formation, [d] These impervious basements are most efl'ective when they are in the form of pitching troughs, thus concentrating the waters from the sides along a well- defined channel. These pitching troughs may be formed by a single one of the above rocks or by a combination of two or more of them. The horizon marked by the uncon- formity between the Upper and Lower Huronian is a great natural zone of percolating waters. Here oftentimes the basement formation of the Upper Huronian is itself a lean ore, having derived its material from the Lower Huronian, but in this case a secondary concentration has occurred in order to produce the present ore bodies. [e] Finally, as a result of folding, the iron-bearing formations have been shattered, thus producing natural water-courses. More frequently than not, more than one of these classes of phenomena are found together where the great ore bodies occur, and in many cases all are combined. The original source of the iron ores has been ascer- tained to be in many cases a lean carbonate of iron, often with a good deal of carbonate of calcium and magnesium, formed as an ocean deposit. Van Hise adds to the above statement that generally the ore bodies, as a result of their methods of concentration, somewhere reach the rock surface. The Mansfield ore body has well-defined foot and hanging walls of normally impervious rock. The iron-bearing formation is much fractured. We thus have certain of the conditions favorable to the concentration of an ore body. Wliether a trough is completed by a slight cross fold in the formation, or possibly by an intersecting dolerite dike, has not been determined. 'Foiirteeuth Ann. Kept. U. S. Geol. Survey, Part I, 1893, pp. 107-108. OKIi DEPOSITS IN MANSFIELD SLATE. 73 EXPLORATION. Exploration has developed no other deposits ulong- the Mansfield slate belt.' It' other deposits exist, it is hig'hly jjrobable that they extend to the rocK surface — that is, are covered by the drift mantle alone. The intervals between possible ore bodies along the strike of the slates ar'^ probably occupied by mixed chert and ore or ferruginous chert. Explora- tions should extend from the impervious slate below the iron-bearing forma- tion to the impervious rock above the iron-bearing formation. In order to explore the belt thoroughly, rows of pits cross-sectioning the formation ought to be made at intervals not greater than 100 feet, and even with such inter- vals an important deposit might be missed, for it frequently happens that at the surface of the rock an ore deposit is smaller than it is at a moderate depth. SECTION III.— THE HEMLOCK FORMATION. This formation, the most interesting petrographically in the Crystal Falls district, consists almost exclusively of typical volcanic rocks, both basic and acid, with crystalline schists derived from them. Sedimentary rocks play a very unimportant role. With one exception they have been formed directly from the volcanics, and occur interbedded with them. Cutting through the volcanics are intrusive rocks, which likewise include both basic and acid kinds. Chemically the intrusive and extrusive rocks show very close relationships. The name Hemlock has been given to this volcanic formation because the river of that name flows through it for a number of miles, and in places affords excellent exposures. DISTRIBUTIOSr, EXPOSURES, AXD TOPOGRAPHY. Beginning in sec. 36, T. 46 N., R. 32 W., the place where the Hemlock formation enters the part of the district studied by ine, the formation has a width of one-half of a mile. From this place the formation has a north- western coui'se for about 5 miles, gradually widening. It then bends to the west, and after a short distance to the south, which course it follows for about 9 miles. In township 45 N., Rs. 32 and 33 W., the belt has a maxi- mum width of 5 miles. At the end of the southern course the formation ' Since tbe above w.as Tvritten I hare been informed that Mr. George .T. Maas, of Negaunee, has, with a diamond drill, located a body of bessemer ore 30 feet thick on lot 6, sec. 20, T. 43 N., E. 31 W., 1 mile sonth of the Mansfield mine. 74 THE CRYSTAL FALLS IRON-BEARING DISTRICT. bends to the southeast, and continues with this general trend for about 16 miles into T. 42 N., R. 31 W., where my field study of it ended. At the north the belt runs into the eastern half of the district described by Smyth, and swings south, which course is followed for some 15 miles. The entire belt thus foiius an oval suiTOunding the sedimentaries, except in the southeastern part of the district. Another area of Hemlock volcanics is found in T. 43, Rs. 32 and 33, just north of Crystal Falls. This area is about one-half a mile wide just north of the city of Crystal Falls, but rap- idly widens as it is followed to the west until at the western limits of the area it is about 3J miles wide. A third small isolated area is found in sees. 17, 18, 19, and 20, T. 42 N , R. 32 W, and sec. 24, T. 42 N., R. 33 W., about 4 miles south of Crystal Falls. The topography of the Hemlock formation is exceedingly rough where- ever erosion has succeeded in cutting through the drift mantle. This occurs only adjacent to some of the streams. The rough topography at these places is due to differential erosion working upon rocks approximately on edo-e, and of varying hardness. The valleys usually indicate the location of beds of tuff and the higher grounds are almost universally occupied by dense rocks forming the lava flows, or of the coarse-grained massive intrusive rocks. In a few places, however, the thoroughly consolidated and indurated tuffs form high hills. In traversing the Hemlock formation one makes an abrupt ascent, followed by a sharp descent into a nan-ow swamp, then another ascent, and so on. Exposures appear for the most part in small areas along the edges of the swamps and scattered over the faces of the hills. These are fairly numerous, but so small and disconnected as to prevent the tracing out of the individual flows, although this might be pos- sible if the traverses were made at very short intervals and the area mapped in ffreat detail. ° THICKNESS. As has been seen, the belt of eruptives varies in width from one- half of a mile to nearly five miles. The dip of the rocks is about 75° W. The enormous thickness of 25,500 feet which these data would give is probably illusory. In the case of the assumption of the thickness of a series of lava flows and tuffs, it is important that the initial dip, which these deposits must have, be considered. This dip varies greatly, depending on the slope of the THICKNESS OF HEMLOCK FORMATION. 75 cone, whic'li iu its tiivn, is dept'iideiit on the viscosity of the lava and the j)resenco of varying' (jnantities of fragmental jirodncts. If we assume these pre-Cambrian volcanic products to have had an initial dip of 15'^, I lielieve we are within limits for products consisting, as these do, of what was i)rob- ably moderately viscous basalt and vast masses of fragmental material. This estimate is based on the assumption that the volcanics here represented were deposited for the most part upon the westward slope of a volcano, or a series of volcanoes. This initial dip of 15° is then to be deducted from the present dip, 75°, of the flows. Taking this into consideration, we get a thickness of 23,000 feet for the volcanics. It is highly probable that the rocks have been subjected to close folding, and for this reason also the apparent thickness would be much greater than the true thickness. The schistose character of some of the rocks shows clearly that they have been severely mashed, and this mashing was probably produced in connection with folding. It is probable that this possible maximum thickness shoiUd be very materially reduced, jDOssibly to one-half or one-third of the amount. However, even the maximum above calculated is probably paralleled by the vast masses of volcanic material accumulated in certain volcanic areas, such as those of Hawaii or Iceland. Geikie writes:^ The bottom of these Iceland Tertiary basalts is everywhere concealed under the sea. Yet their visible portion shows them to be probably more than .3,000 meters in thickness. An especial interest belongs to this Icelandic plateau because volcanic action is still vigorous upon it at the present day. RELATIONS TO ADJACENT FORMATIONS. In the northern part of the Crystal Falls district the volcanics overlie the quartzose dolomite formation known as the Randville dolomite. In the central part of the district, through which the Deer River runs, as shown in section G-H, PI. VI, outcrops are so scarce that it has been found impossible to trace the boundaries of the formations with any degree of accuracy. Consequently this part of the district is mapped as Pleistocene. From the few outcrops of slate, probably equivalent to the Mansfield slate, which have been found in the Deer River area, it has been thought 'The Tertiary basalt-plateaux of northwestern Europe, by Sir A. Geikie; Quart. Jour. Geol. Soo. London, Vol. LII, 1896, p. 395. 76 THE CRYSTAL FALLS IRON-BEARING DISTRICT. higlilv probable that this slate iu an extremely plicated condition may underlie the volcanics of this area, and it is so represented in section G-H, Plate VI. As evidence of this, in T. 43 N., R. 31 W. the volcanics overlie the Mansfield slate unconformably. In places test pits have disclosed an amygdaloidal lava flow immedi- ately overlying the Mansfield slates. At one place, at the northeast corner of sec. 7, T. 43 N., R. 31 W., angular fragments of the underlying black slate have been found in the tufaceous deposits of the Hemlock volcanics. Farther south, along the contact just west of the Mansfield mine, a con- glomerate is exposed, which contains fragments of slate, lava, and rounded grains of quartz with secondary enlargements. The rock is evidently water deposited. There is also obtained from the workings of the mine a conglomerate, taken from just alcove the ore, which consists of lava frag- ments and pieces of chert and ore, as mentioned on pp. 64, 68. From these occurrences it is clear that some of the sedimentaries are unquestionably older than some of the volcanics, and yet the conglomerates bearing the fragments of ore and slate contain also fragments of lava, showing the existence of some of the volcanics before the deposition of this conglom- erate. The only explanation of all of the facts which has occurred to me is as follows: After the ore-bearing Mansfield slate was deposited, an erosion interval occurred. Then followed a volcanic outbreak. It is highly probable that this outburst began far north of the Man.sfield mine, coincident with the upheaval which resulted in the erosion of the Mansfield slate. The volcanic ejectamenta were mixed with the sedimentary fragments and all together were rounded and bedded, forming in places conglomerates. In places along the shore lava flows descended, some reaching into the sea and covering the sedimentaries along the shore where no conglomerate had been formed. At other places deposits of scoriae, etc., including fragments of slates from the sedimentaries through which the volcano burst, were made, and thus deposits of tuff are found overlying the sedimentaries. The various deposits, though really separated by a slight physical break, are practically conformable with the series below, all having a north-south strike and a high westward dip. The formations which underlie the volcanics in the northern and southern parts of the district are of different character. This difference RELATIO^TS OF HEMLOCK FORMATION. 77 niav be oxplaiucd by supposing the voleaiiocs broke out in tlie nortliern ])art, while the Manstielil shite was still being deposited in the south. Gradually, however, the volcanic activity spread toward the south, proba- l)ly following- a fissure along the pre-Cambrian shore, and igneous materials buried the Mansfield slate. Hence, while on the whole these volcanics are younger than the Mansfield slates, some of the lower of them ai-e con- temporaneous with some of the upper Mansfield beds. The volcanics invariably overlie the Randville dolomite, and are unque.stionably of later age than that formation. The Hemlock volcanics are overlain throughout their extent b}- the Upper Huronian series of graywackes and slates. Near the contact line with the volcanics wherever the Huronian outcrops, or has been exposed by exploration, it has been found to be characterized bv a line of magnetic attraction. By means of magnetic observations the line of contact has been traced, where owing to lack of exposures it would have been otherwise impossible to comiect the isolated outcrops. RELATIOlSrS TO USTTRUSIVES. High ridges composed of dolerite are found extending in a general north- west and southeast direction through the volcanics. That these masses were forced up through the Hemlock formation is indicated by the folding which they cause in certain places. Such rocks are unquestionably younger than the volcanic series. There may be seen also on the map, in T. 44 N., R. 32 W., a number of isolated knobs. These are also doleritic, and are presumed to be, like the larger ridges, intrusive in the volcanics. The dolerites have in their turn been cut by acid dikes. These are coarse micropegmatitic granites. Similar acid dikes have been found cutting the surrounding volcanics. This set of acid dikes may be looked upon as the youngest intrusive igneous rocks occurring in the Hemlock volcanic formation. Cutting the volcanics are also basic dikes varying from fine to moder- ately coarse grain. It is well known that during a volcanic epoch the out- poured lavas and clastic volcanic deposits are penetrated by dikes coming from the same magma. Whether or not these dikes are of this origin, and are hence contemporaneous with the later volcanics, or are of later age, and 78 THE CRYSTAL FALLS IRON-BEAEING DISTRICT. correspond to the coarse dolerites, it is impossible to determine with certainty. They are presumed, however, to form an integral part of the Hemlock volcanics, as no connection between the dikes and the unques- tionably intrusive dolerites could be traced in the field. VOLCANIC ORIGIN. In spite of numerous occuiTences of ancient volcanics which have recently become known, the late Professor Dana makes the following statement:^ It is not yet certain that a volcano ever existed on tlie continent of Ifortli America before the Cretaceous period; for the published facts relating to supposed or alleged rolcanic eruptions in the course of the Paleozoic ages are as well explained on the suppo.sition of outflows from fissures and tufa ejections under submarine conditions; and none of the accounts present evidence of the former existence of a volcanic cone, that is, of an elevation pericentric in structure made of igneous ejections. . The presence in the Hemlock formation of a quantity of pyroclastics, great in proportion to the solid lavas, and the absence of any great sheets of lava, so important a product of great fissure eruptions, seem to point to the derivation of the Hemlock rocks from a volcano or volcanoes situated near the border of the contemporaneous Huronian sea, rather than from a simple fissure. While some of the eruptives may have been submarine, the occurrence of large quantities of clearly subaerial deposits shows that the eniptives were largely on the land. Thus it appears that neither a fissure flow nor a submarine volcano will wholly explain the Hemlock formation. However, it is hig'hly probable that this volcanic outburst, which piled masses of volcanic material upon the land, was accompanied, as have been all or nearly all the great outbursts of recent times, by submarine lava flows and tuff" ejections. No such clear evidence of the presence of a Pale- ozoic or pre-Paleozoic volcano on the North American continent has been adduced as that given by the English geologists for certain volcanoes in the British Isles. But while the j^i'esence of a central cone with peri- centric arrangement in the Hemlock district is not conclusively j^i'oven, the presumption in favor of such a cone or cones having existed is certainly strong. 'Manual of Geology, by J. D. Dana: 4th ed., 1895, p. 938. VOLCANIC OUIGIN OF HEMLOCK FORMATION. 79 An attempt was made to locate the vent or vents from wliicli the material was derived, l)iit no evidence could be found, unless we consider tlie vents to have been where the accunudations were the greatest. The coarse-grained rocks which were first supposed to represent the plugs of ancient volcanoes, on careful and detailed examination appear to be later intrusives, or else are indeterminate. CXiASSIFICATIOlSr. The general character and distribution of the Hemlock formation hav- ing been given, we may now proceed to a petrographical consideration of the rocks comprising it. This will be given in more detail than for the other rocks of the Michigamme district because this great pre-Cambrian volcanic formation possesses peculiar interest. The rocks of the Hemlock formation are chiefly of direct igneous origin. Some interleaved sedimentary rocks occur, which, however, with a single exception are composed of fragments of the igneous rocks. For the sake of easy reference, the usual classification into igneous and sedimentary rocks will be used. The massive igneous rocks are subdivided according to chemical composition into acid and basic rocks. The acid rocks include rhyolite-porphyries,' aporhyolite-porphyries, and acid pyroclastics. The basic rocks include altered uonporphyritic basalts, porphyritic basalts, and variolite and basic pyroclastics. The sedimentary rocks are divided into the volcanic sedimentaries and the nonvolcanic sedimentaries or normal sedimentaries. The first include tuffs and ash beds — the aeolian deposits, and volcanic conglomerates — subacpieous deposits. The normal sedimen- taries are represented by slates and limestones. Various schists are locally produced from these numerous kinds of rocks through metasomatic changes and dynamo-metamorphic action. Many of these schists resemble one another very closely, though, as will be seen later, they are derived from both the massive rocks and from the elastics. These have been described in connection with the rocks from which they have been derived. 'According to a late riiliug of the Director of the United States Geological Survey, based on the recommendation of a committee on nomenclature for geologic folios, "porphyry" is to be used only with a textural significance. Hence "quartz-porpbjry," according to this ruling should no longer be used as a rock name. The rhyolite-porphyries here described are what have been kuown as normal quartz-poriihyries. 80 THE CRYSTAL FALLS lEONBEAKING DISTRICT. The following table will show the arrangement outlined above, which will be followed in the descriptions: Classification of the rocks of the Hemloch formation. Igneous . . . , Uavas * Rhyolite^poiphy ry .... J gchistose acid lavas . fAcid .. <'^''''""' ( Aporhyolite-porphyry . S ( Pyroclastics ( Nonporphyritic ( Lavas Metabasalt <; Poipbyritic Basic . .' . ( Vaiiolitic Pyroclastics . . Eruptive breccia Volcauic sediments Flow breccia . Sedimentary Normal sediments , ) Eohan deposits J Ash beds I Subaqueous deposits.. .Conglomerates S Slate ( Liiuestone , Crystalline schists. ACID VOLCAlSriCS. The acid volcanics are comparatively unimportant in quantity. They may be conveniently subdivided into the lavas and pyroclastics. ACID LAVAS. The acid lavas occur in such small quantity as to make it impossible without very great exaggeration to place them upon the accompanying small-scale general maps, though they have been introduced upon the detail maps wherever the scale permitted. They usually form isolated ridges, and their relations to the surrounding basic volcanics are obscured by lack of exposures. The trend of the individual ridges agrees with the general strike of the banding in the basic tuffs. Moreover, in nearl}^ all cases the isolated exposures which are closest together lie in such relations to one another that when connected the large sheets thus formed follow the strike of the tuff banding, as do the individual ridges, and they are there- fore confidently assumed to be the isolated portions of acid flows inter- bedded with the basic volcanic rocks. The rock types represented are the two closely related rocks — the rhyolite-porphyry and the aporhyolite-porphyry. Under the rhyolite- porphyries are included the porphyritic acid lavas, which have, so far as can be determined, an original holocrystalline groundmass. Under the aporhyolite-porphyry, following Miss Bascom's use of cqm,^ I include those 1 structures, origin, and nomenclature of the acid volcanic rocks of South Mountain, Penn- sylvania, by Miss Florence Bascom : Jour. Geol., Vol. 1, 1893, p. 816. ACID V'OLGANICS OF HEMLOCK FORMATION. 81 acid lavas which are now hkewise hoh)cr)stalhue, but wliich owe this cliaracter to the devitriHcatiou of an orig-iual ghissy base, supposing- tliem in tlieir original vitreous condition to have corresponded to tlie modern liyalorhyolite-porphyries. RHYOLITE-PORPHYRY. The rhyolite-porphyries on fresh fracture are dark grayish-blue to lilack. From this the)' grade with advancing alterations through chocolate brown to purplish. The weathered surface A^aries from white to reddish. The weathering has in one case brought out very well the fluxion banding of the rock. Tlieir texture is very pronouncedly porphp-itic. The quartz and feldspar crystals stand out ^ilainl}- from the groundmass, which is usually dense with a somewhat resinous luster. The porphyritic quartzes average perhaps the size of a small pea, and hence are macroscopically very plainly visible. They frequently stand out on the weathered surface and show their cry.>-tal forms, and in other cases we see the angular cavities out of which they have fallen, like the kernel from the nut. Under the microscope the rocks are seen to be typical rhyolite-por- phyries. The phenocrysts are chiefly corroded dihexahedral crystals of quartz. Crystals of plagioclase and orthoclase are less common. These lie in a very fine-grained holocrystalline groundmass, composed largely of feldspar and quartz, with some zircon in small crystals, and here and there magnetite. These are presumed to be the original constituents of the groundmass. Associated with them are considerable quantities of secondary chlorite, epidote, biotite, muscovite, calcite, and reddish to brown alteration products of the magnetite. Included in the groundmass are here and there oval areas of finely crystalline secondary quartz, probably filling former amygdaloidal cavities. In thin section the crystal contours of the quartz phenocrysts are more or less rounded, with here and there embayments of the groundmass projecting into them. The crystal form is, however, always cleai'ly marked. In some cases the individuals have been broken before the cool- ing of the magma, the fragments of an individual, though now separated, being seen to conform to one another. That they have been subjected to 23ressure is shown by the undulatory extinction and also by the separation MON xxxvi 6 82 THE CRYSTAL FALLS IRON-BEARING DISTRICT. of the black cross of uniaxial minerals into hyperbolfe. Embayments of groundmass, and liquid inclusions in which a dancing bubble may be seen, are in places rather thickly distributed through the quartz. The liquid inclusions have very commonly an hexagonal form, corresponding to the contours of the inclosing quartz. These liquid inclusions are certainly in some cases secondary. This character is well shown in some of the crystals, which are broken across, giving along the line of fracture a very wavy extinction. Along this line of fracture the greatest quantity of inclusions are seen, both with and without bubbles. As the distance from a fracture increases, both the undu- latorA^ extinction and the number of inclusions diminish. (See fig. A, PL XIX.) These fractures in the quartzes are but continuations of those which extend in many cases all the way across the section. The fractures have since been healed by secondary quartz. This secondary quartz has also in some cases healed the fractured quartz phenocrysts, and then agrees with them in orientation. The possession of an imperfect rhombohedral parting is very noticeable in a number of quartzes, and especially those wdiich, being on the edge of the section, are very thin. (See fig. B, PI. XIX.) Similar parting in the quartz occurs in various rocks studied in this district. The phenocrysts of the porphyries are traversed by fractures, some of which are more or less circular, and simulate very imperfectly perlitic cracks. With the exception of those in porphyries in two localities, the quartz phenocrysts are surrounded by zones, largely of quartz, of varying widths, and considerably lighter than the remainder of the groundmass. Much of the quartz of these zones has the same optical orientation as the phenocrysts. In those sections in which the zones are observed they occur around every section of quartz. The feldspar phenocrysts are orthoclase and plagioclase, the latter apparently predominating. They occur usually in rounded, badly corroded crystals, with indentations filled with groundmass. They are always altered, and have associated with them as secondary products calcite, epidote, muscovite, biotite, and chlorite No large original ferro-magnesian phenocrysts appear to have been present in the porphyry. Their former presence is at least not indicated by any aggregates of secondary products. Whatever ferro-magnesian min- ACID VOLOANICS OF HEMLOCK FOliMATlON. 83 erals were preseut must have been scattered tlin)Ugli the <>Toiin(lniass, and have been completely altered. The secondary minerals contained in the groundniass are chlorite, calcite, epidote, niuscovite, and biotite. TEXTCRE OF THE I'OUrilYpRIES. The texture of the dense groundniass varies according to the mode of association of the two chief minerals — quartz and feldspar. The commonest variety is the rhyolite-porphyry with microgranitic groundmass (porphyre granulitique of Michel Ldvy). A second variety is the rhyolite-porphyry with micropoikilitic groundmass.' The microgranitic texture is too well known to warrant a description of it here. The micropoikilitic texture presents certain characters which render a further description desirable. This peculiar phase of the micropoikilitic textm-e was briefly described by the writer and illustrated by microphoto- graphs in 1895.^ Shortly after the separates of this article were disti-ibuted, I received from H. Hedstrom, of the Swedish Geological Survey, an article published in 1894 containing a description of what appears to be very nearly the same texture.^ If I have understood the description correctly, however, there seems to be one essential difference. In order to explain clearly this difference, I shall describe the textm-e in detail. In certain of the rhyolite-porphyries, as already mentioned, the quartz phenocrysts are surrounded by certain zones. These zones in the rocks having a micropoikilitic texture possess exactly the same texture as does the groundmass. The zones are composed of minerals which are of suffi- cient size to permit readily their determination. Quartz and feldspar are the essential components, with some chlorite, epidote, muscovite, and iron oxide. The first two are the important minerals, and will alone be referred to in the further description. The chief peculiarity of the zone is in the arrangement of the two minerals, and this character is best shown on the accompanying microphotographs. This textm-e can be seen even in ordinary light. It is brought out better when the field is partly shaded, so ' Eruptive rocks of Electric Peak aud Sepulchre Mountain, by J. P. IiUlings : Twelfth Anu. Eeirt. U. S. Geol. Survey, 1891, p. 589. On the use of the terms poikilitic aud micropoikilitic in petrography, by G. H. Williams: Jour. Geol., Vol. 1, 1892, pp. 176-179. - Volcanics of the Michigamme district of Michigan, by J. Morgan Clements: Jour. Geol., Vol III, 1895, pp. 814-816, flgs. 1 and 2. » Studier iifver Bergarter fran Moriin vid Visby, by H. Hedstrom : Geol. Foren i Stockholm Forhandl, Bd. 16, H. i, 1894, pp. 5-9. 84 THE OKYSTAL FALLS IRONBEAKING DISTRICT. as to exhibit the difference in reHef of the minerals, and, best of all, between' crossed nicols. (See fig. A, PI. XX.) The zones are seen to be made up of reticulating areas of clear quartz, in which lie embedded iiTegular pieces of feldspar. Where two or more of the quartz stringers or needles unite, one sees broad ai-eas of limpid quartz. The network of quartz is best seen when it exhibits its highest polarization color, as then the feldspar is for the most part dark. The pieces of feldspar in such a quartz area for the most part have irregular orientation, as is shown by their varying extinction, although a number extinguish simultaneously. This quartz net is connected with the quartz phenocrysts, as shown by the continuation of the quartz of the phenocrysts and that of the zone, and the consequent agreement in orientation. The lack of a uniform optical orientation of the feldspar grains is made especially apparent when the quartz is cut perpendicular to the c axis, and consequently remains dark under crossed nicols. Under the above circumstances we see certain feldspar grains polarizing in the zone around the quartz, and as the stage revolves other particles lighten as those which polarized in the previous position of the stage become dark. From this description it is evident that the texture is not micropegmatitic according to the generally accepted definition of the term, but corresponds to the micropoikilitic, as described by Iddings and Williams.^ A gradation toward a spherulitic texture was noticed in one instance where a number of long quartz stringers were arranged pei-jjendicular to the periphery of the quartz phenocryst. (Fig. B, PI. XX.) The texture of this micropoikilitic mass, it will be observed, is finer than that before described. The groundmass of the porphyries is formed of irregular roundish areas having exactly the same micropoikilitic texture as the zones surround- ing the quartz phenocrysts. An explanation of the origin of the zones should therefore also explain the texture of the groundmass. Certainly in many cases, probably in most cases, the groundmass areas result from tangential sections through one of the micropoikilitic zones surrounding the quartz phenocrysts. The description given by Hedstrom^ of this same structure as observed by him is essentially the same as the above, if I have understood him cor- rectly. The following difference is, however, to be noted. In speaking of 1 Op. cit., pp. 589 and 179. ^Op. cit., p. 8. ACID VOLCANICS OF UEMLOCK FORMATION. 85 t\w stnic'turt' where tliu (iiuutz i.s siuTuunded by this luicrojjoikilitic zone, he calls it the graiiophyric structure. As I liave already emphasized alcove, the feldspars in the network of (piartz have varying orientation, and tlie structure is, strictly speaking, micropoikilitic, and in no sense granophyric (micropegmatitic). Moreover, he describes in addition to tlie above type one in which are found i)henocr}'sts of quartz lying in a micropoikilitic groundmass with the above reticulating texture, but the phenocrysts abut sharply against the groundmass, instead of being connected with it by means of these zones. The micropoikilitic texture has been held in some cases to be of sec- ondary origin and the result of devitrification. While recognizing that there may be certain unquestionable cases where a ixiicropoikilitic structure results from the devitrification of a glassy groundmass, I can find no evi- dence in the rocks here described that points to this origin for the micro- poikilitic texture under discussion. On the other hand, there is an absence of evidence that indicates its unquestionably primary character. Rather than to regard the quartz as secondary and influenced in its orientation by the phenocrysts, as in the enlargements of quartz grains, it seems natural to suppose that when the lava was extruded after the crystallization of the phenocrysts, there began, consequent upon the diminished pressure and temperature and other factors, a rapid crystallization of the mineral elements from the remaining magma. This resulted in the production of the feldspar in very imperfect and small crystal individuals. At the same time the quartz of the phenocrysts continued to grow, and in so doing inclosed these small feldspars in its meshes. In certain rhyolite-porphyries the micropoikilitic texture is somewhat different from that above described. In these the quartz phenocrysts are surrounded by zones which are illustrated in figs. A and B, PI. XXI. These appear to correspond very closely to the ones described by Michel Ldvy^ and VViUiams,- and since described by many other writers. The zones have a much higher index of refraction than the quartz of the phenocrysts, and hence contrast strongly with it. Examined closely, they are seen to be composed of chlorite, epidote, and black or reddish 'Annales des Mines, Vol. VIII, 1875, pp. 378, 381. =iDie Eruptivgesteiue der Gegeucl vou Tiiberg im SchwarzwaUl, by G. H. Williams: N. Jahrb. fill- Jliu., Bil. II, 1883, p. 605. 86 THE CRYSTAL FALLS lORN-BEARING DISTRICT. ferruginous grains, which He in a white matrix. This matrix shows the following characters: The greater part of it extinguishes and lightens simultaneouslv with the quartz phenocrysts which it surrounds, and is consequently believed to be quartz. When the matrix and quartz pheno- crysts are dark, one sees scattered through the matrix, making up a very small proportion of the total zone, certain irregular areas which show polarization effects. These are believed to be feldspar grains, though this could not be determined. With the highest magnification no radial arrangement of the quartz and feldspar could be observed which would wai-rant the inclusion of these aureoles under Michel Levy's term " sjjhero- lites a quarts (jlohidaire."^ Where two quartz crystals with different orienta- tion are in juxtaposition, each possesses its own zone corresponding with it in orientation. The way in which the zones about the quartzes are confined to the quartz is clearly shown in one case in which a very much altered feldspar phenocryst was found, one portion possessing a typical coarse microjDCgmatitic texture. In this case where the quartz of the micropeg- matitic intergrowth touches the groundmass, it grades into a micropoikilitic area, whei'eas the feldspar does not do so. The texture of the zones about the quartzes is apparently but a fine- grained variety of the micropoikilitic texture, the coarser phases of which are illusti'ated on PI. XX. The groundmass of the rocks showing the texture is composed of roundish areas of exactly the same composition as the zones around the phenocrysts, with a feldspar of small dimensions here and there between these areas. (See fig. B, PI. XXI.) The texture approaches very closely if it does not correspond exactly to the quartz ^pongeuse phase of the quartz-globulaire texture of the French." In one part of a section of rhy- olite-porphjTy the quartz phenocrysts have aureoles and the groundmass has the texture just described. In another portion of the section the quartz phenocrysts have no aureoles and the groundmass possesses an imperfect microgranitic texture (structure microgranulitique of Michel Ldvy). This shows the passage of a micropoikilitic textured rock into one with a micro- granitic texture. I explain the aureoles and the roundish areas in the ' structures et classification des rocbes 6rui)tives, by A. Michel L(?vy, Paris, 1889, p. 21. = lt is found to show exactly the same texture as seen in a section obtained from Paris and labeled " Porjihyre ;i quartz globulaire de la Sartbe.'' ACID VOLGANICS OF HEMLOCK FORMATION. 87 groundmass as original, in exactly the same way as has been suggested hy Williams' tor those which he described. This is essentially the same expla- nation which I have given on a previous page for the less common, coarse micropoikilitic phase. The cause of the formation of the microgranitic phase appears, however, rather difficult to discern. Its development seems to depend upon peculiar local conditions. APORHYOLITE-PORPHYKY. Intimately associated with the rhyolite-porphyries are rocks very similar to them in mineral constituents, both macroscopically and microscopically, so that the description of the rhyolite-porphyries will largely answer for the aporhyolite-porphyries. Flow texture, however, is well shown by the aporhyolite-porphyries. A beautifully developed perlitic parting, fig. A, PI. XXII, is taken to indicate the presence of an original glass; hence the rocks are classed with the aporhyolites. The perlitic cracks are well brought out in ordinary light by the chloritic flakes along them. Between crossed nicols these disappear, and the groundmass resolves itself into a fine-grained mosaic of quartz and feldspar. (Fig. B, PI. XXII.) This groundmass has all the characters of that of a microgranite. No evidence wdiich would point to the devitrification of a glass could be seen other than the presence of a perlitic parting, as described. For recent excellent descriptions of similar devitrified lavas in which various structures characteristic of vitreous lavas have been identified, the reader is referred to the articles already mentioned, and the one by Dr. Bascom,- in which a moderately full bibliography is found. SCHISTOSE ACID LAVAS. The results of the ordinary alterations of the acid lavas, chiefly meta- somatic in character, by which the phenocrysts and the matrix ha^•e been changed, have been briefly described. The results produced by dynamic action are more interesting and perhaps more striking. The mashing, result- ing in chemical changes and schistose structure, has in many cases almost obliterated the porphyritic texture, and in extreme cases destroyed all inter- nal evidence of igneous origin. Even the fluxion banding, as is well known, at times simulates very closel}" sedimentary bedding, and thus increases 'Op. cit., p. 607. -Acid volcanic rocks of South Mouutain, by Dr. Florence B.ascom: Bull. U. S. Geol. Survey No. 136, 1896, p. 87. 88 THE CKrSTAL FALLS IRON-BEARING DISTRICT. the difficulty of determiniug the igneous character of the rock. In the rocks to be described the phenocrysts may still be observed, though more or less deformed, and the fluxion banding has been in one case exceptionally well preserved, so that no doubt is felt as to their igneous character. Dynam- ically metamorphosed rhyolite-porphyry flows have been found in two areas in the Hemlock formation. In the following each area will be described separately, the one in which the original character of the porphyry is least in doubt being considered first. The Deer River schistose porphyries are found in the SE. J sec. 36, T. 44 N., R 32 W., beginning at 400 N., 250 W., and continuing to 600 N., 350 W., of the southeast corner near the bridge on the Floodwood road. They occur in several outcrops which are practically continuous, being separated by very short distances, and are so much alike both macroscopic- ally and microscopically that there is sufficient reason for the conclusion that they belong together. Their field relations to other rocks have not been observed. No data have been found which offer any clue as to the time of eruption of these rocks other than the fact that they are surrounded by the basic volcanics of the Hemlock formation and have undergone the same dynamic action. The porphyries are dense, bluish-black rocks, with porphyritic crystals of red feldspar. A fluidal structure is not present in them. The schistose structure is apparent to the eye, especially upon the weathered surface, and the cleavage of the rock also indicates it. The cleavage face of the rock has a silky luster, due to the sericite and biotite flakes parallel to it. The rock breaks readily in various directions at angles to the cleavage, so that it is impossible to obtain well-shaped hand specimens. The schistosity in these porphyries is clearly brought out by weathering, the weathered rocks showing perfect schistosity, while fresh specimens from the same exposure, although splitting easiest in one direction, appear perfectly massive in hand specimens when broken across the schistosity. That the dynamic action was greatest along certain zones of the rock, other portions being more or less exempt, is shown by the fact that of several specimens collected with the view of obtaining different stages of alteration from different portions of the same exposure some are markedly schistose, while the least altered approach a fairly massive character. I shall give a brief description of tliis least altered phase, and then AOID VOLGANIGS OF HEMLOCK FORMATION. SB" cou.sider the cliiuigx's wliicli liuve tukeii place and tlie character of the rock which has resulted in the more altered phases. The slightly schistose rock, like all the porphyries, is very fine gi-ained and l)lack, with a more or less silky luster on fresh fractures parallel to the schistosity. The porphyritic character is not very strongly marked. Maotroscopically, comparatively few small feldspar phenocrysts are visible. Under the microscope the rock is seen to be a micropegmatitic rhyolite- porphyry in which the silica has not crystallized as quartz phenocrysts, bi\vnish-green and strongly i)leochroic. A few spots of brown iron hydroxide and small heaps of grains of sphene probably indicate the former presence of titaniferous iron ore. The few crystals of apatite present are broken and separated, but otherwise retain the usual characters of this mineral. The groundmass has a very marked schistose structure, brought out especially well by the i)arallel arrangement of the mica flakes. The way in which these lines of schistosity flow around the mashed phenocrysts, one line never coalescing with another, but remaining continuous, may be seen with great distinctness where the lines abut sharply against the crystal at a very obtuse angle. As the angle becomes less and less obtuse, the ends of these lines bend up slightly in the direction which would enable them to pass the crystal, and then end, so that along the face of the crystal one can follow them, as it were, in a series of steps until those lines which strike the crystal near enough the edge to flow around it bend slightly, and passing around continue on the opposite side. The fact noted by Fiitterer^ that an increased amount of sericite occurs on the two sides of the feldspar crystals parallel to the schistosity is very patent in these porphyries. The diminu- tion in grain of quartz and feldspar seems to accompany the increase in the amount of the sericite. The slides are crossed by narrow fractures cutting the planes of schistosity, which are filled with secondary quartz, showing marked sti-ain efi"ects. Associated with the quartz were observed some crystals of brown nitile. In one of the more altered slides these fissures have been filled with calcite, whether or not as a replacement of the quartz could not be told. Schistose porphyries showing the extreme alteration phases are found from N. 300, W. 300, to N. 400, W. 250, in the SE. i sec. 4, T. 44 N., R. 32 W. They form a rough escai-pment upon the southeast side of and near the base of a large hill, and overlook McCutcheon's Lake. The exposure is not continuous throughout, though practically so, but the unexposed parts are sufiicient to prevent a perfect sequence being traced. The appearance of the rock is strongly like that of sedimentary rocks. Difterent bands nearly on edge may be seen, dipping 60°-90° SW. and striking N. 30° W. I O]). cit., p. 40. 92 THE CEYSTAL FALLS IRON-BEARING DISTRICT. At a point about 100 feet higher and three-fourths of a mile distant, oq the very northwest flank of the same hill, at N. 1725, W. 775, from the south- east corner of sec. 4, T. 44 N., R. 32 W., there is a small ledge of schistose jDorphyry, macroscopically and microscopically similar to those to the south- east of it, and with its schistosity striking N. 20° W. and dipping 80° SW. The striking agreement in strike, dip, and general character of these two separated outcrops points to their being merely isolated portions of the same mass. There seems to be no discrepancy between the dip and strike of the schistosity and that given above for the bands. The most striking macroscopical characteristic of these mashed por- phyry flows is the occuiTCUce of phenocrysts in a schistose and beautifully banded rock. These phenocrysts stand ont clearly from the groundmass^ in all cases. The general appearance of the rocks is that of the well-known very dense banded halleflintas of Elfdalen, Sweden. The bands vary in color, ranging on the weathered surface from light creamy white, tln'ough light greenish, to red and almost black. Tlie rocks which have very light colored weathered surfaces are always bluish black on a fresh fracture, and very dense, and those weathering red are usually cream colored on freshly fractured faces; Many of the areas which appear macroscopically to be single phenocrysts are resolved under the microscope into tangled groups of individuals, though in rare cases the individuals show the imperfect radial arrangement rather frequent in medium-grained micropegmatitic rhyolite-porphyries. The feldspar has iindergone considerable alteration. In the least- changed grains there is a cloudiness caused by numerous indeterminable specks, probably of iron oxide, which give a reddish tinge to the mineral. Further changes result in the production of muscovite and epidote, with biotite in rare cases, accompanied by the obliteration of the twinning lamellse. The greater part of the phenocrysts seem to be orthoclase, though associated with them are found pieces which show indistinct traces of the polysynthetic twinning of plagioclase feldspar. The feldspars exhibit marked strain eff'ects, especially in their flattening into long oval and spindle-shaped areas. Some crystals have been broken and separated perpendicular to the direction of the schistosity. The spaces between the fragments are filled with secondarj- muscovite, quartz, and feldspar. Sur- rounding the phenocrysts — that is, between the phenocrysts und the ground- ACID VOLOANICS OF HEMLOGK FOKMATION. 93 mass proper — we find a mass of small angular, finely striated, limpid grains of feldspar, associated with similar grains of quartz, tlie two lumng in places between tliein sericitic fiakes. In one especiall}- clear case, this secondary aggregate fills half the space formerly occupied by an individual feldspar, the other half being still occupied In^ the remnant of the appar- ently simply twinned feldspar from which it was derived. (Fig. A, PI. XXIII.) While no large quartz phenocrysts were observed, a mosaic of qviartz is found in small round or oval areas in various sections. The individual fragments exhibit the usual strain effects of crushed minerals. (Figs. A, B, PI. XXIV.) The groundmass consists of the same preponderant minerals as the scliistose porphyries, which have been previously described. The accessory minerals are apatite, which is present in very small quantity, and rutile, which in one of the slides is present in verj' considerable quantity in the form known as "thonshiefer-nadelchen." Calcite is found in all of the slides, the amount varying very much. Those which contain a great deal have a scoriaceous-looking surface, due to weathering out of the calcite. The flow structure mentioned as having been observed in the schistose porphyries of the Hemlock formation is perhaps of sufficient general interest to warrant a few comments. This is well marked only on one hand speci- men. In this there is an alternation of pink and dark grayish-blue bands which are rarely more than a fraction of an inch thick. Some, especially the thicker bands, are remai-kably persistent. Even macroscopically on the weathered surface the pinkish bands can be distinctly seen to wrap around the pink feldspar phenocrysts and oval areas of the grayish-blue part of the rock. Under the microscope the bands which macroscopically are the darkest are clear and transparent, while the pink bands are much less trans- parent. The microscope shows the diff"ereuce in the color of the bands to be due chiefly to tlie fineness of grain, and brings out the flow structure even more clearly than the weathered surface. (Figs. A and B, PI. XXIV.) Accompanying this variation of grain there is also a diff"erence in mineral- ogical composition. The dark bands are composed essentiall}- of quartz grains, with feldspar, sericite, some magnetite, considerable calcite, and rare crystals of apatite and rutile, the quartz including- many black and indeterminable specks. The pink bands are very fine grained, so much 94 THE CRYSTAL FALLS IRO:J^-BEARING DISTEICT. so that tlie clear white mmeral grains composing it can uot be determined, though probably both quartz and feldspar are present. These bands are darkened by innumerable black indeterminable specks and long rutile needles, with a small amount of biotite. It is possible that some of the minute biotite flakes have been mistaken for rutile needles when viewed on edge, but it is certain that( these bands contain a great deal more rutile than do the others. Whether or not there is a still more essential chemical difference between the bands than that indicated by the increased quantity of rutile, was not determined. It has become more or less common of late to attribute the banding found in metamorphosed eruptives altogether to the pressure to which they have been subjected. In the present instance I can not but consider the banding as being an original fluxion structure, with the slight original difterences between the bands emphasized, as it were, by subsequent pressure. It appears highly probable that the rock was originally more or less glassy and showed a flowage structure, and that the present miner- aloo-ical character of the groundmass is due to the process of devitrification wliicli did not destroy the banding of the original glass. ACID PYROCLASTICS. The only acid pyi-oclastic rock found was formed from the aporhyolite- porphyry. This is a true eruptive breccia. The fragments are angular to rounded in shape, weather to a pure white color, and have an exceedingly rough surface. This roughness is due to a great extent to perhtic partings, which are macroscopically visible, and give the rock an almost scoriaceous appearance. Other inequalities on the surface adding to its rouglmess are caused by the leacliing out of feldspars, and by the fact that many of the quartz phenocrysts have fallen out of the inclosing matrix The fragments are all aporhyolite-porphyry, containing a very large proportion of quartz and feldspar phenocrysts. The cement of the breccia is aporhyolite. This contains far less numerous phenocrysts, and is, therefore, on the whole much finer grained than the fragments. The weathered surface of the cementing aporhyolite appears a bluish gray, and is very smooth compared to the scoriaceous appearing surface of the fragments ah-eady described. Tliis dif- ference in weathering shows the brecciated character admirably, as the finer- grained matrix stands out sharply and delimits the contours of the fragments. BASIC VOLCANICS OF HEMLOCK FORMATION. 95 Movi'iiK'Hts of the inayina are shown by a flowage sti'ucture in the matrix and by the fracturing of the quartz and feldspar phenocrysts and separation of the pieces in ))oth the cement and fragments. (Fig. B, PI. XXIII.) This eruptive breccia can be seen in its best development in the NW.- SK. trending ridge, just west of the small lake, crossed by the Chicago, Milwaukee and St. Paul track in sec. 32, T. 44 N., R. 32 W. BASIC VOLiCANICS. The basic volcanics are considered under the main headings of lavas, pyroclastics, and Bone Lake crystalline schists. BASIC LAVAS. GENERAL CHARACTERS. The basic lavas are so very characteristically developed that no one could for a moment doubt their trae natm-e, even upon the most superficial examination. One of the nearly general characters is the presence of a well-marked amygdaldidal texture. (Figs. A^ B, Pis. XXV and XXVI, and fig. A, PL XXVII.) Some of the lavas are so full of amygdules that they may be correctly said to have been scoriaceous. The amygdaloidal portions of the rock masses — which may be considered the surface parts — grade over into other portions, the interiors of the lava flows, which are, macroscopically at least, nonamygdaloidal. Owing to the homogeneous character of the basic magmas, a fluxion structure is rarely shown macro- scopically, though microscopically it may be more or less well developed. Columnar jointing was nowhere observed. An ellipsoidal parting, on the other hand, is common. NOMENCLATURE. In a preliminary article on the Hemlock volcanics,^ I made a brief mention of the occun-ence on the Upper Peninsula of Michigan of the basic pre-Tertiary equivalents of the post-Tertiary and Recent family of basalts. Following the Danas, Wadsworth, Williams, Iddiugs, Kemp, Darton, and Diller, some of the most influential of the men who, in the ' The volcanics of thj Michigamme district of Michigan, by J. Morgan Clements: Jour. Geol., Vol. Ill, 1895, pp. 801-822. 96 THE CRYSTAL FALLS IRON-BEAEING DISTRICT. United States, have advocated the simplification of petrogTaphical nomen- chxture, I used the term basah, now ordinarily used for the Tertiary or post- Tertiary basic rocks. This term was, however, modified by the prefix "apo," as indicating their altered condition and the presumed presence of a glassy base.^ This was a logical continuation of the use of the prefix as proposed by Dr. Bascom' for devitrified acid lavas. More detailed studies upon the Hemlock volcanics have shown the presence of rocks which were apparently originally holocrystalline, and therefore do uot belong with the altered vitreous basalts, the apobasalts, and others in which some of the glass is apparently unaltered. Consequently, since the apobasalts comprise only a portion of the Hemlock volcanics, the replacement of that term as a general heading by the older, more general, one of basalt was considered. The use of this term is, however, not altogether satisfactory, for the rocks, while clearly recognizable as basalt derivatives, do not possess the mineralogical composition of the basalts. The term "apo" having been restricted, as above indicated, can not be applied to them, for their altera- tion is in many cases metasomatic and dpiamic, "and in most cases not devitrification. If we adopt the prefix "meta" to indicate alteration of all kinds, then these rocks could be called "metabasalts." The terms "metadolerite," "metadiabase," etc., were proposed by Dana^ for metamorphic dolerites, diabases, etc., and first used by Hawes* in the description of the altered rocks around New Haven. Recently these tenns have been revived, but with a very diff"erent significance from that with which they were first used. It is proposed to designate bj^ such terms "rocks now similar in mineralogical composition and structm-e to certain igneous rocks, but derived by metamorphism from something else."^ Fol- lowing this suggestion, an uralitized dolerite (diabase) would be called a metadiorite. Such a use of the term does not seem justified, and the 1 Loc. cit., p. 805. '^The structures, origin, and nomenclature of the acid volcanic rocks of South Mountain, by Florence Bascom. Jour. GeoL, Vol. 1, 1893, p. 82s. 3 Chloritic formation of New Haven, CoDuecticut, by J. D. Dana : Am. Jour. Sci., 3d ser., Vol. XI, 1876, pp. 119-122. ••The rocks of the "chloritic formation" on thevresteru border of the New Haven region, byG. W. Hawes : Am. Jour. Sci., 3d ser., Vol. XI, 1876, pp. 122-126. '■ On a series of peculiar schists near Salida, Colorado, by Whitman Cross : Proc. Colo. Sci. Soc, g). 6, footnote. Paper read Jan. 2, 1893. BASIC VOLOANIOS OF HEMLOCK FORMATION. 97 ol))i'C'ti(iii to it (•iiu not he g'wvn butter tliau by (juotiug the words wliicli Zirki'l uses in the disoussion (tf the uietamorphisni of rocks:' Bei solcheu nietaraorphiscli veriiiulerten Gesteinen ist es nicht zweckmilssig, sie niit dem Nameu desjeiiigeu Typus zu belegen, dein sie durcb die Veriiriderung iilinlich, oft bios scheinbar iihiilicli gewordeu siiid. Eiiie solclie Bezeichnuiig werde iiur za inissverstiiiidliclien Autt'assuug der von dem (iestein gespielten geologisclieu Eolle fiihren, welcbe niemals ausser Acht gelasseii werden darf. Und so ist es deiin ent- scUicdenvorziizielien, der Beneiiiimig soldier Gesteineeiue Form zu gebeu, in wlielcber zuviirderst auch zuin Ausdruck koiiimt, was sie friiber gewesen sind, and nicbt eiueu Nauieu zu wiibleu, der sie iu erster Linie zu etwas stem pelt, mit welcbem sie genetiscb keine Gemeiuscbaft baben. Using these terms in the way suggested by Cross, attention is most pointedly directed to that variety of rock which the secondary product now resembles mineralogically, rather than to the type from which it was derived, and which in all likelihood it still resembles most closely in its chemical constitution. Whether or not a petrographer will use the term "metadio- rite" or the term "metadolerite" (diabase) for a metamorphosed dolerite will depend on whether or not he prefers to emphasize the present miner- alogical composition of the rock, or its original characters, and thereby its chemical constitution and genetical relations. In the present report the terms "metabasalt" and "metadolerite" are used as including all those altered rocks which demonstrably were originally basalts and dolerites. These same strictures hold good in the case of Giimbel's term "epidior- ite," when used, as it is very commonly, in the literature of the Lake Superior region and elsewhere, for rocks avowedly derived from dolerites (diabases), and characterized by the presence of fibrous secondary amplii- bole. It is preferred, in accordance with the above statement, to use the term "epidolerite" (epidiabase) instead of "epidiorite" for such altered dolerites. None of these rocks, unless extremely changed, would resemble chemically a diorite, and we have come of late years to rely more and more upon the chemical composition, combined of course with the mineralogical composition and textures of the rocks, to separate the vanous kinds from one another. As stated above, the term "epi," associated with the rock name, has come more and more to be restricted in its use solely to a rock, the epidiorite, characterized by a specific alteration product, the amphibole. In respect to this restriction to specific alteration, the term corresponds to ' LehrbucU der Petrographie, F. Zirkel: 2(1 ed., Vol. f, \>. 573. MON xxxvr 7 98 THE CRYSTAL FALLS IRON-BEARING DISTRICT. "apo," and it is iiufortuuate that these two terms should have been so nar- rowly confined. As it is, the epi- and apo-basalts would be subordinate to and therefore included under the metabasalts, as this term is used in this report. In the first the production of secondary hornblende is character- istic; in the second the process of devitrification, and hence the original presence of a vitreous base, is the chief characteristic.^ METABASALTS. All of the basalts belong to the plagioclase type. They may be most conveniently divided into nonjjorjjhyritic and porphyritic kinds, according to their most obvious macroscopical characters. There has also been found a single occurrence of a spherulitic basalt, which will be described under the head " variolite." NONPORPHVUITIC METABASALT. The nonporphyritic rocks possess a fine-grained or aphanitic structure and are amygdaloidal or nonamygdaloidal. There are included under this general name the microophitic-textured fine-grained pre-Cambrian basalts (diabases in part), the very amygdaloidal forms of tlie basalts (spilites),^ and the melaphyres in part. In these rocks the former presence of a considerable amount of original glass is probable, and they show the various textures known as navitic, intersertal, pilotaxitic, and hyalopilitic. With the nonporphyritic basalts there have been included some rocks which are to a considerable extent devitrified glasses, and others in which only a few microlites have developed. These last two vitreous types occur more especially in fragments in the tuff's, and are quantitatively unimportant. petrographicai characters. — 111 color the uouporphyritic basalts on fresh fracture show various uniform shades of green, dark olive green usually prevailing. Much less common are purplish-black rocks, and these are much more vari- able in color. In one of them is seen lighter-colored schlieren, which pass over into the ordinary dark colors. The lighter-colored portions are seen on microscopical examination to be due to a smaller quantity of the iron in them and to a greater quantity of chlorite than occurs in the rest of the rock ' The above discussion was written and the determination to use the terms porphyry — without textural significauce, as in rhyolite-porphyry — metabasalt, etc., was reached, in 1896, before the com- mittee on petrographic nomenclature of geologic folios was appointed by the Director of the United States Geological Survey. '^ Microscopic characters of rocks and minerals of Michigan, by A. C. Lane: Kept. State Board of Geol. Survey for 1891-92, 1893, p. 182. BASIC VOLCANIOS OF HP:ML()CK FORMATION. 99 mass. Where weathered, the rock.s are usually covered by a thin crust, in which ^rav, brown, and pinkish tints prevail. The rocks vary in texture very much, from the dense aphanitic kinds to medium fiue-grained varieties. The latter are usually less amygdaloidal than are the aphanitic forms, and approach in ajjpearance both macroscopic- ally and microscopically the coarser-grained basalts or dolerites represented m the Michigamme district by the coarse-grained intrusives. 0\<'ing to the basic nature of the rocks, they have generally suifered mucli alteration, and as a result the original texture is in many cases poorly preserved. On the whole, however, it is remarkable, considering their age and basic character, how well preserved it is. Where it is preserved it varies from the micro- ophitic to the various microlitic textures, such as intersertal, navitic, pilo- taxitic, and hyalopilitic, and lastly glassy. In places a flowage structure is beautifulh" brought out by the jjosition of the feldspar microlites, especially around the amygdules. The constituents present are plagioclase, light-green fibrous hornblende epidote-zoisite, chlorite, calcite, muscovite, apatite, sphene, quartz, mao-net- ite, and pyrite. Of these the feldspar, apatite, and iron oxide alone are original. In some places the hornblende is wanting, the chlorite then appearing in correspondingly greater quantity. The feldspar ordinarily occurs in lath-shaped crystals showing twins of the albite type, but where the texture is fine the feldspars are microlitic, and, while showing their prominent long extension, the edges of the various crys- tals interfere, and the outlines consequently are less sharp. In some of the rocks which appear to have been vitreous the feldspar forms feather and sheaf like aggregates (figs. A, B, PI. XXVI), apparently quite similar to those described by Ransome in rocks from Point Bonita, California.^ No reliable measurements could be made upon the microlites, and consequently their character could not be determined. The feldsj)ar is more or less completely altered to aggregates of epidote-zoisite which have chlorite associated with them or are altered to sericite. In a number of places minute limpid spots of secondary qviartz and albite are present. The very small quantity of apatite present shows its usual characters. Titauif- erous magnetite ore is apparently the only oxide present. It occurs in crys- tals and in irregular grains, which in a few cases are not entirely altered, ' Eruptive rocks of Point Bonita, by F. Leslie Ransome : Bull. Univ. of Cal., ^'ol. 1, 1893, p. 84, fig. 6. 100 THE CRYSTAL FALLS IRON-BEARING DISTRICT. thoug'h iu most cases they are replaced by spheue. In some cases the alteration product is not well enough individualized for one to diag-nose it as sphene, and it should perhaps be called leucoxene. In some of the fine- grained rocks the matei'ial in the angles between the feldspars consists jjre- dominately of grains of magnetite. This abundant magnetite renders the rock very dark, giving the rare purplish-black lavas. The most of the hornblende has a light-green color. A lesser portion shows a decided bluish tinge, and gives fairly strong pleochroism. This resembles the hornblende, which in the coarse dolerites is undoubtedly sec- ondary after the augite and it is considered to be secondary after the origi- nal augite in these rocks. The original augite was presumably in most cases pi-esent in wedge- shaped pieces filling the spaces between the feldspars, and consequently the hornblende pseudomorphs never show augite outlines. No unaltered augite was observed amongst the hornblende fibers. The fine fibers frequently form a fringe beyond the original boundaries of the pieces and penetrate the adjacent felds])ar. Quite frequently the secondary hornblende shows partial alteration to chlorite and epidote. Though careful search was made for olivine or indications of its pres- ence, no traces of it were found, and I have concluded that these basic vol- canics were essentially nonolivine bearing, though it would be rash to state that such rocks did' not contain some olivine. The calcite is usually found in irregular secondary granular aggregates scattered tlirough the rock, and evidently replaces the other minerals. Less commonly it is seen as an infiltration product along fissures. A second form of the occurrence of calcite in the nonporphyritic meta- basalts, and one not so conimon as the granular aggregate, is that of large porphyritic atitomorphic rhombohedra and scalenohedra which lie embedded in the eruptive groundmass. Such a rock, as, for instance, Sp. 32472, shows macroscopically large rhombohedral phenocrysts in a green groundmass. On the weathered surface are ferruginous rhombohedral cavities, once occu ]3ied by the carbonates. The groundmass consists of rather fresh plagioclase microlites, between which are observed some quartz, fresh magnetite crys- tals, and lastly chlorite flakes as alteration products of originally present bisilicates or glass, or l)oth. The texture is undoubtedly that of an eruptive. The carbonate is more or less ferruginous, brown iron hydroxide resulting BASIC VOU'ANICS OF HEMLOCK FORMATION. 101 from its alteration, and as it ctiervesces quite readily with cold IICl, it is supposed to l)t' ferruginous falcite. Sericite is found in minute flakes replacing the feldspars, and it is also found in large porphyritic plates occurring in the eruptive grounihiiass asso- ciated witii the porphyritic carbonate al)ove descrilied. In some cases we find epidote in these altered basalts, in others zoisite. In a great number of instances the same individual exhibits the high interference colors of epidote and the low l)lue interference color of zoisite in different parts. These different portions, formed respectively of the epidote and zoisite mole- cules, are most closely intergrown, and I have therefore used the compound term "epidote-zoisite," indicating this fact. Associated with this, one finds m many of the specimens small mineral aggregates which merit somewhat further notice. These aggregates have a brownish-yellow color and possess a verv high single and also a high double refraction. In these masses the single and double refractions of the granules composing the aggregates appear to be higher than that of epidote. In shape the aggregates vary from perfectly round, zonally arranged spheres and irregular, elongated, rounded aggregates to forms giving oblique quadratic sections. All of these ao-o-reo-ates are found at times included in the epidote-zoisite crystals. In a few cases the oblique quadratic sections were seen to occupy the centers of the epidote-zoisite crystals, having exactly identical outlines. It is believed that they are composed of an epidote much richer in iron than the common variety with which they are associated. This increase in iron explains the darker color and the increase in single and double refraction, as shown by Forbes.' Why it should appear, especially in the aggregates, can not be explained. The chlorite does not appear to be entirely an alteration product of the secondary hornblende with which it is associated. There is usually rather more chlorite than it would seem could pos.sibly have been formed from the alteration of the hornblende alone. In some of the rocks the larger angles, as well as the extremely fine areas between adjacent feldspars, are occupied by a very fine felt-like chloritic mass. The chlorite which is not secondary after hornblende is considered as the product of an altered glassy base. No glass was observed in the nonporphyritic basalts occurring in large masses, but in one of the fragments of basalt in a tuff a dark chocolate- brown glass forms the matrix in which are lying well-developed plagioclase 'Epidote ami its optical proiierties, by E. H. Forbes : Am. Jonr. Sci., 4th ser., Vol. 1, 1896, p. 30. 102 THE CRYSTAL FALLS IRON-BEARING DISTRICT. microlites. The glass where thick appears isotropic, but where thin appears to be full of globulitic de^'itrification products, which show slight polariza- tion effects between crossed nicols. The original ])resence of glass in other basalts is considered to be indi- cated bv the occurrence of amvgdaloidal cavities, with very sharply defined walls marked by accumulations of magnetite. The character of one basalt points strongly toward its glassy condition. It is amygdaloidal, the amygdaloidal cavities being sharply defined. The groundmass contains at present no indication of the existence of any orig- inally crystalline elements whatever. It is now a dense mass of felty chlorite and minute epidote grains. Through this mass and around the amyg'daloidal cavities wind lines which are somewhat difterently colored from the rest of the matrix, and seem to indicate the direction of flowage. The amygdules are not all elongated, though some are, and these agree in direction of elongatiou. It is really impossible to describe the groundmass so as to do justice to its appearance and convince one who has not seen it of its devitrified character. The general impression it makes is that of a devitrified glass, and the photomicrograph (fig. B, PI. XXV) gives a fairly good idea of its appearance under the microscojje, and will j^robably prove more convincing than any description that might be given. Fig. A, PI. XXVII, represents a polished face of the specimen in its natural size. Another kind of glassy basalt is represented in this district. This rock resembles the one just described, but differs from it in that it was not alto- gether glassy. In it one sees long, slender, much-altered feldspar microlites scattered through the matrix. These feldspars occur in needles, which fringe out at the ends. They do not give the groundmass textures usually found in the basalts, but occur in sheaves and imperfect spherulitic forms; the rock thus approaches in texture the variolites. The base in which the feldspars lie is brownish gray, and consists of recognizable chlorite, epidote, some clear mineral in minute particles, probably quartz or feldspar, or both, and aggregates of yellowish granules, which are apparently of a single kind and are so minute as not to permit of determination. The grannies show very slight polarization effects mider crossed nicols, and the groundmass in many places where they occur in great quantity appears almost isotropic. It seems highly probable that a large portion, if not all, of the ground- mass was originally a glass. Further evidence of the originally glassy BASIC VOLCANICS OF HEMLOCK FORMATION. 103 iiiitiire ot" the gTouudiuas.s is afforded by tlie grouudinass, wliicli, under the microscope, shows variable li<»-hter and darker shades of brown, and these poi-tions interpenetrate, forminj^- an imperfect eutaxitic structure. Such structures are especially common in the glasses. The photomicrographs (figs. A, Jl, PI. XXVI) show veiy well the general microscopical characters of this rock. Chemical composition. — Tile followiug couiplcte aiialvsis, for wliicli I am indebted to Dr. Henry Stokes, of the United States Geological Survey, shows the chemical composition of one of these pre-Cambrian nonporphy- ritic metabasalts. The rock is very fine grained and microophitic, with a marked amygdaloidal character. The amygdaloidal cavities are filled with chlorite, into which project crystals of epidote-zoisite and calcite. The altered condition of the basalt is very clearly shown by the high percent- ages of water and carbon dioxide. The other oxides show nothing remarkable except that the percentage of titanium oxide is quite high. On the whole, however, the analysis is very similar to those of recent fresh rocks of the same character. Analysis of pre-Cambrian nonporphyritic metabasalt. Constituent. Per cent. Constituent. Per cent. SiO, 46.47 1.28 16. 28 3.15 8.96 .09 6.56 7.90 3.64 K5O 0.21 .13 1.26 TiO> P2O, CO. AliOi FeOi CI FeO SO, MnO H.O at 110° .28 3.89 MgO H;0 above IIC^ Total . ... CaO 100. 11 Na,0 PORPHYKITIC METABASALT. The porphyritic rocks are fine grained, and may be or may not be amygdaloidal. They include diabase porphyrites and porphyritic forms of the melaphyres. These last in the textures of their groundmass correspond to the labradorite-porphyrites, the equivalents of the andesite-porphyries, though more basic than they. The phenocrysts lie in a fine groundmass which shows the same kinds of texture already mentioned as having been 104 THE CRYSTAL FALLS IRON-BEARING DISTRICT. observed in the corresponding nonporphyritic basalts, the microophitic, inter- sertal, navitic, pilotaxitic, and hyalopilitic. The various basalts are connected by transition phases. The close connection between the different varieties is well shown where one passes from the fine-grained amygdaloidal rock through the fine-grained nonamygdaloidal over to the porphyritic macro- scopicallv nonamygdaloidal type. petrographicai characters. — As statcd lu tlic gcueral descriptlou, these rocks do not differ essentially from the nonporphyritic basalts just described. The most important difference is in the presence of the feldspar phenocrysts, giving them a porphyritic texture. Measurements upon the phenocrysts, made against the albite twinning plane on zone _L to 010, according to the Michel Ltivy method, give an average extinction angle of about 18°, which points toward its character as labradorite. However, angles obtained lower than this indicate the possibility of the association of andesine with the predominant labradorite. The feldspars show the usual alteration products. One very infrequently finds augite phenocrysts which have been completely uralitized associated with the feldspars. Other phenocrysts are now represented by masses of chlorite, with or without epidote, evidently pointing toward the basic and maonesian nature of the original mineral. As uralite is the common sec- ondary product of pyroxene in these volcanics, the altered phenocrysts represented b^s' chlorite masses are not believed to have been pyroxene. The original mineral was perhaps olivine. The very noticeable scarcity of augite phenocrysts in the basalts stamps them as difterent from the great majority of basaltic rocks and as being very similar to the basalts described by Judd,^ from the Brito-Icelandic petrographicai province, in which por- phyritic crystals of augite are seldom if ever seen and in which the pheno- crysts are feldspar and sometimes olivine. The groundmass in which the phenocrysts lie have generally the same mineralogical composition and texture as the nonporphyritic rocks already described, and the two kinds are supposed to have been originally similar.^ ' On the gabbros, dolerites, and basalts of Tertiary age in Scotland and Ireljind, by J. W. Judd: Quart. Jour. Geol. Soc, Vol. XLII, 1886, p. 79. -The groundmass of one of these porphyritic forms dift'ers somewhat in one important respect. In it were observed numerous round areas of small size occupied by a clear white aggregate, polariz- ing in low gray colors. The centers of some of the areas were occupied by clumps of yellow grains, with here and there a minute flake of chlorite. Others contain only the white material, which is apparently secondary. The round areas are not sharply delimited, and hence are most probably not microamygdules. Their general appearance is strikingly like that of leucite in those plagioelase BASIC VOLCANIOS OF IlKMLOCK FOKMATION. 105 Measurements made on the feldspar microlites (»f the g-roundmass gave 17° as tlie maxinmm extinetiou in zone perpendicular to 010. This angle points toward tlie microlites being acid labradorite. The microlites thus seem to agree essentially with the phenocrysts in composition. Flowage structure around the phenocrysts is most distinctly shown by the arrangement of the feldspar microlites. In one case in which the porphyritic texture and the ilowage structure are very good, secondary actinolite crystals have devel- oped parallel to one another and parallel to the flowaee direction, o-ivino- the rock under the microscope a distinctly schistose appearance. Chemical composition. — In the preliminary article upon these Hemlock vol- cauics published in 1895/ the occurrence of andesites as well as basalts was mentioned. This determination was based solely on the microscopical study of the rocks, and the rocks which were presumed to be andesites were those porphyritic forms which have just been described. Since the publication of that article the following analyses (Nos. 1 and 2) have been obtained of the porphyritic rocks. The rocks selected for analysis were those which appeared to be especially rich in feldspar, and, having a rather lighter color than the others, seem to be somewhat more acid than the average. These, it was thought, might have the composition of andesite. The comparison of series of rocks derived presumabl}- from the same magma is more profitable than the study of single analyses. This line of investigation, as followed by Rosenbusch," Iddings,^ Lang,^ Broegger,' Becke,* and Michel Li^vy,' has been very fruitful. basalts iu which it is present in very small quantity, filling irregular but iu general rounded areas between the other constituents. It would, of course, be impossible to base the determination of the former presence of leucite in these pre-Cambrian rocks upon such scant evidence as has been obtained. Still it is worth while to notice even such a doubtful indication of its former presence as has been mentioned above. ' Jour. Geol., cit. pp. 805-806. 'Ueber die chemischen Beziehungen der Eruptivgesteine, bv H. Rosenbusch: Tsch. Min u Pet Mitt., Vol. II, 1889, pp. 144-178. 3 Origin of igneous rocks, by J. P. Iddings : Bull. Phil. Soc. Wash., Vol. XII, 1892, pp. 88-214. Tlie eruptive rocks of Electric Peak and Sepulchre Mountain, by J. P. Iddings: Twelfth Ann. Kept U S Geol. Survey, 1892, pp. .571-664. < Ordnung der Eruptivgesteine uach ihrem chemischen Bestand, by H. Otto Lang: Tsch. Min. u. Pet. Mitt., Vol. XII, pp. 199-252. Beitrage zur Systematic der Eruptivgesteine, bv H. Otto Lang- Tsch Min. u. Pet. Mitt., Vol. XIII, 1892, pp. 115-169. ■• Die Eruptivgesteine des Kristianiagebietes, by W. C. Broegger. I. Die Gesteine der Grorudit- Tinguait serie. II. Die Eruptiousfolge der triadischen Eruptivgesteine bei Predazzo in Siidtyrol. Videnskabsselskabets Skrifter, I Matheuiatiek-naturv. klasse. No. 4, 1894; No. 7, 1895. ; H2O above IIC^ Total 100.26 99. 73 100. 11 aNo. 3 is the analysis by Dr. Stokes of the nonporphyritic basalt, and is given for comparison. I Uber (lie cliemischen Beziehungen der Eniptivgesteine, by H. Rosenbusch : Tsch. Min. n. Pet. Mitt., Vol. XI, 1890, p. 144. 2 Tables Xos. II and III were calcuLated for me by Mr. Victor H. Bassett, assistant in cbeniistry ill the University of Wisconsin. BASIC VOLCANICS OF HEMLOCK FORMATION. 107 TAUI.E II.— MOLKCULAK rKdl'dRTloN Ol' THE CIIIKB- OXIDES. SiO. 50.55 3.54 16. 45 3.28 9.72 5.41 4.50 5.05 1.50 54.07 1.40 16. 38 6.29 4.33 5.71 5.18 5.95 .69 49.15 1.35 17.22 3.33 9.57 8.36 6.94 3. 85 .22 TiO; Al.Oi Fe,0 1 FeO CaO MgO Na^O K Total 100.00 100.00 100. 00 TAHLE III.— ATOMIC PROPOKTION OF METALS. Si 46.97 2.48 18.07 9.87 5.42 6.25 9.16 1.78 49.49 .97 17.73 7.67 5.63 7.09 10.61 .81 45.41 .94 18.80 9.74 8.33 9.58 6.93 .27 Ti Al Fe Ca JI" Na K - Total lOO: 00 100. 00 100. 00 1 As tne calcium increases there is a corresponding increase of magne- sium and a diminution in potassium. A decrease in sodium is also shown if Nos. 1 and 3 alone are compared. The percentage of sodium present is rather high and with the potassium indicates a magma family rich in alkalies. The magnesium is notably high; such high percentages as we have here usually accompanying much lower percentages of alkali. It may also be noted here that the presence of the magnesium in such amounts indicates the former presence of olivine or the presence still of its alteration products, a point to which attention was directed in the microscopic description of the rocks. No. 1 is remarkable for its percentage of titanium, which is very high, even when compared with that contained in the others, which are themselves considerably above the average. All of the rocks contain a large amount of water of hydration. The percentage of CO2 contained in Nos. 1 and 2 indicates also that they are much altered. These analyses show that the rocks can not be classed with the typical andesites. Should they be called andesites at all, they must be classed with 108 THE CRYSTAL FALLS IRON BEARING DISTRICT. the augite-andesites, and placed on the border Hne between them and the plagioclase basahs. It is preferred to hichide them under the basalts, though it can not be doubted but that if analyses of perfectly fresh rocks could be obtained, there would be found some which would incline more decidedly toward andesites than do the above specimens. VAKIOLITIC METAUASALTS. Variolites are spherulitic basalts, usually very vitreous. Since the tend- ency to crystallization is so much stronger in the basic than in the acid rocks, it is not surprising that they should be far less common than the cor- responding acid kind. Moreover, the basic glasses are very susceptible to alteration, which naturally obscures the original characters of the rocks. This probably partly accounts for the fact that they are very nifrequently observed. This spherulitic phase of the basalts is well known in Europe, but there has thus far been found only one reference to its occurrence in the United States. Ransome^ has described a variolite from Point Bonita, California. To this there may now be added a single occurrence in the Crystal Falls district of Michigan. This variolite exposure occurs at N.375, W. 900, sec. 4, T. 44 N., R. 33 W., in close proximity to the remnant of a basalt sti'eam which shows well-marked flowage structure. The relations of the two rocks are not determinable from the exposures. The rock presents a very rough manimillated surface, due to diiferential weathering. The varioles, being more resistant than the groundmass sur- rounding them, form the protuberances. These protuberances vary in shape from round to oval, and very rarely are irregular. The varioles vary also in size from minute ones to those about one-half inch in diameter, and con- stitute by far the greater part of the rock. These general characters may be seen on the photogi'aph, fig. A, PI. X, taken from the hand specimen. The color of the weathered surface of the rock is gray or light brown, while the fresh surface is in general a dark green. Upon the polished surface of a fresh rock the varioles have an olive-green color, with, in the majority of cases, a distinctly darker center of purplish color. Less frequently this center is lighter green than the remainder of the variole. The varioles are usually separated from each other by narrow areas of groundmass, darker than the varioles themselves, with a purplish or very dark olive-green • The eruptive rocks of Point Bonita, California, by F. Leslie Ransome: Bull. Dept. Geol. Univ. of Cal., Vol. 1, 1893, p. 90. PLATE X. 109 platp: X. Fig. .4. (Sp. No. 32273. Natural size.) Photographic reproductiou of the weathered surface of a variolite. This brings out very clearly the uiaumiillated surface of the rock, which is due to the diftereutial weathering of the varioles and of the groundmass between them. The rounded character of the varioles, and their gradation irom those of very small to those of much larger size can readily be seen. (Desc., p. 108.) Fig. B. (Sp. No. 32273. Natural size.) Reproduction of the polished surface of a variolite. This is designed to show the circular character of the varioles, and the fact that each is separate and distinct from the one adjoining it. It can be seen that some of the varioles have very dark and others much lighter centers. (Desc, p. 108.) 110 U. S. GEOLOGICAL SURVEY MONOGRAPH XXXVI PL. X (-A) Weathered surface of variolite. iS) Polished surface of variolite. BASIC VOLCANIOS OF HEMLOCK FORMATION. 1 1 1 color. In places the varioles are in juxtapositiou. However, they do nut coalesce, but each is separate and distinct (tig. B, PI. X). The rock when examined under the microscope is seen to be extremely altered. The only ori<;-iual minerals present are feldspar, apatite, and pos- sibly some magnetite. The groundmass consists of a finely crystalline secondary aggregate of flakes of chlonte, associated with minute limpid grains, some of which are probably qua/tz and others feldspar. Scattered tlu-ough this aggregate are grains of epidote, calcite, a few crystals of original apatite, and mag- netite, and numerous dark reddish brown and black fen-uginous specks. The varioles are readily distinguishable from the matrix. From this, as well as from each other, they are invariably separated by a crack, along which reddish-brown ferruginous matter has been infiltrated. The varioles are in o-eneral much finer grained than the groundmass, and at times exhibit phenocrysts of feldspar. The composition of the varioles is the same as that of the groundmass, except that apatite is more common in them, and that in addition to the minerals mentioned as occurring in the groundmass a small quantity of original feldspar may be recognized, both as phenocrysts and as part of the groundmass of the varioles. Where these feldspars occur, they are to a great extent replaced by a mass of epi- dote, chlorite, sericite, quartz, and feldspar. The phenocrysts are found near the center of the varioles, and the occasional light-colored centers which were observed macroscopically are due to the presence of these altered feldspar phenocrysts. The more frequent dark centers are due to an acctunulation of the dark ferruginous specks in varioles in which the phenocrysts are wanting. No textures could be determined from the remnants of the original minerals. In one variole aggregates of secondary epidote grains and fer- ruginous specks lie in such a position as to produce a distinct radial arrano-ement. With advancing alteration, spherulites in acid rocks are frequently found to have between their radial fibers secondary deposits of epidote and ferruginous matter, which mark very clearly their radial arrangement. The similar radial arrangement in these varioles of epidote and ferruginous matter seems to point to the varioles having possessed the spherulitic character, though it is now impossible to determine the nature of the fibers forming the spherulites. 112 THE CRYSTAL FALLS IKON BEARING DISTRICT. TIIK ELLIPSOIDAL STRL'CTUUE IN THE iMETAIS ASALTS. Upon examining the flat surfaces of many of the lavas one is immedi- ately struck by their resemblance to a conglomerate formed of round bowlders, all of the same kind of rock, lying in a matrix of very small quantit}' and of very different color. Fig. 7 is a sketch showing a portion of such a lava flow. I find that these ellipsoidally parted rocks have been called "massive conglomerates," and the blocks have been spoken of as "bombs" in the manuscript notes of some of the men who have worked among them. The latter term was undoubtedly due to the resemblance of the ellipsoids to the spindle-shaped pieces of lava m ^ - H ■^fmm'^:rA Fig. 7.— Skitch of ihe siirliice dl' Uic «Hitci'ui» of au ellip.suidal basalt, aliuwiufi the gtin^ral rliaiaiter of tlit- ellipsoids find matrix- which one finds around the modern volcanoes. Ellipsoidal basalt is very common throughout the Hemlock volcanic area. It is found most frequently in isolated ledges. However, it is also associated with and grades into non- ellipsoidal varieties. In one good exposure it is overlain by a fragmental scoriaceous mass which separates it from another mass of similar ellipsoidal basalt. While the scoriaceous portion may represent the brecciated surface of a lava flow, it is not so considered, biit is presumed to be a tuff" deposited upon the flow represented by the ellipsoidal basalt. According to this view the ellipsoidal basalt is on the surface. In another exposure an ellipsoidal basalt overlies a bed of water-deposited clastic I'ock. There is no passage between the two kinds of rock. The contact between the two is an undu- lating one, and is marked by a mass of schistose material about 2 inches thick and similar to that which is between the ellipsoids. This particular BASIC VOLOANICS OI'" HEMLOCK FOKMATIOX. 113 biisiilt is \cr\- iK'iisi.', with ouh' otHMsioiiuUy small cliloritc-lilU'd \c.sicli's in it, iiiul tlicic is no true flow structure observable. The facts cited seem to show thai this elli[)soiihil [lortiou was tlic surface of a lava flow, whether the top or the bottom is immaterial In c(n-taiu cases the (■lli|)soi(hil tiu-ies ma\' constitute an entire How. Where tlie direction of How coidd with' anv de"Tee of certuint\' be deterniinc^d, it w;is seen that the two hiu"er axe.s of the ellipsoids are in the })laue of tlie flow. Tlie ellipsoids vary in size from a few inches to G or 8 feet in diameter, and are usually spoken of as spheroids. Attention has already l^eeu called to tlie incorrect usaye of this term by F. Leslie Ransoiue, in lii.s iuterestinu- paper on "The eruptive rocks of Point Bonita, California."' The outlines of the Ixtdies are circular onlv in exceptional cases. Un the other hand, sections in all directions through them give almost in- variably ellipses, aiid there- fore they are more })roperly ellipsoids than spheroids. On the surfaces exposed the long- axes of the ellipses lie in the same general direction. The ellipsoids are formed of a very fine-grained porphy- ritic or nonporphvritic rock. This is amygdaloidal or non- amygdaloidal. Where amygdaloidal, the amygdules are as a rule dis- tributed throughout the ellipsoids, though on the whole the masses are more scoriaceous on the periphery than near the center. In exceptional cases, the amygdules are much more numerous on the west side of the ellipsoids than on the east side (fig. 8). In such cases the west sides are toward the tops of the lava flows. The ellipsoids are very commonly split up by cracks. Some of them have a roughly radiate arrangement. These may be due to the effects of contraction in the early stages of the existence of the Fig. 8. — Sketch showing the couceDtr.^tion of the ain.vgdak)idal cavities ou one aide of au ellipsoid, thia side probably representing the side nearest tlie surface of the How. MON XXXVI- ' Op. cit., p. 75. 114 THE CRYSTAL FALLS IRON-BEARmG DISTRICT. ellipsoids. Others, and by far the greater number, liave one set of lines par- allel and another parallel set which in different cases cut the first set at dif- ferent angles, very rareh' at a right angle (figs. 8, 9). One of these sets is usually tranverse to the long axes of the ellipsoids (figs. 7, 8). The blocks are separated from one another by a thin layer of a schistose matrix, rarely more than 3 inches in thickness, though exceptionally nearly 8 inches thick. (Of figs. 7 and 8 of this pajjcr, and fig. 1 by Ransome.^) Since the above description of the Crystal Falls ellipsoidal lavas was written in 1896, there has appeared Sir Archibald Greikie's valuable work on the Ancient Volcanoes of Great Britain,- in which several similar occur- rences are mentioned. His illustration of this structure on page 184, as can rcadih' be seen on comparison, would answer, but for the absence of a well- defined schistose matrix between the ellipsoids, very well for >^^^^C ^ sketch of a Michigan pre-Cambrian ellipsoidal lava. ImI 4x1/1 'I ll '^riw schistose matrix between the ellijjsoids upon the i|i''Tj+l/ll^/. weathered surface is seen to be made up of layers concentric '/^^^f with the ellipsoids. It is possible that these layers are not EiG. 9.-Eiiipsuids absolutelv concentric in the third dimension. However, no ■with sets of parallel • i r l l • • c ^ • • f^i lines cutting each exposui'e ])ermitted 01 the determination ot this pomt. iTe- quenth' certain layers seem to grade off into others of a some- what different character. The matrix between any two ellipsoids usually separates near the center; where apparently the greatest movement having occurred the schistosity is most developed. One can often as easily knock an ellijjsoid out of its encircling matrix as one can the kernel out of a nut. In some cases there is no absolutely sharp line of demarcation l)etween matrix and ellipsoid, but a gradation from one into the other. At the ^^laces ■where three blocks are in juxtaposition one frequently finds, instead of a triangular space entirely filled by the matrix, in the center of the matrix a triangular area of infiltrated vein quartz (figs. 7, 8). In certain cases the minerals which compose the schistose matrix are not thoroughlv cemented and give it a somewhat friable character, causing it on weathered surface to appear granular. In very rare cases a matrix with a distinctly brecciated character was observed, but in this as well as in the cases above described a certain degree ' Op. cit., p. 76. = Ancieut Volcanoes of Great Britain, by Sir Archibald'Geikie, Vol. I, 1897, pp. 26, 184, 193. PLATE XI. 115 PLATE XI. (Sp. Xii. 2367."). Xatiinil si.e.) Thi.s colmed plate rej>re.seiits tlic pnlisUeil surface of iiu ellipsoiil with a portion of the matrix which surrounds it and .separates it from the adjacent ellipsoids. The dense character of llie center is nicely .shown. Aroiiud this oval area we get narrow concentric zones of alternatiug light-green and dark-green material. The light-green material corresponds to that in the center, and represents the least-altered basalt of the ellipsoids. The dark-green areas are the cliloritized basalt. Beyond this, forming the outermost greeuish-gray zone, one linds the matrix, which possesses distinctly frag- mental cdiaracters, though in spite of this with a marked schistose character. The schistosity of the matrix conforms to the contours of the ellipsoid. 116 us GEOLOGICAL SURVEY MONOGRAPH XXXVI PL XI JULIUS eiENaco lith n v BASALT ELLIPSOID WITH MATRIX. BASIC VOLCANICS OK IlKMLOCK FORMATION. 117 (it' scliistositA' is iiotici'iiblc. Tlif iiiaTrix lictwccii llic ellipsoids viirics very iniicli in (l('<>Tee of" srhistosity, color, aiul coiniiositioii. Tlic most schistose, and 1)\' far tlic most comuKin variety, is tlie dark tiTcen matrix, whicli consists essentially of chlorite, ei)i4. Uelier niechiiuische Gcsteiusuniwandlungeu bet Hainii--heu in Snrbsen, by A. Rotbpletz : Zi:-itschr. (lent. (ieol. Gesell., Vol. XXXI, 1879, pp. 374-397; Vol. XXXII, 18S0. p. 447. Report (lu tlie jjeology of iiortbein New Brunswick, by R. W. Ells: .\un. Rept. Geol. and Xat. Hist. Survey of Canada, 1879-80, D, p. 24. E. Datlie: Jarb. K. pieiiss. geol. Laudesanstalt, 1883, p. 432. K. Dalmer: C'f. Zirkel Pet., Vol. IT, p. lioO. Report on the geology of the Lake of the Woods region, by A. C. Lawson : Geol. and Nat. Hist. Survey of Canada, 1885, CC, pp. 51-.53. The greenstone-schist areas of the Meiioiniuee and Marquette regions, by G. H. Williams: Bull. U. S. Geol. Survey, No. 62, 1890, pp. 137, 166-168, 173, and 203. Ou the variolitic rocks of Mont Genevre, by G. A. J. Cole and J. W. Gregory ; Quart. .Jour. Geol. Soc, Vol. XLVI, 1890, pp. 29.5-332. On a variolitic diabase of the Fichtelgebirge, by .J. W. Gregory: Quart. .Jour. Geid. Soc., Vol. XLVII, 1891, i.p. 45-62. The Kawishiwin agglomerate at Ely, Minnesota, by N. H. Winchell: Am. Geol., Vol. IX, 1892, pp. 359-368. The eruptive rocks of Point Honita, California, l)y F. L. R.ansome: Bull. Dept. of Gecd. I'uiv. of Cal., Vol. I, 1893, pp. 71-114. Editorial note on the above paper, by N. H. Wiuchell : Am. Geol., Vol. XIV, 1894, p. 321. The geology of Angel Island, by F. L. Ransome: Bull. Dept. Geol. Univ. of Cal., No. 7, 1894, p. 202. Variolite of the Lleyn and associated volcanic rocks, by C. Raisin: Quart. .Jour. Geol. Soc, Vol. XLIX, 1893, pp. 145-165. Ou a radiolarian chert from Mullion Island, by H. Fox and .J. .J. H. Teall: Quart. .lour. Geol. Soc, Vol. XLIX, 1893, p. 211. Ou greenstone associated with radiolarian chert, by J. J. H. Teall: Trans. Roy. Geol. Soc. of Cornwall, 1894: C'f. Rosenbusch, Mikroskopische Physiographie, 3d ed., p. 1064. The volcanic rocks of the Michigamme district, by J. M. Clements: .Jour. Geol., Vol. Ill, 1895, p. 808. The geology of Point Sal, by H. W. Fairbanks: Bull. Uept. Geo!. Univ. of Cal., Vol. II, 1896, p. 40. Geology of the Fox Islands, Maine, by G. O. Smith, 1896, pp. 16-18. The Ancient Volcanoes of Great Britain, by Sir Archibald Geikie, London and New York, 1897, |.p. 26, 184, .and 193. BASIC VOLOANIOS OF HEMLOCK FORMATION. 119 ^'il^illus attcnijirs liavc liccn made to explain tliis peculiar structure Honnev, hatlii-, auut whose iudetenninate down- ward extension sliows them to be cohimns. The rounding of the cokimns, which were presumably originally prismatic, he ascribes to dynamic action- He also sug'srests that elliijsoidal masses could result from a similar dynamic modification of a mass of lava parted into shorter prisms, or even ellipsoids. In the description of the eruption at >>antorin, Fouque' mentions a viscous lava exuded in the form of a mass of blocks. These blocks, tum- bling o^'er one another as the mass is pushed from l^ehind, have accumulated in a rough pile, PI. XII. Fouque climbed these piles of block lava shortly after their production, and noticed the breaking off of pieces from the sides, due to the cooling and contraction of the individual blocks.^ In general this character agrees well v.'ith that of the aa lava of Hawaii, as descriljed by tlie late Prof. J. D. Dana.' He describes the formation of the blocks as due to the slow for- ward movement and contemporaneous breaking up of the viscous lava. The surface contrasts with the ropy surface of the more liquid pahoehoe. The aa is as a rule compact as compared with the })alioehoe, though the exterior "is roughly cavernous, horribly jagged, with projections often a foot or more long that are bristled all over with points and angles." From the illustrations of this lava (see fig. 10, taken fi-om Dana) the blocks may be seen to be, while irregular, still in general distinctly rounded. This is the shape which viscous material would naturally tend to take when sitbjected to the rolling action attendant upon the onward motion of the stream of which they form an outer portion, or in certain cases the entire thickness. This is clearly shown from the following quotation from Dana's description of the constitution and concli- ' Santorin t-t ties truptions, Ijy F. Foiujue: Paris, 1879. Clia]). II. Compare especially Pis, VIII and XIII. 2 Op. cit., p. 54. 'Characteristics of Volcanoes, by .I.D.Dana: New York, 1890, pp. 9, 241, and Am. Jour. Sci., 'ii\ ser.. Vol. XXXIV, 1887, p. 362. "An aa or arate lava stream consists of detached masses of lava as far as is visible from the outside. The masses are of very irregular shapes and confusedly piled up to nearly a comuion level, altliough often coveriuj; areas mauy miles long aud half a mile to a mile or more wide. The size of the masses in the coarser kind varies from a tew inches across to several vards." Fig. 10. — Rt-produetion of illustr.4tion of aa lav.i, after Daua (Characteristics of VoleaDoes). S G BASIC VOLCANICS OF HEMLOCK FOh'MATION. 121 tidii of the ;i;i strciini wlu-ii in motion:' "(1) A mass of rouyii blocks outside, jii-fciscly like the cooled aa stream; (2) tlie motion extremelv slow, indi- catiiifi- a semifluid condition heneath: . . . (.")) the Idocks of the upper part of tlie front, as the stream creeps on, tund)lin»' down the liigh slope, owinj^- to retardation at bottom from friction, and tlnis a rollinj^- action in tlie front part." Dana describes the gradation of pahoehoe into aa lava. He writes, "a lava stream may chano'e from the smooth-flowing or pahoehoe condition to the aa and back again to the smooth-flowing."- Platania' describes from Aci-Trezza and Aci-Castello basalts with globular structure. The interspaces between the globes are filled with silt, or silt and tufF, and the exterior of some of these globes presents a thin vit- reous cracked crust (cf. p. 117). These globular basalts are apparentlv but a modification of the block or aa lavas described by Fouque' and Dana, in which the separate portions of the lava have assumed a sufficiently rounded character to be called globes. However, Platania's fui-ther descriptions show this term to be clearly inapplicable unless the word "globe" is used with considerable latitude. The Santorin block lava, the Hawaiian aa lava, and the Aci-Castello globular lava are all products of a slowly-flowing comparatively viscous mass. They will in the further description be included under the general term " aa lavas," as this is the most common form of occurrence of such viscous lavas. The elhpsoidal basalts of the Crystal Falls district appear to be com- parable to the Hawaiian aa lava and block lavas of the kind described b}" Fouqu^. The lavas have subsequent]}- been exposed to great pressure and are considerably altered. The most obvious character of these masses, their rounded outline, is believed to be due to considerable extent to the onw^ard motion of the stream as desci'ibed by 1 )ana. Contraction caused by cooling, accompanied by falling off of fragments from the outside, as observed b}- Fouque'' in the Santorin block lava, would also tend to round blocks which were originally angular. (PI. XI.) In 'Characteristics of Volcanoes, liy J. D. Dana, New York, 1890, p. 242; and Am. .lour. Sci., 3tl ser., Vol. XXVI, p. 100. 2Am. Jour. Sci., 3d ser., Vol. XXXIV, p. 363. ^Geological notes of Acireale, by Gaetano Platauia: The Southern Italian Volcanoes, H. J. Johuston-Lavis, editor, Naples, 1S91, Chap. II., p. 41. <0p. cit., p. 54. 122 THE CRYSTAL FALLS IRON BEARING DISTRICT. some cases the separate portions of the lava may have been originally nearly globular, similar t(i the ones described by Platania. The ellip- soidal basalts, however, are so common in the Crystal Falls district and such globular basalts are so rare that this peculiar form is not considered worthy- of much consideration in the further discussion, the first two kinds being chiefly the forms from which these were derived. The lava blocks rolling over one another as the lava stream advanced, would lie witli their axes in all positions, but pressure and the onward movement of the flow would, in the lower portion of the stream at least, be sure to produce from the blocks ellipsoidal bodies with their two longest axes corresponding — the one to the direction of flow and the other to the lateral extension of the stream. After the stream ceased to flow and the lava solidified, there would be a gradation from the ellipsoidal into the non- ellipsoidal portion of the flow. An aa stream, such as described and shown in fig. 10, when subjected to great pressure subsequent to burial beneath thick deposits, would be compacted bv the breaking up of the jagged outer portions, which, falling down, would fill the spaces between the blocks. This broken material filling the spaces would be most exposed to movement and to the action of percolating waters. It would consequently be very much altered, as in the material described above (p. 119) by Ransome. Such alterations would result in producing a matrix of exactly the same general composition as the altered ellipsoids. It is the common case of inetamorphic action producing from rock masses of essentially the same chemical composition, but of difterent character, similar end products. This brecciated character of parts of this matrix is well shown in parts of PL XI, and fig. B, PI. XXXIV. In this case silica has been introduced, filling the spaces and marking out the outlines of the fragments. Where mashing has been excessive, the outlines of the fragments are obliterated and the matrix rendered schistose. There may even be a gradation fi-oni the schistose matrix into the altered basalt of the ellipsoid, which at the center is massive. Let me recall the statement made on previous pages concerning the distribution of the amygdaloidal cavities in the ellipsoids. This is one of the characteristic features of the lavas. We have (1) amygdaloidal cavities distributed about evenly throughout the ellipsoids, the cavities being some- what smaller in the center than upon the periphery ; (2) the cavities are BASIC VOLCANICS OF HEMLOCK FORMATION. 123 (■(luct'iitrittctl upon the |ifri})lu'ry witli tV-\v (ir only iiiicntscopical ciivitics in the center; (3) tlle^' ;ire concentrated on one side of the ellipsoid, this side rei)re.senting' apparentl\- that side of the ellipsoid tiirucd towanl tlie upper surface of the lavii stream. 'I'he following- explanation is offered for this difference in occurrence. The distribution of cavities is determined by three factors: The viscositv of the lava; the difference in specific gravity between the l)ubl)les tilling' the cavities and the lava; and the expansive action of the gas. In the case of (1) the ellipsoids are considered to have consisted of lava in a viscous condition through wliich the gas [)ores formed, but in which, owing- to the high degree of viscosity, they remained nearly or quite in the positions in which they were formed. Here viscosity was the determining factor. In case (2) the gas pores, influenced chiefl}- by the expansion of the gas, collected upon the periphery — ^^just as, for instance, in the steel ingot while the center is compact the outer surface is porous. The lava in this case was prolialily less viscous than in the former. In the last described condition of distril)ution (3), where the gas cavities are on one side, which is the upper surface, the lava was still less viscous than in the preceding cases. Here specific gravity was the controlling factor, and, as a result of the speciflc gravity and the less viscous nature of the lava, the gas bubbles rose and collected upon the upper surface. The explanation of the ellipsoidal basalts which has been offered — viz, that thev are comparable with aa or block lava — seems to offer a ready explanation for all of the observed characters. On the whole, the e\\i\)- soids owe their origin and certain peculiarities to the viscous nature of the lava. Thev possess also characters which are due to contraction, others which are due to original flowage, and still others which are the result of subsequent orogenic movements. In certain places we may find the ellipsoids only half formed — that is, attached by one side to the main unbroken part of tlie lava flow, the other side showing a rounded outline. This probably represents a place where the aa grades into a pahoehoe or smooth-flowing form. Such an instance is possibly that illustrated by Ransome.' Both Ransome and Teall compare the ellipsoidal basalts studied by them with pahoehoe lava. The latter also suggests a submarine origin for the basalts studied by him. It should be noted that pahoehoe lava in its 1 Point Boiiita, op. cit., fig. 2. p. 77. 124 THE CKYSTAL FALLS lEON-BEAKlNG DISTRICT. typical occuiTeiK-e in Hawaii is fuuiid only in dry places, whereas tlie aa is confined to those jjarts of the lava stream — which in other portions of its course is perhaps developed as pahoehoe — where it crosses moist valleys or other depressions presumed to have contained a considerable amount of moisture.^ In the case of some of the block lava of Santorin described by Fouque," with which this may be compared, the conditions were such that the lava practically welled up through the water. From Dana's description it appears that lava in the pahoehoe form can not exist in the presence of moisture, being changed to the aa form. It would thus seem that Teall's statement of a submarhie origin for the pahoehoe lava is untenable. Wherever the ellipsoids have been studied in the Cr^ystal Falls district, they have been found to exist as separate units, thus indicating the extremely viscous character of the lava. It would seem that the analogy between these basalts and the aa or block lava is much greater tlian that which exists between them and the pahoehoe or smooth-flowing lava. AMYGDALOIDAL STUUCTURK. The amygdules in the basalts are composed of nearly the same min- erals as those which occur secondarily in the rock mass itself Arranged in order of frequence of occurrence, they are as follows: Chlorite, ei)idote- zoi.site, quartz, calcite, feldspar, iron oxide, and biotite. An araygdule may consist entirely of one of the above minerals, or, as is most commondy the case, of two or more of them. In the latter case tlie minerals are usually arranged in concentric layei's. The nonoccurrence of zeolites is very noticeable. Their al)sence from these Huronian volcanics is especially striking since they are so common in their altered modern equivalents, and also occur in basalts as old as those of the Keweenawan of Lake Superior' and of the South Mountain of Pennsylvania.^ ' Cf. Characteristics of Volcanoes, by J. D. Dana : Ne-sv York, 1890, p. 243. = 0p. cit., Chap. II. 'Faragenesis and derivation of copper and its associates on Lake Superior, by Raphael I'lim- pcUy : Am. Jour. Sci., 3d ser., Vol. II, l.STl, ji. 1K8; also Cieol. Survey, Michigan, Vol. I, part 2, 1873, pp. 19-46; Gcol. of Wisconsin, Vol. Ill, l.-SO, p. 31. The copper-bearing rocks of Lake Superior, by R. D. Irving: Mon. U.S. Geol Survey, A'ol. V. 1883, p. 89. ■iTbe volcanic rooks of South Mouutain m Pennsylvania and Jlaryland, by G. H. Williams: Aiu Jour. Sci., 3d scr., \ol. XLl V. 1892, p. 491. BASIC VOI.GANICS OF HEMLOCK FOIIMATIOX. 125 It is also ot' iiitcrcsr to notice tliat there is a total absence of indica- tions ot" copper in these Huroiiian volcanics,as well as in tliosi; ot" the Peiiokee- Gofebic, althoiiiih it is associated with similar rocks in the nix-ita above referretl to as well as in many others. The aniN-^'dn.les, with the exception of those of chlorite and of l)iotite, are of nnich lighter color than the body of the rock, and from a short dis- tance gWe the rock the appearance of a porphyry. AVeathering- gives tlie rock a different appearance according to the materials filling the vesicles. Where these weather readily they are removed and the rocks become scoi-iaceous. Where, on the other hand, as frequently happens, the vesicles are tilled with (juartz, the matrix weathers more rapidly and the rounded quartz cores stand out on the face of the rock like the rjuartz })ebbles from the softer matrix of a conglomerate. In a few cases hematite is disseminated tlii'ovigh the quartz of the amvgdules, giving it the bright-red color of jasper, and by some these amvgdaloidal fillings have lieen taken for included jasper pebbles. Careful study was made of the filling of the vesicles with the object of determining the order of deposition of the minerals. However, it was found that the aniA'gdules in a single slide contain very different fillings, one chlorite, another calcite, a third epidote, and so on; and that even in the same slide the relations are not always the same, a mineral which here occupied the center of an amygdule being found there on the periphery. ]kIoreover, the same mineral species was found at times occupying tlie out- side and the center of the same amygdule. It is clear that the fillings are not the result of a solution connnon to all the lavas, but that the same kinds of solutions were active in the various lavas at different times and even in the same la\a. at different times. How- ever, the conclusions reached were that the chlorite was generally the first product deposited an. 29. Jletasiiuuitio development of the copper-beariuj; rocks of Lake Superior, liy Raphael Pumpelly : Proc. Am. Acad. Arts and Sci,, Vol. XIII, 1878, p. 307. 126 THE CRYSTAL FALLS IRON-BEAKING DISTRKJT. consideration, was quartz, the tendency naturally being to replace more alterable witli less alterable minerals. Flattening of amygdaloidal cavities. 111 SOlllC of the amVgdaloids (fig. B, PI. XXV) the cavities retain their circular shape, as though the rock had not flowed to any great extent. ]\Iore commonly the ca^'ities are di-awn out into irregular (fig. A, PI. XXV) or lenticular shapes, the long axes agreeing with the direction of flowage in case their deformation resulted from this, or with the direction of schistosity in those cases where the rocks have been extensively mashed. In some cases the cavities have been so extremel}' flattened that the amygdules appear almost the shai)e of a melon seed, showing a mere streak of chlorite in the sections cut perpendicular to the schistosity, and in the planes of schistosity large lustrous oval areas. In some few of the basalts the groundmass immediately surrounding the amygdules is characterized by an accumulation of ferruginous matter. In most cases, however, this part of the groundmass does not diff"er in any respect from the rest of the groundmass of the basalts and points to a very _ gradual cooling. Al.TEKATION OK THK BASALTS. The descriptions given are of the freshest and most characteristic basalts. As already explained, the mineral constituents in even these freshest ones have undergone a veiy far-reaching alteration. The rocks which show a more advanced stage of alteration exhibit merely a difference in degree rather than in kind, and the minerals which result are in all cases the same. They are uralite, actinolite, epidote-zoisite, chlorite, white and brown mica, calcite, sphene, quartz, and feldspar. The amount of these secondary minerals varies greatly, showing that the alteration products resulting from the same kind of original rock may differ very materially according to the process of metaniorphism. In a general way the alteration of the basalts, as observed under the microscope, has taken the following course: Even in the rocks nearest their original condition the augite has largely changed to uralite. The vitreous base, if any was present, has become devitrified. Rocks in this stage of change still show the more important external characters of igneous rocks, including in many cases those which are characteristic of glass. Some of the rocks at this stage are light gray to green and exceedingly tough. Many of these break with a ringing sound almost like phouolites. At a ' BASIC VOLCANICS OF HEMLOCK FiJUMATlON. 127 furtliLT stage of (.•luingL- thu t'cld-spars are partly altered to a graiuilar ■io(ire<'-ate of A'arious niinerals. In ordillar^' liylit the textures of iyueous rocks are still preserved, Imt in polarized light none are seen, with the exception of aniygdules which may be present. In some cases even these are obliterated, and the original nature of the rock can only be determined from its mode of ttccurrence and its association. Further changes may produce rocks which consist practically of calcite, and may be nearly white. Ag-ahi, from these basic rocks there may be produced in extreme cases, bv a process of silicification, a rock which consists practically of ])ure silica. Description of some phases of alteration. As iUuStratiug SOme CaSBS iu wlucll the same alteration products, but in different proportions and arrangement, give rocks dift'ering very essentially, there are given the following Ijrief descrip- tions of some of the rocks studied. The flow structure was noted as being exceedingly well develoj^ed in the microlitic rocks, and in some of them the production of amphibole needles and cldorite flakes has taken place parallel with the long direction of the feldspar microlites (the flowage direction), thus develo})ing, in com- bination with the unaltered microlites, a well-marked schistosity. The feldspars are still fairly well preserved. In another case the feldspar microlites have become completely sericit- ized, the interspaces between them being occujjied by epidote, chlorite, and iron oxide. The preservation of the feldspar shapes, showing in ordi- nary light the igneous texture of the rock, gives the only clue to its original nature. (Figs. A and B, PI. XXVIII.) In some of the basalts the feldspar is replaced chiefly by epidote-zoisite, and, as in the above case, such rocks show their igneous character only when examined in ordinar}- light or by uncrossed nicols. (Figs. A and B, PI. XXIX.) In still other rocks calcite is very abundant. Its occun-ence in jjor- phyritic rhombohedra and scalenohedra was mentioned in the description of some of the rocks. These porjjhvritic calcites have thus far been found only in the fine-grained microlitic types of groundmass, the coarser ophitic rocks having it only in the usual granular aggregates. Muscovite, occurring in large porphyritic ^^lates, conforms in occurrence to the calcite. When muscovite is present, calcite is found associated with it in every case, though the calcite may occur alone, and this latter is also b}* far the more 128 THE CEYSTAL FALLS IKON BEARING DISTRICT. ' common. These crystals give a secondary porpliyritic character to the havas, and the microscopical appearance of the rocks varies somewhat according to the occurrence of the calcite. Such rocks, for instance where the rhombohedra occur, look on fresh surface l^y rapid examination like por^jhyrites in which the feldspar sections are all c[uadratic. In the others the scalenohedral sections resemble in general lath-shaped feldspar pheno- crysts Iving scattered in all directions on the surface of the rock. Another case of extreme alteration is shown in a light greenish-gray, much-altered schistose rock from sec. 21, T. 46 N., R. 32 W. Upon the weathered surface long grooves are noticed — one measuring 60 mm. long bv 5 nmi. wide — which on the fresh sm-face are tilled with calcite. On faces perpendicular t(^ the long extension of such grooves they appear as narrow slits, with the long direction of the slit, that is, the width of the groove, agreeing with the schistosity. These are clearly flattened amygda- loiilal pores, and but for them the igneous nature of the original rock could not have been determined. The extreme flattening of these amygdaloidal cavities and the schistose nature of this rock produced from an original volcanic, points toward mashing as one of the causes, if not the main cause, of its present characters. It is now composed of fairly large automorphic actinolite individuals, a very small amount of biotite and chlorite flakes, and masses of grains of cpiartz, calcite, epidote-zoisite, magnetite, with ilmenite and hematite in thick plates filling in the spaces between the actinolites. If any feldspar was originally present, it is now entirely concealed hv tlie calcite and epidote-zoisite. The calcite phenocrysts are found in the ftiirlv fresh lavas. They are beautifully automorphic and are certainly not replacement pseudomorphs of some original phenocrysts, hnt replace the various minerals of the fine- grained mass. Moreover, it is clear that they were formed subsequent to all dynamic action, as their crystal Qutlines are perfect and the}' never show any evidence of pressure. This is so even in those cases where the amygdules which have been markedly elongated are filled with calcite The process of replacement could not be followed, but it is evidently con- nected with the development of chlorite, those rocks in which a great deal of the calcite occurs having chlorite developed instead of actinolite. In other sections in which the amount of porpliyritic calcite or calcite and museovite is much greater than in the rocks just described, the amount BASIC VOLCANIOS OF HEMLOCK FORMATION. 129 of i-lilorite, iron oxidt', rutik', and quartz is also greater. Tlie quartz is in very fine graius. The jn-esence of the feklspar cau only be d(itermiued witli (hffiouhy, and usually only on the edges of the sections, as the laro-e amount of chlorite in the center conceals it. The textures caused by the feldspar and the amygdules still indicate the original character of such extremely altered stages. Figs. A and B, PI. XXX, illustrate such a rock, showing the secondary porphyritic muscovite and calcite, and also the original amygdaloidal character. A still further stage of alteration gives a rock whose groundmass is composed of the finest-gTained quartz and of grains and needles of brown rutile (auatasef). In this lie rhombohedra of ferruginous calcite, plates of muscovite, and irregular flakes of chlorite. The rock is macroscopically gi-ay, hard, and quartzitic, has a ferruginous, brown, weathered crust, effervesces with cold HCl, and yet shows its volcanic character by the numerous beautiful amygdules. These stand out on the surface like pebbles in a conglomerate. In some cases the weathering bnngs out the concentric character of the filling very nicely. For example, some may be seen in which the core is quartzitic, and is standing suiTounded by a ring-like depression, showing by difference in the weathering the diff"erent character of the mineral filling. Under the microscope the only amyg- dules which happened to be cut by the section were found to be filled with fine-grained quartz, with chlorite in automorphic flakes at the center of the amygdules, and lying in the quartzitic mass. The macroscopical appear- ance of some of the amygdules shows that just the reverse condition also exists, that is, that quartz forms the centers and chlorite surrounds it. The extreme stage of such an alteration is a rock which shows no amygdules macroscopically or microscopically, but is otherwise like the groundmass of the above last-described rock. It would be impossible to determine the original character of such a rock except by its association. The extremes of texture obtained in the alteration ^w'ocesses are, on the one hand, a porphyry with eruptive groundmass and secondary pheno- crysts; on the other, a porphyritic schist, in which all elements are secondary. These extremes are connected by gradation varieties, in some of which the calcite and muscovite approach more closely to the size of the elements composing the groundmass, and which consequently approacli the ordinary schists in structure. >I0N xxxvi 9 130 THE CRYSTAL FALLS IRON-BEARING DISTRICT, 111 these rocks the porphyritic characters are unquestionably due to the itroductioii of secondary pheuocrysts of mica (muscovite) and calcite, not by contact metaniorphisni Ijut by dynamic action.^ It has not been found possible to determine definitely from a study of the specimens, in many cases from widely separated exposures, on which the above observations were made, whether the process which has taken place in the production of such rocks has been a combination of calcification and silicification, or a process by which carbonate is being replaced by silica or the reverse. The replacement of carbonate by silica, as shown by Irving and Van Hise,^ has taken place extensively in the case of the ferru- ginous carbonates of the Penokee-Gogebic and Marquette iron ranges of Wisconsin and Micliigan. The automorphic character of the carbonate would seem to point toward calcification as the controlling process in the Crystal Falls rocks. Though the presence of quartz as the last filling of the amygdaloidal cavities points toward silicification as being the process which would eventually predominate, it is most prol^able that both processes of calcifi- cation and silicification are active; but whether the one or the other is the controlling one depends upon the depth of burial of the rocks which are altering. This statement appears to be supported by the facts to be described in the following pages. The following oljservations, which were made upon sections taken fi-oiu an ellipsoidally-parted basalt occurring on top of the bills to the west of and overlooking Mansfield, illustrate the changes which take place in the passage from the massive rock of the ellipsoids into the schistose material of the mati-ix. • The change is one of increasing altera- tion. This alteration is largely one of carbonation followed by silicifica- ' Metamorpbism of clastic feldspar in conglomerate schist, by J. E. Wolff: Bull. Mvis. Couip. Zool., Vol. XVI, 1891, pp. 173-183. Pis. I-XI. Cf. also Wolff on Green Mountains, Mon. U. S. Geol. Survey, Vol. XXIII. Principles of North American pre-Cambrlan geology, by C. R. Van Hiso: Sixteenth Ann. Rept. U. S. Geol. Survey, Pt. 1, 1896, p. 692. Phases in the metamorpbism of the schists of Southern Berkshire, by W. H. Hobbs: Bull. Geol. Soc. Am., Vol. IV, 1894, pp. 169-177. '^ Origin of the ferruginous schists and iron ores of the Lake Superior region, by R. D. Irving: Am. Jour. Sci., 3il ser., Vol. XXXII, 1886, pp. 255-272. The iron ores of the Penokee-Gogebic series of Michigan and Wisconsin, by C. R. Van Hise: Am. Jour. Sci., 3a ser., Vol. XXXVII, 1889, pp. 32-48. The Penokee iron-bearing series of Michigan and Wisconsin, by R. D. Irving and C. R. Van Hise : Tenth Ann. Rept. U. S. Geol. Survey, 1889, pp. 341-507 ; Men., Vol. XIX, 1892, pp. 254-257. BASIC VOLOANICS OF BEMLOGK FORMATION, 131 tioii. It may be characteristic nho of basalts with uo ellii)S()idal parting, but it has l)eeu possible to follow the siiccessive changes only in the ellip- soidal basalts. This is due to the fact that each ellipsoid shows all stages from the comparatively fresh material of the center to the much altered material on the periphery, and to the most altered basaltic material forming the so-called matrix sun-ounding the ellipsoidal bodies (p. 114). The freshest part of the interior of an ellipsoid from this occurrence is a very fine-grained micro-amj-gdaloidal basalt, in which in ordinary light lath-shaped feldspar microlites can be readily distinguished. Upon close examination the feldspars are found to be nmch altered, and in many cases their crystal outlines are almost completely filled out by grains of calcite and flakes of sericite and chlorite in a quartz-albite (!) aggregate. The spaces between the feldspar laths are now occupied by large crystals of epidote-zoisite, grains of iron oxide, a few flakes of chlorite, and innumera- able small round yellowish-brown and greenish indeterminable Ijodies. The epidote-zoisite crystals also include large quantities of the brown and green globular bodies, showing that they were produced previous to the epidote-. zoisite. The substance in which this aggregate is embedded could not be determined, as the aggregate is either so dense that nothing could be dis- cerned or else underlain by feldspar. In the last case the substance is r.oen to be clear white. The minerals mentioned, with the exception possibly of the iron oxide, have evidently been produced secondarily from the sub- stance or substances originally filling the spaces between the feldspars. Nothing points toward the original substance or substances having been crystallized, and I am inclined to believe that it was glass. Toward the exterior of the ellipsoid the rock is more altered. The zoisite and calcite are more abundant. The calcite occurs in the spaces between the feldspars, as well as occupying parts of their outlines. (Figs. A and B, PI. XXXI.) All of the other products drop into the back- ground, owing to the fact of nonproduction, or concealment by the zoisite- calcite aggregates. Still nearer the exterior of the ellipsoid the calcite frequently fills the spaces once occupied by the feldspars with long scalenohedi-al crystals, which in a way maintain the original igneous structure. The calcite is, however, not confined to these feldspar areas alone, but, as stated above, also occurs between them. 132 THE CRYSTAL FALLS lEON-BEAEOG DISTRICT. The matrix, representiug the most altered phase, is a granular aggre- gate of calcite, in which one may here and there discern small clear limpid grains of secondary quartz and feldspar (?) and flakes of chlorite. The calcite includes in considerable quantity the globular bodies mentioned. These are found also in the spaces between the calcite grains, as though pushed away from the gi-ains as thej^ crystallized. Some of the calcite in the first stages of the alteration of the rock may have been derived from a basic feldspai; It is clear, however, that the o-reat mass can not owe its origin to this process, but nmst be the residt of infiltration. The calcite grains derived from, and lying in, the feldspar acted as nuclei, around which the infiltrated calcite was gradually col- lected, producing pseudomorphs after the feldspar laths. Quite recently Dr. W. S. Bayley^ has noted in the Clarksburg submarine volcanic forma- tion of the JMarquette district, Michigan, the occurrence of tuffs, in which calcite has been introduced in such quantity that they may almost be called limestones. In another case in which the alteration of the ellipsoid (PI. XI) appar- ently proceeded along the lines of shearing, and produced the kind of aggregates of chlorite, including crystals and aggregates of epidote- zoisite, which were described (p. 117) as the usual matrix of such elhpsoids, one can see in thin section the calcite entering the chlorite aggregate along minute fissure lines. The calcite literally eats its way into the chlorite, and produces by an interchange of elements a mass of calcite (magnesian 1) and epidote, besides including epidote which originally occurred scattered through the chlorite aggregate. The carbonation of the original basalt or of the secondary chlorite mass results in producing a mass of carbonate which has associated with it some secondary quartz, chlorite, and epidote. This carbonate mass may be almost massive or it may be decidedly schistose. When schistose, the grains of calcite and quartz have a uniform elongation, and the schistosity is matei-ially enhanced by flakes of chlorite, which are not uncommonly found in thin streamers or thick masses in the carbonate aggregate, at times in sufficient quantity to give it macroscopically a decided green tinge. I have used the term "carbonate," although having described in detail above the calcification of the basalt, for the reason that at times, and for no ' Mod. U. S. Geol Survey, Vol. XXVIII, p. 473. BASIC VOLCANICS OF HEMLOCK FORMATION. 133 discernible re.ason, the iron carlionate (siderito) may replace the ealcite, in which case we get a dark bluish-black variety of" matrix (p. 117). The siderite masse.s do not ditier essentially from the ealcite, tliough in some of them a very small quantity of actinolite is found associated with the chlorite. As illustrating the purely local development of these two carbonates, I would mention having observed in one section a band of siderite separated from a band of carbonate, which from its color appeared to be quite pure ealcite. One may also see commonly in exposures areas of pure white ealcite, almost in juxtaposition with areas of siderite. It is a fact generally recognized that carbonation is a process confined to the outer crust of the earth, so that we may perhaps best explain the local occurrence of these carbonates replacing the basalt as products of carlDonate-bearing waters. That such carljonation of the igneous rocks through which these waters percolate is now taking place seems certain. The carbonate grains in the rocks described are shattered and elongated, or at least show undulatory extinction. They thus give evidence of having been more or less mashed since their production, and this mashing probably took place after they had been more or less deeply buried, and was, as a matter of fact, to some extent due to the pressure of the superincumbent rocks. The probability that these rocks have been thus deeply buried subse- quent to their formation is to be borne in mind with special reference to the next process to which they have been subjected, that of silicificatiou. This process is most clearly shown in the siderites, and the phases of alteration noted in their study will be briefly described. The microscope shows the siderite matrix to be a coarsely granular aggregate composed essentially of crushed siderite grains. Between these grains in a few places are small grains of quartz, flakes of chlorite, and very rarely needles of actinolite. Large quantities of black ferruginous specks are included in and also lie between the quartz grains, and such specks are also to be seen included in siderite areas, but close inspection shows that they are also associated with blebs of quartz. The chlorite flakes and quartz grains are generally elongated in the same direction, and the quartz shows wavy extinction. A more advanced stage in the process of silicificatiou was studied in the case of a rock which is bluish-black in color, exceedingly fine grained, and minutely schistose, the schistosity agreeing with the contours of the 134 THE CRYSTAL FALLS IKON-BEARING DISTRICT. ellipsoid from around which it was broken. This is essentially an exceed- ingly fine-grained quartz rock, with chlorite flakes and black ferruginous specks scattered through it, and here and there an irregular oval siderite o-rain remaining. Very few and unimportant grains of epidote were also noticed. This rock represents nearly the last stage in the process of silici- fication by which the siderite has been replaced, and a part, probably the greater part, of its iron content oxidized. Some chlorite and epidote has been produced, clearly from the lime and magnesian impurities in the siderite. Essentially the same process of siliciflcation has been described bv Van Hise in his various articles on the Penokee-Gogebic and Marquette iron ranges, to which refei'ences have been so frequently made. I have desired especially to call attention to it here, however, on account of the fact that it shows the possibility of the production of an iron ore from an original eruptive rock by the combined processes of earbonation and silici- flcation. It is true that the end product in the case described does not contain enough iron to be an ore deposit, but that is a mere detail. May not this serve also to some extent to explain the numerous clearly marked belts of magnetic attraction which occur throughout this area of altered basalts, in Avliich little of the original magnetite remains i;naltered to exert an influence npon the magnetic needle"? To explain this we must suppose the influence to be exerted by secondarj^ magnetite accunudated along cer- tain lines. The magnetic lines traced out agree in a very marked way with what has been determined to be the trend of the lava flows and tuff beds. The condition which would determine the presence of such a line of earbonation, if we may so put it, may be the presence of a scoriaceous lava flow or a bed of tuff, which offers exceptional facilities for the passage of carbonate-bearing waters. It is thus intimated that there is possibility of flnding purely local ore bodies of small size even in the midst of this volcanic area. The process of siliciflcation is generally considered as a deep-seated one, occurring far below^ the outer weathering zone. When the rocks exhibiting these various phases of siliciflcation are exposed in the zone of weathering, certain interesting re.sults are obtained which are worth noticing. Rocks are produced from these which upon the surface strongly resemble amygdaloids, but in which the pseudo-amygda- loidal cavities are of purely secondary origin. For instance, when the siderite mass has become onlj' partially replaced by silica, weathering BASIC VOLCANICS OF HEMLOCK FOKMATION. 135 afcncies It'iu-li out the reinaiuiug siderite areas and leave the tliiu lihns of silica which lie between them standing up, thus giving the rock the appear- ance of a ver}' dark pumice. As the silicificatiou progresses the siderite is very much reduced in (puuitity, the intervening siliceous areas increasing correspondingh'. The pressure exerted upon the rock has caused the isolated siderite areas to take on an oval cliaracter, the longer axes in general agreeing and being perpendicular to the ])ressure. When such siderite areas are leached out, the silica bands remain, and pseudo-amyg- daloidal cavities are produced, giving a very perfect jiseudo-amygdaloidal structure to the hand specimen. This is the origin of that character of matrix which some of the geologists have described in their field notes as like rotten, worm-eaten wood (tig. B, PI. XXVII). Although at present the material between the ellipsoids differs so markedly from the rock forming the ellipsoids themselves, nevertheless there is no reason for supposing the original composition of that part of the rock mass to have been essentially different. The change in the character of the basalt in passing from the ellipsoids toward the schistose matrix is in mineralogical character much as has been described for other basalts from this same district. The reason for the more complete degree of the replacement process in passing away from the ellipsoids may be readily understood from the discussion of the origin of the ellipsoidal parting of the basalts, where the conclusion was reached that the matrix between the ellipsoids resulted from the comminution of basaltic material of the same general character as that of the ellipsoids. This matrix was ot course more porous and probably more vitreous than the basalt, and hence more liable to be altered. PYROCLASTICS. The majoi-ity of the clastic rocks have been derived from the basic volcanic rocks already described. These elastics are veiy characteristic of the Hemlock formation and constitute the greater part of it. They comprise several classes, the more important of which are the eruptive breccias, volcanic sedimentary rocks, and schistose pyroclastics. ERUPTIVE BRECCIA. The term "eruptive breccia" is here used to include those clastic rocks in which angular fragments of an igneous rock are surrounded by a matrix 136 THE CRYSTAL FALLS lEON-BEARlNG DISTRICT. also of igneous origin. In an eruptive breccia the fragments may be like or unlike. Likewise the matrix may be hke or unlike the fragments. Where the fragments have been rounded during the movement of the erup- tive magma surrounding them, the resulting rock may be called an eruptive pseudo-conglomerate. Eruptive breccias are not very common in the Crystal Falls district. "Where they do occur, the fragments, while predominantly angular, are to some extent more or less rounded, and are similar in nature to the matrix in which they lie. Since the rocks which form them preserve the main characters of the massive lava flows which have just been described, they will not be discussed in detail. The exact method of the formation of these breccias could not be told. In one case, in which both fragments and matrix are amygdaloidal, it appears probable that the occurrence represents a true flow breccia in which the broken surface of a lava flow had been recemented by a later lava flow of the same kind of rock, or that it represents a very possible case in wdiich the lava welled up through and floAved over portions of its own crust, cementing the fragments. In such breccias a flow structure around the fragments is quite plainly shown and the matrix possesses a peculiar ropy appearance. In one instance, in which both the fragments and matrix were niacroscopically nonamygdaloidal, it is probable that they were formed under considerable pressure, and that this was a case in wdiich lava was forced up through a previously consolidated mass of rock of like character, and in its passage carried with it various fragments, forming an eruptive "reibimffs-hreccia" or friction breccia. VOLCANIC SEDIMENTARY ROCKS. Under the term "tuffs" have been very generally included all kinds of volcanic clastic rocks.^ This is probably due to the fact that there is fre- quently considerable difficulty in discriminating between eolian deposits and those which have been deposited in water. It seems desirable, wdierever it is possible, to make this discrimination. To that end I shall in the fol- lowing pages restrict the term "tuff" to eolian deposits. The term "volcanic conglomerate," or, for the sake of brevity, simply "conglomerate," will be used for those coarse deposits which have l^een sorted by and deposited ' Text-book of Geology, by Sir Archibald Cieikic : 3d ed., p. 135. PYROCLASTICS OF HEMLOCK FORMATION. 137 in wutor, and wlio.se t'niynientsj show a rounded character. Slntuld the fragiueut!:) be aiiguhir, the rocks may be called "volcanic breccias." It has been found practicable to maintain this distinction in earlier studies on Tertiary volcanics,^ and it is also maintained in the ^wesent study of pre-Cambrian volcanics. I am confident the same distinction could he made more generally tlian it is, and would in that case tend to a greater precision in the separation of rocks of diiferent characters. However, it is rather difficult to separate true eolian deposits of volcanic fragmentary materials from those in which tlie fragments have been deposited rapidly through water without liaAing eml^edded organic remains and. without having undergone sufficient attrition to he much rounded. More or less rounding, it is well iinderstood, results from the attrition of the volcanic ejectamenta during their ascent and descent through the air, so that they may in this respect resemble many of the sedimentaries. The exact mode of origin of many of the volcanic fragmental deposits of the Michigamme district is not clear. The greater portion appear to be of true eolian origin, and where the origin of any is in doulit it has been put with those of eolian origin. COARSE TUFFS. The coarse tuffs include rocks composed of fragments of all sizes, from the large volcanic blocks to the fine-grained particles of sand and dust which fill in the interstices. The ejectamenta may be uKire or less rounded by attrition during their progress through the air, so tliat if a refinement of the nomenclature should be needed one might very properlv be justified in speaking of tuft' breccias and tuff conglomerates. Tuffs are very common and characteristic for the district. The char- acters of the beds is best shown on the weathered surfaces. Here the scoriaceous and dense light-green fragments stand out well from the brownish-red matnx of more altered, finer fragments and cement. On a fresh surface the interstitial material usually has a darker green color than the fragments. The fragments have a prevailing green color, but many, especially in sections, are brown, much darker than anv of the rocks forming the lava flows. The larger fragments are usually sharply angular, but in many cases are more or less rounded because of attrition durino- • Die Gesteine des Duppauer Gebirges in Nord-Bohmen, by J. Morgan Clements : Jahrbucli K.-k. geol. Relchaanstalt, Vol. XL, 1890, p. 324. 138 THE CRYSTAL FALLS IRON-BE ARIXG DISTRICT. their j^rogress tlu'ough the air. (PI. XIII.) They are for the most part not .scoriaceous, though rather commonly amygdaloidal. The macroscopi- cally dense fragments seem to predominate, though the amygdaloidal ones do occur in some specimens in nearly equal quantity. The fragments of the tuffs are derived from the various kinds of basalt already described as forming the lava flows. Among the fragments some of the most typical of these rocks liave been found, and remarkable as it may seem, some of the thin sections from them show the least-altered basalts. In addition to the kinds mentioned under the basalts there are a number which differ slightly from them, and ajjparently represent more glassy modi- fications of the basalt magma. In one of these the amygdules are more sharply outlined by the accumulation of iron oxide around the edges of the amygdule than is the ease in the crystalline flow rocks. An especially well-preserved fragment shows perfectly fresh plagioclase microlites exhib- iting well-developed fluidal structure lying in a dark-brown apparently isotropic glassy base. Where the section is thin, globulitic devitrification products can be seen, and tliere also the base no longer appears isotropic, but very feebly double refracting. There is very frequently found among these tuffs amygdaloidal fragments which appear to have been derived from what was originally a completely glassy rock, no indication of the presence of any original crystals having been preserved. The background of these fragments consists of a fine felt of a green chloritic mineral, dotted with innumerable grains of epidote, in which one may distinctly discern concen- tric circles and arcs of circles outlined liy aggi'egates of epidote grains. These circles probably rejjresent perlitic partings. (Fig. A, PI. XXXII.) The dark-ljrown fragments mentioned as occurring with the prevailing green ones are very dense, appear to be very rich in iron, and may possibly represent a very basic devitrified glass. Should accumulations composed essentialh' of such glassy fragments be found, they could j^roperly be called "palagonite tufts." In addition to the rock fragments, a few rare ones of large plagioclase crystals were found, and also in one case a fragment of a violet-brown augite, the only specimen of fresh pyroxene thus far found in any of the volcanics. The tuffs show in places fairly well-developed banding, caused by the PLATE XIII. 139 PLATE XIII. (Sp. No. 23644. Natural size.) This illustration is a faitliful representation of the appearance of the polished surface of a pyroclastic from the Hemlock formation. It is somewhat doubtful whether or not the fragments composing the rock have been deposited through the mediation of water or air alone. The larger fragments are rather dense. Vesicular fragments are more common among the smaller particles. Pyroclastics similar in appearance to this are of very common occurrence in the Crystal Falls district, and huge clift's of it are readily accessible from the railroad. 140 U S GEOLOGICAL SURVEY MONOGRAPH XXXVI PL XIII ^^^^sr jd i \ "*■■ ^v^^Bb^^^b^H '^SP\ 3L *! ^^n 1^ p^^. ^'^^ >5a IT "** m ^ m4m ■H^ V V .^Cfl^^H^I^^H -4aHilr N f ' JUS B I E N a c BASAI.T TtrFF. I'YUOOLASTICS OF HEMLOCK FORMATION. 141 iiitoi-bcHl(liii'E SCHISTS. 149 soiitlR-iist, it jiTiulually iijirniws, until iu sec. 3G, 'V. 4(i X., li. 32 W., the eusteru limit of the Jirca studied hy lue, it is only about half a mile wide. E.xt-ept iu the vicinity of Bone Lake, where erosion has uncovered some of the knol)s, outcrojis ar(> ^^erN' scarce, since the drift is very lieaNA', and the drainage is poorl\' develo})ed. FIELD EVIDENCE OF CONNECTION WITH THE VOLCANICS. If one examines attentive^' the Hendock formation in its typical dcvidopnient, 1 )eginning, say, in sec. 27, T. 45 N., R. 33 Vs., and f(dlowiiig its northw;n-d exteusion through sees. 22, 1(!, and 15 of the same township, he will observe instances of banding in the tuffs and of schistositv in the amygdaloidal lavas and pyroclastics. The strikes and dips of the jjrimarv and secondary structures approximately coincide, both having a general north-south strike and dipping high to the west. Tlii-oughout this area, h()wever, the unmistakable massive volcanics are the predominant rocks. Continuing the examination farther nc^rth into sec. 34, T. 46 X., R. 33 W., rocks are found which possess almost invariablv a strongly marked schis- tositv, but with their volcanic origin clearly shown by the flattened amyg- dules. This is also true for the exposures east of this place on the under side of the Hemlock belt, in sec. 31, T. 46 N., R. 32 W. The strike of the schistosity of the amygdaloids varies from N. 30° to 70'^ E., and the dip is high to the northwest. Farther along this belt to the northeast, in sec. 24, T. 46 N., R. 33 W., schistose pyroclastics were observed striking N. 80° E. The original characters of these pyroclastics liave been almost entirely obliterated. The exposures next to the east in sec. 16, T. 46 N., R. 32 W., possess all the characters of crystalline schists. Somewhat farther east, however, associated with these schists are isolated outcrops in which traces oi flow structure and remnants of amygdides were observed, and, in some, traces of igneous textures were seen under the microscope. The schistosity of these rocks strikes for the most part south of east, varying from N. 65° to 80° W. and dipping to the nortJieast. Following the belt as it now turns to the southeast, the crystalline schist characters prevail, the volcanic char- acters being obliterated. The schistosity at the same time bends farther around to the southeast, pointing toward the continuation of this area of volcanics to the southeast, outside of the area studied. From field observations the conclusion seems iiecessarv that these 150 THE CRYSTAL FALLS IROX-BEARmU DISTRICT. .schists are metamorphosed volcanic rocks, and this conckision is strength- eued l)y detailed petrog'raphical examination. PETROGRAPHICAL CHARACTERS. ■ The crystalline schists are tine to medium grained schistose rocks which vary in C(ilor from a moderately light green for the more chloritic phases to a ^ery dark green and purplish-l^lack for those in which the hornl)lende, mica, and iron ores are prominent. The minerals of which the rocks are composed, arranged in order of importance, are hornblende, biotite, feldspar, chlorite, epidote muscovite, quartz, magnetite, hematite, ilmenite, and rutile. Under the microscojje the schistose structure is seen to be produced by the general parallelism of the liisilicate constituents. The por])hyritic texture is seen in a few specimens, and hornblende forms the phenocrysts. Hori:l)lende occurs in iine needles and also in coarse crystals which are automorpliic in the prismatic zone, but on which no terminations have been observed. It also occurs rather comm(inly in sheaf-like bundles of ragged crystals. The marked orthopiuacoidal development so common for actinolite is quite noticeable. The crystals show the usual strong pleoch- roism: c = bluish-green, lj = olive green, it r= yellow, whereby i:>>lj>a. The hornblende crystals frequently contain large fjuantities of the minerals of the groundmass, many of them iu such quantity that there are really only skeleton hornblende crystals present. The general character of the hornblende in all these rocks is that of a secondary porphyntic constituent, and seems to be analog'ous to such minerals as garnet, staurolite, etc., which are produced in clearly metamorphic rocks. Brown biotite is rather common in some of the rocks. Though usually subordinate to the hornblende, it is at times the predominate bisilicate. It is light brown and shows the usual characters of biotite. It is present in small irregular flakes, and also in larger individuals which show poor pinacoidal development. In one case such a mica individual in perfectly fresh condition may be seen with its ragged edges interlocking with the fringed periphery of an altering feldspar crystal. The liiotite appears to have derived some of its necessary elements from the feldspar and to be eating int(i it, and consequently to be a secondary product. Feldspar is not found as an original mineral in any of the crystalline BONE LAKE CRYSTALLINE SCHISTS. 151 schists. It occurs as a socond.ary constituent. It is found, however, as a primary constituent in a few rocks which, as they still possess remnants of orifi'inal igneous textures, strictly speaking, should not perhaps be included with the crystalline schists. They represent more properly the transition stao-es to the cr^'staHine schists, but the process of the alteration of the feldspar is so well shown in these that it is considered expedient to mention it at this place. The original feldspar occurs in this transition phase in the large tangled intergrowths connnonly seen in andesitic and basaltic rocks, as individual phenocrysts, and as microlitic lath-shaped individuals in the groundmass. The greatest interest centers in the phenocrysts, as in them the changes which take place are more clearly seen. The feldspar phenocrysts ai-e always cloudy, due to numberless black ferruginous inclu- sions. They also inclose the various secondary dark silicates composing the gTOundmass, grains of epidote, ilakes of biotite, and crystals of horn- blende. These are usually surrounded by very narrow clear zones, appar- ently feldspar. Near the edges of such altered crystals, and especially in the more altered individuals, these inclusions ai-e more numerous, and are accompanied by grains of quartz and new feldspar (albite !). These last two have certainly been derived from the alteration of the feldspar, but that mineral may possibl}^ also have contributed something to the produc- tion of the dark silicates. The secondary feldspar, that of the schists proper, is found in grains usually imstriated, though in a few cases striations were observed. This feldspar was not determined, but is probably albite The chlorite is in flakes scattered through the schists, showing the usual characters. Epidote, muscovite, quartz, and rutile appear as usual. Ilmenite is present in one case as micaceous titanic iron oxide, and is then in extremely thin plates which show a beautiful hexagonal develop- ment, though more frequently the plates are rounded. They are transparent with the characteristic clove-brown color. The thicker plates are thin enough to be transparent only along the edges. The iron oxides, magnetite, and hematite occur in some of these rocks in large quantity. In certain pai'ts of the area underlain by these schists considerable excavations have been made in search of iron, the presence of which was indicated by the magnetic needle, and moderately large bodies of ore have been found, though in no case in sufficient quantity to 152 THE CRYSTAL FALLS IRON-BEARING DISTRICT. admit of successful mining. Such ore bodies probably owe their presence in great part to processes active subsequent to the formation of the schists. (See p. 134.) According to the quantity and association of the minerals described above as occurring- in tlie schists, the following- rocks mav result from the complete metamorphism of tlie basic volcanics : Amphibolites, chlorite- schists, epidote-schists, mica-schists, mica-gneisses, and possibly siliceous hematite and magnetite ore. The complete metamorphism of dense basic lava flows into crystalline schists has been described by Williams^ for the Menominee and Marquette districts, and also by Van Hise and Bay ley ^ for the Marquette district. Williams^ has also described the production of schists from the igneous elastics in the Menominee district and similar products have been described from the Marquette di.strict both by Williams* and by Bayley.^ The above-described schists cover a considerable area, with only iso- lated exposures of rocks associated with them in which volcanic characters are recognizable. They are confidently believed to represent extremely metamorphosed volcanics of the same general original character as those constituting the Hemlock formation and belonging to the same relative period of extrusion. The same conclusions have been reached by Smyth for similar schists along the Fence River to the southeast of those described. It is noticeable that the most intense metamorphism of the volcanics has taken place in the northern and northeastern ^^fii't of the Crystal Falls district, that part in which the crystalline schists have been produced, though the explanation for this can not be offered. NORMAL SEDIMENTARIES OF THE HEMLOCK FORMATION. The normal sedimentaries are in small quantity. It has been seen (p]i. 64, 78) that the Mansfield slate is overlain by a conglomerate in which volcanic material predominates, but which contains partly rounded frag- ments of chert and slate and round quartz grains derived from the under- lying sedimentaries. But for the interming'liug of this normal clastic debris I Bull. U. S. Geol. Survey No. 62, cit. ^ Op. cit., p. 158. •Mod. U. S. Geol. Survey, Vol. XX VIII, cit., pp. 152-159. °0p. cit., pp. 160-169. = 0p. cit., p. 133. NOKMAL SEDIMENTAKIES OF HEMLOCK FORMATION. 153 with the l)vnl(•hl^sti^•s, the coni^loinerate shows iiothiu;^- ditfercut from the volcaiiie con<>louierate already described. It is a transition rock between the tuffs and tlie normal sedimentaries. Similarly, in sec. 34, T. 4.') N., R. 33 W., a gradation occurs in the upper horizon of the Hemlock formation from the ^■ok•anic conglomerates to the true normal sediments. The sediments are slates about 17.5 feet thick, containing- lenticular raas.ses of limestone. These beds dip 80° to the west, generally strike north, but vary in places a few degrees to the west. They are underlain by conglomerates containing well-rounded volcanic pebbles. This volcanic conglomerate grades from the coarse conglomerate up into Avhat might be termed a water-deposited volcanic sand. Tlie peb- bles are all of volcanic material. Between the conglomerates and slates is a small area without outcrop. Overlying the slates is a succession of tuffs and lava flows. The slates in color range from light gray and green to purplish red, and the lenses of limestone vary from cream color to purplish red. In thin section the slates are seen to be composed of a felt of sericite, chlorite, and quartz, with associated innumeralile minute rutile crystals, and here and there a large spot of limpid quartz. A ferruginous carbonate is present in all of them in porphyritic rhombs. Where chlorite is abundant, the slates are a light green. Where iron oxide is abundant and the chlorite less plenti- ful, the slates are purplish. The lenses of limestone are rather pure, consisting mainly of calcite, with some few scattered areas of cherty silica. On the edges of the lenses some of the slate material is found forming bands in the carbonate. These intermediate phases grade on the one hand into the pure carbonate, and on the other hand into the slate beds. From the crust of limonite, which may be seen on the weathered surface of the rock, the calcite is evidently rather ferruginous. The process of alteration is clearly seen under the micro- scope, where many of the grains are surrounded by rims of hydrated oxide of iron and hematite. ECONOMIC PRODUCTS. BUILDING AND ORNAMENTAL STONES. The rocks of the Hemlock formation are not likely to be much used for building purposes. The compact basalts possess in a high degree the 154 THE CRYSTAL FALLS lEONBEARING DISTRICT. two essential features of strength and durability. For trimming in contrast with lighter stones they might be found desirable, and it may be suggested that they are especially suitable for mosaics in which rich greens are desired. They are of too somber a color to l)e used in large quantity for anything else than foundations. Moreover, the difficulty and consequent expense of quarrying them, and their remoteness from cities of large size, will operate strongly against their use. The pyroclastics are natural mosaics, and some of them have a very pleasing appearance (PI. XIII) and are suitable for table tops, wains- coting, etc. ROAD MATERIALS. The importance of good roads in aiding in the material development of a region can hardly be overestimated, and in the building of good roads, especially in thinly inhabited regions, the proximity of good road material is of prime importance. Thus far the 15 miles of good road between Crystal Falls and tlie adjacent mining villages have been covered with the ferruginous chert and slates from the dumps of the mines, and unroll themselves to the traveler like red ribbons laid through the green woods. No rock is better suited for use in building macadamized roads than the liasalt, and of this the Hendock formation offers an inexliaustible supply. The fine-grained compact basalts are by far the best rocks obtainable, and, other things being equal, should of course be chosen rather than the scoriaceous and consequently weaker facies, but these weaker kinds and also the pyroclastics are preferable to the cherts and slates which have been used. The cherts are very hard and durable, but the dust and sand froiu them possess but slight capacity for cementation. Consequently the roadways upon which quartzite and chert have been used are more likely to wash out than are the roads macadamized with basalt, since the dust in this latter case serves as a cement which binds the larger fragments more firmly together. The road commissioners have thus far used very little basalt, chiefly for the reasons that no crusher was at their disposal, and the chert and slates were at hand ready for use. CHAPTER V. THE UPPER HURONIAN SERIES. The upper series of tliis district is connected in tlie northeastern part of the area witli the Upper Marquette series of the Marquette district ah-eady described in the Fifteenth Annual Report and Monograph XXVIII. In these reports the Upper Marquette series is regarded as part of the Upper Huronian. As has been stated, the Crystal Falls district is the southwestern extension of the Marquette district, and consequently we should expect the chief formations of the two districts to be continuous, as they are. Because of the drift and becaiise of a change in the character of the rocks, in mapping the western part of the Crystal Falls district it has not been practicable to divide the Upper Huronian into several formations, corresponding to those in the ^Marquette district. No independent name will be given to it, but it will simply be called Upper Huronian, with the understanding that it corresponds stratigraphically to the Upper Marquette series. DISTRIBUTION, EXPOSURES, AND TOPOGRAPHY, Beginning in the northeastern part of the area discussed by me (see PI. Ill), this series covers the southern parts of T. 46 N., Rs. 31 and 32, where it is only 4 miles in width. It is here a northwest-southeast syncline. From this place it stretches beyond the northern limit of the map. With slight interruptions where intrusives occur, it extends in a broad area to the west and south about the Hemlock volcanics to a point lying beyond the limit of the map. On the eastern side of the district it abuts against and is folded in synelines in the Archean granite. Exposures are scanty for the greater part of the area in the Crystal Falls district underlain by the Upper Huronian series. This is due to two conditions, first, to the soft character of the rocks constituting the series, and, second, to the presence in places of the Cambnan sandstone, and more especially to the deep covering of glacial drift which is found spread over the entire district. The Upper Huronian is composed in great measure of 155 156 THE CRYSTAL FALLS lUON-BEARING DISTRICT. slates, which are interbedded with much smaller quantities of graywackes and chert. The slates are eroded much more readily than the associated harder beds, and therefore, except along valleys, we rarely find the soft slates exposed. The graywackes and cherty rocks are the ones ^^'llich form the striking topographical features of the landscape, the slates forming softly-rounded hills. The drift is also an important factor in the scarcity of outcrops. In the northern and western parts of the districts especially the drift is very heavy. In this portion the youthfulness of the topography is emphasized by numerous swamps, lakes, and generally imperfect di-ain- age. In the southern anstioniil)ly tliu wcsturu (•iiiitiiiuatioii of the .Michiginiinic- fonnatinn, In wliicli tlu' rocks correspond |»etro- LOTVER IjURONlAN" MAN9FI&l.n SLATE HKML OCK FORM ATION I Aim 1 rAJK:^^! lliPP55_5S^oNiAN" INTRUSIVE DIOHITE 18 I T^ US-GEOLOOrCAL SURVEY MONOGPAPH XXXVI PL XVIIl DKTAII. GI:01.0GR:AK MAP Ol-' THK MClXm OK CRVSTAL FM.LS AND NLVN'SMKLO. SlIEP:'!' IT CONTOUW lNTKH\-.'U..ZO I'L^KT cltijja "11" uotDiTiunBi] BinKi- unaiUp 'V OUUit-pn vvilli a ■ Tcsi T""* ''olioiiicil 111 rork AlXJONKTAN LOWKR i rt-nONLW UPPER HLTtONlAW HEMl ^CK VOn UXnOtt L'KPn.TDCU nOLERITE DIORITK OKE DEPOSITS OF UPPEK UURONIA>;. 179 CV'luinlnu, Duuu, Miistodou, juid others to the west (PL XVIII), sire prol)- ablv the westeru conthiuatiou of tliis hue of luiiies, and follow the trend of the main synclinal axis of the district. The position of these mines with reference to the main structural features of the district can be seen on the relief map and the sketch map corresponding to it, PI. XIV. The section made through the closely folded Upper Huronian lieds by w Fill. 12.— Sketcli illustraliiig coiitortioD of Ujiper Huroiiiaii strata. the Paint River affords the best opportunity in the district for studying the rocks, but the rocks are so crumpled that even here the succession was not made out with certainty. The sketch fig. 12, by W. N. Merriam,^ shows the folding of the slate and chert strata as seen in the railroad cut between the Paint River and the Lin- coln mine. The strike of the rocks is about N. 80° E. The sketch is taken looking almost along the strike of the beds. In fig. 13 a sec- ,..-•''" \. ond sketch is given, also hj W. N. Merriam, which illus- trates the rapid change in strike in these beds, due to the con- tortion of the strata. This change is seen near the east end of the wagon bridge just across the Paint River from OrVStal Falls At this nOint ^'^' ^^' — S^^tch showing change of strike of Upper Huronian beds •^ ' i ilue to the folds. the beds bend from a strike ot S. 40° E. to W. 10° S. The change takes place by means of three very sharp bends. The following are the observations made by Rominger upon these expo- sures near Crystal Falls: Among the recently discovered productive fields for irou mining, the vicinity of Crystal Falls has become famous for its wealth in ore. The formation enclosing the ore ' Manuscript notes. 180 THE CRYSTAL FALLS IRON-BEAKING DISTRICT. deposits, has there a great thickness, but its determination by actual measurement is , impossible, on account of the much folded condition of the strata, and for want of con- nected exposures transverse to the stratification. Estimating its thickness to several thousand feet is surely not far beyond the truth. This folded condition of the strata is in many instances an obstacle in the decision, whether in a given locality we have under observation a descending or an ascending succession of beds. If we follow the railroad from Crystal Falls village upward along the bed of Paint River, we find, in the first cut the road makes into rock beds, a series of hard, black slates, transversely intersected in almost vertical positions, and, according to their cleavage planes, dipping in southwest direction. This cross cut is 210 steps long; thence, for the distance of 100 steps, no rock ledges are touched by the roadbed, but on the left side of the road similar slate rocks are denuded, which apparently represent a continuation of the former succession. From here for eighty steps a cut is made through similar slate rocks, but interlaminated with numerous quartzite seams; further on, the intersection of slates in alternation with quartz seams continues for quite a while, but these slate rocks are more graphitic than the former and readily disintegrate, on exposure, into splintery fragments, as they contain a large proportion of iron pyrites and rusty ferruginous seams causing the decay. By this time we have reached close to the river below its falls, and find, laid open in its embankments formed by the bluffs thirty feet high, a further conformable series of graphite-schists, 300 feet wide. Beneath the graphite-schists, close to the water level at the foot of the falls, succeeds an ore belt six feet wide at the surface, but widening to fifteen feet, followed into the hillside.' Below the ore belt follows an immensely large succession of thinly laminated banded ferruginous quartz-schists of dark, rusty color, which beds, in steeply erected position crossing the river bed diagonally, give a cause to falls eight or ten feet in height. The exposed succession of beds amounts at the falls to a thickness of over 800 feet. Intermixture of pyritous shaly seams with the quartzite beds, induces their rapid disintegration on exposure, into shelly fragments covered with an iridescent varnish like coating of oxide-hydrate. These beds are, in the embankments on the opposite river side, remarkably corrugated, describing in their flections perfect coils.^ CHARACTER OF THE ORE. The ore obtained from the Crystal Falls district is chiefly soft red hematite, though iu places it is hydrated aud graded as brown hematite (limonite). The ore is very porous and shows many crystal-lined cavities. At places a hard steel hematite ore is found, which runs as high as 70 per cent metallic iron. This ore occurs in very small quantities associated with the soft ores, and appears for the most pai-t to have formed iu geodal cavities. When the cavities are still partly open, the Ore has botryoidal and stalactitic forms. The ores are very similar to the ores of the Michigamme 'Iron and copper regions of the Upper and Lower Peninsulas of Michigan, by C. Rominger: Geol. of Mich., Vol. V, 1895, Pt. 1, p. 74. -Ibid., p. 75. ORE DEPOSITS OF UPPEK IIURONIAN. 181 slatos of the Upper ^I;ir(|iu'tte series, Ijut (litter very coiisidei'ablv from tliose of the Lower ]\Iai-quette series, in whieh the hard hematites and magnetites arc important ores, and from the ores of the Menominee district, wliieh pro- duces Uirg-e quantitie(5 of soft Ijhie hematite, some inartite, and also some specular ore. The following figures show the average composition of the ores for the district. They were taken from analyses furnished b}- the management of the various mines and from the reports of the State commissioner of mineral statistics of Michigan. The metallic iron of the ores ranges from 54 to 63 per cent, the aver- age being about 59 per cent. Phosphorus in exceptional cases is as low as 0.05 per cent, though usually ranging from 0.1 to 0.7 per cent, most com- monly approaching the higher figure. Silica averages about 3 per cent. These analyses show the ore to be ratlier low grade.' It is due to this that this district has been so sensitive to the prices of iron ores. A low market price makes the cost of production exceed the selling value, and under these conditions work necessarily stojis. Some of the ores in the Crystal Falls district contain a ver}' high percentage of AI2O3, CaO, and also of manganese. It is reported that some very good deposits of manganese have been found, one unauthenti- cated statement being- to the effect that an analysis of the ore runs as follows: Metallic iron, 17.46; manganese, 29.81; phosphorus, 0.064; silica, 0.009(?). ' Brooks states that " the ores are unlike those in the more easterly part of the Menominee region in being richer in iron, freer from silica, and in containing more water." (Analysis 68, p. 302.) Geol- ogy of Michigan, Vol. I, part 1, p. 182. Since the above was written the volume on Mineral Resources of the United States, 1896, Part V, of the Eighteenth Annual Report of the United States Geological Survey, has appeared, and the following analyses of ores from the Crystal Falls district are taken from Mr. John Birkiubiue's article on iron ores in that report. The analyses were prepared for the ore association at Cleveland. Ohio, and show the average cargo analyses of iron ore as shippeil from the various mines. The analyses were made from ores dried at 212^, the amount of natural moisture being added. Iron. Silica. Phos- phorus. Man- ganese. Alu- mina. Lime. Mag- nesia. Orgjinic Sul- and l)hur. ivolatile ' matter. Mois- ture. Crystal Falla ..,. Dunn ^. 58.55 58.61 60.20 60.86 61.00 4.25 3.88 4.71 3.86 4.50 .721 .573 .309 .240 .350 .20 .58 .39 .45 .30 1.16 1.88 2.81 1.67 2.75 2.64 1.80 1.87 2.14 .50 .77 .83 1.32 1.73 .30 . 0U8 t 2. 92 .033 1 5.44 .010 .010 .075 7.20 8.70 6.62 6.50 9.00 Lincolu Mastodon 182 THE CRYSTAL FALLS IROK-BEARING DISTRICT. One of the best results gives as high as 61.5 per cent metallic Mn. The ores thus range from a manganiferous iron ore to a manganese ore. The further statement is made that the bed lies very close to the surface, and is from 6 inches to 3 feet in thickness. From 4 to 6 feet of bog iron is found underlying tlie bed of manganese. RELATIONS TO ADJACENT ROCKS. The ore is associated with white or reddish chert, which in places is jaspery. The cherty iron formation passes into ore by a decrease of the silica. An intermediate phase is chert with "bands and shots" of ore. In places the chert is more or less brecciated, and the ore often has a similar character. Commonly the ore is completely surrounded by the chert beds, or chert and ore, forming the so-called mixed and lean ore. In such cases theA" form both the foot and hanging walls of the ore bodA". But the ore-bearing chert formation is always associated with black carbonaceous slates, which consti- tute the base on which the ore-bearing formation rests. In the Youngstown mine 3 feet of so-called "graphite" was jjassed through before the usual carbonaceous slates were reached.^ The hanging wall is also carbonaceous slate. At places thin quartzitic beds which approach a true quartzite are associated with the slate. The ores occur in the cherts in pockets and lenticular masses, which always agree in greater dimensions with the strike of the beds with which thev are associated. The lenticular character is well shown in the Dunn, Columbia, and Great Western mines. In the Dunn mine the bodies over- lap. In the Great Western mine in 1887 seven different ore bodies in an east-west line, separated by areas of barren rock, mostly slate, were being mined. In following these isolated ore bodies to the east, at various places thev are found to turn around a horse of rock. Their occurrence is illus- trated by the horizontal section, tig. 14. Evidently the ore bodies accumu- lated in westward-pitching synclinal troughs, in which the hanging wall appears to the miners as a horse of rock. The ore bodies in general pitch to the west at varying angles corre- FiG. 14.— Sketch to illus- trate the occurrence of ore bodies. 'This information was furnished by Mr. C. T. Roberts, of Crystal Falls, obtain a specimen of the graphite for examination. It was not possible to ORE DEPOSITS OF UPPER HURONIAN. 183 .spoiiding- ro the pitches of the axes of tlie syiielhies in which thev occur. Tlie pitdu's of these folds in turn correspond to the westward pitch of tlie Crvstal Falls syncliuoriuin, of which the secondary synclines containing rlie ore l>odies are a part. A typical example of the occurrence is shown in the Armenia mine ore body, which is found, according- to Van Hise, "at the bottom and on the sides of a synclinal trough, pitching at an angle of about ^,-o"i rpi^^^ trend of the axis is to the south and west. The dip of the ore bodies is always steep, and generally to the south, but varies in places to a few degrees north. ORIGIN. The fact tliat tlie important mines in tlie district are located in a synclinal basin and that they all possess an impervious footwall of black slate gives very clearly the reason for their existence and indicates their mode of origin. They are concentrates in synclinal troughs. In the ]\Iarquette and Penokee-Gogebic districts the ore Ijodies are frequently found associated with dikes of dolerite (diabase), which have been altered to "diorite "-schists, and so-called soapstone or paint rock.^ Only one such association is known for the Crystal Falls district. Wads- worth mentions having seen a dike in the Paint River mine.^ In the field notes of the Lake Superior survey for 1y C. R. V.an Hise; Am. .Tonr. .Sci., 3d series, Vol. XLIII, 18fl2, pp. 130. -Mernaui also mentions in his notes a dolerite dike found cutting the ferruginous rocks at the Glidden exploration. In this case it does not appear that an ore body was foruu-d. " Sketch of the geology of the iron, gold, and copper districts of Michigan, by M. E. Wads- worth : Kept. State Board of Geol. Survey for 1891-92, 1893, p. 108. ■•Now known as Lamont mine. 184 THE CRYSTAL FALLS IRON-BEARING DISTRICT. this dike, as It occurs in a pitching trough, formed by its junction with the impervious slate. These same relations are well known to be the cause of similar occurrences in the Lake Superior districts above mentioned. The original rock from which the ores were formed was cherty iron carbonate, which in many places is found associated with the iron-bearing formation. The cherty carljonate shows the various stages of alteration from the compact cherty siderite to the banded ore and chert rocks which form the nuclei for the addition of the iron obtained from the higher exten- sions of the beds. Percolating waters have been the agents in this process of replacement and concentration. Consequently where the rocks have been most shattered, we find the water was especially active. Hence it is, also, that we find the deposits in this closely folded part of the Upper Huronian. As to the origin of the cherty carbonate itself, we know nothing definite. Its association with the carbonaceous slates would indicate the ao-ency of organic matter in its prodiiction, possibly in some such manner as is rather generally accepted for the formation of the Carboniferous carbonate ores. The Upper Huronian ores, as well as the Lower Huronian, are supposed to have been formed in this same way, and from the same kind of rock. Under the discussion of the Lower Huronian ores (p. 70) these points were discussed more in detail, and references given to the literature, and the reader is referred to that discussion for further details. SIZE OF THE ORE BODIES. No definite general statement can be made as to the size of the ore bodies, as this varies considerably. None of the bodies which are being worked, so far as I can learn, are less than 30 feet wide. In one of the old mines crosscuts disclosed a width of nearly 200 feet. This same ore body is reported to be at least one-foin-th of a mile long. METHODS OF MINING. The first develoinnent of the iron ores of this district was by the stripping and open-cut method, very few resorting at once to under-ground work. Nearly two-thirds of the product of certain of the mines has been from open-pit work. When the open pits become too deep to be readily worked as such, shafts are sunk and the exploiting of the ore body is carried on under ground, at times both open-pit and under-ground work being carried on simultaneously. The Mastodon presented the unusual ORE DEPOSITS OF UPPER HURONIAN. 185 siflit of ;m o\)(.'U pit c'Xten(liii<4- down 200 feet, part of tlic wdrkinf^'s still boiii"' roofed over by an enormous arch of ro(dc. All work in tlie district is at present under gTound.' As a rule, the nndcr-yroun'E.4SE.Jsec.l9.T.43N.,K.32^V NW.|NW.^SfC.31.T.43N..E.32W Lot3.scc.20.T.43K., B.32W Claire Mining Co 55, 000 70,770 57,351 57.682 9,012 22, 426 Annenia'lit reddisli-browii to dark- ^av and greenisli rocks, ami iu j^-rain from tine to coarse. The structure ordinai'ih' is tliat of a normal granite. In some of them the micropey- niatitic intergrowth of quartz and feldspar may be observed in small quan- tity. In others this forms the characteristic part of the rock, and these rocks may be properly termed micropegmatitic granites. In most of the sections the usual constituents in ordinary proportions occur. In all the rocks the main mass of the quartz forms irregular grains, molded on the other constituents. Sometimes round areas of quartz are included in the feldspars. The quartz contains very commonly, and usually in great quantities, liquid inclusions with dancing as well as stationary bubbles. The feldspar is nearly always of two kinds — orthoclase and plagio- clase. Microcline was also observed, but in neglectable quantity. These feldspars in the great majority of slides show fairly good rectangular out- lines, and in some cases these are strikingly well developed. In some slides the plagioclase is observed in rectangular crystals and the oithoclase is found in large irregular plates which are jnolded on the plagioclase, show- ing conclusively their relative age. The plagioclase is finely twinned according to the albite law, and also in some slides exhibits pericline twin- ning. A case was observed in which two crystals, one showing albite twinning, the other both albite and pericline twinning, were grown together so as to correspond to the Carlsbad twins of orthoclase. The jjlagioclase gives low extinction angles, which show it to be rather acid. The amount of plagioclase in some of the sections — for example, in those of the numer- ous small dikes cutting the schists near Norway portage in sec. 15, T. 42 N., R. 31 W., and the one cutting the gabbro at the SE. corner of sec. 22, T. 42 N., R. 31 W. — is very large, denoting an increase in soda and lime and indicating a relationship to the diorites. The geological relations are not such, however, as to enable this connection to be shown in default of chemical analyses. The orthoclase is for the most part untwinned, or else shows simple Carlsbad twinning. In some sections the feldspars are quite fresh, but in others they are seen to be opaque, porcelain-hke, and iu still others the 192 THE CKYSTAL FALLS IRON-BEARING DISTRICT. original feldspar material is almost entirely replaced by a mass of musco- vite, with some little epidote-zoisite and biotite. The muscovite in these secondary aggregates gives excellent though small rectangular sections, showing its fine cleavage very distinctly. Well-determinable kaolin flakes were not found. The mica is well represented Ijy both Ijiotite and muscovite. Both occur in very well developed crystals, the muscovite showing the most per- fect development. The biotite has j^artly altered to chlorite, with a simul- taneous production of rutile, sagenite, and sphene. Between the chlorite laminae one frequently sees lenticular areas of secondary calcite. Quite commonly the sagenite is found included in these areas. In onlv one specimen was hornblende oljserved. This was from a granite dike whicli cut the dolerite. The hornblende is of a noncompact variety, which upon the edges is finely fibrous. It corresponds exactly to that which is found in the adjacent dolerite. The contact between the granite and dolerite appears irregular, as though the dolerite had been to some extent broken. As the contact is approached from the granite side the hornblende increases in quantity. It is thought probable that the hornblende in the granite along the contact is secondary after pyroxene, and that this pyi'oxene was obtained by the inclusion of fragments of the dolerite. Accessory minerals are not present in large quantity. Iron oxide is not present in great quantity, and when seen it is usually titaniferous, as rutile is found as an alteration product. In one case the hexagonal plates show the presence of ilmenite. Apatite is rare, as a rule, though occurring in some sections in considerable quantity. Zircon is scarce, as are also sphene and rutile. Epidote is rather common. In some cases it is seen in biotite surrounded by a pleochroic halo, and in such cases it is probably original. The secondary minerals, muscovite, biotite, chlorite, epidote, sphene, rutile, and calcite, show their usual characters. Calcite is abundant, and is more or less ferruginous. It is found in rhombohedra and also in irregular masses. In all cases its secondary origin is clear. MICROPEGMATITES. The micropegmatitic varieties of the biotite-granite show the same variations in color, from reddish to gray and greenish, and in grain from ACID INTKUaiVES. 193 tine to mc'iliuiu, as do tlu' biotite-iirauitcs j)r()[)er. In o'euenil tlu;y niuy be described as biotite-granites in wliicli the niicropegniatitie intergrowtii of quartz; ami feldspar instead of being sul)ordinate preponderates. Many of the well-crystallized feldspars are surrounded by a border of micropegraa- tite, which varies from a narrow strip to a very wide border, usually in inverse ratio to the size of the feldspar nucleus. The feldspar in the inter- growth is continuous with that of the nucleus. A coarsely radial arrange- ment of the niicropegniatitie intergrowtii was frequently observed. Where the feldspars and quartz are predominantly poi-jihyritic, and micropegmatite forms the groundmass, the rock grades over into the rhyolite-porphyries with micropegmatitic groundmass — the inappropriately named granophyres of Eoseiibusch. The biotite of the micropegmatitic granites has partly altered to chlorite and sageiiite. In some of these rocks the biotite is collected into large aggregates of imperfect individuals, which surround larg-e pieces of iron ore. In some instances it is included in the plagioclase. The biotite flakes in the feldspar are sometimes so numerous as to conceal almost com- ])letely the feldspar substance. In one instance the feldspar of such a micropegmatitic intergrowtii is completely replaced by biotite. These sec- ondary biotite flakes surrounding the remaining more or less rounded quartz areas of the micropegmatite produce a rock which is strikingly like a mica- schist in places, although it is of -unquestionably eruptive character. Tourmaline is a rare accessory in these granites, a small smoke-brown crystal having' been observed in one section. Titaniferous iron ore alterinff to leucoxene or sphene and rutile is found, as are also the common accessory minerals — apatite, rutile, and zircon. They contain also the same secondary minerals as the normal biotite-g'ranite. MUSCOVITE-BIOTITE-GRANITE. These are medium-grained rocks, and, owing to the fact that the mus- covite is more abundant than the biotite, have a light-gray color. The muscovite is noticeably automorphic with respect to the biotite, though the biotite is also in well-developed automorphic plates. Plagio- clase is present in these granites in very large quantity. It shows an excel- lent zonal development, with diminishing angle of extinction — that is, increasing acidity — from the center outward. The center of the individuals MON XXXVI 13 194 THE CRYSTAL FALLS IRON-BEARING DISTRICT. is nearly always extensively altered, while the outer zones are compara- tively fresh. A maximum extinction angle of 15° against the twinning planes in the zone perpendicular to 010 was observed on an unaltered zone surrounding an altered core. This would indicate the feldspar to be per- haps as basic as labradorite at the center. The other essential minerals, quartz and orthoclase, occur in usual quantity and show nothing of espe- cial interest. Sphene and apatite are the only accessory minerals present. An apatite crystal was observed which was included in quartz, and con- tained the brown apparently vitreous core so frequently seen in the apatites of basic rocks. Where included in biotite, it is surrounded by a pleo- chroic halo. No analyses were obtained of these granites, but from the quantity and character oi the feldspar as noted above, these rocks are thought to be closely related to dioritic rocks. Indeed, it is a (question if they should not be classed as quartz-diorites. RELATIONS OF GRANITES TO OTHER INTRUSIVES. In two cases alreadv mentioned granite dikes cut the diorites. Granite also cuts the gabbro, and dikes of granite were observed penetrating the dolerites, thus indicating that tlie granites are younger than tliese igneous rocks. DYNAMIC ACTION IN GRANITES. An examination of the granites with particular reference to pressure phenomena shows that the}' exhibit a great difference in this respect. Some show scarcely any traces of pressure, while others quite closely associated may have been affected thereby to such an extent that a more or less strongly wavy extinction of their mineral constituents is general. However, but a single instance of a supposed granite possessing an excellent cataclastic structure and imperfect schistosity was observed, and this rock was so extremely altered as to render dcnibtful a determination of its original character. CONTACTS OF GRANITES AND SEDIMENTARIES. The largest intrusive granite mass is found l>etween the Paint and Michigamme rivers, in sees. 19, 20, 29, and 30, in T. 42 N., R. 31 W. The granite is a inusco\'ite-biotite-granite. The sedimentaries are mica- ceous graywackes, which have been described on p. 170. Let it suffice ACID INTliUSIVES. 195 here to repeat that thev li;ivc Ix-cii luiicli mashed and recrystalhzcd, Imt that some of them still show their fraymeiital nriuiii. New hiotite, musco- vite, feldspar, iiiid cjuartz have developed. Where most altered, they are mica-schists and mica-gueisses. These chang-es are presumed to Ije due, for the most part, to the orogeuic forces which were active prior to the iutrusiou of the fi'ranite (p. 170). If su(di be the case, the g-ranite beg-an its metamor- phic action upon a rock already greatly changed from its original couditiou. EVIDENCE OF INTRUSION. The intrusive character of this large mass of granite is indicated by its stratigraphical position, and is further confirmed by the mechanical effects ])roduced by the intrusion of the granite upon the sedimentaries, by the contact effects produced in the sediments, and by the presence in the sedimentaries of granite dikes forming offshoots from the main mass. The mechanical effects are well shown by the inclusion of sedimentary fragments and by the dislocation and folding of the beds. Inclusions are rather common and are usually of considerable size. The dislocation and folding are lieautifully shown at N. 1970, W. 570 paces, sec. 30, T. 42 N., R. 31 W. In general the layers in the graywacke, which are alternately rich and poor in mica, strike N. 15° to 30° W., but where the intrusives are, these layers are found to strike almost due north and south. At the aliove location the beds are folded into small, closely- compressed anticlines and synclhies, which pluiige to the east at an angle of about 80°. At this place the micaceous graywacke is broken into small pieces, which are thoroixghly injected and cemented by the granite, thus forming a typical eruptive breccia. The granite cement is microgranitic, with comparatively little quartz and a small amount of chloritized mica. The fragments of micaceous graywacke in the breccia appear to be rather more feldspathic than usual, but otherwise seem not to have been nuich affected. Owing to the altered condition of the sediments i)rior to the granite intrusion, and to the alternation of sediments of somewhat varying char- acter, we can not expect to find such clearly outlined concentric zones suri'oundinp- the aranite as in cases where the sediments are uniform. In one case a contact was observed between the granites and apparently the main mass of the sediments. Along this line of contact biotite and white 196 THE CRYSTAL FALLS IKON-BEARING DISTRICT. mica have developed in great abundance. Mica is well known as one of the minerals produced in granite contacts, and it evidently here owes its abundance to the presence of the granite. At a considerable distance from the nearest intrusive outcrop (2 miles) a mica-schist was observed which was characterized by numerous small but prominent nodules that stood out upon its weathered surface. The rock contains a considerable quantity of an apparently original chlorite in large automorphic plates. The nodules were produced by large individ- uals of staurolite. The staurolite has almost completely altered, remnants only of the original indi^^duals remaining. These remaining grains show a very poor cleavage, and extinguish parallel to it. These include blebs of quartz and particles of iron oxide. Thev have the usual pleochroism for staurolite, varying from golden yellow for c to yellowish white for a and 6. The alteration products in which the grains lie are fine scaly aggregates of minute leaves of muscovite, with here and there larger plates of the same mineral. A few grains of quartz and a small amount of iron oxide, possilily jiartly original, are found in the mass. This observation of the alteration of the staurolite to muscovite confirms the observations of Thiir- ach^ and Pichler.' A similar staurolitiferous mica-scliist, occurring in the same locality, was described by C. E. Wright for the Wisconsin survey.^ On these particular specimens the characteristic twins of staurolite may be observed macroscopically as well as in thin section. Wright has also described a garnetiferous mica-schist from this area of metamorphic schists.* Both of these schists contain prisms of bluish tourmaline in considerable quantity. It appears highly probable that these staurolitiferous and garnetiferous schists owe their origin to the intnision of the igneous rocks, though no well-marked contact zones could be outlined. The exomorijhic contact effect of the granite is more noticeable where a large body of the granite contains a sedimentary intrusion than elsewhere. The determination of the sedimentary origin of the fragments included in the granite is based primarily ujjon the probability that in its passage I Thiirach, GrotU's Zeitschr., Vol. II, p. 423. ' A. Pichler, Beitriige zur Mineralogie Tirols, Nenes Jahrb., 1871, p. 54. ^Geology of the Menominee iron region, by C. E. Wright: Geol. of Wisconsin, Vol. Ill, 1878, Part 8, 11.695. * Log. cit., p. 695. ACID INTUUSIVES. 197 tlirou"-!! the seiliiii('ntiU-\' rocks whirli now surniuiKl it tliu granite iiichuU-d fraj'-meiits of tliciu. In addition to this a well-defined banding- is still ])resent in these frag- ments. Thougli by no means conclusive evidence, this is considered as an indication of their having been originally deposited through the mediation of water. The sediments have been comi)letely recrystallized into fine- grained mica-gneisses. The sedimentary fragments included in the granite now sliow the foliowiug characters. They are composed of layers of two kinds. The one kind of layer is very fine grained, of gray color, and consists predomi- nantly of biotite in fairly good automorjihic plates, muscovite in small quantity — but in automorphic jilates — even with respect to the biotite, feldspar, quartz, and iron oxide. The feldspar is in small equi dimensional grains. Only one finely striated feldspar was observed, the greater part possibly being orthoclase. It shows in places a well-developed zonal structure, the zones conforming to the outlines of the grains. The zonal structure probably depends upon varying quantities of the soda and potash molecule. Quartz occurs in grains in very small quantity. The second kind of layer which is seen in the fragments is nuicli coarser grained than is the first, just described, and is very much darker. It is composed mainly of muscovite and biotite, in about equal quantities, feld- spar, quartz, magnetite, and ilmenite, with some tourmahne. Both feldspar and quartz occur in grains, the former in small quantity. The biotite is in small plates less well developed than the muscovite, though mostl}- auto- morphic. It is partly bleached and has associated with it here and there some secondar}^ epidote. The muscovite is in very large automorphic plates, some of them twinned according to Tschermak's law, and includes flakes of biotite and grains of quartz and feldspar. This gneiss contains also a large number of crystals of tourmaline, showing strong dichroism from light pinkish to dark grayish blue. Iron oxide is present, both as magnetite and as ilmenite. The quadratic individuals of the one and the hexagonal plates of the other at times are very well developed. No signs of pressure whatsoever are seen. The muscovites evidently represent the last product of crystallization, as shown by their including all the minerals which had been previously formed. (Fig. A, PI. XXXIV.) These muscovites are probably to be 198 THE CRYSTAL FALLS IRON-BEARING DISTRICT. looked upon as the product of mineralizers, dependent upon the presence- of the hot g-ranite injection, to whose action may also be referred the presence of the tourmaline. The line of contact between the granite and the sedimentary frag- ments, though somewhat irregular, is macroscopically very well defined by the difference in size of the mineral constituents of the two rocks. Here and tliere in the fragments there are seen thin masses of the granite which were injected along the planes of sedimentation, and also traverse cracks penetrating the sediments. On the granite side of such a contact there is a nan-KW zone very noticeably richer in large biotite flakes tliau the granite is ordinarily. No diff'erence can be noticed on the sedimentary side of the contact. The microscopical examination of the contact emphasized the endogenous character of the metamorphic action. There is more biotite present than in the normal granite. The feldspar in the normal granite has a decided tendency toward autoniorphic development. Where the feldspars of the granite touch the metamorphosed sediments they are partly rounded, and the mica plates developed in the sediments have in general a parallel structure around that side of the feldspar which is turned toward the fragments. At another point the quartz of the granite, where it comes in contact with the sediments, has developed as an automorphic individual, and looks as though it were pressed into the fragment. The quartz crystal contains near its edge grains of feldspar and flakes of mica, thus making an imper- fect narrow poikilitic zone. Beyond this zone there is the sedimentary rock jjroper, and there the mica plates lie parallel to the contours of the quartz. An illustration of such a contact is shown in figs. A and B, PL XXXV, as seen in ordinary light and also between crossed nicols. It appears that the formation of the quartz and feldspar noted above caused the arrangement of the constituents of the gneiss parallel to their contours. It would thus seem probable that in this particular case the recrvstallizatiou of the original graywacke into the gneiss which we now find in its place followed and was, chiefly the result of the intrusion of the m-anite. ° BASIC INTRUSIVES. The basic intrusives are represented by metadolerites and metabasalts. The dolerites are the most important, and will be treated in detail. Rocks- BASIC INTRUSIVES. 199 very siniihir to tlie l);iisalts have alread}' been described at length under the Hendock volcanics, and since they are found in loniparatively few dikes they will be passed over with very Ijrief mention. METADOLERITE.i GEOGRAPHICAL DISTRIBUTION. The dolerites of the Crystal Falls district for tlie most part form liigh ridges extending in a northwest-southeast direction. Their principal occur- rence is in the area immediately north, northeast, and east of Crystal Falls. Beginning in sees. 32 and 33, T. 43 N., R. 31 W., there extends a great intrusive mass, varying from a mile to a mile and a half in width, due north to sec. 6, T. 43 N., R. 31 W. There it bends to the northwest, and ends in sec. 3,' T. 43 N., R. 3"2 W. The northwestern extension of this mass is much narrower, never exceeding a half mile, and at many places it is only a few hundi'ed yards in width. In the northern part of T. 42 N., R. 31 W., are a number of knobs which are evidently connected below with these large masses, although the exposures are discontinuous. A large dike begins in sec. 1, T. 42 N., R. 32 W., and extends for about 5 miles to the northwest into sec. 19, T. 44 N., R. 32 W. This averages about one-eighth of a mile in width, though in places it is three-fourths of a mile Avide. Another dike begins in sec. 28, T. 44 N., R. 32 W., and runs for 3^ miles to the northwest into sec. 18, T. 44 N., R. 32 W., and, like the above, is narrow, being only about one-eighth of a mile in average width. A narrow dike less than one-eighth of a mile in width extends in a high ridge from in sec. 16, T. 43 N., R. 32 W., to the northwest for 3 miles and ends in sec. 7 of the same township and range. Numerous isolated knobs occur in T. 44 N., R. 32 W. A small boss is in sec. 24, T. 46 N., R. 33 W., and another at 1,600 paces N., 1,000 W., sec. 19, T. 47 N., R. 33 W. PETROGRAPHICAL CHARAf^TERS. iiacroscopicai. — Tlic doleritcs vary in color from greenish to dark olive- green and almost black. The weathered surface is usually of a very light color, rather a light gray, with frequently a reddish tinge. The texture is 'luse the name " ilolerite " here merely to iudic;ite the macrostiuctural dift'erence between the rocks inclnded under it and the fine-grained and aphanitio rocks of the same composition included under the basalts. It is also extended to include jja^fo as well as neo cruptives. As nearly all, if not all, the paleodolerites have undergone great alteration, the prefix meta — indicating alteration without reference to any specific kind — is very generally applicable. 200 THE CRYSTAL FALLS IRON-BEARING DISTRICT. medium to coarse. Probably the most striking textural characteristic is the pecuhar mottled appearance described as "luster-mottling" by Pumpelly,' to which the name poikilitic" has of late years been more generally applied. This texture is almost always brought out macroscopically on the weathered surfaces by the difference in the weathering of the feldspar and the augite or uralite, these lieing the prominent mineral constituents of the rock. This poikilitic texture is most common in the coarsest of the metadolerites. In such rocks the augite or uralite areas are of large size, quite commonly 2 centimeters in diameter, and show their mottled character very plainly to the naked eye. In the medium-grained dolerites the ordinary ophitic texture is pre- dominant, though in these rocks poikilitic areas may be seen. In fact, these poikilitic areas really possess an ophitic texture, according to the definition of that texture by A. Michel Levy, for the feldspars are developed as laths and the pyroxene is the mesostasis in which the feldspars lie.^ It is thus clear that, restricting the statement to these dolerites, the ophitic texture is at times included in the poikilitic, and that under such circumstances the two can not be considered as totally different and independent textures, but are, on the contrary, practically identical. In one dike of dolerite the influence which the conditions of consoli- dation exert upon the texture is well shown. This dike is only 8 feet wide, but the center is developed as a dolerite, while along the edges where cool- ing was more rapid, the rock is a porphyritic basalt. The porphyritic texture is caused by the development of pyroxene and feldspar phenocrysts, which lie in a dense basaltic groundmass. Microscopical. — Tlic OHgiual miucrals of which the rocks were composed were feldspar, quartz, pyroxene, olivine, biotite (?), apatite, and titano- magnetite. The minerals which are now present in the rocks are for the most part secondary. They are hornblende, muscovite, epidote-zoisite, chlorite, biotite, sphene, leucoxene, calcite, albite, quartz, and iron pyrite. Of these, hornblende is by far the most prominent constituent. A study of the isolated specimens of these rocks might result in their determination as ' Metasomatic develoitmeiit of the coppei-beariug rocks of Lake Superior, by Raphael Pumpelly: Proc. Am. Acad. Arts Set., Vol. XIII, 1878, p. 260. 'On the use of the terms "poikilitic" aud "micropoikilitic" in petrography, by G. H. Williams- Jour. Geol., Vol. 1, 1892, pp. 176-179. = Structures et classiljcatiou des roches eruptives, by A. Michel Levy, Paris, 1890, }>. 30. BASIC 1NTKUS1VE8. 201 uralitic dolerites, cpidioriti^s, or even dioritus, as it is impossible, without a stHiiieuce of cliaujj^es, to deteriniue wlietlier the hornhk'iide is original or secondary. Rare rocks contain quartz in sutfieient (juantity to warrant their designation as quartz-dolerites. However, they differ in no essential respects from the other dolerites. The quartz is in micropegmatitic inter- growth with feldspar, filling the angular interspaces of the rock. These intero-rowths were evidently the last elements to crystallize. The feldspar occurs in large automorphic lath-shaped crystals, which in most cases show poly synthetic twinning. In a few cases unstriated crys- tals were obseiwed. Owing to the alteration of the feldspar, which has in most cases almost completely destroyed the striations, it has been impossible to make many accurate measurements. Measurements ou the zone perpen- dicular to 010 gave equal e.xtinetion angles against twinning planes of 37° as maximum, showing the feldspar to be bytownite. The chief alteration products of the feldspar are epidote and zoisite. With these are usually associated more or less muscovite, some chlorite, and, more rarely, scales of biotite. Accompanying these alteration products one very fi-equently finds hmpid spots of secondary albite or quartz. Some few of the feldspars are smoke-colored, and as the coloring appeared homogeneous even under the highest powers, it Avould seem to be due to some pigment in the mineral and not to minute inclusions. Pyroxene is very rare, having been observed in only a few sections and in the majority of these is present merely as small remnants surrounded by its secondary product, uralite. The pyroxene possesses the usual char- acters of common augite. The augite is quite free from inclusions. Along the edge its alteration to the light-green hornblende, uralite, can be readily followed, and in one case an octagonal basal section of augite was observed which was completely occupied by uralite fibers. The former presence of olivine is based upon verj^ slight proof, viz, the existence in some of the pyroxene and uralite crystals of areas which are oval or round in shape and are occupied by pilite. The presence of this pilite in the altered augite might possibly be explained as an alteration product of the augite itself, but it is difficult to explain why pilite should develop in one part of the augite and secondary coarse hornblende in the other part. Moreover, the general characters of the rocks are such as to lead one to expect to find olivine present in some of them. 202 THE CRYSTAL FALLS IRON-BEARING DISTRICT. Biotite occurs in large irregular masses whicli are considered to be primary, as well as in the scales which occur within secondary products of the rock and are considered to be secondary. It is scattered throughout the rocks in irregular pieces, usually associated with iron oxide. Where fairly fresh, it is brown and shows its ordinary character. By weathering it becomes green, having still a high double refraction. By further weath- ering it passes into a nearly colorless mass that has the faintest tinge of green and scarcely polarizes light. Such masses are crossed by lines of hair-like crystals, some of which intersect one another at angles of 60 degrees, extinguish parallel to their long directions, and show high single and double refraction. These are taken to be rutile. Other crystals, somewhat coarser, also lie irregularly in the biotite masses. They show the same inter- sections as the rutile. They are very faintly greenish, have a high single and double refraction, are positive in the long direction, and have a maximum extinction angle of 46 degrees. These characters were not sufficient to determine the mineral by, and no other characteristics could be observed. Ilmenite and titano-magnetite are in irregular grains. These minerals are more or less altered to leucoxene or to sphene. Very frequently the alteration product incloses bands of the iron oxide, which intersect one another at angles of 60 degrees, pointing toward the hexagonal character of the ore. In one case a beautiful example of the alteration of such an ilmenite to rutile was observed. It is exactly similar to that described by Williams in the case of the greenstones of the Menominee district.^ By low power the mass has a semimetallic luster, and, as it seems to. be almost solid, has very much the general appearance of an ore, but by higher power it is resolved into a mass of small golden-brown crystals. These frequently intersect one another at angles approximating 60 degrees (120°), similar to the fine needles in sagenite. The hornblende is mainly in large xenomorphic plates inclosing the automorphic feldspars. This is the variety of hornblende known as uralite and is all presumed to be of a secondary nature. In no case does it pos- sess the compact nature of original hornblende, but is always more or less fibrous, its fibrous nature being best seen along the edges, and less clearly shown, though still observable, where the sections are thicker. It varies 1 A letter to Neiies Jahrbiich,Vol. II, 1887, p. 263. The greenstone-schist areas of the Menominee' and Marquette resious of Michigan. Bull. U. S. Geol. Survey, No. 62, 1890, p. 99. BASIC INTRUSIVES. 205 iVoin scarc't'lv colored needles to those wliieli are stroujilv pleoeliroic. The pleocliroism varies from yellowish for a to yellowish oi- dlixe •'•reen for ItJ, and in many cases to a dark hluish-green for c. In a few cases mucli of the hornblende has frequently a darker shade in the center than at the border, althon0" Fe O3 2.53 .45 Trace. .13 .04 .82 1.66 .06 2.10 None. FeO MnO CaO BaO SrO M"-0 1.35 5.73 .54 2.10 .15 6.57 KG Na.O Li,0 H at 100° — . . .60 3.62 .03 None. .97 .05 .52 .08 .09 .18 HoO at 100°+ PjO, CO, c . SandSOi None. None. Trace. None. None. None. CI F . . Total 99.57 99.72 99.76 100. 76 XHoO at 110°. +HiO above 110°. No. l = Clay slate (.Sp. 32497). Sec. 17, T. 43 N., R. 31 W., 450 N., 1620 W. ; George Steiger. No. 2 = Spiiosite (Sp. 32861). Sec. 7, T. 43 N., E. 31 W., 750 N., 1380 W. ; H. N. Stokes. No. 3=Spilosite (.Sp. 32827). Sec. 7, T. 43 N., R. 31 W., 250 N., 325 W. ; H. N. Stokes. No. 4 = Aillnole (Sp. 32465). Sec. 8, T. 43 N., R. 31 W.,500 N., 475 W. ; George Steiger. In these analyses the usual increase of silica as the dolerite is approached is at once noticeable, and hand in hand with it goes the diminu- tion in percentage of alumina and iron oxides. The content of water and carbonaceous matter also suffers a diminution. The most iKiteworthy differ- ence between the clay slate and the contact rocks is shown in the relations of potassa and soda. This is well Ijrought out in an examination of analyses Nos. 1 and 2. It will be seen that there is only about one-eighth as much potassa in the contact rocks as in the normal clay slate; while, on the con- CONTACT METAMOUPHISM BY INTKUSIVES. 211 triir\', aljout 12 times as much soda as there was in tlie shitt^ has hceii added to the contact rock. This causes a reversal of the relations of the soda and potassa, so that, whereas in the clay slate there is present 10 times as nuicli potassa as soda, we find in the contact rock taken as a sample very nearly 1») times as nuich soda as potassa. The very considerable change in chemi- cal t-omposition, especially in the amount of silica and soda, seems to lend great weight to the supposition that in such contacts an actual transfer of material (soda-silicate) takes place from the basic intrusive to the slate. This idea is upheld by Roth,^ Zirkel," and others. W. Maynard Hutchings^ advocates this view, and has described some interesting products as a result of the contact of the Whin Sill which still further support it. XO KNDOMUUPIIIC EFFKCTS OF DOLKRITE IN'TUUSION. Although the exoraorphic contact effects of the dolerite intrusion were so obvious, no evidence is found that the dolerite itself suffered any change consequent upon its intrusion. METABASALT. Basalt has been described at length under the volcanics, where it plays an exceedingly important role. Basalt as a dike has been found in only two places, and therefore very little remains to be added. The two basalt dikes occur within a very short distance of each other, in sees. 15 and 16, T. 42 N., R. 31 W., and are found penetrating the crystal- line schists of the Upper Huronian. Their reUitions to the other intrusive rocks of the same region are not known. The}' are probably of the same age as the dolerites, of whicli they should most likely be considered offshoots. These dikes are a porphyritic basalt. The phenocrysts were of augite, olivine, and labradorite. They were in a very fine groundmass of feldspar, augite, and iron oxide. However, the former existence of the augite and olivine phenocrysts is determinable only by means of their outlines. They are in very small quantity and are entirely altered to pilite. The feldspar phenocrysts are in coarse, heavy crystals and are remarkably fresh. The groundmass is very fine grained, and ranges from an exceedingly fine micro- ' Chem. Geol., by .1. Roth, Berlin, 1890, Vol. Ill, p. 145. ^ Lehrbuch tier PetrograpUie, by F. Zirkel, Vol. II, 1894, p. 722. ' Notes (III thu composition of clay slates, etc., and. on some points in their contact metamorpliism, by W. Maynard Hutchiugs : Geol. ila','.,Vol. I, Dec. 4, 1894, p. 75. Cheni. Geol., Vol. Ill, p. 145. An interesting contact rock, with notes on contact metamorphism, by W. Maynard Hutchlngs : Geol. Mag., Vol. II, 1895, pp. 122-131, 163-169. 212 THE CRYSTAL FALLS IRON-BEARING DISTRICT. ophitic texture to the pilotaxitic texture. The feldspars in it are in small lath- shaped individuals, and, like the phenocrysts, are fresh. The augite of the groundmass is to a great extent altered to uralite, and the iron ores to spliene. One of the dikes is about 5 feet wide. In the center it is a moderately fine-grained rock; on the edges it is a dense aphanitic basalt. Even in thin section the gradation from the rock with microophitic groundmass to the one with a dense pilotaxitic groundmass is well shown. A dike of larger size might readily have cooled sufficiently slowly to have crystallized at its center as a dolerite. UliTRA-BASIC INTRUSIVES. '^ Under this head are the descriptions of the picrite-porphyries (porphy- ritic limburgites). PICRITE-PORPHYRY (PORPHYRITIC LIMBURGITE). GEOGRAPHICAL DISTRIBUTION AND EXPOSURES. The picrite-porphyries occur in isolated outcrops of comparatively small size in sees. 9, 22, and 27, T. 44 N., R. 32 W., in the area supposed to be underlain by the Lower Huronian Hemlock volcanics. They are surrounded by outcrops of the altered poikilitic dolerites, but the exposures are not such as to allow their relations to be determined. Their occurrence points to an intrusive character. It is on account of their field occui-- rence alone that we feel justified in describing them here under the general heading for this chapter, "Intrusives," instead of under the volcanics with the basalts, their proper place from a strict petrographical standpoint. PETROGBAPHICAL CHARACTERS. The picrite-porphyries are medium-grained rocks, which varj^ in color from gray to dark green and almost black. In general they have a por- phyritic character. This is, however, not so well marked in the gray as in the darker-colored rocks. The gray ones have a spotted appearance. The sppts are- gray in color, fibrous, ver}" rarely larger than 3 or 4 millimeters in length, and lie in a finely fibrous, dark-green matrix. In the dark rocks the porphyritic crystals reach a length of 1 centimeter, and are bluish to black, with silky luster. They lie in a fine-grained, more or less fibrous, green groundmass. In one of the dark rocks the magnetite is very notice- able. The crystals project from the weathered surface and the rock is strongly polar-magnetic. ULTHA BA81G INTIUTSIVBS. 213 Tlic rocks ()rii;iiiall\' cousisti.'il l;ii';ic'ly i>t (ilixiiic, i)\i'().\ciic, linruhlciiiU', biotite, inaynetitt', and ilinenite. Tliey now contain alsso, in consideralile quantity, a cliloriti<' product wliicli seems to lia\e been dei'ived from the alteration of an original vitreous base. All of the specimens are exceed- in>a. This hornblende contains inclusions of iron oxide and has all the appearance of an original mineral. By alteration it passes through a compact greenish amjjhibole to a much lighter colored, reedy, actinolitic variety of amphibole. In the secondary amphibole occur certain golden-brown grains with high single and double refraction, which are supposed to be rutile formed from the hornblende, and also some brown transparent plates of ilmenite. The orientation of the secondary horn- blende is the same as the original. No further alteration of this amphibole was observed, but it is believed that the prismatic crystals altered to chlorite, calcite, and magnetite, as described above, are the extreme cases of alteration of an automorphic form of a brown hornblende very similar to the part described. The biotite between the phenocrysts is in ragged areas either surround- ing iron oxide or associated with it or with the hornblende. It is very pleochi'oic, the absorption jjarallel to the basal cleavage being so strong as to render the section opaque. Perpendicular thereto the color is a dai-k chocolate brown. The mica does not show its usual bright polarization colors in sections cut parallel to crystallographic c. This may be due in some measure to the very strong absorption. In some cases the biotite is seen to have a strong blue to violet metallic luster in incident light. The biotite has partly altered to chlorite. The alteration proceeds along the basal 216 THE CRYSTAL FALLS IROJf -BEARING DISTRICT. cleavage. As this alteration progresses there is a lighteuiiig of the color of the biotite, and, as a consequence of this the whole cause of the metallic luster and the partial cause of the color of the biotite is disclosed. In the lighter biotite one by careful examination can see innumerable small plates of a brown or smoky color. At first sight they remind one strongl}- of the inclusions so common in man}' hj^perstlienes. Closer examination only emphasizes this reseiiiblance, and they are believed to be micaceous ilmenite plates. These inclusions were studied by means of an oil immersion objective givmg a magnification of about 1,250, and were found to have mainly a roundish or hexagonal outline. In addition to these, some plates of long, irregular form were observed. These are all isotropic and lion- pleochroic. These minute plates lie parallel to the biotite lamellse. The consequence of this is that in sections parallel to c one sees, for the most part, onl}^ short black streaks — the edges of the plates — whereas in the basal sections of the biotite one can determine the irregular or rounded contours of the plates. The plates ai-e too small to allow the metallic luster to be seen on an isolated one. En masse they j^roduce a very decided blue metallic shimmer, as seen in some of the biotite fragments. Numerous apatite crystals occur. They are usually clear white, but one crystal was seen showing a dichroism from faintest brownish for rays perpendicular to crj-stallographic c to light smoky brown for rays parallel thereto. This crystal contains a core of brown glass. • Some of the iron oxide is in roughly rectangular masses, and appears to be magnetite. This is associated with an iron oxide, which occurs in opaque, ragged masses formed of long, irregular, and knotty stringers. These at places are parallel to one another and at other places cut one another at various angles, and at still other places meet at a common cen- ter, forming an opaque mass of varying dimensions, but usually small. Now and then one of the large magnetite masses constitutes a center from which extend the knotty, irregular stringers. The general appearance of these ragged masses is that described by Grerman petrographers as mrhackt. When these stringers pierce the section at an oblique angle, the ends are translucent, with a brown color, becoming more opaque as the section gets thicker. Such masses have all the appearance of ilmenite, and are believed to be that mineral. Similar ilmenite stringers are included in the chlorite, wliich results from the alteration of the biotite ULTRA BASIC INTUUSIVES. 217 The cliloritc in tlio iiiirainorplis after liiotite sliows extrenielv low blue polarization color, and the characteristie pleoehroism — yellowish, tinged with red, when the rays vibrate perpendicular to the cleavage, and green when i)arallel thereto. Apatite needles included in the l)iotite are unaltered in the secondar}' chlorite. Some of the minute octahedral crystals in the amphibole pseudomorphs after olivine appear to be slightly pellucid, with a brown color. If so, thev might be referred to picotite, but there is doubt of the correctness of the observation, in view of the high power used, the oil immersion lens, and the fact that the search was for picotite. Close search was also made for perovskite, but none could be found, unless the transparent crystals very doubtfully referred to picotite are realh' iierovskite. Forming the matrix in which the pseudomorphs after hornblende, olivine, and pyroxene of these rocks lie, and frequently surrounding isolated crystals, one sees an aggregate composed chiefly of a fine felt of chlorite fibers. This alteration product contains a few apparently original apatite needles, some secondary grains of magnetite, and crystals of amphi- l)ole which are colorless or else show but the faintest tinge of green, and are larger than the amphibole crystals in the pseudomorphs. It is a secondary amphibole very poor in iron, probably highly calcareous, and approaching tremolite in composition. This chlorite aggregate shows no indication whatever of crystal forms. It seems to be the product of a homogeneous mass, such as would result from the decomposition of a vitreous base. Such a base the agg-regate is presumed to represent, althoug'h no trace whatever of the glass has been observed in the rock, nor in view of the altered con- dition of the rock could such a glass be reasonably expected to still remain. DARK SERI'ENTINIZED PICRITE-rOIiPH YRY. The second variet}' of the jncrite-porphyries is very dark greenish- black, and represents the results of a slighth' different process of alteration from that by which the gray forms just described were produced. These dark picrite-porphyries show a very much better developed porphvritic structure than do the gray ones. This is due to the fact that the oliAines in these rocks were well developed and reached a length of a centimeter. The olivines are completely altei-ed, serpentine, pilite, and magnetite, being the products which form the pseudomorphs. The characteristic mesh structure 218 THE CRYSTAL FALLS IRON-BEARING DISTRICT. of altered olivine is well broug-lit out l^y the serpentine and iron ore. In the centers of the meshes there remain small masses of a felt of tremolite needles (pilite). This alteration of the olivine corresponds to that first described by Lewis/ and more recently by Professor Bonney and Miss Raisin,^ from a rock — kimberlite — very similar to the picrite-porphyries here described. He writes as follows: "It frequently happens that while sei'pentinization begins at the outside of a crystal, iibrous tremolite begins growing- within, finally forming a mass of asbestiform fibers surrounded by a zone of green serpentine." The minerals which composed these black picrite-porphyries were the same as those constituting the gray ones. These minerals were olivine, pyroxene, hornblende, biotite, magnetite, and ilmenite. They were cemented by a glass matrix. The glass is completely altered. All of the minerals are represented by pseudomorphs. Remnants of the original hornblende and biotite alone are preserved. The contours of the original pyroxene crystals are filled with pilite, serpentine, and magnetite. The serpentine is present in greater quantity in these pyroxene pseudomorphs than it was in tlie pyroxene pseudomorphs in the o-ray picrite-porphyries. The alteration of the hornblende results in the production of an aggregate of chlorite inclosing grains of caleite, some sphene, and iron oxide, similar to that in the gray picrite-porphyries. The biotite, magnetite, and ilmenite also show those characters which have been described for the same minerals in the first-described picrite-porphyries. Between all of the foregoing minerals we find a fine felty chlorite mass eontaimng grains and dendritic masses of iron ore and a few needles of tremolite. This corresponds to the material forming the cement for the minerals in the gray porphyries, and, hke that, is believed to represent an original vitreous matrix. In one of the dark picrite-porphyries the magnetite is present in large quantity and is very noticeable, crystals of it standing out upon the weath- ered surface. This rock did not affect the magnetic needle very powerfully, though it was expected that it would do so. However, another one of 1 Ou a dlamondiferous peridotite and the genesia of the diamond, hy H. C. Lewis: Geol. Mag., 3d ser., Vol. IV, 1887, p. 22. Papers and notes on the genesis and matrix of the diamond, hy the late Henry Carvill Lewis, edited hy Prof T. G. Bonney, London, 1897, p. 14. 2 Notes on the diamond-hearing rock of Kimherly, South Africa, Part 11, hy Prof T. G. Bonney and Miss C. A. Raisin : Geol. Mag., 4th series, Vol. II, 1895, p. 496. ULTKA-BASIC INTRITSIVES. 219 tliosc porphyrios, in wliicli, by tlio way, tlic! iron content is relatively low, is unitpie, in tiiat it is very stronj^ly \)o\av niag-netio, and in this, as well as its probable original niineralogical composition, may be cdinpanMl with the polar magnetic wehrlite from the Frankenstein, Hesse-Darmstadt, Germany.' The German rock shows tremoliti? scattered through the serpentine result- in"- from the olivine. It is a coarse, evenly granular rock, differing in this respect from the Crystal Falls rocks which are porphyritic. Au analysis (No. 1) of the polar magnetic serpentinized picrite- porphyry, in which great abundance of olivine was originally present, is here given, and there is placed with it for comparison an analysis (No. 2) of a very similar rock described by Darton and Kemp," from New York. Both analyses were made by Dr. H. N. Stokes, United States Geological Survey. In No. 1 Ba, Sr, Li, CI, S, and SO' Avere not looked for. Analyses of picrite-porphyry. SiO- 37.36 TiO, : "t^ Al.O:, ITO Cr,03 62 Fe,03 n.61 FeO 6.12 MnO Trace. XiO ; .04 I CaO [ 1.19 BaO SrO MgO I 31.11 K^O ;1 Trace, i Na^O 1 I P.O5 1 -06 CO. None. SO3 ! S I HiOatllO^ -65 HnO above 110° 10-37 36.80 1. 2G 4.16 .20 Less 0=S . . Total 8.33 .13 .09 8.63 .12 Trace. 2.5. 98 2.48 .17 .47 2.95 .06 .95 .51 6.93 99.68 100.22 .47 99.75 1 Der raagnetstein von Frankenstein an der Bergstrasse, by Audreae and Kfinig: Abhaudl. der Senkenberg. naturf. Gesell., Frankfort a Main, 1888, pii. 59-79. Cf. Above article, p. 66, footnote, for references to other occurreuces of tremolite as.soclated Tvith serpentine. - Newly discovered dike at Dewitt, near Syracuse, New York, by N. H. Darton and J. F. Kemp: Am. .Jonr. Sci., Vol. XLIX, 1895, p. 461. 220 THE CRYSTAL FALLS IRON-BEAKIiSG DlSTKiUT. CLASSIFICATION. These rocks just described, from their miiieralogical composition, if we admit the presence of a vitreous base, would belong- with the picrite- porphyrites of Rosenliusch.' This designation does not seem, however, to be appropriate, as he states" that he uses the term "porphyrite" onlv for certain textural phases of rocks containing lime-soda feldspar. He has evidently extended that definition so as to be' able to iise it for these picrites, considering that the glass possesses the necessary ingredients for the formation of such lime-soda feldspar, provided the conditions under which it cooled had been favorable for the feldspar development. The porphyritic texture of these Crystal Falls picrites and the presence of a vitreous base^ show them to be closely related to rocks of effusive char- acter. Those which they most closely resemble among the younger basaltic lavas are the porphyi-itic forms of the limburgites (magma basalts). One of the best-known rocks with which this may be closely compared, as far as association is concerned, is the rock first described by H. Carville Lewis as a saxonite-porphyry,* later called kimberlite. This was described by him as volcanic, and as associated with dolerites and melaphyres. He described it as a basic lava.^ Other occurrences of very closely related basic rocks havino- a vitreous base have been described from the United States bv Diller, Williams, Merrill, Brainier and Brackett, Kemp, and Darton and Kemp." ' Microscopische Physiograpliie, by H. Roseiibusch: 3fl ed., Stuttgart, Vol. II, 1896, p. 1191. -Op. cit., p. 436. 'Should the vitreous baso be considered as not having been pre.sent and the rocks be pnt among the peridotites, then they would correspond very closely to the wehrlite described on p. 2."i4. ••Papers and notes, cit., p. .">0. '■The genesis of the diamond, by H. C. Lewis : Science, Vol. VIII, 1886, p. 345. On a diamonerhaps also be compared with these ULTRA-BASIC INTRUSIVES. 221 Hatch' lias also described a very similar pre-Tertiary rock from Kug- laiid as a limburgite. Kemp- emphasizes the resemblance of the Dewitt dike to liiiil)urgite, and states that it should be called limburgite.^ If we attempt to extend the use of the term " limburgite" to include the pre-Tertiary vitreous basalts, we shall have to include under it the rocks heretofore desig- nated as picrite and picrite-porphyrite. Rosenbusch has now put the jiicrites and picrite-porphyrites with the effusive rocks, and if of these two sets of terms there is one to be discarded, it should be the name "limburgite." It seems preferable luider the rules of priority to retain the name "picrite." It would then seem very suitable to apply to these pre-Tertiary porphyritic limburgites Hussak's old term, "picrite-porphyry," using the term "por- phyry" simply witli a textural significance.* SECTION II.— A STUDY OF A ROCK SERIES RANGING FROM ROCKS OF INTERMEDIATE ACIDITY THROUGH THOSE OF BASIC COMPOSITION TO ULTRA-BASIC KINDS. Beginning near the town of Crystal Falls, in isolated knobs, and extending southeast toward the Micliigamine River, where the exposures are larger and better connected, there is found a series of rocks wnose charac- ters are of such interest petrogeiietically as to warrant a detail descrijjtion of them. These rocks are all intrusive in character, with few exceptions are medivim to coarse grained, and, while the granitic texture is predominant, there are certain facies in which the texture is porphyritic and even parallel. They have been only slightly affected by dynamic action, and these cases are purely local. Analyses show them to vary in chemical composition from those of intermediate acidity to those of ultra-basic character. The prevailing rocks are, on the one hand, diorites of intermediate acidity ranging to more acid rocks, tonalites, quartz-mica-diorites, and rocks, re[)re8enting as it probably does the oliviae-free form of the limburgite (augitite). Geology of Angel Island, by F. L. Ransoine : Bull. Geol. Dept. Univ. of Califoruia, Vol. 1, 1894, p. 200. ' The Lower Carboniferous volcanic rocks of East Lowthian (Carlton Hills), by F, Hatch: Trans. Royal Acad. Edinburgh, Vol. XXXVII, 1892, p. 115. -Op, cit.,p.460. "'Taking plutonic rocks as practically the granitoid, and volcanic as the porphyritic, the Dewitt rock is a basaltic dike of the same composition and texture as limburgite, and should be called limburgite, even if it is not a surface flow." (Loc. cit., p. 460. ) ■•I believe E. Hussak was the first to use this term for a somewhat similar rock. Pikrit-pliorphyr von Steierdorf im Banat, by E. Hussak : Verhandl. K.-k. geol. Reichsanstalt, 1881, pp. 258-262. 222 THE CRYSTAL FALLS IRON-BEARING DISTRICT. granite (plagioclastic), and, on the other, hornblende gabbros, gabljros, norites, and, kxstly, peridotites of varying mineralogical character. These rocks thus resemble in their -sanations those Scottish plutonic rocks so well described by Messrs. Dakyns and Teall.^ The rapid changes in mineralogical composition and texture in a single rock, and the changes from one kind to another through intermediate facies, show verv clearly the intimate relationship of these rocks to one another, and warrant the assumption that they all belong to a geologic unit, a con- clusion reached a luimber of years since by Williams for a somewhat sim- ilar series, the Cortlandt series, from New York. Granite is present as a local facies of the diorite. Howevei', it is very subordinate in quantity and not altogether typical, and as no analysis has been obtained, its jjosition is still more or less doubtful. In the following pages only those kinds of rocks of Avhich analyses have been obtained will be included in the final discussion. Others will be described in detail or merely mentioned, as representing facies of the main types, according to their petrological interest. The rock tyj^es of which analyses have been made are as follows: Diorite, gabbro, norite, and peridotite. DIORITE. NOMENCLATURE. Diorite, according to the generally accepted definition, is a granular rock consisting essentially of hornblende, which must be primary, and a soda-lime feldspar.- The term has been used in a difi^erent sense by many writers on the Lake Superior and other regions. It has been used to com- 2irise rocks which contain hornblende and plagioclase as preponderating constituents, it is true, but in which the hornblende is secondary, therein diff'ering from a true diorite. These so-called diorites have been regarded as derived from an original dolerite (diabase) by uralitizatiou of the pyrox- ene. By some writers these rocks have been classed ^yh]^ the epidiorites, thus recognizing their secondary nature, but by this name, epidiorite, unfortunately implying a false relationship. In this paper, following Brogger, I restrict tlie name to the granitic ' On tlie jiltitonic rocks of Garabal Mil! and Meall Breac, by J. K. Dakyns, es(|., M. A., and J. J. H. Teall, esq., M. A., F. R. S., F. C. S.: Quart, .hmr. Geol. Soc, Vol. XLVIII, 181W, pp. 104-121. -Lehrbuch der Petrojjrapbic, by F. Zirkel, Leipzig-, Vol. II, 1891, p. 465. DIOKITE INTUUSIVES. 223 rocks of iiitermediiite acidity, in which the feldspar is ])hiy'iochiso and the bisilicate constituent is mica or primary hornblende. The feldspar is a lime- soda plagioclase.' DISTRIBUTION AND EXPOSURES. The dist.ril)ation of the diorite is limited to a few localities, all of which are in the area underlain by Upper Huronian sedimentaries. The most typical occurrences, and those showing greatest variations, form knobs beginning near Crystal Falls and continuing to the south and south- east. Especially large outcrops form the hills in sees. 19 and 20, T. 43 N., R. 31 W. The smaller occurrences are not indicated on the map. These diorite exposures are always good, so far as getting fairly fresh specimens are concerned, but their contacts with other rocks are almost invariably deeply covered with drift; hence their relations in many cases are doubtful. PETROGRAPHICAL CHARACTERS. The diorites are holocrystalline rocks of medium to coarse grain. In texture they show some variation from those which are granitic to those in which the texture is imperfectly ojjhitic. The color is, on the whole, moderately light gray or reddish, but at times when the dark minerals become more prominent in the basic facies, especially where we g"et basic schlieren, the rock is very dark gray or greenish brown. The important mineral constituents are feldspar, quartz, biotite, and hornblende. The accessory minerals are epidote, apatite, zircon, sphene, and iron oxides. The secondary niinerals, white and larown mica, chlorite, biotite, epidote-zoisite, calcite, and rutile are also present. Feldspar. — Plagioclasc feldspar, orthoclase, and microcline occur together. The plagioclase is found in individuals which are fairly automorphic. In the ophitic textured diorites, the plagioclase is the best developed of all the essential constituents. In the granular rocks the degree of automorphism is highest where orthoclase and quartz are present in the largest quantity, and diminishes as these diminish, when the plagioclase individuals beghi to interfere with one another's development. For the most part the plagioclase ' Die EruptivgeBteine des Kiiatiauiagebietes. I. Die Gesteine dor Grorudit-Tiiigiiait-Soiic, by W. C. Brogger, 1894, No. i, p. 93. II. Die Eruptionsfolge der tiiadischen Eruptivgesteine bei Predazzo in Siidtyrol, 1895, No. 7, p. 35. Videnskabsselslvabets Skrifter, I Mathematiskuaturv. Klasse. 224 THE CRYSTAL FALLS IRON-BEARING DISTRICT. gives rather thin sections, thoug-li they can hardly be correctly called lath- shaped. No other form of plagioclase, showing a uniformly better or poorer development, or any other difference in character indicating the presence of two kinds of lime-soda feldspar, was observed. The plagioclase sections almost invariably show polysynthetic twinning according to the albite law, with twinning lamelhie which vary fi'om very thin to moderately thick plates, the thinner being the more connnon form. Very common is the combination of the albite and Carlsbad twinning laws in one individual. Less commonly we find individuals twinned according to the pericline as well as the albite law, and sometimes a Carlsbad twin is made up of individuals twinned according to the albite and pericline laws. In determining the character of the feldspar, the Levy method was followed.' A great number of measurements made on the zone perpen- dicular to 010 gave equal extinction angles, varying chiefly around 15 degrees, but running as a maximum to 19 degrees. From this it appears that the plagioclase is andesine, probably a somewhat basic kind. That these andesines vary slightly in composition is shown by a very slight but noticeable zonal structure, the moi-e basic character of the center of the individuals being most admirably brought out by the more advanced con- dition of alteration of the center as compared with the periphery. The andesine is for the most part very much altered, to such an extent ~ that in many sections the boundaries of the twinning lamell:3e are so blurred that measurements are rendered impossible. Muscovite, which appears in minute rectangular sections showing good cleavage, is the chief secondary product from the feldspar, with epidote-zoisite next in importance. Calcite and biotite are present, but in comparatively small quantities. In some cases muscovite almost replaces the feldspar; in others epidote-zoisite does so. In such a case one sees in the center of- the feldspar only a mass of secondary mineral. As the examination is carried from the center toward the outside, the original feldspar material is distinguished as a thin film between the secondary minerals. This increases in mass until we reach the ontside narrow rim of practically unaltered feldspar. orthociase. — This is prescut in large quantity in irregular plates which form a part of the mesostasis for the plagioclase and bisilicates. Less com- monly we find it in micropegmatitic intergrowth with the quartz. It is ' fitude sur la determination des feldspaths, by A. Michel Lf'vy, Paris, 1894. DIORITE INTRFSIVES. 225 iiivariahly more or less (IcconiiKiscd, and shows innuiiicralile mimite dark specks scattered through it. The (juantity (if orthocdase varies in tliese dioritic rocks consideral)ly; at times it almost equals or even exceeds the plagioclase, when the rocks apjjroacdi the granites, and at times it sinks to a few large plates in each section, when the rocks are a normal diorite. Microciine. — Thig mineral is not abundant. It is in individuals which frecpiently, though not in all cases, are automorphic with re.spect to the orthoclase and (piartz. It is remarkably fresh. Quartz. — Quartz, at times, is an essential constituent, and again it dimin- ishes in amount until it is present only in a few grains, or even disappears altogether. Like the orthoclase, it is completely xenomorphic, and with the orthoclase constitutes the mesostasis. Undulatory extinction in the quartz gives indication of slight pressure. Biotite. — The original bit)tite in the granular dioritic rocks is automorphic with respect to all minerals but the hornblende. In the ophitic forms it has a development about equal to that of the hornblende. It shows a dark rich chocolate-browu or greenish-brown color for h and C, and a lio-ht yellowish-brown for a. The biotite includes small ei)idote crystals with })leochroic courts and some grains of sphene. Both of tliese are original. The biotite is almost invariably more or less altered, lileaching in s(ime cases to a very light colored mica with exceedingly high polarization colors. This bleaching follows along the laminae of the biotite and results in givino- sections parallel to the vertical axis a banded appearance resembling parallel intergrowths of muscovite and biotite lamiiuie. More commonly it alters to chlorite, rutile (often present as sagenite), sphene, epiclote-zoisite, and calcite. There is also a distinct banding of the biotite and the chlorite in places. In the alteration of the biotite we very commonly find lenses of calcite produced between the lamiure. In some cases the epidote-zoisite is clearly a product of the alteration of the biotite, for in many cases it is found in the rectangular shape of the biotite section, and in other instances in spaces between the feldspars in the ophitic rocks, which in fresher specimens are found to be occupied by the biotite. Moreover, in the epidote-zoisite are minute grains of sphene similar to those contained in the original biotite. Where it is present as a secondary product, it occurs with the musco- vite and is xenomorphic with respect to it. The green tone is absent from the secondary biotite. MON xxxvi 15 226 THE CRYSTAL FALLS IRON-BEARING DISTRICT. Hornblende. — Tlic lioriibleude in the diorites shows a most excellent devel- opment ill the prism zone; very nmch less well developed are the termi- nating planes. The color varies from dirty green to a reddish-brown. The brown hornblende occupies the center of the crystals, while the green occu- pies the outside, the green agreeing perfectly, optically, with the brt)wn. A zonal structure is indicated bv the difference in the character of the horn- blende, though the zones are not sharply delimited, but grade into one another. In a few cases the greenish hornbleude grades into one which is almost colorless. The pleochroism is as follows: Brown hornblende : a, light yellow or light reddish-yellow; b, light reddish-brown; C, darker reddish- brown. Green hornblende: a, light yellow; Hj, bright green; c, dull or olive green. This green hornblende is clearly original and not to be con- sidered as a secondary product after the brown hornblende. Both kinds are free from inclusions. Accessory minerals, — Tlic cpidote Is observed very frequently inclosed in the altered biotite, and is surrounded by pleochroic halos. In such cases it is considered a primary constituent. The accessory minerals, apatite, sphene, and zircon, show none other than their usual characters. Titaniferous magnetite is present in the diorites in very small quantity. According to the relative proportion of the important minerals just, described — plagioclase, orthoclase, quartz, hornblende, and biotite — com- posing the diorites, we get the following- varieties: Mica-diorite, quartz- mica-diorite, quartz-diorite, and tonalite. These grade into one another, as stated above; and, as will be shown later, certain of them grade into o-ranites. On account of these variations these dioritic rocks are especially interesting. DESCRIPTION OF INTERESTING VARIATIONS. SEC. 15, T. 42 N., R. 31 W. A dike of rock 4 feet wide, occurring at 425 paces N., 1050 W., sec. 15, T. 42 N , R. 31 W., near Norway portage, shows the following min- eraloo-ical variation. A specimen taken from the center of the dike shows the rock to be there a typical fine-grained granitite with little or no plagioclase. (Photomicrograph, fig. A, PI. XXXIX.) Along the sides the dike rock is a mica-diorite consisting of mica and plagioclase without any (piartz. Measm-ements on zone perpendicular to 010 gave equal extinction angles DIORITE INTRUSIVES. 227 with ii luiixiiiuiiu (it" If) (legrties. Only oiw kind of plajiioclase i« distin- •fuisliable by its mode of development, and this is rich in Ca( ), as shown by its alteration products. The feldspar rang'es at most from all)ite to andesine. No chemical anahsis has been obtained of either the "ranitite or flic mica- diorite phase, but the mineralogical composition is sufficiently marked to show conclusively that we have here a gradation from a granitic to a dioritic rock rich in CaO. The idea has been suggested by Johnston-Lavis ^ that in some cases the variation in chemical composition of intrusive rocks, especiall}' where this variation is one between the center and the peripher^- of an intrusive mass, may l)e due to resorption by the intrusive of parts of the rock intruded. The sharp line of demarcation which exists between the dike and the intruded hornblende-gabbro in the occurrence described above seems to preclude the possil)ilitA' of a fusion and mingling of the two rocks. ACROSS RIVER FROM CRYSTAL FALLS. Near Crystal Falls, just across the river from the town, are a number of small knobs of granite grading into quartz-mica-diorite. They are medium-grained rocks, reddish to gra}- in color. They take a very fine polish and are well adapted to ornamental stonework, as is shown by the columns made from them which are used in the court-house at Crystal Falls. When examined under the microscope, the rocks are found to con- sist of autoinorphic biotite and plagioclase, with xenomorphic orthoclase and quartz, these last forming the cement. Some of the slides show beautiful micropegmatitic intergrovvths of quartz and feldsjjar. The amount of quartz, plagioclase, and orthoclase varies so that, depending upon the specimen examined, one would call the rocks forming the knobs granite or quartz-mica-diorite. Most coimnonly the rock is a plagioclase-bearing' granite. No analysis has been obtained of the granite, but it is confidently believed that the chemical composition would sustain the microscopical diagnosis. Within the granite there are found lenticular schlieren of considerably darker color than the main mass, in which the plagioclase is the preponderant feldspathic constituent. The rock of these lenses is essentially a quartz-mica-diorite. ' The basic eruptive rocks of Grau (Norway) and their interpretation ; a criticism by H. .1. Johnston-Lavis: Geol. Mag., 4th ser., VoL I, 1894, p. 252. The causes of variation in the composition of igneous roclis, by H. J. Johnston-Lavis : Natural Science, Vol. IV, 1894, pp. 134-140. 228 THE CEYSTAL FALLS lEOI^-BEARma DISTRICT. These knobs are cut bv a nuinl^er of small dikes from a fraction of au incli to 3 inches in width. In all of these dikes the rock shows the same characters. It is very light gray Xo pink in color, and aphanitic. An examination under the microscope enables the separation of each dike into a verv compact fine-grained saalband and a somewhat coarser- grained porphyritic central portion. In the central part of the dike pheno- crvsts of quartz, feldspar, and biotite lie in a very tine groundmass of quartz and feldspar. Tlie texture of this groundmass is microgranitic. The saalband is composed of the microgranitic groun(hnass without the phenocrysts. The (juartz phenocrysts show the usual characters. The feldspar phenocrysts are in most cases so completely replaced by a musco- vite aggregate as to preclude any exact determination of their original character. In some cases indistinct remains of poly synthetic twinning are seen. Even when the main mass of the feldspar phenocrysts is entirely altered, there is a narrow zone of very fresh feldspar material surrounding it. Twinning in the center is also continuous through this zone. More- over, this zone itself shows a very noticeable zonal structure by the change in extinction angle observed in passing from the inner to the outer portion. This less altered zone of feldspar contains numerous inclusions of quartz from the groundmass. The character of the feldspar phenocrysts could not be determined, but the presumption is that they are of the same character as the feldspar in the coarser main mass — that is, andesine — with a more acid feldspar, possibly oligoclase, surrounding them. The further presump- tion is tlien wan-anted that the feldspar of the groundmass agrees with this outer feldspar zone in character — that it is also oligoclase, or at least is more acid than the phenocrysts. Automorphic biotite plates are now repre- sented bv chlorite pseudomorphs, with here and there some secondary epidote. The groundmass consists chiefly of quartz and feldspar, but contains disseminated through it many minute plates of white mica and a few crystals of zircon. The feldspar of the groundmass is too small to permit of its accurate determination. A plagioclase feldspar in sections indicating an approach to automorphism was observed. Its character as oligoclase (f), ( )r at least a feldspar of a more acid character than that of the centers of the ithenocrysts, is sunnised for reasons given above. Microcline in sec- tions showing characteristic twinning and in more or less rectangular out- lines was obsei-ved in considerable quantity. An untwinned feldspar was DIORITE INTKUSIVKS. 229 deterniined as ortlioclasc l)y tlic (lirtereiicci shown ))v its i-efractive index and tliat of the twiuucid plajriorlase. Quartz was also recognized in tliis way in the g-roundniass. The quartz and ortliorlase form tlie cement for tlie other constituents. The muscovite in tlie groundmass is ])resumably secondary, as is tliat in the phenocrysts. (Figs. A and />', I'l. XL.) The rock is here inserted as showing an exceedingly line grained j)or- l)hyritic form of the quartz-mica-diorite. It may compare to tliis mica- diorite as does the tonalite-porphyrite of Becke' to the tonahte described by Inm, and one miglit call it a quartz-mica-diorite-porphyry. No analysis of this rock has thus far l)een obtainable. Possibly its chemical composition may indicate it to be more closely allied to the true granites than is believed to he the case, judging from its mineralogical composition and its association with the rocks on the border line between granites and diorites. SOUTHEAST OF CRYSTAL FALLS. Southeast of Crystal Falls, in sec. 16, northwest of Lake Tobiu, and extending southwest into sec. 28, T. 42 N., R. 32 W., is a range of hills upon which are numerous exposures of a uniformly medium-grained rock. The main mass of the knobs is of tonalite, which shows several facies. A miarolitic texture is very common in this massif. The cavities are now filled with calcite, quartz, and epidote-zoisite alone or togetlier. This last mineral occurs in single large individuals or in tufts of individuals, which radiate from one side of the cavity. Li one case a cavity incompletely filled by such a tuft has been completely filled by a later infiltration of quartz. The color of the rock varies from light pink to very dark greenish gray. The areas of the lighter-colored rocks may be seen extending in finger-like projections into the darker-colored phases. There are, however, no sharp lines between these varieties, but a gradual passage from the lio-hter to the darker rock. These diff"erent phases evidently belong to a single rock mass. Under the microscope, however, important variations in the textural and mineralogical character of the rock masses are seen. The main mass of the rock is granular tonalite. The essential constituents are plagioclase, orthoclase, quartz, hornblende, and mica. The most common association of minerals is hornblende and mica in automorphic crystals, ' Petrographische stndien am Tonalit der Rieserferner, by F. Hecke: Tschermak.s mineral Mittheil., Vol. XIII, 1892, p. 435. 230 THE CRYSTAL FALLS IRON-BEARING DISTRICT. witli plag-Ioclase somewhat less well developed. Between these there is found the quartz, with some accessory orthoclase, and microcline as the last products of crystallization. In some cases these two minerals are present in micropegmatitic interg-rowths. A textural variation, which the facies mentioned below also undergo, is from a granular to an im]:)ei'fectly ophitic texture. In such cases the order of crystallization of the hornblende- mica and plagioclase is reversed, the plagioclase being most automorphic in the ophitic varieties. The rock resembles the tonalite described by Becke from the Rieser- ferner.^ It also closely resembles some slides of the typical Adamello tonalite with which I have been able to compare it. The chief difference between them is that the plagioclase and hornblende have a better crystallographic development in the Crystal Falls rocks than in the Adamello tonalite, and that the accessory allanite of the Adamello rock is wanting in the Crystal Falls tonalite, though the normal epidote may represent it. The horn- blende also differs slightly from that of the Adamello rock in that it is not throughout reddish brown. Tlie central portion of some of the crystals shows this color, but the outer portion is a dirty green, even grading into an almost white hornblende. The tonalite grades, by diminution of biotite, with corresponding increase of hornblende, into a quartz-diorite, and by diminution or disap- pearance of the hornblende and increase of the biotite into a quartz-mi ca-diorite. Hornblende never occurs alone in the rocks, whereas biotite may occur as the only l^isilicate C(instituent. It is a very common thing to find in the diorites rounded basic segregations consisting chiefly of mica with hornblende suljordinate and just a little accessory feldspar. When the orthoclase and quartz diminish, we get the mica-diorites. Orthoclase is always present in all of these dioritic rocks. In certain facies orthoclase and quartz are very abundant and the plagioclase is correspondingly dimin- ished. Such rocks are clearly plagioclase-bearing granites, and represent gradations between the ordinary tonalite and granite, and point to close relationship of this occurrence with the occurrences nearer Crystal Falls alread}' described, in which the granitic facies predominates and the dioritic facies is subordinate. ' Petrographlsche studien am Tonal it iler Rieserferner, by F. Becke: Tschermaks mineral. Mitt- heil., Vol. XIII, 1892, pp. 364-379. BIORITE INTKUSIVES. 231 Similar "Tiuliitions have l)eeu noted \)y lierke in tlic tonalito Iruni the Riesei-t'einu'i-.' Tlie diorite massif (if the Crystal Falls district seems to cor- respond very closely to the granodiorite masses of Becke, Turner, and Lind- gren,- which on the one hand grade into the granitites and a. Green hornblende: a, greenish- yellow; Jjs, yellowish- or brownish-green; c, dull olive green, frequently with bluish tinge; C>I)>-a. This hornblende, with respect to its rather exceptional pleochroi.sm and its general characters, seems to agree very well with that described b}^ van Horn from ver}^ similar rocks from Italy, and, like that, is probably a very basic hornblende.^ Twinning parallel to 100, co P oo, is very common. An imperfect parting parallel to the orthopinacoid 100, oo P oc, was also observed in some cases. It is also indicated by the jjlaty inclusions which lie in this plane In some of the sections where the green hornblende is not intergrown with the brown the green kind shows very commonly a system of fine striations parallel to the positive orthodome 101, P oo. In rare cases the brown hornblende is intergrown with almost colorless hornblende, one end of a crystal being brown, the other faintly yellowish. Irregular mottled intergrowths of the two were also found. The normal brown-green hornblende is rendered poikilitic in some specimens by a few rounded grains of perfectly fresh pyroxene, and also by plagioclase crystals which it includes. This same kind of hornblende is frequently rendered very dark by the number of exceedingly small inclu- sions which it contains, and in this, and also in its reddish-brown color, resembles so strongly many hypersthenes as to be readily mistaken for them upon cursory examination. These inchisions are of several kinds, all dis- tributed throughout the same individuals. It is impossible in studying them ' Petrographische Untersuchungcn fiber die noritischen Gesteine der Umgegend von Ivrea in Oberitalien, by F. R. van Horn: Tschermaks mineral Mittheii., Vol. XVII, 1897, Paxt V, p. s.J9. 236 THE CRYSTAL FALLS IRON-BEARING DISTRICT. to get any optical tests, except that of extinction, owing to the minute size of the inchisions and to the fact that where large enough for exaniinatiou the tests were vitiated b^s' the presence of the hornblende. Of these inclusions some are readily distinguishable as rutile. Some of the larger of the crystals reach a leng-th of 0.04.5 nun. and a thickness of 0.0125 mm. Numbers of them show the characteristic heart-shaped and geniculated twins of rutile, so that there is no doubt as to the determi- nation. Associated with the rutile are other crystals 0.019 mm. long, which show the typical 2)ointed pyramidal development of octahedrite (anatase). Still others show a flat tabular development somewhat similar to that of brookite, though these could not be jjositively determined as that mineral. The hexagonal plates of clove-brown color so frequent in hornblende and also in hypersthene occur also in this hornblende. They are believed to be micaceous ilmenite. The thin ^ilates are translucent, thicker ones are less so, and those which are still thicker are opaque and metallic. The thin plates appear when on edge as fine, hairlike streaks. The thick ones appear in the same position as more or less rectangular bars or rods. Often these small plates are associated with masses of iron oxide, also included in the hornblende. This iron ore occurs in the plates and bars characteristic for ilmenite. These ilmenite masses are translucent only on the edges, where the slide has cut the mass in such a manner as i o give an exceedingly thin section of the ore. At such places the ore is translucent with the same brown color as the thin plates. Another rare variet}' of the inclu- sions occurs in round grains of rich green color, and may possibly be a spinel. In those sections in which both original brown and original green hornblende occur the inclusions are confined to the brown kind. Where the brown kind is surrounded by the green hornblende the inclusions grad- ually diminish in quantity as we approach the green zone. With this goes also, hand in hand, a lightening of the color of the including mineral (brown hornblende), and there is thus an imperceptible change from the brown to the green hornblende. Where the green hornblende occu.rs alone it is frequently as full of inclusions as is the brown hornblende of other sections. Individuals of the same sections difter from one another with respect to the quantity of the inclusions, some being crowded with them, while others are practically free from them. GABBRO AND NOKITE INTRUSIVES. 237 Thi.s lirowii li(.nil)lcU(lLs on alturatiou, breaks up into aggregates of ei)i(lote-zoisito and light-green chlorite. The second kind of hornblende is the perfectly fresh, compact, coni- nidu dark-green kind, with pleochroism varying from yellow for a, to yelloAvish-green for Jji, and to bluish-green for C; C>b>a. This is found in very few cases. It ap])ears in every instance to be a primary constituent. The third kind of hornblende may ht-. primary, although the evidence obtainable points to its secondary nature. It has a light-green color, and when examined for pleoclu-oism exhibits a scarcely noticeable change. This hornblende differs very nuich from the other two hornblendes described, in that it is not compact, but occurs in aggregates of coarse reed- like (schilfaehnliche) individuals. Such aggregates do not at all resemble uralite. The individuals are far too coarse and wedge out at short distances within the aggregates. The aggregates occupy irregularly shaped areas. The aggregates consequently have a coarse patchy polarization. They are frequently surrounded by ragged pieces of biotite, just as are the plates of compact hornblende. Moreover, they occur in rocks which show pressure effects, and are best developed in those in which such effects are most marked. The aggregates rather frequently occur with irregular pieces of the greenish or brown compact interposition-bearing hornblende bordering the aggregates or in the midst of them. The light-green reedy hornblende never contains such interpositions, but does have associated with it fairly large grains of rutile, which may perhaps be considered as having been derived from the various titanium-bearing microlites in the original brown hornblende. The general appearance of these aggregates and their asso- ciation with the original hornblende seem to point toward their secondary origin from the latter through the effects of pressure. Pyroxene. — Thc pyroxeue comprises both a monoclinic and an ortho- rhombic kind. These are the first of the bisilicates to crystallize in these o-abbros. The monochnic kind is of two varieties. The first is in colorless to faint-pink grains included in large plates of original brown hornblende. These grains have a well-developed prismatic cleavage. One basal section shows very nicely the characteristic pyroxene cleavage. The extinction measured against the prismatic cleavage reached as high as 50 degrees. This pyroxene is presumed to be augite. It never shows diallagic parting. In other sections the monoclinic pyroxene is a clear white to faintest 238 THE CRYSTAL FALLS IRON-BEARING DISTRICT. green malacolite or diopside. This is in roundish grains included in the original green hornblende, which it equals in quantity. The orthorliombic pyroxene occurs in individuals which show fairly good prismatic development, but with rounded terminal faces. The pris- matic cleavage is very well developed. A transverse parting is also com mon. The pyroxene is usually colorless, but in some cases a scarcely noticeable pleochroism was observed, varying from a faint-greenish tinge for rays ^'ibrating parallel to C, to a yellow for those parallel to a and h. It contains small, dark, streak-like intrusions, some of which under exceptionalh' favorable conditions are transparent,, with a faint-greenish tinge The exit of the bisectrix in basal sections, as Avell as the parallel extinction, renders it easily distinguishable from the monoclinic pyroxene. The optical angle could not be measured, but was clearly very large, as the hyperbolas passed completely out of the field of view. The orthorliombic pyroxene is evidently enstatite or bronzite, and the pleochroism clearly points to its position near bronzite. A few crystals of rutile and also some of the ilmenite plates, so common in hypersthene, were found occurring in the bronzite. The ilmenite plates are in rather rare irregular patches in the crystals. In many cases along the cleavage lines or around the edges of the crystals or along the transverse parting planes are narrow zones of a sec- ondary yellowish-green, finely fibrous, serpeutinous mineral. Beyond these zones is a pure white aggregate of secondary talc scales (Fig. B, PI. XXXVIII). Among these scales are a few minute rutile crystals, and also a few Ijlack ferruginous specks, these products being probably derived from the inclusions in the bronzite, and the ferruginous material possibly to some extent from the mineral itself In some cases, instead of the intermediate serpentine zone, the rather rare occurrence is observed of the passage of the bronzite directly into the talc aggregate. Olivine. — The determination of the original presence of olivine in the gabbroic rocks is based upon very slight evidence. In some of the sections containing augite almost every individual of this augite has near its center a rounded, very rarely irregular area of yellowish-green fibrous serpentine. These areas are sharply delimited from the surrounding pyroxene, and the conclusion seems warranted that it resulted from the alteration of some mineral included in the pyroxene. Tlie only bisilicate found in the rocks GABBliO AND NOltlTE INTKUSIVES. 239 of the district, which ciystalhzud before the jjyroxene is ohviiie. In the peridotite, to l)c described in the next section, this is usunllv surrounded bv nionochnic or orthorlioinl)ic pyroxene. This idtered niinend is not important in (punitity. Iron oxide. — Jlinouite aud titaniferous magnetite occur in some of tlie rocks in considerable quantity. Both alter to sphene and rutile. Apatite. — Among- the accessory minerals apatite is jjerhaps the most common, and, as usual, one of the very earliest minerals to crystallize. It is contained in all the essential constituents, and in biotite is surrounded by pleochroic hales. In some cases it has even crystallized before sphene. It is noticeable in some sections that great numbers of apatite crystals are arranged along lines representing sections of planes between the jjlagioclase plates, thus practically outlining the feldspar individuals. Sphene. — In many cases in these gabbros sphene is found contained in some of the freshest rocks as an original accessory constituent. It is present in largest quantity in the very finest-grained gabbros, which «how a parallel texture. In these rocks sphene in some cases surrounds an iron ore, which, to judge from the rod-like sections which are so common, is ilmenite. One might be led to think that the sphene was secondary in such cases, but the iron ore is perfectl}' fresh, and, considering that in the same thin section crystals of apatite are also surrounded by sphene, it seems clear that we may consider such sphene as original. It thus appears that a portion of the titanium oxide combined with the iron oxide first, forming the titanic iron ore. This was followed by the crystallization of the calcium-titanium compound, thus giving the sphene. In these rocks sphene is not in crystals, but in grains. These grains are arranged in long chains hing between the other mineral constituents and with the long direction of the individual grains, as well as of the lines of grains, parallel to the long directions of the other constituents of the rock. Zircon and rutile. — Zircou is iu vcry small quantity. Rutile shows its usual characters. It is most commonly associated with the octahedrite (anatase) and brookite (?) as inclusion in the hornblende. The iron oxide is chiefiy present as ilmenite, with some titanic magnetite. The secondary minerals have alread}^ been mentioned and their char- acters described under tlie description of the minerals from which they are derived. 240 THE CRYSTAL FALLS IRON-BEARING DISTRICT. DESCRIPTION OF INTERESTING KINDS OF GABBRO. The minerals described above as the leading essential constituents of the rocks to be described may be combined in varying quantities. Accord- ing to these combinations a number of different mineralogical types of rock may be produced. The wide range in mineral composition of the gabbroic rocks is equaled, if not surpassed, by similar variations noted by Fairbanks in certain rocks from Point Morrito, California.' It may cause the further description to be more readily understood if we preface it by the statement that all of these types, however, are simple facias of a single magma. The important phases which will be described are, in the order of their impor- tance, hornblende-gabbro, consisting essentially of liornblende and lab- radorite; gabbro, consisting of monoclinic pyroxene and labradorite, and bronzite-norite, consisting essentially of bronzite and labradorite. The various mineralogical types exhibit very interesting ranges in texture in certain cases, t-o which attention will be called. HORNBLENDB-GABBRO IN SEC. 1.5, T. i'2 N., R. 31 W. A hornblende-gabbro forms a large knob in sec. 15, T. 42 N., R. 31 W., just at the foot of the Norway Rapids, on the west bank of the Michigamme River. This exposure shows very prettily a change in texture. The change in texture is also accompanied by a slight mineralogical change. The knob is composed partlv of a fine-grained granular, but more largely of a coarse- grained porphyritic, gabbro. The fine-grained portion is a pure gabbro composed of plagiocla^e and brown hornblende, with very little brown mica. No quartz was observed, nor was any orthoclase definitely deter- mined. The plagioclase is in fairly well-de 'eloped automorphie plates. The hornblende is the brown variety, with numerous minute inclusions, which has already been described, and is not always so well developed as is the plagioclase. In places it plays rather more the role of a cement. This relation of the two minerals results in forming an imperfect ophitic texture in places, though on the whole the two minerals are about equally developed, and produce a granular structure. (Fig. ^, PI XLIV.) ' The geology of Point Sal, by H. W. Fairbanks : Bull. Dept. Geol., Univ. California, Vol. II, 1896, p. 56 et seq. GABBRO AND NORITE INTRUSIVES. 241 The coarsse-iiTniiU'd })()r])liyriti('. gabbro foriiiing- tlie greater ])art of the knol) consists of phigioclase, liornbleiide, biotite, and iron oxide, witli a very small amount of pyroxene. The hornblende occurs in phenocrysts which have irregular rounded shaj)es instead of being well crystallized. Some of the largest phenocrysts have a diameter of slighth^ less than 1 centimeter. They are poikilitic, rendered so by inclusions of lath-shaped plagioclase and rounded grains of pyroxene. (Photomicrograj)h, fig. A, PI. XLI.) This por- phyritic liornblende is a dark reddish-brown -variety containing such great numbers of minute inclusions as to be opaque in many places, which grades over into, and is in many places in optical continuity with, a dirty green hornblende. This green hornblende is in anhedra and forms the cement for the feldspar, and the two together the groundmass for the brown horn- blende phenocrysts. Tlie plagioclase is most commonly in broad, well- developed crystals, which frequently give quadratic sections. Some few grains of a pink monoclinic pyroxene are included by the hornblende. SECS. 15, 22, 28, AND 29, x. 42 n., b. 31 w. Exposures of a hornblende-gabbro with interesting facies associated with it occur in the southeastern corner of sec. 15, at tlie southeastern corner of sec. 22, extending east and west through the northern part of sec. 28, at the southeastern corner of sec. 28, and on the west bank of the Micliigamme River in sec. 29, T. 42 N., R. 31 W., at the location N. 100, W. 1,250 paces. This is medium to coarse grained and of a gray color from a short distance. Examined at moderately close quarters, one distinguishes very readily a milky white feldspar and a black or dark-green hornblende in about equal quantities. The microscopical examination adds to these two minerals in a very subordinate quantity biotite, pyroxene, and orthoclase. The labra- dorite plagioclase is in medium-broad, irregular plates, though at times approaching a very distinctly lath-shaped form. The orthoclase is present in a few rare individuals in the form of irregular plates. The hornblende constituent is in irregular plates and varies in character. It may be the brown or the green variety ah-eady described, or the two together in sepa- rate individuals, or even the brown grading into the green. This green is original and not the alteration product of the brown. Biotite is the normal reddish-brown kind in irregular plates. The pyroxene is usually absent MON XXXTI 16 242 THE CRYSTAL FALLS IRON-BEARING DISTRICT. from the sections of these rocks, but when present it is ver}- rare, and occurs in small irregular grains not uncommonly intergrown with the hornblende and evidently older than the hornblende. It is light green in color, 'with a scarcely noticeable pleoclu-oism. Its monoclinic character was readily deter- minable; but a more exact determination was not made. It does not, how- ever, show diallagic parting, and is diagnosed as possibly diopside. The feldspar shows the best development of the accessory minerals. It can rarely, however, be said to be automorphic. The texture is, on the whole, granular. From a mineralogical study of the rocks alone, one would unhesitat- ingly place them with the diorites, especially if those facies were seen in which the pyroxenic constituent was wanting. The following analysis (Sp. 23354), obtained from Mr. George Steiger, of the United States Geological Survey, shows the chemical composition of one of these rocks : Analysis of hornhlendegahhro. SiO, . TiO, . Al.Oa FeiOa FeO . MnO. CaO . MgO. Per cent. 49.80 .79 19.96 6.32 .49 11.33 7.05 K,0 NajO H.O 100^— H2O 100^-f, P2O5 CO2 Total Per cent. .61 2.22 .13 1.71 .07 .15 100.63 An examination of the analysis shows that the microscopical determi- nation of the rock as a diorite would be incorrect if we accept, as has been done in the preceding pages, Brogger's characterization of the diorite and gabbro families.^ The rock analyzed is hornblende-gabbro, as shown by the relatively high content in the characteristic alkaline earths, esj^ecially magnesia, which usually appears in inverse proportion to the silica, and in the low percentage of alkalies. ' Op. cit., Part II, pp. 35, 39. c GABBRO AND NOIUTE INTKUSIVEB. 243 HOUNBLKNDK-CSABUKO DUCES. One (tt" tlic exposures of the above-descTibod. liornlilende-g-abln-o — tlie one on the west bank of the Michigamnie River in section 29— is very interesting on aeeount of at least two different series of dikes which cut it. The coarse-grained liornblende-gabbro forms the main mass of the knob. The dike rocks may be divided into the fine-grained hornblende-gabbro forming the earhest dikes and the coarse-grained bronzite-norite forming the latest dike. Some of the specimens of the fine-grained rock are massive, but the o-reater portion possess a distinctly parallel texture. These are distinctly micaceous. The rock of the dikes has in general very much the macro- scopical appearance of a biotite-mica-schist. The dikes are very narrow, never more than 18 inches wide. The larger ones send off branches, and in places inclose angular pieces of the coarse diorite. Thus the relation of this fine-grained rock to the main gabbro mass is perfectly clear, though in places it so closely resembles, macroscopically, pieces of mica-schist that in spite of the branching of these dikes, indicating their intrusive nature, they were supposed by one observer to be curiously shaped stringers of the mica-schists included in the main mass of the gabbro. The notes do not indicate from just what portions of the dike the specimens were taken; hence it is impossible to state positively that the more schistose parts are nearest the edges and the more granular portions nearer the center, as one would naturally expect. However, in all but one of the specimens which show a contact between the dikes and gabbro which they penetrate, the rock nearest the contact shows a parallel struc- ture. Hence it may be stated that in some cases the edges of the dikes are the more schistose portions. The one specimen referred to above is granular and much finer grained along the contact than farther from it. The microscope shows the rock of these dikes to be a gabbro differing little in character from the main mass. The plagioclase is well developed in rectangular, more or less lath-shaped crystals. Mica of a rich brown color is rather more abundant than usual, and is in about equal quantity with the hornblende. The hornblende is brown or of a dirty greenish color, containing the inclusions mentioned in the detail descriptions of the minerals of the gabVjros. Some irregular grains of a light-green augite (diopside) were observed in the sections. Orthoclase (?) in grains is present simply 244 THE CEYSTAL FALLS lEON-BEAElNG DISTEICT. as an accessory. Ilmeuite occurs in irregular masses in rod-shaped pieces: Sphene is present in original grains and also as a secondary ])roduct after ilnienite. Apatite occurs and is sometimes included in the sphene. The most striking feature of the rock is its textural variation. Some of the sections show very good granular texture; others have a fair ophitic texture ; the most common is a striking parallel texture which macroscopic- ally gives to the rock a schistose appearance. This may be seen even in the same section with the ophitic texture, the two grading into each other. The parallel texture is occasioned by tlie arrangement in a common direc- tion of the long diameters of nearly all of the minerals (figs. A and B, PI. XLII). The grains of sphene often lie in long trains agreeing with this general parallelism. One's first idea would probably be that the texture was due to the cause which has produced parallel structures in most other ancient rocks — pressm'e. However, it can not be referred to this, as the minerals — with some individual exceptions — show slight or no pressure efi"ects. Apparently the only explanation borne out by the facts in this case is that we have to do with a fluidal texture, produced by the movement in the magma consequent upon its injection along the fissures in the gabbro, the parallelism of the minerals agreeing with the bounding sides of the fissure. BRONZITE-NORITE DIKE. The main hornblende-gabbro and the fine-grained dikes just described are cut by a dike, about 3 feet wide, of coarse rock which resembles that forming the main mass of the knobs in every way except that it contains a very much altered orthorhombic pyroxene (bronzite) in greater quantity than the hornblende. The rock is a very pure kind of bronzite-norite (fig. B, PI. XLIV). The following analysis (No. 1, Sp. 23755), by Mr. George Steiger, shows the gabbro affinities of the rock. The high percentage of magnesia gives a clear indication of the important role played by the bron- zite in the composition of the rock. With the analysis of the bronzite-norite there is placed for comparison an analysis (No. 2) of a norite from Ivrea, Italy, which is essentially the same as the above in mineralogical as well as chemical composition. In the Italian rock hypersthene, instead of bronzite, is the chief pyroxenic constituent.^ ' Petrographische Uutersuchiingen iibcr die uoritischeu Gesteine der Umgegcnd von Ivrea in Oberitalien, by F. E. van Horn: Tschermaks mineral, Mtttbeil., Vol. XVII, 1897, Part V, p. 404. GABBKO AND NORITE INTRUSIVES. Andh/scs of iiorites. 245 SiO, TiOj ALO, Fe,0., FeO MnO CaO MgO KaO Na^O H;0 100°—. HiO 100° + . PjOa CO, 1. 48.23 1.00 18.26 1.26 6.10 Total 9. .39 10.84 .73 1.34 .26 2.00 .07 .43 99.91 49.95 .69 19.17 4.72 6.71 Trace. 9.61 .5.03 .74 3.13 .09 Trace. 9.84 SEC. 29, T. 42 N., R. 31 W., 1,200 N., 200 W. On the east bank of the Michigamme River at 1,200 paces N., 200 W., sec. 29, T. 42 N., R. 31 W., there is an outcrop which shows even better than the occurrence just described the variation in mineralogical character and the relative ages of these varieties. At tliis place there is a knob com- posed of hornblende-gabbro essentially like the general type described as typical for this district. This knol) is cut by a rock which is coarser in grain and a trifle darker than the variety which it intrudes. Examined under the microscope, it is seen to differ from the normal phase also in min- eralogical composition, and resembles rather closely in this respect the porphp-itic variety described on p 240. Like that, the hornblende is reddish- brown, containing a large number of inclusions and grading over into the green hornblende. This hornblende includes rounded grains of a white to pinkish monoclinic pyroxene in sufficient quantity to be characteristic. The pyroxene is never automorphic, as one would perhaps expect to find it, though it was evidently one of the first minerals to crystallize. Contained in the pyroxene individuals are rounded areas of yellovvish-greeu serpentine. These areas are sharply outlined from the pyroxene and do not appear to be the result of its alteration; cousequenth' the conclusion is reached that the 246 THE CRYSTAL FALLS IROX-BBARING DISTRICT. serpentine results from the alteration of a mineral older than the pyroxene. Most 'irobably this mineral was olivine, though no positive statement to that effect can be made. This pyroxene I'ock also contains an exceedingly large quantity of apatite. No analysis was obtained of this facies, liut the microscopical characters enable us to place it as a gabbro (possibly olivine- bearing) and to consider it, like the bronzite-norite, as a facies of the pre- dominant hornblende-gabbro. This same exposure of gabbro is cut by a coarse peridotite (wehrlite), a description of which will be found on p. 253. In this peridotite there is found a narrow strip of rock, about 2 inches wide, which is presumed to be either an inclusion or a yerj narrow dike in the peridotite. The exposure does not admit of its relations being determined more accurately. The presumption is that it is a dike. Whether an inclusion or a dike, it is younger than the massive hornblende-gabbro forming the main exposure. In this respect it corresponds to the gabbro just described (p. 243) as cutting the normal gabbro. This dike rock is macroscopically a fine-grained, granular, dark-gray rock. The microscope shows it to be very fresh, porphyritic in texture, and composed of phenocrysts of bronzite lying in a finely granular ground- mass of plagioclase, hornblende, and orthorhombic and monoclinic pyroxene. No quartz whatever was found associated with these minerals. The pla- gioclase is the usual kind, labradorite. Some unstriated feldspar, possibly orthoclase, was also observed, though in very small quantity. The bronzite is in narrow prisms which reach a length of 1.23 mm. Commonly they have well-developed transverse partings. It is worthy of note that a few of the crystals contain the brownish platy inclusions so common in hyper- sthene. There is, however, no relation between the color and pleochroism of the mineral substance and these brownish plates. The bronzite is very clear, with weak color, showing a scarcely distinguishable greenish tinge for the rays vibrating parallel to C, with yellowish for the rays parallel to a and I). In one case the bronzite was seen altering to a greenish-yellow fibrous aggregate of serpentine. In the groundmass this same orthorliombic pyroxene is represented by irregular grains. The hornblende is the usual reddish-brown kind, but differs frOm that seen in the other gabbros of this district in that it contains ilmenite inclusions only, without any rutile oi* anatase. It is in anhedra. A faint-greenish pyroxene occurs in irregular p GA15BKO AND NOHITR INTUUSIVES. 247 xenoniorpliic individiuils, ioi'iiiiiiy a soiiu'wliat smaller proportion of the "Touiuliiiass than does the hornblende, l)ut i.s more abundant than the bronzite of the groixndmass. Sphene is pi-esent in considerable (quantity, and likewise ilmenite (fig'. A, PI. XLV). This rock stands between the hornblonde-gabbros and the norites. In texture it might be compared with the norite-porphyrite (enstatite-porphyrite of Ivosenbusch) with holocrystalline groundmass. But it differs essentially from this rock in the presence of a large quantity of hornblende. This hornblende connects it with tlie hornblende-gabbros, from which its content of pyroxene, both orthorhombic and monoclinic, tends to separate it. Owing to its obvious relation to the bronzite-norites, which, like it, occur as differ- entiation facies of and dikes in the hornblende-gabbro, I shall call it a "bronzite-norite-porphyry," using the term "porphyry" purely in a textural sense. The rocks may be compared, in their variation, to those described by G. H. Williams from Maryland,' by Chester from Delaware," and by Fair- banks from California.' A series of basic rocks similar in many respects to those of Crystal Falls has also recently been described in two interest- ing papers by Van Horn * and Schaefer.^ DYNAMICALLY ALTERED GABBRO. Near the junction of sees. 26, 27, 33, 34, T. 43 N., R. 31 W., there is a large gabbro mass which shows raai-ked evidence of dynamic action, and may well be cited as an example of a metamorphosed gabbro, or, perhaps more clearly, as a rock intermediate between a hornblende-gabbro and a hornblende-gneiss. None of tlie gabbros thus far described show any evi- dence of having taken part in very extensive orogeuic movements. The minerals of but few of them show more than the common phenomenon of slight wavy extinction. Hence it is clear that their intrusion took place 'The gabbros and associated hornblende rocks occurring near Baltimore, by G. H. Williams: Bull. U. S. Geol. Survey No. 28, 1886 ; Oiitliue of geoIo;,ry of Maryland, Baltimore, 1893, p. 39. -The gabbros and associated rotks in Delaware, by F. D. Chester: Bull. U. S. Geol. Survey No. 59, 1890. ^The geology of Point Sal, by H. W. Fairbanks: Bull. Dcp. Univ., California, Vol. XI, 1896, p. 56 et seq. ^Petrographische Uutersuchungen iiber die noritischeu Gesteine der Umgegend von Ivrea im Oberitalieu, by F. R. van Horn: Tscliermaks mineral., Mittheil., Vol. XVII, 1897, Part V, pp. 391-420. ^ Der basische Gesteinzug von Ivrea im Geliiet desMastalloue-Thals, by K. W. Schaefer: Tscher- maks mineral., Mittheil., Vol. XVII, 1898, Part VI, pp. 495-517. 248 THE CRYSTAL FALLS IRON-BEARING DISTRICT. after the occun-ence of the inountaiu-building inovemeuts to which the intruded rocks have been subjected. It is not beheved that the mass referred to in this description is an exception to the general rule, but that its metamorphosed condition is referable to local causes, such as have pro- duced the shearing planes to which allusion has been made (p. 234). It may perl laps occur near or along some minor fault plane, though no indications of a fault have been observed. The outcrop is composed in part of massive gabbro and in part of the metamorphosed kind. The gabbro in its typical massive form (fig. A, PI. XLIII) is a medium to coarse grained granular rock, composed essentially of plagioclase and dirty brown-green compact liornblende, the latter being quite full of the inclusions mentioned above as commonly occurring in the hornblende of the rocks in this region. In the metamorphosed rock the grain is much finer, the rock possesses imperfect schistosity, and the color is a lighter green than in the original. The component minerals are chiefly hornblende and feldspar, with some quartz, chlorite, epidote, calcite, and rutile. The hornblende has a light-green color. It occurs mainly in aggre- gates of small irregular grains. In some cases these surround an angular nucleus of dirty brown-green original hornblende containing the same inclusions and in every way similar to that of the coarse-grained uncrushed rock. The light-green hornblende contains none of these minute inclusions and onlv here and there grains of rutile, which, if the diagnosis of the interpositions as titanium minerals is correct, may possibly be considered as havino- been derived from them. This hornblende is believed to be sec- ondarv. It is produced by a process of mechanical Ijreaking down of the original hoi'nblende, accompanied by recrystallization. The process is somewhat similar to the crushing of ordinary plagioclase and the produc- tion of a more acidic variety. As may be seen from the photomicrograph (fig. B, PI. XLIII) of a section, the feldspar exhibits signs of intense crushing. The twinning lamellse are strongly bent, and the pieces possess wavy extinction. Many, in fact most, of the feldspars are fractured. Wherever these fractures occur the feldspar along the edges of the fractured portions has been altered, producing secondary epidote, muscovite, plagioclase feld- spar, and quartz; the last, however, in small quantity. The biotite has been crushed. This has partly altered to chlorite, and the latter contains many grains of epidote. Some granular calcite and considerable iron GABBttO AND NOUITE INTRUSIVES. 249 pyritt's are also toiind in the altered rocks. These secondary products are easily distinguished from the original minerals by the total absence m them of wavy extinction. The effect of the crushing upon the texture has been to render it more or less schistose, the resulting rock resembling in its mineralogical character, and in texture, an ordinary honiblende-gneiss, or, when (piartz is present in but small quantity, a plagioclase amphibolite. RELATIVE AGES OF GABBROS. An attempt to determine the relations of the varieties described resulted in the establishment of the following order of intrusion: The typical coarse-grained hornlilende-gabbro was the first formed, and seems to be the most widely distributed. This was then intruded by dikes of gabbro, which contain both monoclinic pyroxene and hornblende in association and rep- resent a gradation to the normal gabbro. These hornblende-gabbros and normal gabbros were then cut by the dikes of bronzite-norite and bronzite- norite-porphyry. PERIDOTITES. Until recently all of the ultrabasic rocks were included by most petrographers under the family name of peridotite. ■ In the last edition of his Mikroskopische Physiographie, Rosenbusch has separated these rocks into the plutonic rocks and the volcanic rocks. The term " peridotite " is restricted to the first, and the last are called the "picrites." The charac- teristic differences between these two types are very clearly shown in the specimens from the Crystal Falls district which I have had the oppor- tunity of studying. The rocks here described agree with Rosenbusch's narrower definition of the peridotites.^ DISTRIBUTION, EXPOSURES, AND RELATIONS. The peridotite dikes were all found near the Michigamme River, in sees. 29 and 22, T. 42 N., R. 31 W. Typical wehrlite with very little green amphibole is found on the east bank of the Michigamme river, near the center of sec. 22, T. 42 N., R. 31 W. It shows no relations to the other rocks. The amphibole-peridotite, exhibiting marked variations to wehrlite and olivine-gabbro, forms an outcrop on the east bank of the Michigamme River in the NE. ^ sec. 29, T. 42 N., R. 31 W. This dike cuts the gabbro. ' Op. cit., p. 343. 250 THE CRYSTAL FALLS lEON-BBAKIXG DISTRICT. A rock to be described last, connecting the diorites, gabbros, and peridotites, was taken from near the northeast corner of sec. 22, T. 42, B. 31, where it is found cutting the gabbro. PETROGRAPHICAL CHARACTERS. The peridotites are very dark green to black coarse-grained rocks, showing in almost all cases a granular texture. In one case an excellent poikilitic texture was observed, and in another the same texture in a very imperfect condition was seen. The almost total absence of any pressure phenomena in these rocks excludes the idea of their having been subjected to any powerful dynamic action. The chief mineral constituents, arranged in order of relative amounts, are pyroxene (monoclinic and orthorhombic), olivine, hornblende, and biotite. The following minerals also occur associ- ated with these: Feldspar, apatite, green and brown spinel, and iron oxide. Pyroxene — Thcrc is prcscut in these rocks both orthorhombic and mono- clinic pyroxene. The orthorhombic has a 3'ellowish tinge, and contains a few tabular inclusions. Only a few grains of this pyroxene were found, but upon one section a figure was obtained and the positive character of the mineral was determined. It appears to be bronzite. The bronzite grains are found included in the hornblende. Its relations to other minerals are not shown in any of the thin sections. The monoclinic p\TOxene, though one of the earliest minerals to crys- tallize, is likewise in xenomorphic individuals, many of them twinned, some polvsynthetically. It is usually of a light yellowish color, and is then nonpleochroic; in some sections, however, it is sufficiently colored to show pleochroism from light yellowish to brownish. It includes a number of the clove-brown tabular interpositions like those in the hornblende, and these at times give the pieces a decided violet tinge. Less commonly than the tabular interpositions one observes green needles and laths, more rarely rounded grains or plates of larger size and having a brownish-green color. They are so minute as to defy positive determination, but are presumed to be hornblende microlites. The orthopinacoidal parting is in some cases well developed in the augite. This diallagic augite is in large quantity, and though usually inchided in the hornblende, nevertheless includes both hornblende and biotite in a few ragged plates. Such sections are pi'obably those which have passed tlu-ough tlie outer edges of the augite crystals. PERIDOTITK INTRTTSIVES. 251 Olivine. — 'riiis is present most commonly in round iuilit'dra, and is usually almost entirely altered to yellowish-green serpentine tihers. When the olivine is completely altered, the mesh structure affords a ready means of" recogniz- ing the original mineral, especially when taken in conjunction with grains of a rich-green isotropic mineral, and also octahedra and grains brown in color, probably spinels, which are found included in the ])seudomorphs. Hornblende. — Tlic luost of tlic hoHiblende in the peridotites is a brown variety showing strong pleochi'oism. a is light cream-yellow, c is yellowish brown, and h is reddish brown; tJ>>C>H. Patton ' has already called attention to the pleochroism of the hornblende, which "is exceptional, inasmuch as the brownest color is that of rays of light vibrating parallel to the orthodiagonal axis." The brown hornblende is accompanied by a very small quantity of green hornblende. Moreover, the brown hornblende grades over into a light green, the two being in perfect crystalline conti- nuity. In rare cases this brown hornblende is also intergrown with a light- green pyroxene in such way as to give a mottled polarization effect. The pinacoidal cleavage of the hornblende continues through the pyroxene, and the extinction angle of the hornblende against this cleavage is 19 degrees, while that of tlie pyroxene runs up to 34 degrees. The hornblende includes numerous anhedra of pyroxene, somewhat fewer grains of olivine, and, less commonly, ragged pieces of biotite, giving it a poikilitic character. It is also very full of opaque metallic or brownish translucent plates of ilmenite. In some individuals minute clear microlites, similar to those described above in the hornblende of certain gabbros, are noticed in small quantity. These are irregularly distributed in the hornblende, giving the crystals a patchy appearance. In respect to its color and these inclusions, this horn- blende, as noted by Patton, bears a rather striking resemblance to hyper- sthene on superficial examination.^ Iron ore in large masses is fairly frequent as an inclusion, and it is very commonly noticeable that where such inclusions occur the zone of hornblende immediately surrounding them is free from the platy inclusions mentioned above. Such a clear zone is also observed at times surrounding the inclusions of biotite and olivine, but never in case of pyroxene. Whei-e these clear zones surround the ' Microscopic study of some Michigan rocks, by H. V. Patton: Rept. State Board of Geol. Surv. for 1891-92, 1893, p. 186. ' Loc. cit., p. 186. 252 THE CliYSTAL FALLS IRON-BEARING DISTRICT. included biotite or olivine, these two minerals have associated with them numerous small grains of ore, which probably represent the iron that would have been incorporated in the sun-ounding hornblende but for some selective influence exerted by the olivine and biotite. Biotite. — This mineral is present in flakes of very irregular outline. The pleochroism varies from cream color to yellowish red or brown. Although one of the last minerals to crystallize, its crystallization began before that of the pyroxene or hornblende had entirely finished. Hence we find flakes of it included in these minerals, but near the edges of the crystals. The biotite itself is almost free from inclusions, containing only a little hematite and mag- netite. It alters to a brilliant green, strongly pleochroic, chloritic mineral. Feldspar. — Tliis is prGscut lu specimcus from two outcrops, and in these hardly reaches the rank of an essential constituent. It was the last mineral to crystallize, and is consequently in anhedra, forming the mesostasis. All of the feldspar sections were tested, but no determinative measurements could be made. It is probably very basic. Apatite. — Apatite is present in small quantity. It exhibits its usual characters. Spinel. — There is a spinel found in round grains which are included in the olivine (serpentine). It is g-reen in color, and is possibly pleonaste. A second spinel, probably picotite, occurs in small brown grains and octahedra in the olivine. caicite. — This mineral, derived partly, if not wholly, from the altering minerals, is found in lenses between the biotite lamellse and in minute veins which traverse the slide. Iron ores. — Irou Ore IS rcprcseuted by hematite, magnetite, and pj^rite. The hematite is in blood-red transparent flakes inclosed in the biotite. Magnetite is included by all of the chief minerals, and is in irregular masses without good crystal development. The iron pyrite is found in good crys- tals, though not in large quantity, and is scattered here and there through the slides. PERIDOTITE VARIETIES. The relative proportions of the minerals described above diff"er very much, and we have different kinds of rocks corresponding to these min- eralogical variations. These kinds are not sharply sepai'ated, but are seen under the microscope to grade into one another. rEKlDOTlTE INTKUSIVES. 253 Tlie purest form of peridotite is welirlito, which is composed essentially of olivine and augite. When, besides these minerals, hornblende is present in large quantities, the rocks Ixdong to the ampliil)ole-peridotite type. In some specimens biotite is almost in sufficient abundance to warrant the naniini;- of them biotite-peridotite. Ag-ain, in other specimens feldspar is present in considerable quantity and the rock approaches an olivine-gabbro or oliviue-hornblende-gabbro. WEHRLITE. This is represented by a coarse-grained rock which is mottled and has a dark-green color. (Specimen 23763, from sec. 22, T. 42 N., R. 31 W., N. 1,500, W. 900 paces.) Under the microscope the mottling is seen to be due to the association of very dark greenish-black serpentine pseudo- morphs after olivine with a light-colored augite. Olivine and augite were present in about equal quantities. They are in anhedra, and therefore must have crystallized at about the same time. The olivine is, with very few exceptions, completely altered to serpentine. Augite has a very poor development. Between the olivine and augite are small quantities of irregular plates of biotite. A few small irregular pieces of a very light colored greenish hornblende were observed. Thev are intergrown with the pyroxene and give it an imperfect poikilitic texture. This wehrlite is unquestionably the same as specimen 1247 of the Geological Survey of Wisconsin, described by Dr. A. Wichmann as serpen- tine, consisting chiefly^ of serpentine with some unaltered olivine and augite. AMPHIBOLE-PERIDOTITE. This variety of peridotite was obtained from the outcrop N. 1,260, W. 200 paces from the southeast corner of sec. 29, T. 42 N., R. 31 W., on the east bank of the Michigamme River. The rock is very coarse grained, and possesses poikilitic texture. It is composed of hornblende, pyroxene, olivine, biotite, and iron oxide. The hornblende equals in quantity all of the other constituents. Some of the hornblende individuals measure 3 cm. in length, and include all of the other constituents except the biotite. The pyroxene and olivine seem to have crystallized at about the same time, as 'Microscopical observations of the iron bearing (Huronian) rocks from tbe region south of Lake Superior, by Dr. Arthur Wichmann, Leipzig, 1876: Gool. of Wisconsin, Vol. Ill, 1880, p. 619. 254 THE CRYSTAL FALLS IRON-BEAKmG DISTRICT. they never include each other. They are both, however, included in the hornblende, which with the biotite forms, as it were, the mesostasis. Biotite is present in this specimen in very small quantity, and is essentially the same kind as that above described (p. 252), except that it shows a trifle higher absorption parallel to the cleavage and becomes a yellowish-red. The rock is very fresh and shows scarcely any traces of alteration. This is partly due to the erosive action of the Michigamme River having removed the weathered crust, thus making fresh specimens obtainable. This rock, from the description just given, would be classified as an amphibole-peridotite, with accessory diallage, bronzite, and biotite. It approaches Williams's cortlandtite. In some specimens the biotite is pres- ent in very large quantity, though hardly in sufficient quantity to waiTaut the designation of any of the rocks as biotite-peridotite. GRADATIONS OF AMPHIBOLE-PERIDOTITE TO WEHRLITE AND OLIVINE-GABBEO. There were taken from the same exposure whence the above-described amphibole-peridotite came some specimens which macroscopically can not be distinguished from those of the amphibole-peridotite except in that they are a trifle finer grained. Examined under the microscope, however, we find difl^erences. In some the hornblende is very much reduced in quantity, and varies from the brown kind just described to a light-greenish color, the two being in optical continuity, and the augite and olivine are increased in quantity. These are good types of a wehrlite. In some of the wehrlites there is a variable percentage of feldspar. In certain cases it reaches an amount which would almost warrant the classing of the rock as an olivine - gabbro. Patton described a rock from the same outcrop in which the horn- blende still predominated, but in which there was also a certain amount of plagioclase.^ He called it a hornblende-picrite.^ According to the ter- minology here used, if the plagioclase is to be neglected, it would be au amphibole-peridotite. The thin sections of the feldspathic phase of this rock seem to show that it approaches more closely to a gabbro — that is, to be more feld- spathic than the one described by Patton. They certainly contain far less hornblende than his, judging from his description, and more feldspar. The ' Mikroscopisctie Physiographie, by H. Roseiibiisch ; 3a ed., Stuttgart, Vol. II, 1896, p. 352. -Op. tit., p. 186. PERIDOTITE INTKUSIVES. 255 constituents ai'c tlir same in tlie two rocks, and with some few modifications Ins description would answer. Aiigite is the chief constituent, and following it, in order of im])ortance, come olivine, hornblende, biotite, and feldspar. The diallagic augite is more automorphic (see fig. B, PI. XLV) as the feldspar increases in quantity. It is the only one of the minerals which shows any marked degree of autt)morphisni. The augite present in the sections which I have studied has a light-brownish color, differing from that described by Patton, which is green to colorless. The augite contains the inclusions occurring in hyper- sthene, as well as the green (hornblende?) ones already described. It is invariably surrounded by a narrow rim of light-brown hornblende, and includes in places on the edges irregular patches of the same brownish hornblende. J. Romberg describes the augite in Argentinian gabbros,^ both with and without olivine, as being almost always surrounded by a rim of green hornblende. In one case, however — that of the olivine gabbro from the island of Martin Gai'cia, in the La Plata River^ — both brown and green hornblende is present around the augite. The brown hornblende forms part of the periphery of a crystal; the green the remaining portion. Of the green hornblende some is fibrous, and is considered by Romberg to be certainly secondary. The olivine possesses its usual properties. It is in annedra, with the exception of three or four individuals, which show a fair degree of auto- morphism. The olivine includes rounded grains of a brown spinel, and is traversed by anastomosing veins of the iron oxide. It shows the usual alter- ation to serpentine, and the iron oxide is the result of this serpentinization. The olivine is of exceptional interest on account of the fact that it is sur- rounded by certain zones Avhere it is close to the feldspar (figs. A and B, PI. XLVI). The characters of zones observed in sections from this same locality, and which are almost, if not quite, identical with these which I shall proceed to describe, have already been described by Patton.' There are two of these zones. An inner one is composed of a mineral which is probably an orthorhombic pyroxene. It was so determined by Patton in ' Untersuchungen an Diorit-Gabliro-und Amphibolitgesteiuen aus dem Gebiete der Argentini- Bcaen Republik, by J. Romberg: Neues Jabrbuch fiir Miueral., BB. IX, 1894, pp. 320-321. -Op. cit., p. 322. »0p. cit., p. 168. 256 THE CEYSTAL FALLS IRON-BEARING DISTRICT. the specimens collected and studied by him. I can obtain no positive proof for or against this statement. If it is an orthorhombic pyroxene, it agrees with the inner zones of related occurrences wliich have been described by Tomebohm, G. H. Williams, Adams,^ Romberg,^ and others. This zone is at any rate composed of a colorless, compact mineral, with high single and moderately high double refraction. Its single refraction is nearly equal to that of olivine. The mineral, as a rule, extinguishes parallel to the lines of cleavage. In a few instances the line of extinction made a scarcely noticeable angle with the cleavage. It is separated from the olivine by a shai'p line. At times this inner zone seems to disappear, and at others becomes considerably broader than the average. The width is usually about 0.02 mm., thoug-h it becomes at times 0.08 mm. Outside of this pyroxene zone there is a very much broader zone of light-green hornblende. This is compact, and is in optical continuity with the ordinary brown hornblende, which is the dominant hornblende in the rock. This, in its compact nature and in its relation to the compact brown hornblende of the rest of the slide, differs from the short fibrous actinolite zone ordinarily described as taking part in such " reaction rims." This hornblende zone reaches an extreme width of 0.15 mm. The outer edge of this zone is penetrated by tubular ramifying growths of a colorless mineral, which usually extend inward, perpendicular to the periphery, and which appear to be continuous with the feldspar. This portion of the hornblende rim is about 0.05 mm. wide No such intergrowth of feldspar with the brown hornblende was found, nor have I been able to find elsewhere any description of such an outside zone.^ However, Romberg describes the iutei'esting occurrence in an olivine-gabbro from the Argentine Republic of zones around the hornblende which are very much like those above described, except that the pseudopodia-like growths, as he describes them, consist of a dark-green spinel instead of a clear white feldspar, as in the Michigan rock. In some cases, where the olivine and augite are in juxtaposition, the inner orthorhombic pyroxene zone completely surrounds the olivine. The outer hornblende zone, however, surrounds both the augite and the olivine ' Uber das Norian oder Ober-Laurenti-ites in the g-roundmass, or to those cases which are so roninion in plutonic rocks — even in this rock described — where hornblende is found surrounding- the pyroxene. A general explanation which would account for tlic successive crys- tallization of hornblende and pyroxene in this rock should be applicable to such a zonal g-rowth as occurs around tlie olivine, taking- into consideration, of course, the jirobability that a factor of slight importance in the one case may be the controlling factor in the other. Such occurrences seem clearly to indicate a change in the chemical composition of the mag-ma as the chief factor in the crystallization of the diflPerent minerals, in the pressure, in the temperature, and also in other factors, either one alone or more of these combined. ANALYSI.S Ol' I'ERIDOTITE. The peridotite just described was analyzed by Dr. H. N. Stokes of the United States Geological Survey, and his results are here given (No. 1) : Analysis of peridotite. SiO;.. TiO, . A1,0:,. Cr:0,. FejO, FeC. MnO . NiC. CaO MgO KcO NajO... HjO at 110^ . . . HjO above HO'-^ PiOs COi Total 1 (23353). 44.99 .97 .5.91 .25 3.42 8.30 Trace. 8.79 21.02 .74 .91 .63 3.19 .05 Trace (f). 2 (22981). 37.36 .79 4.76 .62 6.61 6.12 Trace. .04 1.19 31.11 Trace. .65 10.37 .06 None. 99.17 99.68 260 THE CEYSTAL FALLS lEON-BEAEING DISTEICT. It will be seen from the analvsis that the silica is somewhat too hig-h for the typical peridotites. This same fact is also emphasized by the teiidenc}' manifested in some facies of the peridotite for feldspar to develop, and thus for transitions to norite and g-abbro to be produced. With this analysis of the peridotite there is placed for comparison the analysis (No. 2) by Dr. H. N. Stokes of the picrite-porphyry already descril^ed. The close resemblance chemically liecomes at once manifest, althouiih the latter is more nearly a typical peridotite in composition. It can not be denied that possibly this picrite is but a further differentiation product of the same magma to which the 2:)eridotites belong, although its occurrence is so remote from these that it is impossible to connect them in the field. PERIDOTITE PROM SBC. 22, T. 42 N., E. 31 W., N. 1,990, W. 150. Just west of the northeastern corner of sec. 22, T. 42 N., R. 31 W., there is a bold outcrop of hornblende gabbro, which is cut by a dike, about 10 feet wide, of a very massive, coarse, granular black peridotite. Macro- scopically one can readily distinguish in the peridotite flakes of Ijiotite, poikilitic plates of hornblende, and a smaller amount of white feldspar. Under the microscope the constituents are, in order of imjjortance: Horn- blende, augite, feldspar, biotite, bronzite, olivine, magnetite, and quartz.-' Hornblende. — Tlils Is thc rlch browu kind, full of inclusions, g-rading into the green variety which was described on p. 234 as occurring in the gabbros of this district. It is jiresent in anhedra inclosing liiotite, pyroxene, and olivine. Pyroxene. — Tliis is represented by mouoclinic and orthorhombic varieties. The mouoclinic pyroxene, augite, is most abundant, and is in light-yellow to pink-colored anhedra, except where it touches the feldspar; there the augite is automorphic, and is surrounded by a narrow border of light- brown hornblende. The orthorhombic pyroxene is present in a few anhedra, which are colorless or have a faint cream tint. It is presvimed to be bronzite. Feldspar. — This fills tlic intcrspaccs between the other constituents, and occurs in grains which are polysyntheticalh' twinned after the albite law. ' Only one section has been prepared from this specimen, and it may not give a correct idea of the true proportion of these minerals in thr rock mass. In tlie macroscopical examination of the hand .specimen the biotite seemed to he subordinate only to the hornblende. PERIDOTITK INTKUSIVES. 261 Mfiisurements ji-avo a, syninietrical t'xtiiiction of 32 eaeli side of the twiniiiii"- plane on zoiie_L01(». I tlierefo.- ; eonclude tlie feldspar to be labradorite. Biotite. — This is the ordinary yellow to brownish kind, and is in irre"'n- lar plates. It shows its usnal characters and is inclnded in the hornblende. Magnetite. — This mineral occnrs in (;rystals and gTains, inclnded in all the other constitnents. Quartz. — A few grains of (jnartz were fonnd associated with the feldspar. The presence of dihexahedral liquid inclusions easily gave a clew to the orientation of the grains. The rock composed of the above-described nainerals offers a good illus- tration of that gradation which is one of the fundamental laws of nature and is nowhere better exemplified than in the rocks. On the one hand, from its texture and from the presence of the dominant hornblende, with the small quantity of quartz, this rock may perhaps be considered to be closeh' related to the diorites. On the other hand, the presence of the pyroxene and olivine seems to point toward its connection with a gabbro. Its geological occurrence points most satisfactorily toward its corre- spondence in age and its intimate relationship to the peridotites of the dis- trict. The predominance of the bisilicates indicates it to be of very basic character, and for these reasons I have called it " peridotite," althouo-h I have not succeeded in getting an analysis to prove its ultrabasic nature. RELATIONS OF PERIDOTITES TO OTHER ROCKS. The peridotites occur in such small quantity that general conclusions concerning their relations to other rocks occurring in their vicinity are scarcely warranted." However, from the fact that they are so intimately associated with the gabbro — cutting it in two cases where the contact ^\'as observed — and from the fact that among the peridotites themselves certain phases approach in mineralogical composition certain of the gabbros (see p. 254), it seems advisable to conclude that they represent ultrabasic differ- entiation products of the same magma from which the gabbro types were derived. The inappreciable differences in grain between the portion of the i-ock nearest the contact between these basic rocks and the gabbros and those farther away can be explained by supposing their intrusion to have taken place while the main mass of the gabbro retained considerable heat and thus prevented their rapid cooling. 262 THE CRYSTAL PALLS lEON-BEARING DISTRICT. AGE OF PERIDOTITES. The only statement which can be made concerning the age of the peri- dotite dikes is that they are younger than some of the gabbros, and that, not having suffered the defonnation of the pre-Keweeuawan orogenic move- ments, they are Keweenawan or post-Keweenawan. GEKERAL OBSERVATIOXS ON THE ABOVE SERIES. TEXTURAL CHARACTERS OF THE SERIES. There are represented in the above series rocks with moderately fine grain as well as those of very coarse grain. They vary from those with parallel texture, through those vith porphyritic, poikilitic, and ophitic texture, to those with granular texture. There is, howevei", tlu-oughout a clear preponderance of the medium to coarse granular rocks. The rocks are evidently not of effusive character, though some jDOSsess the textures prev- alent in effusive rocks. The order of crystallization of the minerals in the rocks of granular .exture, excluding the iron ores and the accessory minerals, is as follows. Tlie order in the imperfectly ophitic and porphyritic rocks is not considered, as those are rather exceptional occurrences. In the lists those minerals are hyphenated of which it has not been possible to determine accurately the order of crystallization. It seems that either they were formed at the same time or, in some cases, their formation has overlapped. In such cases the one placed first is the one presumed to have begun its crystallization first. DioRiTE. Gabbroand horn- Bronzitb-noritb, Pkridotite. BLKNDE-GABHRO. Hornblende. Olivine. Bronzite. Olivine. Biotite. Monoclinic pjToxene. Monoclinic pyroxene. Orthorhombic pyroxene. Plagioclase. Biotite-hornblende. Blotite-bornblende. Monoclinic pyroxene. Microcline. Plagioclase. Plagioclase. Biotite-hornblende. Orthoclase-nuartz. Plagioclase. For the entire series the order may be arranged as follows: Olivine, bronzite, monoclinic pyroxene, mica-hornblende, plagioclase, orthoclase, quartz. This is the same order that is exhibited by the most basic rock represented in the series, the peridotite, so far as this rock contains the minerals. SERIES OF INTKUSIVES. 263 The order of" crystallization of the minerals throughout the series is due to their relative solubility in the eruptive raagnui. Among- various factors affecting solubility the fusion point of the chemical compounds constituting the different minerals, the temperature of the magmas, and the pressure under which the minerals crystallized, are important. The porphyritic and the ophitic textured rock facies, having crystallized under different condi- tions of pressure and of temperature from those under which the granular rocks were formed, show, as is to be expected, a different order of crystal- lization of minerals. CHEMICAL COMPOSITION OF THE SERIES. In the following tables there are reproduced the analyses which have been obtained of the various types. They are arranged according to dimin- ishing acidity. Nos. 1 and 4 were analyzed by Dr. H. N. Stokes, Nos. 2 and 3 by Mr. George Steiger, both of the United Slates Geological Survey : Table I. — Analyses of Crystal Falls rocks. SiOj TiOj Al,03 CrjOa Fe^Os FeO MnO NiO CaO MgO K2O NaaO HjOatllOo .. HjO above 110' P2O5 CO, Total... 1 (26023). 58.51 .72 16. .32 None. 2.11 4.43 Trace. None. 3.92 3.73 4.08 3.11 .23 2.00 .30 None. 2 (23354). 3 (23755). 49.80 .79 19.96 6.32 .49 99.46 11.33 7.05 .61 2.22 100°— . 13 1000+1. 71 .07 .15 100. 63 48.23 1.00 18.26 1.26 6.10 9.39 10.84 .73 1.34 100°— .26 100° +2. 00 .07 .43 99.91 4 (23353). 44.99 .97 5.91 .25 3.42 8.30 Trace. None. 8.79 21.02 .74 .91 110° . 63 110°-j-3. 19 .05 Trace. ( ?) 99.17 (1) Mica-diorite (quartzitic); (2) Hornblende-gabbro ; (3) Norite; (4) Peridotite (wehrlite). 264 THE CRYSTAL FALLS IRON-BEARING DISTRICT. Table II. — I'ercentaijes of chief oxides reduced to 100. 1. 2. 3. 4. SiO. 60.36 .75 16.83 2.17 4.57 4.04 3.85 4.21 3.21 50.52 .80 20. 25 6.41 .50 11.50 7.15 .62 2.25 49.64 1.03 18.79 1.30 6.28 9.67 11.16 .75 1.38 47.33 1.02 6.22 3.60 8.73 9.25 22.11 .78 .96 TiO- Al,Oi FeO, FeO CaO UsO K,0 Na.O Table III. — Atomic 2}''oportions of metals. Si.. Ti . AI . Fe. Cii. Mg. K.. Na- 55.85 46.53 .53 .56 18.41 22.03 5.08 .834 4.04 11.42 5.32 9.85 4.99 .74 5.78 4.04 45.27 .71 29.26 5.70 9.51 15.22 .88 2.45 42.48 .70 6.60 9.02 8.98 29.67 .91 1.67 The analyses show that all of the rocks contain a moderately large amount of water. Nevertheless,' they are sufficiently well preserved to warrant a discussion of their analyses for classification purposes. This is especially true of No. 4, which is remarkably fresh for so basic a rock. The chief rock-makiiio- oxides in the above analyses appear in Table II reduced to 100. The molecular proportion of these oxides was then obtained. From these data the atomic proportions of the metals were derived, and are given in Table III. These calculations were kindly made for me by Mr. V. H. Bassett, assistant in the chemical laboratory of the University of Wisconsin. If we examine Table II we see that, in passing from the more acid to the basic end of the series, in correspondence with tliis decrease in silica the alumina increases rapidly, then decreases until it reaches the extreme basic rock, when it cbops suddenly to 6.22 per cent. The analyses also show an increase in iron, which is best brought out in Table III. The alkalies decrease with diminishing silica, whereas the MgO, which for rocks of this character is very characteristic, shows a decided increase. Within the gabbro-norite-peridotite series (Nos. 2, 3, and 4) the lime shows SERIES OF INTKUSIVES. 265 a constant diniinution f()iTes))()ndin») CONTACT OF GRANITE AND SEDIMENTARY. (S) CONTACT OF GRANITE AND SEDIMENTARY SEEN BETWEEN CROSSED NICOLS. THE MFR DEN GRAVURE CO- PLATE XXXVI. 301 PLATE XXXVI. Fig. a. (Sp. No. 32827. Without .analyzer, x 38.) Photomicrograph illustratiug a rather exceptional form of siiilositc with white spots lying iu the fine-grained groundiiKiss. Albite, with a very small amount of chlorite and epidote, forms the spots. (Described, p. 206.) Fig. B. (Sp. No. 32827. With analyzer, x 38.) This is the same section as above, viewed between crossed nicols, showing the aggregate char- acter of the spots of this spilosite. (Described, p. 206.) 302 U.S. GEOLOGICAL SURVEY MONOGRAPH XXXVI PART I PL. XXXVI '^l""' wa r^^??^ ■ »', (^) (-6) (>») SPILOSITE. (S) SPILOSITE SEEN BETWEEN CROSS NICOLS. IFRIDEN GHAVOKE CO. PLATE XXXVII. 303 PLATE XXXVII. Fig. a. (Sp. No. 32958. Without analyzer, x 18.) Normal spilosite. Oval spots of xuacroscopical size, in which chlorite is the predominant con- stituent, with some quartz, feldspar, rutile, and muscovite, are sharply defined from the surrouuding fine-grained grouudmass, consisting of the same constituents, but with the muscovite in large quantity and the chlorite very subordinate. (Described, p. 206.) Fig. B. (Sp. No. 32861. Without analyzer, x 38.) Spilosite in which the spots are of microscopical size and consist predominantly of chlorite aggregates lying in tbe fine-grained quartz-albite groundmass. (Described, p. 207.) 304 U. S. GEOLOGICAL SURVeV MONOGRAPH XXXVI PART I PL XXXVII ■ .•^►>* '•^te* ■ ^'■^'\'' ■'•if;'-'' ■ (^) (/I) SPILOSITE. (B) SPILOSITE. nEN GSAVUHE CO. PLATE XXXVIII. MON XXXVI 20 305 PLATE XXXVIII. Fig. a. ( Sp. No. 32826. Without analyzer, x 38. ) Photomicrograph illu8trating the passage of spilosite to a desmosite. lu the upper jiortiou, especially iu the upper left-hand corner, of the figure, chlorite .aggreg.ates similar to those illustrated in lig. i', PI. XXXVII, are seen. These hecome united, and thus there is a passage into the banded product. This banded character is well shown in the lower half of the photomicrograph. (Described, p. 207.) Fig. B. (Sp. No. 23755. Without analyzer, x 90.) Occurrence and alteration of bronzite in bronzite-norite. This illustration shows the way in which the bronzite occurs in the bronzite-norite. It is frequently included in the hornblende. The bronzite alters around the edges and along the cracks to a yellowish-green fibrous serpentine mineral, which is represented in the section by the dark fibrous material next to the unaltered bronzite. This secondary mineral then .alters to a scaly aggregate of talc. These two secondary products can be seen bordering tba bronzite, especially well where it is traversed by a crack. (Described, p. 238.) 306 U.S. GEOLOGICAL SURVEY MONOGRAPH XXXVI PL. XXXVIII :»^' M) (5) (/») PASSAGE OF SPILOSITE INTO DEMSOSITE. (S) ALTERATION OF BRONZITE IN BRONZITE NORITE. THE MtRIDEN ORAVURE CO. PLATE XXXIX. 307 PLATE XXXIX. Fig. a. (Sp. No. 23321. With analyzer, x 35. ) Pliotomicrograph nf a section of biotite-granite from the center of a dil^e 5 feet Tvido. On the horders of the dike the magma has crvstallized as a normal mica-diorite -.vithout qnartz and orthoclase. (Described, p. 226.) Ficf. B. (Sp.No. 26023. With analyzer, x 3.5.) Mica-diorite showing tendency toward an opliitic texture. Plagioclase is the most automorphic mineral- Biotite is next, but it is poorly developed. Orthoclase and quartz fill irregular areas between the plagioclase. (Described, p. 231.) 308 U. S. GEOLOGICAL SURVEY MONOGRAPH XXXVI PL. XXXIX (^) (B) (») BIOTITE GRANITE. (B) MICA DIORITE. THE MERinEN GHAVURE CO. PLATE XL. 309 PLATE XL. Fig. a. (Sp. No. 32643. Without analyzer, x 35.) Qnartz-mioa-diorite-porphyry. The phenocrysts of feldspar aud quartz stand out clearly from the fine miciogranitic grouudmass. Mica phenocrysts are not seen in this figure, which is intended chiefly to illustrate the character of the feldspar phenocrysts. Muscovite has resulted from theii alteration. The zone which surrounds the altered center is very fresh, and is rendered poikilitic hy inclusions of minutt- grains ot fpiartz. (Described, p. 229.) Fig. B. (Sp. No. 32643. With analyzer, x 35.) Qnartzmioa-diorite-porphyry. The same section, viewed lietweeu crossed nicols. The micro- granitic texture of the groundmass is well shown. (Described, p. 229.) 310 U. S. GEOLOGICAL SURVEY MONOGRAPH XXXVI PL. XL -0^-^^. '^ ■k: J" w,' "' •y V^/'^ •^^r^; --*v- i. M) (B) (A-) MICA DIORITE PORPHYRITE. (S^ MICA DIORITE PORPHYRITE SEEN BETWEEN CROSSED NICOLS. TMt MERlOEN QRAVUBE CO. PLATE XLI. 311 PLATE XL I. Pm. A. (Sp. No. 23320. Without aualyzer, x 18.) Pnrphyritic poikilitic hornblende-gabbro. Tho brown honibk-ndo occupying tho center of the phenocryst grades over iuto a dull-green hornblende. Feldspar and pyroxene are included in the hornblende. (Described, p. 241.) Fig. B. (Sp. No. 23344. With analyzer, x 18.) Hornblende-gabbro showing a poikilitic texture. The tendency of the feldspars toward a lath- shaped development is very evident. Were they lath-shaped tho texture would agree with the L^vy definition of tho ophitio texture. (Described, p. 233.) 312 U.S. GEOLOGICAL SURVEY MONOGRAPH XXXVI PL. XLI (/» PORPHYRITIC POIKILITIC HORNBLENDE CABBRO. CS) POIKILITIC HORNBLENDE. HE MERIOEN QRAVURE CO. PLATE XLII. 313- PLATE XLII. Fig. a. (Sp. No. 23754. Without analyzer, x 38.) A moderately fine grained hornbleude-gabbro showing parallel texture. This liornblende- gabbro occurs in dikes cutting the coarse forms of hornblende-gabbro. The sppcimen shows very clearly the parallel texture rather commonly found in sections from these dike rocks. This texture is best developed nearest the contact, and i.s presumed to be a flow texture. The chief mineral constituents — plagioclase, hornblende, and mica — can be readily distinguished in the section. (Described, p. 244.) Fig. B. (Sp. No. 23754. With analyzer, x 38.) The parallel texture in the hornblende-gabbro is brought out somewhat better when viewed between crossed nicols. (Described, p. 244.) 314 U. S. GEOLOGICAL SURVEY MONOGRAPH XXXVI PL. XLII (^) (B) (/» HORNBLENDE CABBRO. (S) HORNBLENDE CABBRO SEEN BETWEEN CROSSED NICOLS. The meriden qravore co. PLATE XLIII 315 PLATE XLIII. Fig. a. (Sp. No. 26070. Without analyzer, x 17.) Normal granular hornblende-gabbro. This illustrates the normal medium-grained hornblende- gabbro -which occurs in this district. The mineral constituents hornblende, feldspar, aud iron oxide can be readily distinguished. (Described, p. 248.) Fig. B. (Sp. No. 26069. With analyzer, x 18.) Schistose hornblende-gabbro. This illustrates a crushe INCLUDING THE FELCH MOUNTAIN RANGE By HENKY ULOYD SMY'm A CHAPTER ON THE STURGEON RIVER TONGUE By WILLIAM SHIKLET BAYLET 323 CONTENTS. Page. Chapter I.— Geographical omits and physiographv 329 Intvoduction 329 Preliminary sketch of geology 331 Character of surface 331 Drainage 334 Chapter II. — Magnetic observations in geological mapping 336 Section I. — Introduction 336 Section II. — Description of the magnetic rocks 338 Section III. — Distribution of magnetism in the magnetic rocks 339 Section IV. — Instruments and methods of work 341 Section V. — Facts of observation and general principles 344 (1) Observed deflections when the strike is north and south and the dip vertical 344 (2) Deflections of the horizontal needle 345 (3) Deflections of the dip needle 347 (4) Horizontal and vertical components when the magnetic rock dips vertically 349 (5) Horizontal and vertical components when the magnetic rock dips at any angle 350 (6) Determination of depth 354 (7) Summary 356 Section VI. — Applications to special cases 356 (1) The magnetic rock strikes east or west of north and dips vertically 357 (2) The magnetic rock strikes east and west 3.59 (3) Two parallel magnetic formations 361 Section VII. — The interpretation of more complex structures 366 (1) Pitching synclines 367 (2) Pitching anticlines 370 (3) Formations split by intrusives 371 (4) Summary 372 Chapter III. — The Felch Mountain range 374 Section I. — Position, extent, and previous work 374 Section II. — General sketch of the geology 383 Section III. — The Archean 385 Topography 386 Petrographical characters 387 Section IV. — The Sturgeon quartzite 398 Distribution, exposures, and topography 398 Folding and thickness 399 Petrographical characters 401 Section V. — The Randville dolomite 406 Distribution, exposures, and topography 406 Petrographical characters 408 Section VI. — The Mansfield schists 411 Distribution, exposures, and topography 411 Petrographical characters 412 325 32(i CONTENTS. Chapter III. — The Felch Mountain Range — Continued. Page. Section VII. — The Groveland formation 415 Distribution, exposures, and topography 415 Petrographical characters 417 Section VIII. — The mica schists and quartzites of the Upper Huronian series 423 Petrographical characters 425 Section IX. — The intrusives 426 Chapter IV. — The Michigamme Mountain and Fence River areas 427 Section I.— The .\rchean 428 Section II. — The Sturgeou formation 430 Section III.— The Randville dolomite 431 Distribution and exposures 431 Folding and thickness 432 Petrographical characters 434 Section IV. — The Mansfield formation 437 Distribution, exposures, and topography 438 Folding and thiclsness 438 Petrographical characters 439 Section V. — The Hemlock formation 440 Distribution, exposures, and topography 440 Folding and thickness 441 Petrographical characters 442 Section VI. — The Groveland formation 446 Distribution, exposures, and topography 446 Folding and thickness , 448 Petrographical characters 448 Chapter V. — The northea.stern area and the relations between the Lower Mar- quette and the Lower Menominee series 451 Chapter VI.— The Sturgeon River tongue, by William Shirley Bayley 458 Description and boundary of area 458 Literature 459 Relations between the sedimentary rocks and the granite-schist complex 461 The Basement Complex 463 The gneissoid granites 463 The amphibole-schisls - 465 Origin of the amphibole-schists 466 The biotite-schists 467 The intrusive rocks 469 Comparison of the Sturgeon Eiver and the Marquette crystalline series 470 The Algonkian trough 471 Relations liet ween the conglomerate and the dolomite series 472 Relations between the dolomites and conglomerates and the overlying sandstones 473 The conglomerate formation 473 Important exposures 474 Petrographical descriptions 477 The dolomite formation 479 Important exposures 480 Petrographical description 481 Slates and sandstones on the Sturgeon River 481 The igneous rocks 482 The intrusive greenstones 482 Petrographical description 482 The banded greenstones 485 Petrographical description 486 ILLUSTRATIONS Page. Plate XLVII. Relaiious of maguctic beds to variation auil dip 352 XL VIII. Relations of magnetic beds to variation and dii> 362 XLIX. Geological map of the FeUb Mountain range 374 L. Geological map of a portion of the Crystal Falls district 450 LI. Geological map of the Sturgeon River tongue 458 LIT. Map of exposures in sec. 7 and portions of sees. 8, 17, and 18, T. 42 N., R. 28 W 474 LIII. Schist conglomerate from dam of Sturgeon River 476 Fig. 15. Magnetic cross section in T. 45 N., R. 31 W 345 16. Circles of attraction 346 17. The forces acting on the dip needle 347 18. Curves showing the relations between the horizontal components at the points of maxi- mum deflection, for rocks dijiping at various angles and buried to various depths 354 19. Truncated anticlinal fold with gontly dipping limbs 364 20. Truncated anticlinal fold with steeply dipping limbs 365 21. Plan and cross sections of a lutchiug syncline 367 22. Magnetic map of the Groveland Basin 370 23. Plan and cross sections of a pitching anticline 371 24. Magnetic map of a single formation split by an intruded sheet 372 327 THE CRYSTAL FALLS IRON-BEARING DISTRICT OF MICHIGAN. part ii. the eastern part of the district, i:n^cludixg the felch mountain range. By Henry Lloyd Smyth. With a Chaptek on the Sturgeon River Tongue, bv William Shirley Bayley. CHAPTER!. GEOGRAPHICAL LIMITS AND PHYSIOGRAPHY. INTRODUCTION". The teiTitory to be described in this and the four following chapters is situated in the Upper Peninsula of Michigan, between the Marquette and Menominee iron ranges, and is all embraced within T. 42 N., Rs. 28-30 W., and Ts. 42-47 N., Rs. 30-31 W. The area of about 300 square miles included within these townships had for the most part been covered hastily by previous reconnaissances of the Lake Superior Division of the United States Geological Survey, the results of which were placed at my disposal. Our task was to go over with especial care those portions in which outcrops had been found by our predecessors, or which seemed likely to contain the iron-bearing formations. At the same time much of the rest was examined more hurriedly. The tract surveyed in detail comprises a continuous belt about 30 miles in length, and of width varying from 2 to 5 miles, lying wholly within the drainage basin of the Michigamme River and its principal upper tributary, the Fence River, and extending southward from the northern end of the Republic tongue, where it was connected with rocks of well- determined Marquette types, as far as the south line of T. 43 N., R. 31 "W. From this line we passed southeast (leaving a gap of 5 miles) across the low 329 3B0 THE CRYSTAL FALLS lEON-BEARING DISTRICT. di^■ide between the Michigamme and the headwaters of the Sturgeon, to the Felch Mountain range, which was then carefully studied for a distance extending 13 miles to the east. Until within the last few years the larger part of this area had been very difficult of access, and much of it is dithcult still. The rock surface is almost wholly concealed by a cover of glacial deposits of various kinds; dense forest and great swamps also obscure the rocks, and make traveling difficult and slow. It is therefore not a field to invite geological study. While exploration for iron ore has here and there passed the frontiers of the productive ranges on either side, the general ill-success which attended the early enterprises has discouraged the active search that would at least have resulted in important additions to geological knowledge. For these reasons the area as a Avhole, with the exception of the Felch Mountain range, has remained almost unknown geologically, until our work in 1892. The i-efer- ences to it in geological literature are consequently but few In number, and are for the most part merely the records of the unrelated observations of casual visitors. The district, nevertheless, deserves attention from both the economic and the geological standpoint. The iron-bearing formations of the Mar- quette range extend into it from the north, those of the Menominee range from the south. On the west the ore deposits of the Crystal Falls area are connected geographically at least with the western extension of the Menom- inee range. Between these boundaries the area stands as the largest one remaining in Michigan in which iron-bearing formations are known to occur, but as yet not yielding important bodies of ore. Here, too, if anywhere, the questions of the equivalence or nonequivalence of the individual for- mations of the Marquette and Menominee iron-bearing series are to be answered. It is proper to state that the field study, in consequence of the condi- tions under Avhich this work was done, was almost wholly directed to eco- nomic questions, and that it was not originallj- anticipated that the results were to be published as a monograph on the district. This will explain the very brief space devoted to the Archean in the following pages. The field work was begun and ended in 1892. Since that time there has been no opportunity to revisit localities, and the conclusions now stand essentially as they were reached in the field. Considering both the obscurity and com- INTRODUCTION. 331 plexitv of the area, it is very probable* that further study of important local- ities would clear away many of the difficulties, as well as uiodify certain of the opinions now held. The writer was efficiently aided in the field work by Messrs. Samuel Sanford and Charles N. Fairchild for nearly the whole period, and E. B. Mathews and H. F. Phillips for part of it, as assistant geologists, and by Messrs. Lewis and Forbes as skilled woodsmen. PRELiIMINARY SKETCH OF THE GEOL,OGY. The rocks of the Michigamme and Felch Mountain areas range in age from Archean to early Paleozoic. North and west of the Michigamme River, where geological boundaries are most susceptible of determination, the granites and gneisses of the Archean come to the surface in three oval areas of great regularity of outline, from 10 to 12 miles long by 2 to 6 miles wide, while the intervals between the Archean ovals are occupied by highly tilted sedimentary and igneous rocks of Algonkian age. The lower member of the Algonkian has derived its materials from the wasting of rocks lithologically similar to the underlying granites and gneisses. In the southern and eastern portions of the district the edges of the tilted older rocks are partially covered by a blanket of gently dipping sandstones of Cambrian age, very soft and easil}" disintegrating. These rocks first appear near the ]\Iichigamme River as detached outliers. In going south and east from that river the separated patches become larger and more a,bundaut, until finally a few miles beyond the eastern limit of our work in the Felch Mountain range they unite and entirely cover the pre-Cambrian formations. CHARACTER OF THE SURFACE. In its most general aspect the surface throughout this area is a plam, somewhat rolling indeed, which slopes gently upward from the southeast toward the northwest. The surface is formed partly by the soft and gently inclined Upper Cambrian sandstones and partly by the much harder and highly tilted pre-Cambrian rocks of diverse physical and mineralogical characters, and yet over all it maintains a very uniform slope. On the southeast, in the Felch Mountain range, the plain has an average elevation above the sea of 1,200 to 1,300 feet. In the northwest, in the southern 332 THE CRYSTAL FALLS IRON-BEARING DISTRICT. sections of T. 47 N., R. 31 W., the average elevation is 1,800 to 1,900 feet. Since the intervening' distance is somewhat more than 30 miles, the gen- eral slope is therefore less than 20 feet to the mile. The minor topographical features based upon this plain are multitudi- nous in variety and detail, but generally quite insignificant in relief. The maxiinum difference of elevation between the top of the highest hill and the bottom of the neighboring valley is less than 300 feet, and this is reached in Ijut two cases. The country possesses no commanding emi- nences, and in the widest panoramas now and then obtainable from the summits of glaciated knobs the background is restricted to a radius of a few miles. In these the general evenness of the sky line is usually broken only by the remnants of the old forest, which have not yet succumbed to fire and the lumberman. These lesser features have been shaped mainly by the work of the conti- nental ice-sheet, both through the materials which it brought in and through those which it earned away. In the areas underlain by relatively massive rocks, particularly the Archean crystallines, the surface has been left mam- millated with rocky knobs,' which doulitless were the unattacked cores rising into the pre-Grlacial zone of disintegration. These are separated by the similar inverse forms, now for the most part occupied by swamps. In the Archean borders of the Felch Mountain area, where the glacial cover was originally thin, the periodical fires that have followed lumbering operations have burned out the organic matter from the soil and so loosened it that, on the steeper slopes, it has been entirely washed away and the rock surface laid bare. The hummocks and bowls are generally elongated east and west, which is the direction both of the gneissic foliation and of the ice movement. The elevations rise, often with steep, smooth walls, for 5, 10, 20, or even in some cases 60 feet, above the intervening depressions. The latter hold muskeg to the rims. In the wet season they fill with water, which ovei-flows to the next bowl below, but permanent lines of minor di-ainage, here as elsewhere in the Archean areas, are very infrequent. Over most of the area, however, the ice has spread a sheet of till, and has here and there deposited the materials swept along in the subglacial streams in characteristic complexity of form and grouping. The more prominent elevations are, in fact, deposits of modified drift, although occasionally TOPOC.KAPHY. 333 small rock masses like Michig'amine Mountain, which is composed of mate- rial that offers a most stubborn resistance to all degrading- agents, reach an elevation of 100 to 200 feet above the general level of the surrounding country. The fact that the name "mountain" has been applied to hillocks ot this order by the surveyors and woodsmen, who have the widest knowl- edge of the Upper Peninsula, conveys perhaps the clearest idea of the generally level character of the surface. While the details of the topography are thus mainly glacial in origin, the broader features of the next order of importance have often clearly been determined by the presence of the more resistant rocks. The large structural domes of the Archean, which are such characteristic geoloo-ical features, are also indicated by a general upward swell of the surface of the areas which they occupy. The topographical transitions at the margins of these swells are frequently abrupt, and sometimes for considerable distances are marked by scarp-like slopes in the granites, caused by the almost ver- tical contacts with the softer Algonkian formations. Considerable portions of all three of the Archean ovals in the northern part of the district display this slight topographical prominence. Marginal scarps are particulai-ly well shown in the oval west of Republic, in sees. 19 and 30, T. 47 N., R. 30 W., and along the south side of the oval which lies between the Fence and Deer rivers, near their junctions with the Michigamme. The more impor- tant bodies of greenstone also are generally expressed by a noticeable degree of elevation. Thus the great intruded sheets folded in with the Lower Mar- quette series in sees. 24, 25, and 36, T. 47 N., R. 31 W., give rise to long broad ridges that closely follow the changes in the strike. But in all these cases the topographical emphasis is very slight, and the plain as a whole may truly be said to maintain its general slope with practical indifference to the weather-resisting differences in the underlying rocks. These broader swells of the harder rocks are separated by broad, slightly lower-lying plains, in many of which a valley character is still dis- tinctly recognizable in spite of the fact that they especially have been favored with deposits of modified drift. The present drainage, in its main lines, largely follows these older valleys, although much confusion, which is especially noticeable in the details, has of course resulted from their partial choking by the drift. 334 THE CRYSTAL FALLS IRON-BEARING DISTRICT. DRAINAGE. Nearly all the surface water ot" this district finds its way to Lake Michisran tlu'ouo'h the Michiijamme and the Sturcreon riA'ors, which are inde- pendent bi-anches of the Menominee — the largest river flowing- into Lake Michigan from the west. A few square miles along the eastern boundary, however, are tributary to the Ford, which flows into Green Bay north of the Menominee. Of these the Michigamme drains by far the largest j^art of the district. This river heads in Lake Michigamme, which it leaves in sec. 9, T. 47 N., R. 30 W., near the northeast corner of the area shown on the general map (PI. II), at an elevation above the sea of 1,580 feet. Thence it flows for 8 miles southeast to Republic, in a synclinal valley cut out of the soft schists of the Michigamme formation. This valley, which is nearly a mile wide at the northern end and less than .half as wide at the southern, is bordered on both sides by the harder Archean granites, which rise with rather steep slopes to the general level of the plain. Tlu-oughout the length of the valley the river flows over glacial drift, but at Republic, where the soft rocks come to an end, it breaks across rocky barriers in a succession of rapids, and continues first nearly due south (leaving the dis- trict covered b}' our map), and then flows southwest over glacial deposits, which completely mask the bed rock for 10 miles. South of the Archean oval, which occupies the western part of T. 44 N., R. 31 W., and the east- « ern part of T. 44 N., R. 32 W., the limestones and slates of the pre- Cambrian are again exposed, and over these the Michigamme flows in close conformity to the general strike as far as the range line. In the southern sections of T. 44 N., R. 31 W., the Michigamme receives two tributaries from the north — the Fence River, which comes from the eastern side of the Archean mass just mentioned, and the Deer River, which comes from its Avestern side. The headwaters of the Deer and of the west- ern branch of the Fence flow through the same section (21) in T. 46 N., R. 32 W., north of the Archean oval, but farther south they diverge to an extreme distance of 10 miles, and afterwards converge so that their points of junction with the Michigamme are but 4 miles apart. The area thus inclosed is broadly concentric with the Archean oval. In the case of the Fence, at least, the river is placed within a wide depression coincident with the softer stratified rocks of the Algoukian, and follows very faithfully their DRAINAGE. 335 general .strike. Depo.sits of glacial sand and gravels are very abundant witliin this valley, and for these the river often swings aside across the strike for a mile or more. In sees. 21 and 29, T. 45 N., R. 31 W., and in sec. 10, T. 44 N., R. 31 W., excellent rock sections are afforded by such digressions. Tlie old valley between the two Archean ovals west of the Republic tongue (see PI. Ill) is on the south entirely filled with glacial gravels to the level of the old divides, and the large brook known as the east branch of the Fence is diverted to the till-covered western of the two Archean ovals. The valley is clearly indicated, however, by an interesting series of lakes, of which Squaw, Trout, and Sundog, each about 1 mile in length, are the most considerable. The area drained by the Sturgeon lies in the extreme southeastern part of the district, wholly within the marginal fringe of sandstone. The rela- tion of its course to the geology is known in detail only within portions of the Felch Mountain range. This it first enters in Ihe northern portions of sees. 35 and 36, T. 42 N., R. 30 W., in a loop into the Algonkian, from the northern Archean margin, to which it again returns. Five miles farther east it crosses the trough from north to south, transverse to the strike of the Algonkian formations, to the contact with the southern Archean mass. It follows this contact eastward for 2 miles, and then strikes southward across the Archean to the Menominee River, not again returning to the Felch Mountain range. The river valley in the Felch Mountain range is very distinct, and where bordered by Potsdam outliers is rather deep, with pre- cipitous banks. It is but slightly affected by drift deposits. Its course shows an almost complete disregard of the structure of the Algonkian and Archean rocks, and so has the usual characters of a superimposed stream. The Michigamme River, as was early noted by Pumpelly, has practically no eastern branches within this district. The Escanaba and Ford rivers, which reach Lake Michigan directly, and the Sturgeon, which joins the Menominee below the mouth of the Michigamme, all head within 2 or 3 miles of the latter, the course of which is transverse to their general direc- tion. The Michigamme thus flows along the eastern edge of its drainage basin. This fact — the most striking in the general distribution of the streams of the district — is the result of causes which, in part at least, go back to very remote geological periods. CHAPTER II.i MAGNETIC OBSERVATIONS IN GEOLOGICAL MAPPING. SECTION I. INTRODUCTION. As has been said already, the area in which our work was done is largely drift covered, to somewhat varying but usually considerable depths; the mantle on the whole is so evenly spread that outcrops of any rocks except those belonging to the Archean are in many sections few and scat- tered and sometimes are almost entirely lacking over whole townships. Under these circumstances, and since also the pre-Cambrian rock structure is complex, even a general outlining of the old formations would be impossible by the usual geological methods, and if we were restricted to these there would be no alternative but to map most of the territory as Pleistocene. It happens, however, that the Algonkian rocks of Michigan contain a large amount of magnetite, which is known from observation in the developed ranges to be characteristic of certain geological formations. It undoubtedly occurs in more or less amount in all the sedimentary rocks and is also present, sometimes in considerable quantities, in rocks that are not sedimentary, as, for example, around the margins of the old intrusive diorite bosses. But generally speaking, its occurrence in large quantities is confined so closely to definite geological formations, in which it is found in characteristic association with certain other minerals, or to horizons within those formations, that it can be guardedly used in identifymg them, and in tracing them from localities where they outcrop through areas in which they are buried. This use is not only justified, from an empirical standpoint, by the presumption in favor of analogies to which no exceptions are known, but it has a rational basis, in the view of the late Professor Irving,^ which is ' This chapter is abridged from a paper of the same title presented at the Colorado meeting of the American Institute of Mining Engineers in September, 1896. ^ Classificatior of early Cambrian and pre-Cambrian formations, byR. D. Irving: Seventh Ann. Kept. U. S. Geol. Survey, 1888, pp. 451-452. 336 MAGNETIC OBSEUVATIONS. 337 steadily j^aiiiinji' j^rouiul, tlint at least luucli of the iron of this iiuvgnetite was originally Iniricd in thesanu' t\tnnations in which it now occurs, through the ageiic}' of organic life. From this point of view the magnetite is in a cer- tain sense a fossil, but with the important practical advantage over other organic remains, that it need not be dug up in order to prove its existence. These magnetite-bearing rocks always produce disturbances in the compass-needles held in their neighborhood. By a systematic location and comparison of these disturbances the position of the rocks which produce them can be determined with a considerable degree of precision, even when they are deeply buried. Besides their position on the map, the magnetic observations may, and often do, indicate certain other geologically impor- tant tacts, such as whether the rocks are flat lying or highly tilted, the direction of strike and dip, and, in some cases, the depth to which they are buried. The methods employed in the field work were based on those described by Maj. T. B. Brooks,^ who perfected the dial compass and pre- dicted the importance of magnetic methods in geological mapping; but the results reached in interpretation were gradually developed in the progress of this work, as we were daily brought face to face with phenomena which called for explanation. It must be clearly understood at the outset that in the iron ranges of the south shore of Lake Superior magnetite is i-arely concentrated in large bodies, and that, in fact, its known occurrence as such is restricted to a small part of the western Marquette district, where in one producing mine it now forms practically the whole product and in another a variable but usually important part of the whole. It is therefore well understood in the Upper Peninsula that disturbances of the magnetic needle, however great, do not mean workable deposits of magnetite. Whatever significance such disturbances possess is stratigraphical, and properly interpreted may lead to discoveries of rich ore, other than magnetite, in formations to the position and attitude of which the attractions may furnish a clew. But it may be asserted as a general proposition, the essential truth of which has been established by the experience of many years, that in the region referred to magnetic disturbances usually mean that magnetic iron ore in a workable deposit does not exist in the area of disturbance. ' Geological Survey of Michigan, Vol. I, Part I, 1873, Chapter VII. MON XXXVl 22 338 THE CRYSTAL FALLS IRON-BEARING DISTRICT. SECTION II. DESCRIPTION" OF THE MAGNETIC ROCKS. The magnetic rock of the Lower Huronian series of the western por- tion of the Marquette area, which is of special importance to notice, since it forms one of the chief horizons of reference to which our work is tied, is the Negaunee iron formation. It is finely exposed at the south end of the Republic trough; but farther north has been greatly reduced in thick- ness, or locally cut out altogether,' by the Upper Marquette denudation, and, where present at all, is usually drift covered. This rock often possesses a very distinct banding, caused by the alter- nation of layers, in which one of the constituent minerals predominates over the others, sometimes, indeed, to their total exclusion. In the lower part of the formation, quartz and griinerite constitute the bi;lk of the rock, with magnetite scattered somewhat indiscriminately through them. Higher up, the magnetite and quartz relatively increase, until near the top, but below the jasper, the griinerite goes out almost entirely, and the rock consists of quartz bands, heavily charged with magnetite, in alternation with bands of nearly pure magnetite. In the Negaunee formation, as exposed at Republic, the magnetite therefore occurs concentrated in some of the parallel bands and disseminated through the others. In the same district there is another much less prominent locus of mag- netite at and near the base of the Upper Marquette or "hanging-wall" quartzite. Along the strike of this zone, which is of small thickness, the distribution of magnetite is very irregular; and for this and the additional reason that when the magnetic portion of the Negaunee formation comes up to it the disturbances which it produces can not be discriminated from those produced by the latter, the position of the plane usually can not be inferred. In the Menominee district and its extensions there are two horizons in the lower series, characterized by the presence of magnetite. The lower of these is not known to outcrop, but it occurs somewhere near the juiac- tion of the dolomite and the underlying quartzite. The magnetic disturb- ances due to this formation ai'e feeble, but they are quite persistent in the Felch Mountain area, and have thrown some light on the geological structure. ' The Marquette iron-bearing district of Micliigan, by C. R. Van Hise and W. S. Bayley, with a chapter on the Republic trough, by H. L. Smyth : Mou. U. S. Geol. Survey, No. XXVIII, 1897, pp. 531, 537. MAGNETIC OBSERVATIONS. 339 The other formation which produces disturbances is that wliicli I liave correhited witli the Neg-aunee formation and named in a former paper^ the Michiganmio jasper, but which is here renamed the Groveland formation. This rock, while varying a great deal in character, is generally much like that magnetic phase of the Negaunee formation in which the griinerite is rare or absent. From the fact that it now survives for the most part only in shallow and shattered synclines, it often lacks the regular banding; and hematite is always present in greater or less amount. The relative propor- tions of the two iron minerals vary along the strike also. The rock as a whole, however, is very magnetic, but not so strongly so as the Negaunee formation in the Republic trough. In the Felch IMountain rang-e there is still a third magnetic formation, which seems to overlie unconformably the lower series, and is therefore provisionally assigned to the Upper Huronian. This formation consists of ferruginous schists, interstratified with layers of ferruginous fragmental quartzite. It is generally much less highly inclined than the magnetic rocks of the lower series as well as less rich in iron, and the disturbances produced by it are correspondingly small. Besides these rocks of sedimentary origin, with which this paper pro})- erly deals, it may be mentioned that along the Fence River there is a considerable area of metamorphic eruptives, which are often exceedingly magnetic. These are restricted to a definite geological horizon, within which the magnetic disturbances are remarkable for their complexity and irregularity, no doubt as the result of a very irregular distribution of magnetite and of the formations which chiefly contain it. The rocks in portions of this belt outcrop freely, and the disturbances can therefore easily be assigned to the proper causes. SECTION III. THE DISTKIBUTION OF INIAGNETISM IN THE IMAGNETIC ROCKS. Magnetite occurs, therefore, in these Algonkian rocks in different ways. In some instances it is mainly concentrated in nearly pure parallel layers ; in othei-s, it is more or less evenly disseminated through non- magnetic material; and still again it is present in both forms. Moreover, 'Relations of the Lower Menominee and Lower Marquette Beries of Michigan (Preliminary), by H. L. Smyth: Am. .Jour. Sci., Vol. XLVII, 1894, pp. 217, 218, 223. 340 THE CRYSTAL PALLS IRON-BEARING DISTRICT. these rocks have all been folded, more or less strongly, at more than one period ; and wherever they are exposed, they are seen to be inclined to the horizon, often at high angles, and to be traversed by intersecting sets of joint-planes and cleavage-planes, some of which always cut the bedding, and often have been the seat of the development of secondary minerals. By the crossing of these various surfaces, the rocks are divided into small unit masses, at the boundaries of which there is either an actual physical parting or a break in the continuity of the magnetite. It is well known that when a bar magnet is broken and the severed ends are again joined, the two pieces do not unite to form one magnet, but remain as two. It may be conceived, therefore, from the manner of distribu- tion of the magnetite, and the secondary partings existing in these rocks, that their magnetism is seated in an enormous number of small separate magnets, at least one for each of the physicallj^ distinct unit volumes. It is a fact of observation, as will appear hereafter, that the upper surfaces of these magnetic rocks invariably attract the north end of the compass needles and, of course, repel the south end. From this it must be inferred that the small magnets are generally similarly oriented, and have their north ends, which would repel the north end of the compass needle, pointing downward, and their south ends, which attract it, pointing upward. As this is the arrangement that would result from induction from the earth's magnetism, it can be (joncluded further (as, of course, might be assumed) that these rocks are magnetic from the earth's induction. It is also well known that when sevei'al bar magnets are joined in line at opposite poles, the effect upon a compass needle within the range of influ- ence is nearly the same as if the joined magnets were replaced by a single magnet of the combined length. For each member of the pairs of inter- mediate poles, one attracting and the other repelling, is about the same distance away, and their effects so balance each other. The result, there- fore, is to leave one pole unchanged in position and to remove the other to the end of the last magnet added. If enough magnets are added, the iinal result is to carry the moving pole so far away that it has no appreciable influence upon the needle. This is a condition which, from the distribution of the magnetite and the parting surfaces which run through the magnetic rocks, must always be realized more or less completely. It is a necessary consequence of such an ai-rangement of the small magnets that, in the case MAGNETIC OBSERVATIONS. 341 of a tliiu sheet of mao-netic rock lying at a low angle itf dip, the l)urie(l north poles would not be much farther removed than the upper south poles, and consequently the compass needle should be relatively only slightly dis- turbed. This is j)recisely what is found to be the ease. Thus there is firm ground for the conception of the magnetic rocks as made up of sheaves of small magnets, all similarly oriented in a general way and all having their south poles upward at or near the rock surface, while the effective north poles, by the continual addition of similarly ori- ented sheaves below, are carried down, when the rocks are vertical or nearly so, to such depths that their influence is greatly diminished or altogether imperceptible. In equal small areas the individual magnets are no doubt of -very unequal number or strength. This can be proved by holding a swinging needle close to the surface of a magnetic rock, shifting its position without moving it out of the parallel plane and observing the changes in the pointing. These are almost always large and are undoubtedly due to the variations in strength of small areas of the upper poles. In consequence of the law of magnetism, by which the attraction (or repulsion) varies inversely as the square of the distance, the areas immediately surroiuiding tlie needle are very much more important factors in the resultant than those farther removed. When the needle is held higher up, or, what is the same thing, the magnetic rock is buried, the effects are much more regular, since a larger number of the unit areas enter into the resultant with equal weight due t<^ equal distance, and the extremes of indi^'idual ^•ariation are lost in the general mean. Since successive magnetic cross sections over bui'ied rocks showr on the whole a great degree of regularity, we can finally con- clude that the magnetic force of these rocks is seated in an immense, prac- tically an infinite, number of small magnets, which furnish free magnetism at the upper and lower bounding surfaces of the magnetic formation, and that on the average there is about the same number, of about the same aggregate strength, or, in other words, equal intensity in equal areas of these surfaces, if the areas are taken large enough. SECTIOIf IV. THE INSTRUMENTS AND METHODS OF WORK. The instruments used in this work are simple and well known. The dial compass is an ordinary compass, carrying a 2^-inch needle swinging inside a cii'cle graduated to degrees, which is further supplied with a grad- 342 THE CRYSTAL FALLS IRON-BEARING DISTRICT. uated hour circle. It is therefore a portable sundial. Tlie gnomon is a thread, which is attached at one end to the center of graduation of the hour circle near the rear sight and at the other to a point in the forward sight so taken that the angle made by the thread with the plane of the hour circle is equal to the latitude of the place. When this instrument is leveled and set up in the meridian on a sunny day, the thread will cast a shadow on the hour circle at the correct apparent solar time, from which mean time may be determined by applying the equation of time. Conversely, if it is so set up that the shadow of the thread falls on the correct apparent time, the sights of the instrument are in the true meridian. In this position the declination of the horizontal needle maj" be read off from the graduated circle. At work, this instrument is mounted on a light Jacob's staif, or it may be held in the hand. The Jacob's staff, although often inconvenient to carry, is preferable, as with it the readings are all taken at the same height above the ground and the leveling is more exact and steady. In a correctly constructed instrument, Avith good time, the readings may be made to half a degree. Coirectness in the time, however, is indispensable to good work, and this is best secured by keeping a standard watch in camp and referring the working watches to it daily. The dip needle needs no description. In geological work that form known as the Norwegian, in which the needle is pivoted on a universal joint, is not so useful as the type in which the needle is rigidly confined to one plane. In taking the readings, this plane in which the needle is free to swing is made to coincide with the vertical plane determined by the pointing of the horizontal needle. The circle is graduated to single degrees, and with skillful work the readings ai'e reliable to one or two degrees. It ma}- be added that the south end is weighted, in order, either partly or completely, to balance the vertical component of the earth's force. It was found better not to balance it completely, but only to such an extent that the north end of the needle would dip about 10° (the graduation zero being horizontal) in an area removed from local disturbances. It is no doubt desirable that all the dip needles used in the same work should be brought to approximately the same index error, in order that the readings may be more directly comparable. In practice, however, it was found quite impossible to keep our three needles in unison, on account of the rough usage to which they were unavoidably subjected. As, however, the form MA(}NET1C 015SEKVATI0NS. 343 of the (lip curves is veiil.y the subjeet souyht, and since these, in the pres- ence of consideraljle disturbances, are sensiljly indejjendent of .small differ- ences in the index error, it is not indispensable that the needles should be exactly together. These instruments are simple, and, of course, do not give precise results. But the observations are rapidl}' and chea])ly made, and to a sufficient degree of acctn-acy for the end in view. It may be stated again that the object is to detect and compare relative magnetic disturbances, and to find out the bearing of these disturbances on the distribution and attitude of the rocks which produce them. For this purpose the instrmnents are exceed- ingly well adapted. The field work was canned out by parties of two men each, one of whom, a skilled woodsman, carried along the line and observed the hori- zontal needle, while the other read the dip needle, kept the notes, and atteiided to the geology. According to the general plan of the field work, a series of parallel lines was run either north and south or east and west across each section. The direction of the lines of travel was chosen so as to cut the strike of the rocks at the largest angle. The jn'obable direction of strike for each day's work could be inferred in advance from what had ji'one before. If it were more nearly north and south than east and west, the traverse lines were run east and west, and vice versa. These directions were in many cases not the most desirable for the magnetic work alone, but the choice was practically limited by the lines of the United States Land Survey, which give for each square mile eight points of departure (at the four corners and four quarter posts of each section), which are generally identifiable on the ground. On these lines of travel the instruments were read at various intervals, from 6 to 10 or 100 paces, depending upon the local complications. The intervals Ijetween the lines varied from one- sixteenth to one-fourth of a mile, and were determined not onl)' by the magnetic complications, but by the character of the surface, it being especially desirable that the ground should be so closely covered that no outcrop could escape detection. The distances along and off the lines of travel were measured by pacing. The general accuracy of the pacing is remarkable, and is essentially within the platting error of the scale of the maps. The aA^erage closing error for August, 1892, during which about 100 miles' of traverse lines were run, was 20 paces per mile, or 1 per cent. 344 THE CRYSTAL PALLS IRON-BEARING DISTRICT. Two-thirds of tlie errors averaged 10 paces per mile, or 1 in 200, while the maximum was 1 in 30. But this was better thau the average for the season. The observations at each station consisted in a reading of the horizon- tal and dip needles. When there was no local magnetic disturbance, the horizontal needle would come to rest in the magnetic meridian, which in this region is about N. 2° to 3° E., or almost coinciding with the true merid ian. The dip needle, when held in the same meridian, would indicate the index error. When, however, disturbing mateiial was present, the horizon- tal needle would point to the east or west of the magnetic meridian, at an angle determined by the direction of the resultant of the horizontal com- ponents of the earth's and the local forces. The dip needle would come to rest in the same vertical plane, at an angle with the horizon determined by the amount and direction of the three forces, the whole pull of the earth's force, the whole pull of the local forces, and the balancing weight, and in general would show a downward deflection. After making and recording the set of observations at a station, the party proceeded to the next, and so on to the end of the day. At the end of each day, or as soon as possible afterwards, the day's work was jjlatted on a large-scale map, on which the readings of the horizontal needle were represented by short arrows drawn through the stations, turned east or west of the true meridian, as the case might be, and carrying the amount of declination written at the arrow point. The dij) observations were laid off to scale immediately below the stations, measuring all from the same horizontal line, and the })oints thus established were connected by a free-hand curve. 8ECTIOX V. FACTS OF OBSERVATION AXD GET^ERAIj PRINCIPI.ES. I. OBSERVED DEFLECTIONS WHEN THE STRIKE IS NORTH AND SOUTH AND THE DIP VERTICAL. If a magnetic rock, striking north and south and dipping vertically, is crossed by an east-and-west traverse, it is found, as the disturbing belt is approached, say from the western side, that the horizontal needle points toward the east of north, and that this easterly pointing gradually increases to a maximum. Continuing east from the maximum point the eastward declination decreases, and soon a station is reached at which the needle points due north. Still farther east the declination changes to west- ward, and soon thereafter reaches a westward maximum, beyond which MAGNETIC OHSERVATIOTSrS. 345 again tlio westward pointing in its turn gradually decreases, until finally the needle reaches its normal eastward declination, after passing through a second zero. The dip-needle readings at the same stations generally mcrease slowly at first, and then rapidly, and soon reach a maximum at the first zero point between the converging arrows ; beyond this to tlie east they decrease corresj)ondingly, so that the dip curve is symmetrical east and west of the maximum These statements will l)e made clear bv a reference to fig. 15, which represents an actual traverse in T. 45 N., R. 31 W. 2. DEFLECTIONS OF THE HORIZONTAL NEEDLE. It is evident that in crossing- a rock belt which stretches away indefi- nitely in both directions, only a limited part of it will affect the readings on » 10 10!^ II IS 15 14 II 8K 13 7 93< II 13% ISii 13 11>J ■0>i _b:p Cui V .-^ Fig. 15.— Magnetic cross .sectiiin in T. 45 N., R. :il W. a given cross section. Since the })ull of the poles of a magnet on a com- pass needle diminishes with the square of the distance of separation, it follows that the limits to the material that would noticeabl}' disturb com- paratively insensitive instruments would soon be reached. If we consider for the moment only the horizontal components, and call the distance a (fig. 16) at which the needle would respond to the attraction of material possessing the magnetic force of that with which we are dealing, then at any station, P, the material inside a circle di'awn with P as a center and radius a (shaded in the figure) would exert force on P, the material outside would not. If the circle drawn from a station, P', does not reach tii the magnetic belt, the needles at P' will not be disturbed.^ For reasons of symmetry, it is seen that the attraction of the magnetic ' The actual distances at which disturl)eiieral Hr has its iniiiimum vahie when H':= — H sin /?. When the rock strikes north and south or /?rr(), 11^ is a minimum when H':rrO. 4. HORIZONTAL AND VERTICAL COMPONENTS WHEN THE MAGNETIC ROCK DIPS VERTICALLY. If we assume that the magnetic rock has a uniform strike in any direc- tion, a vertical dip and a surface width or thickness equal to 2a, it is easy to show that the horizontal and vertical components of the rock force are given by the following equations, where x is the horizontal distance of the station of observation from the middle plane of the formation, h is the depth of surface covering, assumed to be uniform, and a? is a constant. GO -'^^li^J^(x-af ^^ ^'-'> ^tan->^±^-tan-i'-^^' J (7) GO TT/ In equation (6 ) — rr when ;r — ; therefore a point of no deflec- GD tion of the horizontal needle is found vertically over the middle jjoiiit of the magnetic rock. It is also evident that at corresponding stations on opposite sides of the middle point, the horizontal components are equal, but act in opposite directions. To obtain the points of maximum or minimum values of the horizontal component, we differentiate the right-hand side of equation (6) with respect to X, and place the result equal to zero. This gives irrr±V'F+^2 (8) which determines two points, at equal distances from on opposite sides of the rock, at which the horizontal component has maximum values. Writing for X the measurable distance d, and squaring, we have d''=}i'+a^ (9) The thickness of the magnetic formation is therefore always less than the distance between the points of maximum horizontal deflection, except when 7irzO, or the rock is uncovered, in which case the thickness and sepa- ration of the maxima are the same. 350 THE CEYSTAL FALLS IRON BEARING DISTRICT. By differentiating tlie right-hand side of (7), with respect to x, it is easy to show that V has a maximum value when .czrO. When the rock strikes north and south, this also corresponds to a minimum value of H„ as has already been shown; and, therefore, by a reference to equation (5) it is readily seen that a point of maximum dip coinciding with a point of no horizontal deflection is in that case found over the middle plane of the bui'ied magnetic rock. Where the strike is inclined to the meridian, the points of maximum dip and zero deflection will not coincide, since the maximum value of V does not occur at the same station as the minimum value of H^. As has already been shown, H^ is a minimum when H'=H sin fi (/? being the ang-le of the strike), and this is in general on the side of the rock on which the angle made with an east and west traverse is obtuse. The point of maximum dip will be situated on the same side of the rock between this station and the point of no horizontal deflection, and will approach the latter as the strike approached the meridian, and also as V increases relatively to H'. With strongly magnetic rocks the points of no deflection and maximum dip practically coincide on maps platted to the scale of 4 inches to the mile, except where the strike is nearly east and west. 5. HORIZONTAL AND VERTICAL COMPONENTS WHEN THE MAGNETIC ROCK DIPS AT AN ANGLE. Under the last heading it was assumed that the magnetic rock dips vertically, and that it continues indefinitely downward at this angle; In consequence of this assumption, and also of the conception of the manner in which magnetism is distributed through magnetic rocks, it has been con- cluded that the north poles of the rock, which repel the north end of the compass needle, are situated so far below the surface that their effect may be neglected. Therefore we have taken into account only the south poles, which are situated at the rock surface. In the case of rocks which do not dip at high angles this assumption can not safely be made, and the influence of the bottom poles must be taken into account. Since the force of these poles acts in opposite direc- tions from that of the upper poles, and since they are more deeply buried, it would seem that their influence in general must be to diminish the total force which acts upon the needles at any station, and therefore that the MAGNETIC OBSERVATIONS, 351 deflections botli of tlie horizontal and dip needles caused by the same rock should be less in amount, ceteris paribus, where that rock dips at a low ang-le than wliere it dips at a high angle. In tlie course of the field work certain peculiar deflections of the needles were encountered in traverses across rocks dipping at moderate or low angles. These were not thoroughly understood at the time, but the cause was believed to be connected with the angle of dip of the rock. For example, it was found along traverses crossing certain north-and-south- striking rocks, which were known to have a westward dip that may have been either high or low, that the two points of maximum deflection of the horizontal needle were not situated at equal distances from the point of no deflection between them, but that the distance of the western maximum was much the shorter. It happens in this region that no east-dipping rocks occur which are so far removed from other magnetic formations as to be out of range of their possible influence, but, so far as they go, traverses across these showed that the nearer maximum was situated on the eastern side of the point of no deflection. It therefore seemed probable that the cause of the inequality in the distances from the zero point to the maxima was the dip of the rock, and that the dip was in the direction of the nearer maximum. If the magnetic formation has a surface width nz b, is uniformly buried to the depth h, and dips at the angle ^, then, if A — tan z/, it may be shown that the horizontal and vertical components at any point P, the hori- zontal distance of which from the lower edge of the formation is x, are given by the following equations: H' _ \' , h'+x- 2 A S 4. -1 x—h — Xh , ,x — Xh < tan '- — — =- — tan^ h > Xx—Xl) + h Xx+li] ^^^^ -1 — - =2 j tan-i^-tan-i^ w ( h h V+x"" . 2 + i-nr2lo&i tan-Kli^tML±^)_tan- i+ii^ [. . . (11) A-(l+u;) Ji—XJjj^x x) 352 THE CRYSTAL FALLS IRON -BEARING DISTRICT. If A:iicc> , and the coordinates are referred to axes in the middle of the rock, these equations reduce to equations (6) and (7). By differentiating the right-hand side of equation (10), placing the result equal to zero, and solving for x, the positions of the stations at which H' is a maximum may be determined. This gives: ■^-"^^=^ 2l ^ ^^ Calling the difference of the roots, or the measurable distance between 2a the maxima, 2d, and substituting for h its value -■ — -7- 2a being the true thickness of the rock, we have: d2_^jL+^' ■. . . . (13) sm ~/l For rocks of high dip, therefore, the distance between the maximum points is but little greater than it would be were the dip vertical, and it increases inversely as the angle of dip. A general algebraic determination of the points at which H' is and V is a maximum is impossible, since it involves the solution of equations of a degree higher tlian the fifth. However, hj assuming numerical values XT' Y' for A, h, and a (or li) curves expressing the relations between — and — and GO GO X can be plotted, from which the maximum and zero points can be deter- mined in any desired number of special cases. Let us first assume that A zr 3 (or that the rock dips at an angle of about 70° 34'), I1 — 2, and a — 6. The ordinates to the curves of fig. 1, PI. XLVII, TI/ Y' give the values of — and — corresponding to different values of x. The GO GO ordinates to — do not reijresent the deflections S of the horizontal needle GO from the meridian, but quantities that are connected with those deflections by equation (1). The deflections, however, vary as H' varies, and will have maximum and minimum values at the same points. From this figure it appears, first, that the nearer maximum is situated on the dip side of the rock; secondly, that the point of no deflection is not over the middle plane, but is nearer the upper edge; thirdly, that the hori- zontal force of the rock is numerically less at the nearer than at the more U. 6. OeOLOGlCAL SUXVE^ MONOGRAPH XXXVI PL. XLVII (^) (^) \, RELATIONS OF MAGNETIC BEDS TO VARIATION AND DIP. MAGNETIC OBSERVATIONS. 353 distant m;ixinium, and, fourthly, that tlie distance between the maximum points is nearly the same for the inclined I'ock as for the vertical. In I'l. XLVII, fi<>-. 2, the constants have the same numerical values as before, e.\ce]it //, which now zr4 instead of 2. The rock is thus bm-ied to twice the depth of the former case. The same conclusions are true for this case as for the first. The zero point is still nearer the upper edge of the rock, and the maxima are farther apart. Let it next be assumed that A = 0.5 (or that the rock dips at an angle of al)out 26° 34'), /(rr2, and (i-^6. These data lead to the curves of PI. XLVII, fig. 3, in which, as in the case of the rock ot higher dip, the maximum })oints are unsymmetrical to the }joint of no deflection, the nearer lying on the dip side. In PI. XLVII, fig. 4, we have a rock of the same thickness and dip at a depth /;^4; and the same conclusions hold true. From these four curves, which represent formations dipping at high and moderately low angles, and buried to depths which are in the one case small and in tlie other great, relative to the thickness, it is probably safe to draw the following general conclusions: («) The direction of dip of a magnetic formation is toward the nearer and (for north-and-south-striking rocks) the numerically smaller maximum. (li) The point of no deflection between the converging maxima is not situated over the middle plane of the formation, but is nearer the upper edge. But with increasing depth and diminishing angles of dip, this point may pass beyond the upper edge. (c) With slightly inclined rocks, for moderate deptlis of siu-face cover- ing, the disturbances are spread out over a much wider zone on each side, and the maxima are less sharp, particularly the maximum on the dip side. Under these 'circumstances irregular and anomalous deflections would be expected in practice, as will be seen in the following sections. (d) The curves of the vertical component show maximum values near the zero value of the horizontal component only in the case of the rock of high dip. In the case of the rock of lower dip, the vertical component has a negative value, or is directed upward over a wide zone on the side of the rock opposite to the dip side. Over this zone the readings of the dip needle will be less than normal, or even negative if V > H tan 9. This is in accordance with the facts of observation. MON XXXVI 23 354 THE CRYSTAL FALLS IRON-BEARING DISTRICT. It is also interesting- to determine the relative values of the horizontal components at the two maximum points, for different angles of dip, and for different ratios of /; to a — that is, for different depths of burial. Fig. 18 shows in graphical form these relations for all angles of dip between 90° and 0° for seven different ratios between h ^ ^7. and h =z 4«. The ordinates 50 to the dotted curves (concave upward) give the relative values of the hori- zontal component at the maximum point on the dip side (B maximum); those to the full cur^'es (convex upward) the values at the maximum point 90^ 80" 70J 60° 50° 40° 30° 20o IQo 0° Flo. 18.— Curves showing the relations between the horizontal components at the points of luaximum deliectioo, lor rocks dipping at various angles and buried to various depths. on the other side of the rock (A maximum). These curves show that the horizontal component at the B maximum is always numericallv much less than at the A maximum, and that for moderate depths of burial it diminishes very rapidly at both points with small ang-les of di]). 6. DETERMINATION OF DEPTH. The relations between the dip and thickness of a magnetic rock, the distance between the horizontal maxima, and the depth of covering are given in equation (13). By assuming numerical values for ^, and either for a, the half thickness of the rock, or for r/, the half distance between the maximum points, it is easy to plot the curve which expresses the relations ma(ini<:tic observations. 355 between the other two quantities. It' d is taken as tlie constant, the equa- tion re})reseuts a circle; it' (i, it represents a hyperbola. This e(iuati(in wvax have a useful application in making it possible to judge, in advance of actual test pitting, of the probable depth of surface covering over a magnetic rock, for which the original assumptions are ful- filled, and the numerical values of ^, «, and d are determinable. It will be remembered that the assumptions upon which equation (13) rests are the fundamental ones of Section III, and also that the rock has a uniform strike. In practice, the uniformity of the strike can be established Ijy other traverses on each side of the one in question, and J, the angle of dip, may usually be determined by the outcrop of other formations in the same series. For any jjractical application it is also necessary that a (half the thickness of the rock) and d (half the distance between the stations at which the horizontal component is a maximum) should be known. From an inspection of the equation it is evident that any close determination of h, except for great depths of covering, depends upon very precise knowl- edge of the ratio between a and d. The practical difficulties in the way of the measurement of a and the ever-present probability that the rock may vary from point to point, not only in actual but in effective magnetic thick- ness (which is what a actually signifies), make it clear that for the most part h can only be found approximately. Also, in the case of a rock striking due east and Avest the methods fail, from the fact that 2d can not be determined on the ground. The determination of h is therefore hedged in with important limita- tions; yet in many cases the information supplied by the equation may be very useful. The difficulties in the way of measuring a are disposed of in the event that along one traverse on the strike of the rock li is known, as it may be, by the sinking of a test pit. This value of li at once gives a value of a, which may be used on other traverses across the same rock with much more accurate results, in the lack of disturbing factors, than if a were known only by measurement It should be added that when the traverse crosses the strike of the magnetic rock at the angle ,)', the distance d meas- m.*ed on the line of the traverse must be multiplied by sin y in order to get the value of d to be ixsed in the determination of /;, and also that h must be . corrected for the height of the instrument. 356 THE CRYSTAL FALLS IRON-BEARING DISTRICT. More general inforinatiou as ti) relative depths of Ijurial is also given by the dip curves. It is easily seen that where the superficial covering is small the vertical component of the rock force must remain small, except immediatel}' over the rock. This condition is, therefore, indicated bv steep slopes in the di^) curves. On the other hand, where the depth of covering is considerable, the vertical component increases slowly and steadily, Ijegin- ning at stations at a distance from the rock, and the resulting dip curve approaches the maximum with gentle slopes. 7. SUMMARY. (1 ) The strike of a magnetic rock is given by the line joining the points on successive traverses, at which the horizontal needle is not deflected from the local meridian between the converging aiTOws, or at which the dip angles are a maximum. When the rock is vertical, this line lies in the middle plane of the rock and fixes its position. It may be called a line of mag- netic attraction. (2) The dip of. a magnetic rock is toward the nearer horizontal maxinuini. (3) The thickness of the magnetic formation must, if buried, always be less than the distance between the maximum points. (4) Where the superficial cover is not very great, a change in the dip of a magnetic rock from moderate or high angles to low angles is attended with a rapid decrease in the values of the horizontal component, with a corresponding decrease in the deflections of the horizontal needle. SECTION VI. APPLICATIOT«rS TO SPECIAIL, CASES. In the preceding section certain general conclusions have been estab- lished with regard to the relative positions of the stations at which the horizontal and vertical components of the force of a magnetic rock have maximum and zero vakies. The deflections produced by these comjjonents from the ^Jositions which the magnetic needles assume under the action of the earth's force have maximum and zero values at the same stations at which the components have maximum and zero values, and therefore the conclusions as to the relative positions of these points are true for any anffle of strike. But certain numerical relations between the deflections depend upon the orientation or strike of the magnetic formation and ujion the direction of dip, and these will now l)e considered. MAGNETIC OBSERVATIONS. 357 1. THE MAGNETIC ROCK STRIKES EAST OR WEST OF NORTH AND DIPS VERTICALLY. Let US first take the case of a rock strikino' east of north. At the stations within range of the h)cal influence on the east side of such a rock belt the liorizontal needle is pulled west of the meridian, reaches a west- ward niaxiiuum, tlieii points north, then on the west side of the belt, east of the meridian, and reaches an eastward maximum. It is observed, how- ever, that the westward deflections on the east side of the belt are generally not so great as the corresponding eastward deflections on the west side of the belt. The reason for this is easily seen. At each station east of the belt the local })ull acts along the normal to the belt di-awn through the station. This normal makes with the local magnetic meridian an acute angle. The needle will come to rest within this acute ang-le alonar the line of the resultant of the horizontal components of the two forces, the earth's and the local force, which determine its position. However strong the local pull may be, the horizontal needle can not be deflected past the normal. At the corresponding stations on the west side of the disturbing belt the local pull also acts along the normal from the station to the belt, and has the same numerical value. But in this case the normal makes an obtuse angle with the magnetic meridian. For two points equall}' distant from the magnetic belt, one on the east and the other on the west, the resultant for the western point will, therefore, make a larger angle with the meridian than that for the eastern. On the other hand, when the rock strikes west of north, it is observed that the horizontal deflections are greater on the east side than on the west, and the explanation is entirely similar to that given above. The dip-needle observations at the same stations show general })he- nomena quite like those in the case in which the strike of the rock coincided with the meridian. They gradually increase to a maximum near the sta- tion, where the horizontal needle stands at zero between the converging arrows, and gradually decrease from this maximum on the other side. It . is noted, however, that the readings are not equal at corresponding stations on opposite sides of the maximum. When the strike is east of north, the western station shows a higher dip than the eastern; when the strike is 358 THE CRYSTAL FALLS IRON-BEARING DISTRICT. west of uorth, the eastern station shows a higher dip than the corresponding' station on the west. Generally stated, then, the stations on that side of the magnetic rock on which the angle between the strike of the rock and the line of traverse is obtuse show greater dip angles than the corresponding stations on the side on which this angle is acute. As the angles of dip are represented graphically by a continuous curve, this is the same thing as saying that the dip curve is steeper on the side of the acute than on that of the obtuse angle. These facts are easily explained by the following considerations. The vertical components tend to lower the needle, and would cany it to a ver- tical position except for the action of the horizontal forces, which tend to keep it horizontal. At any station on the acute-angle side of the magnetic belt the resultant of the two horizontal components is larger than at the corresponding station on the obtuse-angle side, the two being represented by the longer and shorter diagonals of a parallelogram. Since the vertical forces are the same at the two stations, it follows that on the obtuse-angle side the angle of dip must be larger than on the acute-angle side. Or, expressed algebraically, since H,. is the only variable on the right-hand side of equation (5) it is evident that tan a, and therefore a, the angle of dip, increases with a decrease in H,. If the rock dips at an angle less than 90°, these results are either intensified or gi-eatly modified, depending upon the direction of dip. It 4 was shown in the last section that the horizontal component of the rock force is smaller on the dip side. If the strike and dip are both toward the same side of the meridian (e. g., if the strike is northwest and the dip south- west), it is evident that the numerical difference between the deflections of the horizontal needle on the two sides of the rock will be still greater than if the rock were vertical. On the other hand, if the strike and dip are toward opposite sides of the meridian (e. g., if the strike is northeast and the dip northwest), the diff"erence between the deflections on the two sides is less than for a vertical dip, or may even be reversed. The deflections of the dip needle in the case of rocks dipping at angles less than 90° are also greatly influenced by the direction of dip. If strike and di}) are toward the same side of the meridian, the diff'erence noted above in angle of dip on the two sides of the rock is neutralized, and the dip curve tends to become symmetrical; while if thev are toward opposite MAGNETIC OBSERVATIONS. 359 sides of tlu' meridian, the difference is increased. It is to be noted that the intiuencc on hotli instnnnents of tlio direction and angle of dip of the rock becoini's weakened with an increase in surface covering. 2. THE MAGNETIC ROCK STRIKES EAST AND WEST. When a vertically dipping magnetic rock strikes east and west, or nearly so, the traverse lines must be run north and south so as to cross it as nearly as possible at right angles. In approaching such a belt from the south the instruments give little warning. The readings of the horizontal needle .show either no deflections, or else very slight deflections, from the mairnetic meridian. Past the middle of the foi'mation the horizontal needle is strongly deflected, often through an angle of 180°, so that it may point due south. But as the magnetic rocks having this strike which were encountered in our work were not deeply buried, and had also quite irregular upper surfaces, generally the needle pointed either east or west of south on account of the weight which the nearness to the surface gave to the adja- cent material, either from the irregular distribution of magnetite or from the protrusion of small masses above the general level. Continuing noi-th, the horizontal deflections gradually diminish and eventually disappear. The behavior of the horizontal needle is explained in the same way as in the preceding cases. The position of the needle at any station is deter- mined by the resultant of the horizontal components of the two forces — the eai-th's force and the rock force — that act upon it. South of the magnetic rock these two components act in the same direction and essentially in the same line, since the magnetic meridian practically coincides with the true meridian. The resultant, therefore, is equal to their sum and coincides with them in direction, and consequently there is no deflection. North of the magnetic rock the two horizontal components act in opposite directions, and when they are in the same line the needle takes up its position in the direc- tion of the greater, which determines that of the resultant; when H' is greater than H (which often happens near and north of the rock) this direc- tion is due south. When the two components do not act in exactly the same line, the needle will point east or west of south at an angle which depends on the angle between the two forces and their ratio. Still farther north the horizontal component of the rock force dimin- ishes rapidly and we consequently first pass tln-ough a zone of large and 360 THE CRYSTAL FALLS lEOX-BEAEING DISTRICT. diminishiug deflections to the east or Avest, depending on the side of the meridian on which this component falls ; and finally, when it becomes insensible, the needle rests again in the meridian. In the case of a rock striking east and west, the points at which the horizontal component of the magnetism of the rock has maximum values become indeterminate by the methods hitherto described, from the fact that throughout the traverse the two components act in or nearly in the same line, and the deflections from the local magnetic meridian, therefore, do not indicate the relative strengths at difi"erent stations of the horizontal com- ponent of the rock force. The dip-needle readings for an east-and-west-striking rock are as fol- lows: At some distance south of the rock the angles are constant at the index error. As the rock is approached, the angles of dip depend upon the depth of burial. If the surface covering is considerable, an increase in the dip angles begins at a considerable distance away, and progresses continu- ously as the magnetic belt is approached. If the rock is near the surface, the dip needle shows either the constant index error or else angles of dip less than the index error for all stations south except those very near the southern margin of the rock. The maximum reading is attained north of the middle plane of the rock, at a distance from it which also depends upon the depth of covering. Farther north the dip angles decrease slowly and are in general greater than at the coiTesponding stations south. The form of the dip-curve, therefore, shows a steeper slope south of the magnetic rock than north of it. The reasons for these differences will be evident from the following considerations. Let it be supposed, for the sake of simjjlicity, that throughout the north-and-south traverse the two horizontal components act in the same line in the meridian. At any station south of the magnetic rock they act in the same direction, and their resultant will be their numerical sum. At the cori'esponding station north they act in opposite directions, and their result- ant will be their numerical difl^erence. The angle of dip is giveii by equation (5): V'+ H tan & tan a rz !— yt For the two corresponding stations, V will be the same. The other quantities are all constants except H^. For the south station H^rz H' -f H; MAGNETIC OBSERVATIONS. 361 for t\\v north station, 11^ = 11 — 11', where 11 ;iufl H' are, respectively, the horizontal components of the magnetism of tlie earth and of the rock, as before. The numerator of the right-hand side of the etjuation will be the same for both stations, while the numerical value of the denominator will be less for the north station than for the south. Consequently tan a, and therefore «, will l)e greater for the north station. For great depths ' of superficial covering, however, these differences become almost imperceptible, owing to the fact that H' is so small that Hr is essentially tKe same at the two con-esponding stations. The tendency, therefore, as h increases is for the dip curve to become symmetrical. In the special case in which H'r= — H, H,.=: 0, and the dip needle stands at 90°. This can only take place north of tile rock, and may, depending on the strength of H', be found at two stations, one on either side of the station at which H' is a maximum. At the same stations the horizontal needle is not acted on by any unbalanced force, and rests indifferently in any position. The dips less than normal which are often observed at stations south of a magnetic rock which lies A^er}^ near the suiface are also easily under- stood by a reference to equation (5). At these stations the resultant pull of the rock is so nearly horizontal that the vertical component V is very small in comparison witli the horizontal component H'. If V is a negligible quantity, equation (5) becomes TT tan a z= jj — Yn ■ *^^^ ^ rl-|-xl In sucli cases the angle of dip is therefore less than the index error. With north or south dipping rocks, where V is negative, tan a becomes negative when V'>H tan 0. 3. TWO PARALLEL MAGNETIC FORMATIONS. The cases so far considered have involved only one belt of magnetic rock, which has been assumed to ha^•e a uniform dip in one direction, or, in other words, to be a monocline. In practice, however, owing to complexi- ties of structure and other causes, which will be considered hereafter, it frequently happens that two or more approximately parallel belts are Ibund within the range of one another's influence. Under these circumstances 362 THE CRYSTAL FALLS IRON-BEARIKG DISTRICT. the effects produced upon the magnetic needles are corresponding!}^ com- pHcated. For the purposes of ilhistration it is sufficient to consider a few extreme H' V cases, and to represent the vahies of — and — for these graphically. Let it tirst be assumed that the two parallel belts are vertical, that the distance between them is 8, and that a^3, and /?zr2. This represents the conditions when the distance of separation is large compared Avith /;. The ordinates to XT' the curve of PI. XL VIII, fig. 1, give the values of — which correspond to the different stations of observation. Those parts of the curve which are above the horizontal axis of •coordinates represent the portions of the trav- erse in which H' is directed toward the west; the parts below, those in which it is directed toward the east. It is seen that besides the middle point there are two other points of no horizontal deflection, which do not exactly correspond with the points A'ertically over the middle of the magnetic forma- tions, but are somewhat nearer the adjoining edges; and also four points of maximum deflection, one on each side of each rock. The maximum points inside of the two formations have smaller deflections than those outside, as shown by the relative lengths of the ordinates to the curve, and also between the inside maximum points the horizontal components are directed away from the middle point. V The curve of — , represented by the dotted line, has two maximum valvies, which fall nearly over the two rocks. If next we assume that the distance between the rocks is 8, and that fflrz3, and h-=8, we obtain the curve of PI. XLVIII, fig. 2, which represents TT/ the value ol — when h is large compared with the distance of separation. This case shows but one point of no horizontal deflection between the two rocks and but two points of maximum deflection, one on the outside of each. V The cvirve of — , represented by the dotted line, shows the interesting feature of three maximum points, one at the center and one over each rock. If h were relatively a little greater, these would evidently coincide at the center. U. 8. GEOLOGICAL 8URVEV MONOGRAPH XXXVt PL. XLVIII RELATIONS OF MAGNETIC BEDS TO VARIATION AND DIP. MAGNETIC OBSERVATIONS. 363 It" the two parallel toriuatious are not vertical, l)ut dip in tlie same direction at the same angle, the resulting curves are somewhat ditlerent. I'l. LXVIII, tigs. 3 and 4, show two cases in which the elements are the same, excejit the depth of covering and the thickness of" the intervening non- magnetic material. Here the rocks dip at an angle of 71 ^ 34'. and the width at the rock surface is 6.3 for each. In fig. 3, PL XLVIII, where Ji-=2 and the width of the nonmagnetic bed is 10.7, and the covering is, therefore, relatively small, the presence of two rocks is distinctly shown by the curves of both components, and the chief result of their interaction is to introduce an additional point of no hori- zontal deflection between them, on each side of which the horizontal arrows diverg-e. The positions of the maximum and of the other zero points are hardlv disturbed, and consequently the direction of dip is very clearly indicated. In fig. 4, PI. XLVIII, where /?rr4 and the formations are separated by nonmagnetic material 4.7 wide, there is but one zero point, nearly over the middle of the upper formation, toward which the pointings of the hori- zontal needle converge. West of this are two points of maximum eastern deflection, between which a faint minimum represents the backward pull of the lower formation. If the two magnetic formations are parallel in strike, but dip toward each other at equal angles, the resulting curves of the two components are shown in PI. XLVIII, figs. 5 and 6. Fig. 5 illustrates the eff'ects on a s^Tichne with steeply dipping sides, the supei-ficial covering being relatively shallow. These conditions result in a point of no horizontal deflection over the mid- dle of the trough with diverging arrows on each side, and besides a point of no horizontal deflection over each rock, toward which the aiTows con- verge. The positions of the two maximum points for each rock, and of the zero between them, is nearly the same as if the other rock were absent, and conseqiiently the fact that the rocks dip toward each other is clearly indicated by the unsymmetrical distances. In fig. 6, PI. XLVIII, the depth of the rock surface is much greater relatively to the inside distance between the legs of the syncline, and the dip is flatter. In this case there are but two points of maximum deflection, one on each side of the syncline, and but one point of no deflection, over the middle of the trough. The maximum points represent the outside maxima 364 THE CEYSTAL FALLS IRON-BEARING DISTRICT. of the former case, and the result of the interaction of the two legs is to increase the numerical values of these, as" well as to bring them nearer too-ether. It is evident that the deflections of the horizontal needle in this case could hardly be distinguished from those that would be produced by a single vertical fonnation buried to a considerable depth. Let it next be supposed that the two rocks dip toward each other at different angles, the strikes remaining parallel, and also that the rock of lower dip is buried to the greater depth. This, then, is a case in which the magnetic effect of one limb of the syncline is much stronger than that of the other. In PI. XLVIII, fig. 7, the curves of the two components are given for the special case in which the right-hand limb of the synclinal dips at an angle of 90°, and has a surface covering /i=r2, while the left-hand limb dips at an angle of 26° 34', and has a surface covering h=i4:. It is interest- Fig. 19 — Truucated anticlinal fold with gently dipping limbs. ing to compare the theoretical results of this figure with the curves of PI. XLVIII, fig. 8, which represent deflections actually observed, and not components. In the latter figure the strike of the two rocks is represented by the heavy lines. The two rocks are the same formation, brought up by folding on opposite sides of a synclinal trough. The synclinal is slightly pushed over, so that the eastern rock dips nearly vertical, while the western has a much lower dip toward the east, and is also more deeply buried. These facts rest on independent evidence, yet they might all be inferred from the observations recorded in this figure. The dip curve in this case shows two distinct maxima, a smaller under MAGNETIC! OKSERVATIOXS. 865 the zone of retardation and a lar«iCT over tlie ))oint of no liorizontal deflec- tion, uliicii correspond i-espeotively to the two niaji'netic i-ocks. If the two formations are parallel in strike, l)nt diii away from each other, the cnrves of the horizontal and vertical comjjonents for different anjrles of dij) and different relations of thickness and depth of coverino- are shown in figs. 19 and 20. In fig-. 19 the formations are widely separated, k is relatively small, and the angles of dip are equal and low; the inter- action of the two rocks therefore extends over a narrow zone only, and the curves of the components clearly indicate the presence of two formations and the direction of dip of each. In fig. 20 the anticlinal is so truncated that magnetic material occupies the whole space on the rock surface between the outer boundaries of the two formations. The angles of dip are equal, and are hig-her than in the preceding case, while the depth of cover- ing is relatively much greater. The horizontal component is zero in the axial plane of the anticlinal, and has maximum values at two points, one on each side of the zero. The ver- tical component is a maximum at one point, also in the axial plane. The deflections pro- duced by these conditions could not be distinguished in practice from those produced by a single vertically dipping fomiation. In general, therefore, when two magnetic formations lie within range of each other's influence, the deflections are determined by the relative magnetic strengths of the .two rocks, hj their distance apart, by their strike and dip, and by their depth of burial. It is evident that for certain given relations among these factors the special cases above described will occiu-, and it is found that they really do occur in practice. For other relations it is not possible to make a general statement either as to the number or the position of the maximum and mininuini points. rig, 20.— Truncated anticlinal fold with steepl.y dipping limbs. 366 THE CRYSTAL FALLS lEON-BEARING DISTRICT. SECTION VII. THE INTERPRETATION OF MORE COMPIjEX STRUCTURES. The existence of two parallel belts of magnetic rocks may be accounted for geologically in more than one way. They may represent two distinct formations occurring at different horizons in the same series, or they may represent the same formation either duplicated by folding or faulting or separated into two parts by the intrusion of a sheet of igneous rock parallel to its bedding. Since, then, two magnetic lines, the existence of which has been established b)' observation, ma}' have more than one interpretation, the discrimination of these cases, when possible, is of special importance. The question whether any given case belongs to the first of these cate- gories can generally be settled only by following the lines of attraction into a district which affords a geological section across the formations involved, or by the occasional outcrop of the rocks which give rise to the disturl^ances, in which case lithological resemblances or differences, the relations to other formations, and the observed structure will decide the matter one way or the other. In the special case in which either or both lines can be followed completely round an anticlinal dome or a synclinal basin, which of course can only rarely happen, the question would be settled affirmatively, e^^en if outcrops were entirely lacking. In the other instances the magnetic observations themselves often give means of discrimination, even when the outcrops are so few or so obscure as to be in themselves indecisive. It is characteristic of the folds in the pre- Cambrian rocks of this region that the axes are not usually parallel with the horizon for long distances, but are often inclined to it; in other words, when followed for greater or less distances they pitch. The outcropping edges of any formation involved in an anticlinal or synclinal fold which has been cut by a plane of denudation will be parallel to each other wherever the axis of the fold is horizontal, but will approach each other where the axis is inclined. In an anticlinal fold they converge in the direction in which the axis sinks, while in a synclinal the}" converge in the direction in which the axis rises. If the formation is a magnetic one, conformably placed between beds of nonmagnetic character, the magnetic lines to which the outcropping edges give rise will therefore run parallel to each other when the axis is horizontal and will converge or diverge when the axis pitches. The convergence or divergence takes place gradually, since the angles of pitch usually are not large. MAGNETIC OBSERVATIONS. 367 111 the case also of a «iu<>-le tbrinatiou which stands on edge and has been spUt b}- the intrusion of a sheet of eruptive rock parallel to the bed- ding planes, the magnetic observations will often show two parallel lines, which, at the extremities of the eruptive rock, where it wedges out, merge into one. In general, therefore, two parallel magnetic lines which represent two distinct formations preserve their identity, and do not pass into each other; when, however, they represent the same formation, they will often come together if followed far enough. The principles which have already been applied to the analysis of simpler cases are useful in discriminating among the three cases of convergence. I. PITCHING SYNCLINES. Let us first consider a pitching synclinal fold, which is repi*esented in plan and by successive cross sections in fig. 21. It is evident that on the lines of traverse along Sections I and II the deflections of the needle will observe the usual sequence for two parallel belts, the details depending upon separation and depth of cover- ing, while on lines along Sec- tions III, IV, and V the phe- nomena will be those caused by a single belt of magnetic rock. Also, on account of the rise in the axis, the south poles of the rock are brought continually nearer the surface on these successive cross sections, and therefore the two components of the rock force will be smaller for each traverse than for the one preceding. Since the magnetic material comes to an end at A, it is no longer true that there is as much magnetic material on one side of these sections as on the other. Conse- quently the horizontal component due to the pull of the rock does not become zero at any point along these sections, but for every station has a positive numerical value and acts in the general direction in which the syn- clinal pitches. At the station in the plane of symmetry of the fold this V^ CRCSS CCCTIONS Fig. 21 . —Plan and cross sections of a pitching syncline. 368 THE CRYSTAL FALLS IRON-BEARIKG DISTRICT. component acts parallel to the axis. The direction and amount of the deflec- tion depend upon the direction of strike and pitch of the synclinal. Let us suppose, first, that the axis of the synclinal strikes north and pitches north. In this case Section I, in fig. 17, is the most northern, Section V the most southern. The traverses along Sections I and II display the usual phenomena for two parallel belts. East of the eastern limb and west of the western the horizontal needle will be deflected toward the syncline. Between the two limbs there will be at least one point of no deflection, and "frequently, depending upon the relations between the depth of burial and the thickness of the intervening nonmagnetic material, either two other points of no deflection or two zones of retardation, one on each side of this middle zero. Along Sections III, IV, and V there will be but one point of no deflection of the horizontal needle, which will correspond with the axial line of the fold. Since this axis is north and south, and so coincides with the magnetic meridian, the horizontal component of the rock force coincides in direction with the horizontal component of the earth's pull, and consequently there is no deflection of the horizontal needle. For other stations east and west of the central station the deflections are toward the west and east, with the usual maximum points. The deflections on successive sections south grow smaller, since the angle between the two horizontal components progressively diminishes. The relative value of the horizontal component of the rock force also progressively diminishes, since the thinning of the magnetic material due to the rise in the axis of the fold brings the buried north poles into promi- nence. Therefore the deflections of the horizontal needle after the mag- netic rock has been left behind very soon become imperceptible. The dip needle deflections for the northern sections, I and II, reach their maximum values at the usual points, over the central zero and near the outside zeros or points of retardation. For the southern sections the dips grow less, since the horizontal restoring couple due to the rock has always a positive immericai value, and also because the vertical component of the rock force diminishes, owing to the nearness of the south poles. As the section approaches the limits of the magnetic material the points of maximum dip become less and less clearly defined, and the dip curve [)asses into an irregular line, slightly depressed below the line of no deflection. MAGNETIC OBSERVATIONS. 369 Tlie reasons for this iire, of course, obvious from what has been said above. In the case of a synchnal fohl pitcliing soutli, Section I (fig. 17) becomes the most southern, Section X the most northern, hue of traverse. Sections I and II present the same general phenomena as before for both needles. In Sections III, IV, and V the horizontal component due to the rock has a positive value for all stations as before, but in this case acts in a generally opposite direction to that of the horizontal component of the earth's force. Therefore, on these sections we should expect at first greater deflections of the horizontal needle, which would diminish rapidly as the sections approached and passed beyond the northern limit of the magnetic material, but whicli, for corresponding sections, would be greater than for the northerly pitching fold. The deflections of the dip needle would also ' be gi'eater for the same reasons. For a synclinal pitching west, Section I is the most western, Section V the most eastern, traverse. In this case, along I and II, the deflections of the horizontal and dip needles are dependent for their details upon the ratio of de^^th to distance of separation, but if far enough to the west will show clearly two belts of magnetic material, aj^proximately parallel, and striking approximately east and west. For Sections III, IV, and V, in which the distance of separation is either nothing or relatively small, the phenomena will indicate but one belt. On these sections, owing to the fact that the horizontal component of the rock pull is nowhere zero, and has everywhere a general westerly direction, the deflection of the horizontal needle will be westerly throughout, and will reach a maximum north of the east-and-west axial plane of the material, where the ang-le which it makes with the magnetic meridian is more than 90°. In accordance with the general principles stated in the discussion of a single belt with the same strike, the angles of dip are in general smaller south of the sjaicline than north, and the maximum dip is reached at a point north of the axial plane. On sections farther to the east, near the limits of the rock and beyond them, the dip-needle deflections, like those of the horizontal needle, rapidly diminish and soon become imperceptible. These facts are well shown in fig. 22, which represents a series of north- and-south traverses across the Groveland basin, the limits of which are defined by outcrops on the eastern side. MON xxxvi 24 370 THE CRYSTAL FALLS IRON BEARING DISTRICT. In this figure it is instructive to notice tlie small dip angles in the sections east of the end of the syncline. In the first of these the dip curve shows a hollow near the axis of the fold or angles of depression less than the normal. This is easily understood upon considering that, since the surface covering is here small, the vertical comjjonent of the I'ock force becomes very small at these stations compared with the horizontal component. For an eastward-pitching syncline it is obvious that the facts will be entirely similar to those stated above, except that tlie deflections of the horizontal needle will be toward the east instead of toward the west. This ; ( ',' 'I ,43 -^.(^ ^■'-- ■IS6 Its' S6, »71 y" 1/. ^ /_ „,. I_ lis. \ 24 >'. --»V I I I ■>, I 1 I 'I I Y I I I I I 1 I 1 Fig. 22. — Magnetic map of the Grovelantl Basin. is also well shown at the western end of the Groveland basin in fig. 22. This basin does not show the jihenomena of two lines, however, from the fact that it is so narrow and shallow that it does not include in its interior any overlying nonmagnetic material, and there is accordingly no sej^aration of its rims. 2. PITCHING ANTICLINES. In the cases of pitching anticlines (fig. 23) the sequence of observa- tions in the area of separation of the rims is very similar to that of pitching synclines. In the zone of coincidence the structural diff'erence in the two cases is that the material does not come to an end, but continues as one band, which, as the axis sinks, is ^progressively buried to a gi'eater depth. MAGNETIC OBSERVATIONS. 371 Thereftiro, in yoneral, tliL- l)uried north poles of the magnetic formation are not brought nearer the surface; and this, together with the fact that the material continues on in the line of the axis, jiroduces characteristic phe- nomena in the magnetic sections. These phenomena, "the details of which can be easily followed out for any given direction of pitch, and need not here be described, show in gen- eral two lines of attraction luerging into one, which continues in the same direction as a strong line, showing, as it is followed, the peculiarities due to an increasing deptli of burial. The points of maximum deflection of the liorizontal needle continue to separate from each other on suc- cessive sections. The dip curve shows a definite maximum closely corresponding, except for due east-and-west strikes, to the point of no horizontal deflection. Where the axis of tlie fold is so oriented that these points can be establi.shed, they indicate the nature of the fold. If the strike is east and west, in which case they become indeterminate, the continuity of the line and its A'ery gradual decrease in power may give an excellent basis for inference as to the nature of the fold. r^7n\ A Fig. 23. CROSS SECTIONS -Pliiii aud cruss sectioBs of a pitching anticline. 3. FORMATIONS SPLIT BY INTRUSIVES. When a single formation has been split into two by the intrusion of a nonmagnetic igneous rock, there are in the area in which the igneous rock occm-s two pai-allel magnetic formations, which give rise on cross traverses to phenomena the precise features of which depend upon the strike and dip of the formation and upon the relation which the width of the intruded mass bears to the depth of burial. To describe these would involve a mere repetition of what has been said before. Such intruded masses always have a definite limit in length, which is usually not very great. When the limits are reached, the two parallel lines pass into a single line which continues on 372 THE CRYSTAL FALLS IRON-BEARING DISTRICT. in undimiilislied vigor. Also, such intruded masses are seldom confined to definite horizons for great distances and seldom split the formation into symmetrical halves. Two nearly parallel lines of unequal strength, which, as they, are followed, become equal, and then again become unequal, with the stronger on the opj^osite side, are often, therefore, characteristic phenom- ena of this case. A good illustration of the unequal division of a magnetic rock at diff'erent points along its strike by an intruded sheet which wedges SS 84 3S 39 40 40 40 CO 61 IS 59 IS 69 64 60 C3 f > ^ >■ f f r f f r \ \ \ \ \ 1- .^i ,^ Jl 34 36 40 44 45 48 ST II. 46 VA'^i " 08 M 68 V. , f t f * f 'Y WX" ' ' * 31 81 34 31 41 47 49 69 16 -^i 3«4 66 86 M " t r t t ' r f f > ^V \ * ' * 36 34 40 43 46 46 ST^S «* 68 « 19 63 6J N '\vC>v 3? 35 4. .} 4; 48 6^66 ^9>;jVl » «T 39 41 43 41 •jjg eiSji > 84>«1 « •} ' » J. / » « y '^y^^ — s — i- ^ Fig. 24. — Magnetic map of a single formation split by an intruded sheet. out at both ends is given in fig. 24. Between the second and third trav- erses from the north end the existence of this sheet has been proved by drilling. 4. SUMMARY. The means of discrimination among these cases of convergence are therefore founded on the deflections in the critical areas, where the separated bands of magnetic material merge into one. Strong deflections toward the point where they run together, with a rapid disappearance of all disturb- ances within a short distance of this point, indicate a pitching synclinal MAGNETIC OBSERVATIOXS. 373 fold. A louy (.•(iiitiiuiaiu-e of the disturbances, with the cliaracteristic phe- nomena attending deeper burial, beyond the point of coincidence, indicates an anticlinal fold. A coincidence in both directions and the continuation of the disturbances without diminution indicate an intrusive sheet. To these may be added the delicate criterion which the unsymmetrical distances of the horizontal maxima from the central zero may afford. If in the area of separation the two belts depart from each so far as to be out of range of each other's influence, and it is found on successive cross sections that the nearest maxima are inside the lines of no deflection which directly indicates the position of the rock, it can be concluded that the rocks dip toward each other, and, on the other hand, if the nearer maxima are outside these lines, that they dip away from each other. In the one case a syncline and in the other an anticline would be indicated, and, of course, in either case it would be certain that tlie phenomena could not be due to an intruded mass. OHAPTEE III. THE FELCH MOUNTAIN RANGE. SECTioisr I. posiTio:^, extent, and previous work. Our map (PL XLIX) of the Felch Mountain range includes 12 sections in the southern tier of T. 42 N., Rs. 28, 29, and 30 W., beginning- with sec. 33, T. 42 N., R. 28 W., on the east, and ending with sec. 34, T. 42 N., R. 30 W., on the west. The range is known to extend beyond these limits both to the east and to the west. Rominger states^ that it has been traced 4 miles east of our eastern boundary, and also west of our western boundary to the Menominee River north of Badwater Village. From a hasty recon- naissance of the country to the east it seemed probable that but few addi- tional facts could be determined, because of the swamps and the extensive cover of the Paleozoic sandstone, and these sections were therefore not studied in detail. We were not able to continue the work to the west, on account of the lateness of the season, but it is desirable that this should l)e done at some future time. The sections surveyed include, however, that portion of the range in which outcrops are most abundant and which has been the principal seat of exploration for iron ore. The strong magnetic attractions in several of these sections and the prominent outcrops of ferruginous jaspers at Felch Mountain in sec. 32, T. 42 N., R. 28 W., and in sec. 31, T. 42 N., R. 29 W., Avere early noticed by the United States land surveyors and indicated on the township plats. With the rapid development of the Marquette range after the close of the civil war the attention of miners was quickly drawn to these as to other outlying prospects, with the result that vigorous exploration was begun on this range even earlier than on the Menominee range proper. 'Geological report on the Tipper Peninsula of Michigan, by C Koniinger: Geol. Survey, Mich., Vol. V, 1895, p. 35. 374 49 us GEOLOGICAL SURVEY MONOGRAPH XXXVIPLXLIX. I R 28 W /R^r US BIEN SCO L .\RCHEAN Granite GEOLOGICAI. MAP OF THE FELCH MOUNTAIN RANGE TOPOGRAPiri'.\XU GEOLOGY BY H.L.SMYTH SCALE 2 INCHES - 1 MILE ^ "^'TOIIR INTERV'AL 20 FEET HORIZONTAL SCALE OF SECTIONS :: INCHES -1 MILE \'ERTICAL SCALE 1 INCH - 1320 FEET ELEVATION OF BASl. LINES 600 FEET W Oulcrops wilhoi' t-bscH-ed Bli-jlu. or rtiu ; Trel pa« botlomed mrock o Drill holes ALGONKIAN CAMBRIAN LOWER HUHONIAN StiirqoonQiinrtzilr I Als I Rand^■ille Dolomito Mansfield Schist AlrH Aim Grovelaiul Formntion UPP ER H URONIAN Ulldmded INTRUSIVE Diabase Granite Au X D PREVIOUS WORK ON FELCH MOUNTAIN RANGE. 375 Tlio following abstracts of the published literature upon the Felch jrouiitMin range are given as far as possible in the author's words: 1850. Bi'BT, Wm. a. Geological report of the survey, "with reference to mines and minerals," of a district of townsbip lines in the State of Michigan, in the year 184^0, and tabular statement of specimens collected. Dated March 20, 18-1:7. Thirty-flrst Congress, lirst session, 1849-50. Senate Documents, Vol. Ill, No. 1, pages 8J:2-875. With maps. The earliest mention of the part of the Upper Peninsula included within the Felch Mountain range was made by Burt in describing the dis- tribution of the talcose and argillaceous slates of the area covered in the coui-se of his land survey in 1846. He states that the argillaceous slates "are developed in ])arts of township 42, ranges 29 to 30 west" (p. 846). The existence of the iron ore in this area was discovered l)y this explorer. The flrst bed discovered of this ore was found while traveling from the Pesha- kumme Falls, near the Meuomonee River, east to Fort River, before it was surveyed in May last, but was not discovered again during the suivey. It is believed, however, that this bed of iron ore is not far distant from the corner of townships 41 and 43 N., between ranges 29 and 30 W. It was found in a low ridge about 3 chains wide, course WNW. This ridge appeared to be nearly one mass of iron ore, stratified and jointed ; consequently it may be quarried with ease. This ore has generally a granular or micaceous structure, but specular varieties sometimes occur; color, iron black, pass ing into a steel gray; luster when fresh broken, metallic, but soon oxidizes when exposed to the atmosphere. This is supposed to be an extensive and rich bed of iron ore. The variation of the needle was taken on the east side of the ridge at the cross- ing of a hunter's trail, and its north end stood S. 82° E. Three or 4 miles west of this, on the north side of a ridge, near a cedar swamp, the variation was N. 45° 30' W. Probably iu this vicinity may be found another extensive bed of similar iron ore (p. 849). 1851. Foster, J. W., and Whitney, J. D. Report on the geology and topography of the Lake Superior land district. Part II. The iron region, together with the general geology. Dated November 12, 1851. Thirty-second Congress, special session, 1851. Senate Documents, Vol. Ill, No. 4; 406 pages; with maps and plates. Foster and Whitney, in sketching the distribution of the rocks of their Azoic system, which comprises "for the most part gneiss, hornblende, chlorite, talcose and argillaceous slates, interstratified with beds of quartz, saccharoidal marble, and immense deposits of specular and magnetic oxide of iron" (p. 8), after describing the main area of these rocks to the north- west and north of the Felch Mountain range, say: "Another arm about 376 THE CRYSTAL FALLS IRON-BEARING DISTRICT. 18 miles in length and 10 in breadth extends easterly into T. 42 and 43, R. 28" (p. 14). This east-and-west trending area corresponds in part with the Felch Mountain range, but also includes two .adjacent troughs, one to the north of it, and the other to the south. Subsequent to the above reports, nothing is known to have been pub- lished containing any matter concerning this iron-bearing area until 1869. 1869. Ceednbr, Hermann. Die vorsiluiiachen Gebilde der "oberen Halbiiisel von Michigan" in Nord-Auierika. Zeits. der deutschen geol. Gesell., Vol. XXI, 1869, pp. 516-568. With map and three plates of sections. In Credner's article we find a considerable advance in knowledge concerning the relations of the rocks of the Felch ]\Iountain area. This advance is indicated by his general map (PL IX) and by his two profiles (fig. 3, PI. IX, and fig. 1, PI. X). In profile No. 3, going from south to north, the granite is represented as overlain by quartzite, separated by a narrow interval from the marble, which in its turn is overlain by a great thickness of the ore formation. The dip is steep to the noi'th. Unconformably upon the last two formations — the marble and the iron formation — there are caps of Potsdam sandstone, with the beds dipping flat to the north. In fig. 1, PI. X, the gneiss overlain by the quartzite, dipping steep to the north, are the only pre-Paleozoic rocks shown. The same profile shows the unconformable Potsdam, with Silurian dolomite resting conformably upon it. In the text there is mention, with an illustration (fig. 5, PI. IX), of the granite dike cutting across' the iron formation in the upper course of the Sturgeon River. Beyond this no reference to the area under discussion occurs in the text. 1873. Brooks, T. B. Iron-bearing rocks (economic). Geol. Surv. of Michigan, Vol. 1, 1869-1873, New York, 1873, Part I. 319 pages. With maps. In the year 1873 Maj. T. B. Brooks's very important study of the Michigan iron ranges appeared, and in it the Felch Mountain area for the first time is distinctly separated as "the North Iron Range" from the "South Iron Range" of the Menominee River iron region, both together, however, constituting the Menominee. The north iron range, about 12 miles from the other in the south i>art of T. 42, Rs. 28, 29, and 30, is in places a prominent topographical feature. The capping of PREVIOUS WORK OJf FELCH MOUNTAIN RANGE. 377 horizontal sandstones, wliicli -has already been mentioned as characterizing- the Menominee liills, gives a somewhat more even character to the crest lines, and in places produces a strikingly ditterent prottle (p. 72). As stilted by Major Brooks in a footnote on p. 157, Hie facts contained in the chapter on the Menominee iron reg'ion were derived largely from the sm-veys and explorations of Prof. R Purapelly and his assistant, Dr. H. Credner. The following passage is the most important statement conceni- iug the Felch Mountain area, or the "North Range:" The north iron belt or range has a coarse nearly due east and west, and is all embraced, so far as known, in the south tier of sections of T. 42, Rs. 28, 29. and 30. The most easterly discovered exposure of ore, known as the Felch Mountain, is in the N. J of sees. 32 and 33, T. 42, E. 28. Traveling due west, fragments of iron ore are found in NE. ^ of sec. 31, T. 42, R. 28; after which no absolute proof of the presence of iron is found (although it is probably continuous) until we reach sec. 31, T. 42, R. 29, where, in the center of the section, is an immense exposure of iron ore in an east- west ridge, which can be traced westerly halfway across section 36 of the next township. The natural exposure of ore on section 31 is larger than at any other point in the Menominee region, and the quality is as good, if not better, so far as can be judged by surface indications. Magnetic attractions and iron bowlders found farther west and southwest on this range prove its extension in that direction. Whether the westerly course continues, or whether it curves to the southwest, as seems probable from the position of the lower quartzite and local magnetic attractions in the northwest part of T. 41, R. 30, has not been determined. The latter hypothesis is most in accordance with the known facts, although the southeast dip of the quartzite on sections 17 and 18, observed by Dr. Credner, is not explained. If this hypothesis is true, the iron range should cross the Menominee somewhere in sees. 24 or 25, T. 41, K. 31, into Wisconsin. There can be little doubt but that the north and south belts belong to one geological horizon, hence somewhere come together (pp. 159-160). A geological section through the north range on the line between Rs. 29 and 30, T. 42, is given, and is also represented as section CC on Atlas Plate IV. The succession from south to north is as follows: Granite. Quartzite. Interval. Marble. Iron-ore formation. Interval. Marble. Interval. Granite-gneiss and hornblende and mica schist. As represented in the section, the beds all dip toward the north at 378 THE CRYSTAL FALLS IRONBEAKIJ^G DISTRICT. high angles. In attempting to correlate these vhrions beds with those of the sonth iron belt, Brooks expei-iences difficulty with the uppermost formation of granite-gneiss and schist. He says: The gueiss and granite outcrop above described may be almost regarded as a typical Laurentian rock in its appearance. If future investigations prove them to be Laurentian, a very trouble.some structural problem would be presented here, as we would have Laurentian roclss conformably overlying beds unmistakably Huronian (p. 175). As will be seen, this objection which Brooks had anticipated was raised bv later workers in the area and was explained by Eominger. The main points of Brooks's conclusions may briefly be summarized : (1) The iron-bearing rocks of the . Menominee region occur in two approximately parallel east-and-west belts (the north belt being the Felch Mountain range and the south belt the Menominee range), separated by a broad granite area which narrows toward the west by the convergence of the iron belts. The north-and-south belts were not traced into each other, but their probable connection was inferred from their bending toward each other and from the occurrence of rocks of the iron-bearing series west of the granite area. The equivalence in age of the two belts was inferred from the lithological and stratigraphical similarity exhibited by the great quartz- ite and marble formations, by the probable continuity above referred to, and by the similar relations of these formations to the basement granites. (2) The iron-bearing formations of the Felch Mountain range were believed to occur at two horizons. That of Felch Mountain itself in sec. 32, T. 42 N., R. 28 W., was held to be a ferruginous phase of the lower quartzite. On the other hand, the exposures of sec. 31, T. 42 N., R. 29 W., were rea'ai'ded as belong'ing' to a horizon above tlie lower marble, and as the close equivalent of unimportant lean ores of the Menominee range. (3) In geological structure the Felch Mountain area was held to be a northward-dipping monocline. (4) As a consequence of this conception of the structure, Major Brooks supposed that there were two marble formations. 1880. Brooks, T. B. The geology of the Menominee iron regioii, east of the center of Range 17 E., Oconto County, Wisconsin. Geology of Wisconsin, 1873-1879, Vol. Ill, published in 1880, Part VII, pages 429-599. PKEVIOUS WORK OX FELOU MOUNTAIN RANGE. 379 In Vol. Ill of tlie Geoloy-y of Wisconsin, published in this year, Brooks reiterates the views previously published in the Michigan reports. The only new material atlded is a table (p. 447) giving the estimated minimum thickness for the north belt as 5,200 feet. 1881. RoMTNGBR, C. Geol. Survey of Michigan, Vol. IV, Part II, Menominee iron region. New York, 1881, pp. 155-241. With map. . This, the first of the reports of the geological survey of Michigan pub- lished while Dr. Rominger was in charge, contains a great number of details concerning explorations in the Felcli Mountain range, as it is thus called (p. 194) for the first time in the chapter on the Menominee iron region. The iron-bearing belt along the Menominee River is referred to as the Quinnesee range. Rominger criticises Brooks's position with reference to the age of the granite and gneiss along the northern border of the Felch Mountain range as follows: Major Brooks declares the granites and the gneisses north of the Felch Moun- tain ore range as younger than the ore formation, which like them dips northward; but their superposition upon the ore formation is nowhere observable; on the con- trary, the south side of the ore range exhibits in several places the direct superposi- tion of the ore formation on the granite. This fact is known to Major Brooks, but he solves the dilemma by identifying the granites on the south side of the ore forma- tion with the Laurentian; those on the north side, he claims, represent the youngest Hurouian rocks. How he can do so I can not conceive, as the concerned granitic and gneissoid rocks north and south of the ore formation are so absolutely identical that no one who ever sees them can doubt for a moment the quality and age of these rocks. Moreover, this identification of the northern granite with the Upper Huronian, and of the southern with the Laurentian, implies another abnormity; groups of rocks, usually separated from each other by thousands of feet of intervening strata, are in this case thought to be in immediate superposition, which does sometimes occur, but not in coincidence with another improbability like the one stated in this instance (p. 207). ^ 1887. Irving, R. D. Is there a Huronian group? Am. Jour. Sci., Vol. XXXIV, 1887, pp. 204-263, 365-374. Read before the National Academy of Sciences April 22, 1887. In discussing this question. Professor Irving takes occasion to refer to the structure of the Felch Mountain range: In the case of the Felch Mountain belt, which does not exceed a mile in width, all of the strata are described by Dr. Rominger as dipping at a high angle to the 380 THE CRYSTAL FALLS IRON-BEARING DISTRICT. northward; and in crossing the belt from the south to the north, after passing the middle, one traverses a repetition of the belts crossed farther south, but in an inverted order. It would seem that we have to do here with a case of a synclinal, whose sides are folded close together (p. 256). The relations of the Felch Mountain range to the rocks of the Menom- inee River and the Marquette district are shown on a profile. The facts are mentioned as having been derived from an unpublished manuscript report (1881 to 1884) of Rominger to the Michigan geological survey.^ 1888.' Irving, R. D. On the classification of the early Cambrian and pre-Cambrian formations. Seventh Ann. Rept. U. S. Geol. Survey, for 1885-86, Washington, 1888, pp. 365-45-1. In this article the statements made above are repeated (p. 435). 1891. Van Hise, 0. R. An attempt to harmonize some apparently conflicting views of Lake Superior stratigraphy. Am. Jour. Sci., Vol. XLI, 1892, pp. 117-137. The author of this paper, after discussing the significance of the various unconformities observed and after dividing the rocks of the Marquette dis- trict into a Fundamental Complex and a Lower and an Upper series, attempts to correlate the rocks of the Menominee region with these divisions. Passing now to the Menominee and Felch Mouutain districts, our information is less exact. It is, however, clear that in both of these areas we have the Fundamental Complex — that is, the granites and the gneisses associated with crystalline schists having the usual "eruptive contacts" — the equivalence in every respect of Lawson's combined Laurentian and Coutchichiug period. Above this complex. Professor Pum- pelly, with whom this whole subject has been discussed, and who has great famil- iarity with the entire Lake Superior region, suggests as exceedingly probable that in the Felch Mountain iron-bearing series only the equivalent of the Lower Marquette occurs, the Upper series, if it once existed, having been removed by erosion (p. 133). 1892. Van Hise, C. R. Correlation papers — Archean and Algonkian. Bull. U. S. Geol. Survey, No. 86, Washington, 1882, pp. 549. In a summary of the literature on the Lake Superior region, the state- ment quoted above is incorporated without change (p. 190). I Published iu 1895, Vol. V. PKEVIOUS WORK ON FELGH MOUNTAIN KANGE. 381 18»3. Wadsworth, M. E. Report of tlie State geologist for 1801-92. State Board of C;eol. Siirv. for the years 1891 and 1892, Lansing, 1893, pp. 01-73. Dated October 17, 189L'. In this brief report Dr. Wadswortli calls attention to the granite which is intrusive into the sedimentary series in the Felch Mountain area. In tbe Menominee region, especially in the Felch Mountain district, these granite dikes are well exposed. Here they are seen not only to cut the gneiss, but to pene- trate the Republic or iron formation. Mr. Wright, in 1885, pointed out one of these dikes cutting the iron series near the Metropolitan mine, on sec. 32, T. 42, R. 28 W, (p. 101). He includes the Felch Mountain dolerites (p. 104) in his Republic for- mation, and correlates this formation with Van Hise's Lower Marquette series. 1S95. RoMiNGER, C. Geological report on the Upper Peninsula of Michigan, exhibit- ing the progress of work from 1881 to 1884. Iron and copper regions: Geol. Surv. of Michigan, Vol. V, Lansing, 1895, pp. 1-94. Chronologically, this is one of the latest reports upon the Upper Penin- sula of Michigan, yet in justice to the author it should be considered as having priority over any articles published since Irving's, m 1887, to which reference has already been made, as in that article Irving made use of the observations recorded in the manuscript of this report, giving Rominger full credit for them. In Rominger's report, in the chapter on the Granitic Group, the granite dike cutting the iron-bearing formation in the Felch Mountain range is described (p. 7). He also describes a wedge-like intru- sion into the highly contorted strata in sec. 33, T. 42 N., R. 28 W., as follows: We have here evidently before us a series of strata plicated into a synclinal and another anticlinal fold, the latter ruptured by an intruding granite mass, which rock is^there the general surface rock and comes on the south end of the exposure in contact with the uppermost ferruginous strata of the overtilted anticlinal fold (p. 8). • The portion of the chapter on the iron-ore group which deals with the Menominee region is devoted to the Felch Mountain range and to outlying prospects as far north as Michigamme Mountain. The description of the Felch Mountain area is in great part a condensation of earlier scattered observations. The strata are given as dipping high to the north and consisting of the following succession: The underlying rock of the iron formation is always formed of crystalline rocks, granite or diorite. The lowest strata are generally heavy light-colored quartzite 382 THE CRYSTAL FALLS IRON-BEARING DISTRICT. beds, with interlaininated thinner ledges and schistose seams, amounting to consider- able thickness. Above this belt an equally large succession of well-laminated, even-bedded, often fissile, slate like micaceous quartz-schists follow, which have a silvery luster. Next above them comes a series of micaceoas argiUites, amounting to a belt even larger than the former, which varies greatly in shades of color, firmness of grain, etc.; some layers are whitish, others gray or bluish and greenish, but the greatest portion of them is intensely red colored by hematitic pigment. A part is a fatty, impalpably fine mass of silky or also pearly luster, according to the size of the mica shales incorporated with them. Another part is rough and gritty from the prevalence of arenaceous constituents. At this horizon and rather in the lower part of it occur locally large bodies of crystalline limestone ledges, some snowy white like Italian marble, but of coarser crys- talline grain and intermingled with radiating clusters of asbestine fibers and larger prismatic crystals of colorless tiemolite, which sometimes forms larger concretionary seams in the lime rock, and are then intimately associated with crystal masses of sahlite, one mineral penetrating the other in a manner which suggests either a process of paramorphosis in progress, changing the sahlite into tremolite, or the original conditions, when the calcareous material combined with the silica by a slight modifi- cation, induced simultaneously the crystallization of the almost identical chemical combinations iu one and the other form ; which latter suggestion is more sustained by the actual condition of the mingled minerals than the first, as some of the crystals of both minerals tightly grown together are so perfect in form peculiar to each and so sharply defined that they must be considered as crystal individuals which formed side by side and altogether independent of one another. In other localities where such crystalline limestone belts occur, the tremolite is only sparingly intermingled, but iu its place colorless mica scales of nacreous luster are plentifully disseminated. In the place of marble-like limestone, sometimes also ordinary lime rock of dull aspect with conchoidal fracture, and variously tinged, occurs; it is then usually full of flinty siliceous seams, resembling the limestones of the Quinnesec range; the quartzose seams, locally even, prevail over the calcareous. Incumbent on the above- mentioned micaceous argellites succeeds a belt, about 800 feet in width, composed of thinly laminated, banded, ferruginous, quartzite ledges of dark purplish tints or hav- ing a metallic luster from intermixture of specular ore granules. The banded portions are formed of an alternation of narrow seams of specular ore with siliceous seams not so richly impregnated with the oxide. Other strata in the succession are porous cherty rocks charged with ochreous yellow or brown oxide of iron and inclosing i)ockets of the limonitic ore. Also blood-red argillitic seams occur in the succession, and with them sometimes pockets of soft crumbly hematite ore. Within the first-mentioned banded alternation of narrow ore seams with quartz seams, larger deposits of specular ore in slaty or in compact granular, or also in the soft friable condition of the so-called blue ore of the Quinnesec mines occur, which constitute the principal storage of ore sought for by the miner, besides the hematitic and limonitic deposits mentioned before. The first impression of every observer exam- iiiing this above described rock series will induce him to consider it as an ascending succession, as the layers follow one another in apparent conformity; but in some local- ites, after having crossed this succession so far, if we proceed farther in the same GEOLOGY OP FELGH MOUNTAIN RANGE. 383 direction, we intersect the same series again in an inverted order, but retaining the same di|), until we have reached again a hirge beltof compact fjuartzite ledges in close contiguity with granite or also diorite, as it may happen, which latter rocks then form the surface rock of large areas on the luirth side of the Felch Mountain ore formation (p. 33). The only satisfactory explanation which I can give of this repetition of the rock beds in an inverted order is the suggestion of a folding of the beds and the overturn of tlie fold by a pressure acting principally from the north side. If this is the case, we would have to consider the light-colored quartzite next to the granite as the most recent deposits and the dark, ore-bearing, banded quartz beds as the oldest, which would bring the structure of the Felch Mountain ore formation in perfect harmony with that of the Quinnesec ore range (p. 34). SECTIOX II. GENERAL SKETCH OP THE OEOL,OGY. The rocks of tlie Felch Mountain range extend from the Archean to the early Paleozoic. The Paleozoic is represented by the Lake Superior sandstone, of supposed Upper Cambrian age, and the overlying Calciferous limestone. These formations were originally laid down over the upturned edges of the older rocks in flat sheets or with low initial dips, and have not since suffered relative displacement to any notable degree. As has already been stated, subsequent erosion has to a great extent removed this over- lying blanket and laid bare the older rocks, except for the covering of recent glacial deposits. The Cambrian sandstone, and to a less extent the Calciferous limestone, still, however, occupy considerable outlying areas, detached from one another throughout most of the district, but gradually coalescing beyond the eastern end, where they completely cover the older rocks and limit all further geological study of these in that direction. The Paleozoic rocks will not be considered further at present. On the detailed Felch Slountain map (PI. XLIX) their known outcrops are repre- sented by appropriate symbols, but except in the larger areas, where they so completely conceal the older rocks that the distribution of these can not be determined, they are assunaed not to be continuous, and to be non- existent, and in this respect stand upon the same footing as the Pleistocene glacial covering. The Archean, which is here made up of granites, granitic gneisses, and various kinds of crystalline schists, is the basement group of the region. The areas in which these rocks are now ex])osed at the surface represent the cores of the larger arches which were constructed over the whole region by the early manifestations of mountain-building activity, and subsequently 384 THE CRYSTAL FALLS IRON-BEAEING DISTRICT. truncated by the deep Cambrian denudation. Our studies have dealt with the Archean only in narrow marginal zones, and have included little more than the location of its outer boundaries, except when it was necessary to go deeper in order to complete the work over a full section. Consequently no attempt at classification can be made upon the map. The rocks, chiefly of sedimentary origin, which are intermediate in age between the Archean below and the Paleozoic above, and therefore fallwithin the system to Avhich the name Algonkian has been given b}' this Survey, occupy a narrow strip nowhere more than a mile and a half and usually less than a mile wide, which as a whole runs almost exactly east and west for a distance of over 13 miles. This strip constitutes the Felch Mountain range. On the north and south it is bordered by the older Archean. The lowest member of the Algonkian occupies parallel zones next to the Archean both on the north and south, and is succeeded toward the interior of the strip by the younger members. While the general structure, therefore, is synclinal, a single fold of simple type has nowhere been found to occupy the whole cross section of the Algonkian formations, but usually two or more synclines occur, separated by anticlines, which may have different degrees and directions of pitch and different strikes, or may be sunk to diff"erent depths, and complicated besides both by subordinate folds and by faults. Among the Algonkian rocks we distinguish two main divisions or series, which are probably separated from each other by an unconformity. Owing mainly to the peculiar lithological and weak physical character of the younger of these two series, actual contacts between them have not been found, and the evidence of unconformability consequently consists not so much in observed discordance of structure as in an inferred discordance based upon their relative surface distribution. This evidence will be fully stated hereafter. In the lower of these two series are included four formations which clearly appear to be identical in lithological character and order of super- position with the four formations that, so far as is known, make up the lower iron-bearing series along the Menominee River. These are, reckoning from the base upward, (1) the Sturgeon quartzite, (2) the Randville dolomite, (3) the Mansfield schists, (4) the Groveland iron formation. Above this series follows the younger series, which lithologically and in AHCHEAN IN FELCH MOUNTAIN DISTRICT. 385 its art'iil relations is very incompletely known. It inclndes mica-schists, ferruginous schists, iuul thin interbeddetl ferruginous quartzites. These rocks, which from our imperfect knowledge must for the present be grouped as a single formation, are believed to have been deposited contempora- neously with the somewhat similar rocks that occur in the Menominee area, at Iron Mountain, but are most extensively exjKised west of the Menominee River, and especially in the Commonwealth and Florence district in Wisconsin. SECTIOX III. THE ARCHEAX. The Archean occurs in the Felch Mountain district in two belts, which limit the Algonkian rocks on the north and on the south. The north- ern belt for the most part does not fall within the limits of the detail map (PI. XLIX). It occupies a triangular corner in sees. 34 and 35, T. 42 N., R. 30 W., at the extreme western end of the area sm-veyed, and even in these it has not been directly observed, but its presence is inferred from outcrops in the adjoining sections west and north and from the observed strikes in the overlying Algonkian formations. For the next 11 miles east its southern boundary lies in the tier of sections next north of those majDped in detail and probably always less than a mile away. This boiindary is therefore not very accurately drawn, as only enough outcrops were visited to permit its position to be fixed in a general way. Our work first touches the southern area of the Archean, which is much better known on the west in sees. 3 and 2, T. 41 N., R. 30 W., a short distance south of the township line. Thence for 3 miles eastward the boundary follows the township line, and in sec. 31, T. 42 N., R. 29 W., crosses it with a trend somewhat north of east. From the west line of sec. 31 to the east line of sec. 36, T. 42 N., R. 29 W., the Archean occupies the southern third of the south tier of sec- tions. Thence for a mile and a half it bends northeast, and in sec. 32, T. 42 N., R. 28 W., reaches its farthest north in the center of the section. From this point the boundary runs southeast, with a sinuous embayment to the south, and passes outside the limits of the map a little north of the southeast corner of section 33. Throughout the Felch Mountain range the southern Archean is much better exposed than any of the other terranes. In the western portion of the range, where hardly more than the contact zone falls within our limits, MON xxxvi 25 386 THE CRYSTAL FALLS IRON-BEARING DISTRICT. outcrops are not especially numerous; but in the six eastern sections, which include a belt from a quarter to half a mile wide, a very considerable por- tion of the surface is bare rock. This exceptional degree of exposui-e has been brought about by the forest fires, which, by loosening the thin soil and destroying the protecting cover of vegetation, have facilitated its removal from the steep-sided knobs that are such characteristic featm-es of the Archean topography. TOPOGRAPHY. The Archean areas, particularly the southern, are distinguished by a characteristic rough topography. The surface is exceedingly uneven on rather a small scale, and has already been described as consisting of hunnnocky elevations alternating with bowl-shaped depressions. Both hummocks and bowls are elongated in an east-and-west direction, in accord- ance with the prevalent gneissic foliation. While the surface is thus so full <>f small details that an adequate delineation of it is the despair of the topogra]>her, the actual relief is insiii'nificant. To men bred on the flat ])lains of the lower lakes, as were most of the early surveyoi's and explorei-s, it may naturally have appeared mountainous, since roughness is a quality particularly noticeable in a wil- derness like this, that can be traveled only on foot. But from a broader point of view the irregularities are almost wholly lost. The higher summits in the same neighljorhood rise to within a few feet of each other. The distant sky line is even in all directions. There is, however, a gentle ascent from east to west, quite imperceptible on the ground, and made evident only by the general course of the streams or by leveling. Along the contacts between the Archean and Algonkian systems there usually but not always exists a topographical depression, occupied by swamp or streams. North of the southern Archean mass this depression is a well-marked linear valley, extending with some interruption from sec. 33, T. 42 N., R. 28 W., on the east, for 6 miles west to sec. 33, T. 42 N., R. 29 W. For 2 miles in tlie middle of this stretch the valley is occupied by the Sturgeon River; thence west for 2 miles by a small feeder of the Sturgeon, while the eastern third holds swamp with ill-defined drainage. On the south the Archean boundary of this valley generally rises with AKCHEAN IN FELCH MOUNTAIN DISTRICT. 387 steep slopes, which are frequently escarpmeut-like iu cliaracter, and tor short distauces present smooth faces to the valley. In sec. 33, T. 42 N., R. 28 W., the mural face which runs southeast across the eastern half of the section with the regularity of a ruled line is a true fault scarp. Toward the western end of this valley, in sec. 32, T. 4-2 N., R. 29 W., the Hoor gradually rises ;uid the swamp area broadens, penetrating the Archean in a network of thicker and thicker mesh about the hiaher hum- mocks, until these are finally oveitopped. PETROGRAPHICAL CHARACTERS. The rocks of the Archean areas niay be di^dded into four quite distinct types, namely: (1) Granites or granitic gneisses, (2) gneisses with banding or distinct lamination, (3) mica-schists, and (4) hornblende-gneisses or amphibolites. Between the first two divisions there is an extremely close mineralogical and chemical lUieness, while in these respects the fourth division stands against all the others in strong contrast. (1) The granites of the first division are, as seen in the field or in the hand specimen, holocrystalline rocks of fine to medium grain, in which the eye can readily distinguish the presence of quartz, pink feldspar, muscovite, and biotite. In color they are prevailingly of pink or reddish tints of light shades. Structurally, they frequently appear in small areas to be entirelv massive, but even in the most massive occurrences the hammer can usuallv part them along roughly parallel surfaces which glisten with spangles of mica, indicating a certain degree of alignment in these constituents. Gen- erally, however, a rude foliation is more or less distinctly visible, and is sometimes exceedingly well developed, even to the point of fissilitv. It is always apparently due to the parallel arrangement of the micas, which are more abundant as the foliation becomes more distinct. The field relations show that the massive and more or less Ibliated varieties of this divi.sion are closely bound together by indistinguishable gradations and, indeed, often constitute a visibly integral mass. The usual arrangement of the micas is not parallel to a surface but parallel to a line which is generally inclined to the horizon at angles varying between 10° and 35°. A hand specimen when turned about the direction of foliation as 388 ■ THE CRYSTAL FALLS IRON-BEARING DISTRICT. an axis, shows a parallel arrangement of the micas on all sides, and a con- tinuous glisten follows the revolution; while on a surface at right angles to this direction the micas are not parallel and wind about the other constitu- ents indifferently. In the more fissile varieties the outcrops often have a rough, channeled surface, suggestive of the surfaces familiar in closely crenulated mica-schists, or on the corrugated walls of a fault. Similar cor- rugated surfaces frequently j^art more massive from more fissile parts of the same outcrop. Under the microscope the essential constituents of the granites and granitic gneiss are seen to be quartz, orthoclase, microcline, plagioclase, biotite, and muscovite, with the iron ores, titanite, and occasionally apatite and zircon as accessories. In the massive phases the general relations of these minerals to one another, and their order of crystallization, in no respect differ from those of igneous granite. The quartz, which is the last mineral to form, contains numerous fluid and gas inclusions, the former often with a mo^^ng bubble. (Jf the feldspars microcline is nmeh the most common, then plagioclase, while orthoclase is generally comparatively rare, although sometimes it is more abundant than the microcline. The plagioclase, from its relief and extinction angles, is probably not lower in the scale than oligo- clase. The orthocUise is usually clouded with alteration products, and some- times the dull interior is surrounded with a narrow unattacked rim. Botli micas are always present as original minerals, and on the whole biotite is the more abundant. They occur in small stout crystals, often as inclusions in the quartz and feldspars. Magnetite is rare, but occurs in idiomoi-phic forms in the later constituents, as do also minute crystals of zircon and apatite. Thin sections of even the most massive-looking specimens invari- ably show the effects of pressure in the undulatory extinction of the quartz and in the bending and occasional fracture of the feldspar. In the foliated varieties with which these massive varieties are closely associated the effects of mechanical stresses are the striking microscopic phenomena. The constituent minerals are essentially the same as in the massive phases, but the micas are relatively more abundant. The quartz and feldspar individuals are fractured and strained, and occur in irregular cores sepai-ated by anastomosing zones of a fine quartz-feldspar mosaic. In these last, new micas, in long curving individuals and clusters, have been developed in great numbers. AROHEAN IN FELOH MOUNTAIN DISTRICT. 3H9 The following analyses give tlie clu'Diical fouijUKsititiu of these granites: Analyses of granites. [liyUr. H. N. Stok es, U. S. Geol. Sun-ey.l 1.' 2.» 3.» SiO. 76.10 .07 Noue. .02 12.95 Noue. .65 .09 Trace. None. .12 .14 6.50 2.36 .17 .48 72.17 .37 69.69 .29 TiO; CO. P.O. AljO, CfiOj 14. 44' 15. 64' Fe 1O3 1.02 .99 .90 1.62 FeO MnO NiO CaO .69 .70 4.84 3.65 1.22 .66 5.30 3.34 MgO K2O NaO H.OatllO'^ H^O above 110^ Total 99.65 I Ba, Sr, Li, CI, S, SO3 were not looked for. ' Water not determined. ' Includes PiOr,. No. 1. Specimen 34677, Lake Superior Division, U. S. Geol. Surv., 1,935 N., 1,040 W., sec 2, T. 41 N., R. 30 W., Upper Penin.sula of Michigan. No. 2. Specimen 34828, Lake Superior Division, V. S. Geol. Surv.. 300 N., 1,8.50 W., sec. 36, T. 42 N., R. 29 W., Upper Peninsula of Michigan. No. 3. Specimen 36081, Lake Superior Divi-sion, U. S. Geol. Surv., 15 N., 1,025 W.. sec 31, T. 42 N., R. 28 W., Upper Peninsula of Michigan. No. 1, which is rather low in alumina, iron, and lime, is a granitic gneiss in which the abundant secondary mica, wdiich has grown in long curving plates in nearly parallel zones of granulation, is wholly muscovite. Nos. 2 and '6 are fine- and coarse-grained pink granites, which show comparatively little crushing and development of secondary minerals in thin section. The rocks of this division therefore have the chemical composition as well as the physical and petrograjjhical characters of igneous granites. The positive proof of igneous origin, hoA^ever — actual injection into older rocks — we have not found. Irruptive contacts may possibly exist, and may have escaped us, since neither the Archean as a whole nor its internal relations were the objects of especially rigid scrutiny. Igneous granites of Algonkian or later age ought to be found within the Archean areas, for several granite dikes are known to penetrate various members of this over- lying series. Whether the known granites within the Archean are really lower-lying and larger masses witli which such dikes are genetically con- 390 THE CRYSTAL FALLS IRON-BEARING DISTRICT. nected is not known, but the possibility must be admitted. The banded gneisses are often so faintly foliated and resemble the granites so closely in color and grain that the distinction can be made with only the microsco])e, and igneous contacts between them might easily be overlooked. It is certain, however, that if the granites have been injected into the banded o-neisses it has not been in the form of narrow dikes, and the fact remains that no case of an igneous contact is recorded in our notes. The gneissic members of this division are merely crushed granites, and owe their foliation partly to the crusliing and partly to the growth of fresh mica in the fractured zones. The}' differ from the banded gneisses in fur- nishing l»oth field and microscopic proof of the way in which the foliation was formed and of the rocks from which they were derived. (2 ) The banded gneisses of the second group have essentially the same mineral composition as the granitic gneisses of the first. They are distin- guished by the eye mainly by the fact that the component minerals occur in ra(^re or less distinct layers, from a fraction of an inch ujjward in thickness. The lamination, which only rarely is very regular, seems to be caused in most if not in all cases by the alternation of darker layers, which are relatively rich in biotite, with lighter layers, which are comparatively and sometimes wholly free from it. The light layers are almost always coarser in texture than the darker, and frequently are coarsely pegmatitic. The individual bands are not indefinitely persistent, but wedge out to knife-edges. The banding is sometimes so indefinite as to be lost in the hand specimen, the large surface of an outcrop being necessary to bring out the slight differences in shade. In color these rocks are light gray, dull ^vhite, or pink. The banding shows great variations in angle of dip, but the strike is usually fairly constant within a few deo-rees of east and west. In a few localities distinct contortion was observed in the gneis.sic banding and pitching folds. The lamination of these gneisses is, so far as observed, of tlie plane-parallel type. The bands are thoroughly welded together, and as a rule, the rock breaks indifferently across them. Under the microscope the composition of these rocks does not differ from that of the granitic rocks of the first division. The structural charac- ters, however, are in strong contrast. Even in those specimens which possess the most indistinct foliation all the minerals are elongated in a common direction. While the individual grains in most cases show more or less strain and are frequently fractured, their mutual boundaries are usually sharp and clear, and it is evident that the forms are not the direct AKGHBAN IN FELGU MOUNTAIN DISTURIT. 391 result of the pressure tluit has ati'eeted their optical properties. The evidence is quite clear that tlic niinerals now present have crystallized in parallel elongated forms, and it is to this they owe their prevalent lamina- tion even when the color 1 landing- is indistinct or wanting'. Subsequent to the time of crystallization they have been exposed to the action of great stresses, which not onh^ have left a record in the strains now frecpieutly perceptible in the minerals of the early crystallization, but also in many cases have produced roughly parallel fractures and fracture zones sometimes coinciding with and sometimes oblique to the early lami- nation. In these zones coarse micas have grown, reenforcing the old lamination when parallel to it, and when oblique producing a less regular secondary foliation, which is entirely analogous and probably contempo- raneous with the foliation of the cnished granites. The following analyses of these gneisses are interesting as showing their striking chemical relationship to the granites (analyses of which are given on p. 389), with which they are intimately associated : Analyses of gneiss. [By Dr. H. N. Stokes, U. S. Geol. Survey.] 1.1 2.2 3.2 SiO, 74.37 .07 None. .01 13.34 71.79 .35 74.63 .09 TiO, CO P,0, AID, 14. 79' 13. 951 CrO, Fe,0, FeO .92 .21 Trace. 1.10 1.09 .35 .32 MiiO NiO CaO .50 .27 6.70 2.50 .12 .44 1.11 .71 3.79 4.29 1.08 .22 6.73 2.55 MgO K,0 Na.O H.2O at 110" H.:0 above llO"^ Total 99.45 ' Ba. Sr, Li, CI, S, SO3 were not looked for. - Water not determined. ' Includes P»0.r,. No. 1. Specimen 34826, Lake Superior Division, U. S. Geol. Surv., 240 N., 1,250 W., sec. 35, T. 42 N., R. 29 W., Upper Peninsula of Michigan. No. 2. Specimen 36058, Lake Superior Division, U. S. Geol. Surv., 325 N., 1,225 W., sec. 36, T. 42 N., R. 29 W., Upper Peninsula of Michigan. No. 3. Specimen 36080, Lake Superior Division. U. S. Geol. Surv., 15 N., 1,025 W., sec. 31, T. 42 N., R. 28 W., Upper Peninsula of Michigan. 392 THE CRYSTAL FALLS IRON-BEARING DISTRICT. (3) The mica-scliists are not widel}' distributed in the portion of the Arcliean areas included in the Felch Mountain map. They are well rep- resented in the northern Archean area beyond the limit of the area mapjjed, but within this limit they are known only in sees. 34 and 35, T. 42 N., R. 29 W., where an overthrust fault brings them into successive contact with the Randville dolomite and Sturgeon quartzite for a distance of three- fourths of a mile. An excellent section, which includes the faulted con- tact with the dolomite, is exposed along the Sturgeon River below the dam in the northern portion of section 35. Though so feebly represented, they possess an unusual interest both in their field relations and in their microscopic characters. The mica-schists when fresh are dark gi'ay, rather soft rocks, of fine to medium grain, with a generally well-developed schistose structure. The HKist noticeable constituent, in spite of the dark color, is muscovite, which occurs in pearl}' flakes of large size plentifully sprinkled along the cleav- age surfaces, and is especially characteristic of thin seams, which are much more fissile than the rest of the rock and part it into parallel bands of nmeh regularity. Biotite, however, is the more abundant mica, although in smaller and less conspicuous plates, and to it the dark color of the rock is due. Quartz and sometimes feldspar may also be recognized. These rocks ofter little resistance to the weather. The biotite gives up its iron with great ease, staining the outcrop a dull red. The final product is a slightly coherent ferruginous mixture in which the large muscovite plates alone are recognizable. At a less advanced stage of weathering the alterna- tion of layers more rich in biotite produces color banding in reds and grays. The mica-schists contain many intruded dikes and sheets of flesh-colored pegmatite and also of amphibolite, both of which are generally parallel to the foliation. The pegmatites are typical " schrift-granits," the feldspar being microcline. Both pegmatites and amphibolites show ragged and intrusive con- tacts with the schists when these are examined in detail. Both also are foliated. Under the microscope the mica-schists are thoroughly crystalline aggre- gates of quartz, biotite, and muscovite, always with more or less microcline. Magnetite is always present as a primary mineral, and hematiteorsomehydrous oxide of iron between hematite and limonite is very abundant in the zone of weathering. Besides these, tourmaline is an abundant accessory in some slides, and apatite, zircon, titanite, pyrite, and chlorite also commonly occur. AKCDEAN IN FELCH MOUNTAIN DISTRICT. 393 Quartz occurs in small and often partly rounded areas, some of which have a very clastic appearance. Except as stated below, it is generally free from inclusions of the micas, which suiTOund and terminate against it in such a way as to indicate that it crystallized the earlier. It is often crowded with fluid and gas inclusions, and an occasional grain bristles Avith radiating clusters of rutile needles. Minute crystals of magnetite are also frequently inclosed. The inclusions of all kinds are frequently grouped in roughly oval areas near the centers of the grains, while between the nuclei and the wandering perimeters the quartz is relatively free from inclusions. Biotite, varying in color from dark brown to light yellowish green, is the predominant mica. It occurs in irregular plates, generally much larger than the quartz; the great abundance and uniform alignment of these plates produce the schistose structure. As already stated, it includes and is there- fore younger than the quartz generally, but it is also found, though rarely and always in very minute plates, included in the small quartz grains which are so abundant in the fresh microclines. The latter occurrences belong to an earlier generation than that of the larger biotites: The chief interest attaching to the biotite is in its alteration under the attack of the weather. The iron separates out along the cleavages in little spheroidal drops and flattened plates, which are red and translucent, but not quite of the deep color of hematite. Doubtless they contain some water, and are possibly close to gothite in composition. Between the red globules the biotite sub- stance becomes paler, its pleochroism diminishes, and double refraction increases, and finally, in a slide containing no basal sections, it can not be distinguished from muscovite. The separated ferric oxide remains in the mica, and while the rock remains firm does not travel and stain the other constituents. In these stages the slide contains a very faintly colored bleached biotite, which is sprinkled through and through with the little dots of bright red iron ore. Muscovite is not very abundant. It is sometimes intergrown with the large biotites, and occurs under similar conditions, but it chiefly comes in little ragged inclusions in the secondary microcline. In the form of aggre- gates of sericite it composes the macroscopically conspicuous pearly micas, and also is an abundant constituent, and sometimes the only representative, of the partly absorbed and older feldspars included in the microcline. Microcline is always a secondary mineral, and is present in variable 394 THE CRYSTAL FALLS IRON-BEARING DISTRICT. amounts in different sections. It incloses quartz, the micas, niag-netite, and an older feldspar. These inclosnres are usually small; they often lie in par- allel alignment in the same and adjoining microclines, and the lines in which they are disposed sometimes bend, apparently indicating that the original i-ock was minutely puckered. The inclosed quartz sometimes incloses .smaller flakes of biotite and muscovite, as well as magnetite and rutile needles. The inclosnres in the little grains of quartz are frequenth^ con- centrated in the centers, as in the case of some of the quartzes outside the microclines, as described above. The microcline sometimes occurs in a few scattered grains; sometimes with its inclusions it makes up almost the whole rock. In its manner of occurrence, its inclusions, and the way in which these are disposed within it, it is strikingly like the secondary albite of the Hoosac schists of western Massachusetts, described by Prof. J. E. Wolff.^ The microclines are distinctly elongated in a direction parallel to the foliation, to which they thus contribute. In a few cases the elongation is parallel to a line, and does not appear in thin sections cut normal to this direction. But in most cases the crystals are flattened parallel to a plane. These forms are those of crystallization ; except along the secondary fracture planes the microline is entirely free from breaking or granulation. The following is a complete analysis of a representative specimen of the mica-schist : Analysis of mica-schist, [By ])r. H. S. Stokes, U. S. Geol. Survey.] SiO. 64.71 .72 Noue. .02 16.43 1.83 3.84 Trace. TiOj CO, PC- Al,03 Fe-Oj i FeO MnO 1 CaO MgO K-O Na.O H.jOat 110° ... H,0 above 110° Total 0.08 2.97 .5.63 .11 .31 2.79 99.44 No. 1. Specimen 34822, Lake Superior Division, U. S. Oeol. Surv., 1900 N., 1310 W., sec. 35, T. 42N., R. 29 W., Upper Peninsula of Michigan. ' Mon. IT. S. Geol. Survey, Vol. XXIII, pp. 59-63. ARCHEAN IN FELCH MOUNTAIN DISTKK T. 395 In its low silica and lime, ami high iron and magnesia, this rock differs in important partienhirs from the granites, to which in its mineral com- position it is allied. In tliese respects, as well as in the great excess of potash over soda, it closely approximates the composition of certain clay slates.' The original character of the mica-schists is indeterminate. They may be altered sediments, as the chemical analysis indicates, but if so they no longer contain any material which can be proved to be in its original form, and in view of the complete recrystallization, for which the evidence is clear and striking, this could not be expected. Their mineralogical relationship and close association with the granites and gneisses is perhaps a reason for regarding them as autoclastic rocks, derived from originally massive granites by dynamic metamorphism. If this be true, then the crust movements which crushed the parent granite l)elong to pre-Algonkiau time, for the later stresses which folded and l:)rought the schists into faulted contact with the Randville and Sturgeon formations found them with a parallel foliation which it bent and crumpled, and no period of great stress earlier than this is known in Algonkian time. The complete recrystallization may be referred with probability to the period of quiescence following the faulting and folding, during which also occurred the recomposition of the older Algonkian formations. (4) The amphibolites or hornblende-gneisses are widely and abundantly represented in the Archean. Macroscopically they are black or dark- green rocks of medium to fairly coarse grain, the fresh fractures of which glisten with the cleavage surfaces of hornblende, which is much the most abundant and often the only recognizable constituent. They are universally foliated parallel to the foliation of the associated gneisses, and exhibit, but in a more marked degree, the same varieties of structure. The folia- tion is easily recognized by the eye as due to the parallel arrangement of tlie hornblende prisms. Depending mainlj^ upon the position of the hornblendes relative to the other constituents, the structure is either " of the plane-parallel or lineai'-parallel type, the latter often superbly developed. The essential constituents of these rocks are common ffreen horn- blende, plagioclase, biotite, and quartz. The structure is thoroughly crys- ' See analyses quoted by Kemp, Handbook of Rocks, p. 107, noB. 4 and 5. 396 THE CRYSTAL FALLS IRON-BEAEmG DISTRICT. talliue. The lioniblende occiu-.s iu long prisms 3 to 10 mm. in length, which lie close together, and inclose, partially surround, and abut against smaller angular grains of plagioclase. The plagioclase is quite unstrained and is iisually fresh and clear, and entirely without crystal boundaries. Brown biotite is iiniversalh' present in small amount, in long plates parallel with the foliation. It does not seem to be an alteration product from the horn- blende. Quartz is the least abundant constituent. It is crowded with fluid cavities and needles of rutile, and often incloses minute crystals of horn- blende. The plagioclase, from its high extinction angles and alteration products, is evidently basic. A little magnetite is present, but titanite has not been observed. The structural features are well brought out in thin section In the linear-parallel type the hornblendes all lie with their crystallographic axes parallel to a line. A thin section parallel to the foliation cuts essentially all in the zone of the prism or near it; one across the foliation gives only sec- tions across the prism. The grains of plagioclase are general!}" elongated without strain. Their outlines are most irregular and quite independent of the twinning lamellae. Their g-eneral apjiearance is that which would be presented if numerous crushed contiguous grains had united by some proc- ess of annealing or absorption to form the new individuals. In the plane- jjarallel type the only difference is that the hornblende prisms have grown parallel to a plane, in which, however, they may have any orientation An indistinct banding is also often observable in this tyjie, caused by a partial grouping of the light and dark constituents in parallel layei's. The order of crvstallization seems to have been plagioclase first, but nearly contempo- raneous with the hornblende and biotite, and the quartz last. The amphibolites occur in comparatively narrow bands of indefinite length in the granites and gneisses. The width usually does not exceed 8 to 10 feet, and their dip is always at high angles. The boiindaries are invariably sharp, and frequently cut the foliation of the ampliibolite within and of the gneisses without somewhat obliquely. There is a general imi- formity of grain throughout the width; the wider bands are not coarser than the narrower. ARCHEAN IN FELGH MOUNTAIN DISTRICT. 397 The IbllowiugcoiupU'tc iuialysi.s sliows tliu cliemical character of a rep- resentative specimen ofainpliiboHte: Analysis of amphibolite. [Hy Dr. H. N. Stokes, U. S. Geol. Survey. 1.1 SiOj 50.36 TiO, 1.77 CO: None. PsOj. .20 Al.O, 13.26 Cr,0;, Fe .0:, 6. 30 FeO 9.34 MnO Trace. 1.' NiO 1 CaO 7.85 5. 55 1.14 2.11 .16 1. 55 MgO K.O Na.O HjO at 110- H-,0 above 1 10^ Total 99.59 i ' Ba, 8i-, Li, CI, S, SO3 were not looked for. No. 1. .Siiecimeii 36407, Lakr .Superior Division, V. S. Cieol. .Snivey, 1140 N., 1000 AV., sec. 32, T. 42 N., R. 28 W., Upper Peninsula of Michigan. From this analysis it appears that the rock has essentially the compo- sition of diabase or basalt. The composition of the amphibolites, as shown by the above analysis, and their field relations leave little room for doubt that they are old dikes of basic rock. Their present crystallization is of course not that due to original cooling, since among other reasons it bears nO relation either to their thickness or t<:) distance from the walls. The evidence of complete recrys- tallization in place after consolidation which they thus afford, and the unquestionable community of origin between their foliation and that of the gneisses, are significant facts in the metamorphic history of the Archean of this district. It is for this reason tliat they are described with the Archean and not with the intrusives. Whether they are really Archean intrusions and not of Algonkian age can not, perhaps, l)e known with certainty. Basic rocks having approximately the same composition are known to have penetrated the Algonkian, but they have not undergone the same recrystallization. These last besides have their known analogues, equally unmetamorphic in the Archean itself. For tliese reasons it seems probable that the amphib- olites were intruded into the Archean before the Algonkian rocks of this district were deposited. 398 THE CRYSTAL FALLS IRON-BEAEING DISTKIOT. SECTION IV. THE STURGEOK QITARTZITE. The lowest member of the Algoiikiaii in the Felch Mountain range is a formation consisting mainly, but not exclusively, of coarse vitreous quartzite. Typical exposures of this formation, as well as one of the rare contacts between it and the underlying Archean, occur along the Sturgeon River, and it is therefore named the "Sturgeon Quartzite." DISTRIBUTION, EXPOSURES, AND TOPOGRAPHY. The Sturgeon formation, next to the Randville dolomite, is the most widespread member of the Algonkian series in the Felch Mountain range. Its general distribution throughout the area mapped is in two ])arallel zones, of varying width, immediately adjoining the northern and southern Archean, except wlien displaced from this position for relatively short distances by faults. These zones extend east and west for the whole length of the range. Their surface width varies with the complexity of the structure and the depth of erosion. In part of sec. 35, T. 42 N., R. 29 W"., the higher formations have been entireh' removed, and the two zones come together, leaving the (][uartzite as the onh^ Algonkian rock at the present surface. On the whole the Sturgeon formation is fairly well but A^ery unevenly exposed. Beginning at the west the zone in contact with the southern Archean furnishes frequent outcrops from the south quarter post of sec. 34, T. 42 N., R. 30 W., to the south quarter post of sec. 36, T. 42 X., R. 30 W., a distance of 2 miles. Then follows a gap of a mile in which no outcrop . have been found. Near the north and south quarter line of see. 31, T. 42 N., R. 29 W., they l)egin once more, and are supplemented by test pits as far as the West sixteenth line of section 32, next east. Then follows another gap without exposures, 2^ miles in length. Near the north-and-south quarter line of sec. 34, T. 42 N., R. 29 W., outcrops begin again and continue for a mile to the east, with frequent inter- ruption, as far as the north-and-south quarter line of section 35, where in the valley of the Sturgeon the southern zone broadens and jt)ins the north- ern, in consequence of the general westward pitch which has carried the higher formations above the present surface of denudation. East of this point the quartzite is known in only a few scattered localities. In the southern part of sec. 36, T. 42 N., R. 29 W., it is in contact with the Archean on the south l)ank of the Stur"-eon River. South of Felch Moun- STURGEON QUAKTZITE IN FELCH MOUNTAIN RANGE. 39y tiiiu, in src. .'52, T. 42 N., Ji. 2S W., it oiitcntps iiniiieili;itel\- soutli of the abiUidoiied Northwcstci-ii mine, iiud lifis also Ix-eii t'ouiiil iu (lrilliii<>' (in the west and in test pits on llie cast of the natnral exposnres throngh a distance of lialf a inih-. In sec. 33, T. 42 W., R. 28 W., a small ledge, a few feet square, occurs between the overlying- dolomite and the Archean, 200 feet east of the road to the Calumet and Hecla (iron) mines. East of this the contact between the Archean and the Algonkian is a faulted one, and the quartzite is buried beneath the overlying- formations. The northern zone of the Sturgeon formation is not nearly so well exposed, nor for the most part does it fall within the artificial line that bounds our detailed work on the north. Sees. 34 and 35, T. 42 N., R. 30 W., on the west contain a few scattered outcrops, one of which is of exceptional })etrog-raphical interest and to be noticed later. The next exposures are o miles east, along- and just north of the north line of sec. 35, T. 42 N., R. 29 W. The main northern zone of the Sturg-eon formation coming from the west lies south of these exposures and is entirely covered. Between the two the tongue of Archean scliists already described is faulted up. Two miles farther east quartzite again appears in test })its, low-lying outcrops and drill holes along the northern border of sec. 31, T. 42 N., R. 28 E., and in section 29, immediately north of section 32, is well exposed in a broad belt that reaches north almost to the east-and-west quarter line. The quartzite often forms distinct linear ridges, which in spite of the chemical staljility and apparent homogeneity of the rock seldom rise to the mean altitude of the neighboring Archean areas. An exception to this rule is the succession of ridges formed by the southern zone in the 3-mile stretch west of sec. 31, T. 42 N., R. 29 W.; these frequently overtop the adjacent Archean plateau. Very frequently, also, the quartzite zones occupy lower ground not only than the Archean but even than the immediately overly- ing dolomite. The southern zone, for some unknown reason, is a distinctly weak belt east of sec. 32, T. 42 N., R. 29 W., and for several miles forms the bed rock of the Sturgeon and the connecting- valleys. FOLDING AND THICKNESS. It is extremel}' difficult in most cases to determine directly the attitude of the Sturgeon formation, owing- to its generally massive and homogeneous character. This is due, as will be shown hereafter, to the completeness of 400 THE CRYSTAL FALLS IRON-BEARING DISTRICT. the recrystallization, in consequence of which the ordinary sedimentary features that it originally possessed have been almost entirely obliterated. Faint color banding, itself of secondary development, but no doubt pre- serving a distinction in original composition, alone remains, and only here and there, as a guide to the former stratification. By scattered indications of this sort, and by the better evidence afforded by the overlying dolomite, often very distinctly banded, it is known that the southern zone of quartzite on the whole dips toward the north. Southward dips also occur in this belt, by which it is known that subordinate folds occur within the quartz- ite itself. From the considerable variations in the surface width of the forma- tion we are led to suspect the existence of more of these little folds than we are able to prove. However, the secondary syncline, which extends from the offset already referred to in sec. 35, T. 42 N., R. 30 W., for 6 miles to the east to sec. 35, T. 42 N., R. 29 W., and includes no formation higher than the quartzite, is very definitely determined. In the northern belt of the Sturgeon formation the indications of dij^ are generally northward at very high angles. These indications, not in themselves conclusive, are reenforced by a corresponding attitude in the overlying dolomite, and it is therefore probable that there is a general, or at least widespread, overturn in the dip of the northern belt. Since the contacts of the Sturgeon formation with the underlying Archean and with the overlying dolomite are (except in one case) covered, it is impossible to obtain the data for very accurate determination of its thickness. The uncertainty in most outcrops as to the dip of the quartzite introduces an additional difiiculty. However, in sec. 35, T. 42 N., R. 30 W., on the west end of the range, and in sec. 33, T. 42 N., R. 28 W., 11 miles farther east, the covered intervals to the limiting formations are not great, and if the contacts are not faulted (which is far from certain), the minimum thickness is determinable within a reasonable limit of error. In the western locality the surface width of the zone probably under- lain by quartzite is about 500 feet. The quartzite itself is structureless, but the overlying dolomite dips northward at an average angle of about 70°. If the same dip holds in the quartzite, its true thickness is about 470 feet. In the eastern locality similar data lead to a thickness of nearly 430 feet. In these two sections the quartzite zone is much narrower than it is else- where, either because undetected faults have reduced it, or because it is STURGEON QITARTZITE IN FELOH MOUNTAIN RANGE. 401 uncomplicated by subordinate folds. It is probably safe to conclude, in view of the uncertainties, that the average thickness of the formation is not less than 450 feet, and may be considerably more. In a preliminary paper on the district,' written before the field notes were fully analyzed, I have placed the thickness of the quartzite at about 700 feet; but this figure is probably too large. PETROGRAPHICAL CHARACTERS. The Sturgeon formation includes a few very closely related rock varieties, of whicli quartzite furnishes the great majority of the exposures. The quartzites are usually light gray in color, and break with a coarsely granular or glassy fracture. To the eye quartz is often the only recogniza- ble constituent in the body of the rock, although the numerous joint and shearing planes shimmer with little silvery plates of muscovite. Occasion- ally a weathered surface is dotted with minute specks of an opaque pinkish substance, which leads one to suspect the presence of feldspar. Chlorite also is now and then visible in the darker varieties. The quartzites are almost uniformly massive, except for the secondary fractures above mentioned. At scattered localities, however, a faint color- banding, due to the presence of layers of a pinkish hue, which are inde- pendent of the secondary fractures, seems to indicate the original stratifica- tion. The color bands are generally only vaguely defined ; occasionally, however, they are numerous and shaq). Closely associated with the massive quartzites are sheared quartzites, or micaceous quartz-schists. These rocks are merely varieties of the quartzite in which secondary shearing planes, with their attendant growths of new muscovite, are more abundant than usual. The shearing surfaces almost invariably intersect, with the result that the new structure tends toward the linear-parallel type, and is often as similar in appearance as it is in origin to the structure already described in connection with the sheared granites. In a locality already referred to, on the south bank of the Sturgeon, in sec. 36, T. 42 N., R. 29 W., where the Sturgeon formation is in visible contact with the Archean, the quartzite is underlain by a considerable thickness of very fissile musco^-ite-biotite-gneiss, which incloses rather sparingly obs cure pebbles of granite and quartz. This gneiss, which no 'Relations of the Lower Menominee and Lower Marquette series in Michigan (Preliminary)- Am Jour. Sci., Vol. XLVII, 1894, p. 217. .> ^ • MON XXXVI 26 402 THE CRYSTAL FALLS IRON-BEARING DISTRICT, doubt was formerly an arkose rich in feldspar, has recrystallized and after- wards been sheared; the coarse micas to which the fissility is due, together with other new minerals, have grown between the fractured surfaces and recemented the broken mass. It affords beautiful examples of foliation parallel to a line. The thin sections of the Sturgeon quartzite are of exceptional interest. The principal constituent is, of course, always quartz. With the quartz are associated, in nuich smaller amounts, and not necessarily all in the same section, numerous accessories, including muscovite, biotite, chlorite, micro- cline, orthoclase, plagioclase, titanite, rutile, zircon, apatite, and the ores. The relations of the quartz to the other constituents present very unusual features, and indicate that the metamorphic changes by which the present completely crystalline rock has been made from an original granitic sand have proceeded along lines not hitherto distinctly recognized in the forma- tion of rocks of this character. Among the large number of slides examined, a broad distinction can at once be made between those which show the effects of stress in a pro- nounced degree and those in which such effects are subordinate or hardly noticeable. Connecting these two classes is a pei*fectly graded series; and it is therefore certain that those of the first are merely the more or less modified varieties of an earlier stage, represented more nearl)^ by the second. In the slides in which the effects of pressure are least apparent the micro- scopic characters are as follows: The background is composed of large irregular grains of quartz, the edges of which interlock with the most minute and sharp interpenetrations. The longest dimensions of these grains range from 1.5 to 6 mm., avei'aging perhaps 2.5 or 3. The"\' often have a rather vague parallel elongation, which corresponds to tlie alignment of the minerals which diey inclose. Scattered very abundantly through these large quartz grains are the accessoiy minerals, some predominating in one slide, others in another, but the micas and chlorite occurring in all. Through each slide the accessory minerals, with the exceptions noted below, lie with their long axes in a common direction, and frequently cross the serrated boundaries between adjacent quartzes. The inclusions in many cases have the form and other characters of clastic minerals, and thus preserve the only microscopic evidence of the original nature of the rock. The included micaceous minerals are usually in small plates, ranging STURGEON QUARTZITE IN FELOH MOUNTAIN RANGE. 403 from 0.05 to 0.75 mm. in longest dimensions, but few, however, exceeding 0.2. Many of tliese are bent and split, the clejir mistrained quartz of the host jiene- trating from the frayed edges into the interior between the partly separated leaves. Biotite and muscovite, and sometimes chlorite, occur in the same individual, indicating alteration before inclusion in the quartz host took place. Besides its conuuon occurrence as an alteration j^roduct of the biotite, a few rounded areas of chlorite, made up of little radiating tufts, seem to be pseudomorphs of garnet. Inclusions of titanite and magnetite^ or a related ore, are not uncommon in the larger micas, and the biotite and chlorite sometimes inclose beautiful sagenite webs. Many of the smaller micas, however, have clear sharp edges and depart from the general paral- lelism of the other inclusions. These are either contemporaneous crys- tallizations or else, perhaps, were primary inclusions in former grains of clastic quartz which has since disappeared. Some of the clastic jjlates of biotite are bleached and include spheroidal blebs of red iron ore, similar to those described in the case of the Archean mica-schists. The microcline inclusions are usually elongated in form, and frequently, jjarticularly in the cases of the larger, have well-rounded clastic outlines. The long dimension, which usually coincides with one of the cleavages of the mineral, rarely exceeds 0.5 mm. or falls below 0.08 mm. The periphery is frequently partly sin-rounded by a thin film of biotite. Within the micro- clines are sometimes contained little blebs of quartz, which are not oriented optically with the host, and also, more rarely, small plates of biotite. The microcline individuals are sometimes broken into two or three differently oriented parts, which may be separated from each other, in which cases the quartz of the host has completely filled the interspaces. Fracture in the feldspar is often unattended with the slightest appearance of strain in the inclosing and cementing quartz, which extinguishes as one individual, and is therefore unmistakably to be attriljuted to stresses previous to the crystalli- zation of the quartz. Besides microcline, both orthoclase and plagioclase are sometimes inclosed in the large quartzes, but much more sparingly. They are invari- ably more or less decomposed, and are sometimes surrounded partially or wholly by a film of ferruginous material. They show the same phenomena of fracture, and occasionally of separation with penetration of the host, as the microcline, and occur in grains having a similar range in size. 404 THE CRYSTAL FALLS IRON-BEARING DISTRICT. Titanite is of frequent, zircon of rather rare, occuiTence. The titanite is found not only inek)sed, as ah-eady stated, in biotite and chlorite, but also in well-rounded clastic grains which are often bordered with an opaque ore. Zircon occurs in broken grains, without doubt clastic, and also in small crystals which show no signs of wear. These -last were probably entirely embedded in original clastic grains of quartz. Besides the above minerals of usual occurrence, small quartz grains of different orientation from the matrix are ^'ery rarely found included in the large quartzes of the general background. Only two or three such cases have been oljserved, and in these the included grain is surrounded almost wholly with thin plates of mica. It is believed that these are original clastic grains which, perhaps because protected by a film of material now represented by the micas, have escaped the general fate of their neighbors. One or two composite inclusions, made up of microcline, the micas, and quartz, have also been noticed. These seem to represent original pebbles of granite or a crystalline schist. The pressure effects begin with the appearance of optical strain and decided elongation in the large quartzes of the groundmass. This is fol- lowed by fracture, either along or quite independent of the original sutures, the crack often halting in the interior of a grain. The fractures preserve very roughly the same general direction, but frequently intersect at -very acute angles, or come together in sweeping curves. The breaking is fol- lowed by movement, and this results in the production of a fine-grained quartz mosaic between the parted surfaces. In the final stages shown in the series of slides in my collection, the rock is made up of long, narrow lenses, each of which is an enormously strained qviartz individual, separated by narrow anastomozing zones of very finely subdivided quartz. After the fracturing took place there seems to have been no further distortion of the lenses, for the edges of adjacent individuals follow similar curves, which are often reversed, and in many cases could be brought together with an accurate fit. If the Sturgeon quartzite represents an original sandstone, it is evident from the facts stated alcove that the old quartz grains have undergone com- plete recrystallization. The usual conception, since the time of Sorby, of the process by which quartzites are formed from original deposits of sands is that new quartz is deposited around each original fragmental quartz grain, STURGEON QUARTZITE IN FELCH MOUNTAIN RANGE. 405 ill similiu- crystallogTaphic orientation with it, and that neighboring grains thus enlarged finally interlock bv mutual limitation of one another's growth. Tliis explanation evidently can not account for the background of large interlocking qnai-tz areas in these rocks, for if it were true it would be nec- essary to assume that the quartz grains were less numerous in the original deposit than those of almost any other mineral, in some slides even than the titanite or chlorite. There seems to be but one escape from the conclusion that the large quartz areas nuist each represent a nnmber of original frag- mental quartz grains, which, as deposited, nuist have lain in the rock with their crystallographic axes disposed entirely at haphazard; and that is the hy])othesis that this quartzite was not originall}' a sandstone, but consisted mainly of soluble and easily replaceable material, such as limestone, with the fragmental particles scattered through it, and that the large quartzes of the background have replaced this soluble substance. I have been able to tind no positive evidence to support this hypothesis, and I am com- pelled to believe that the rock was a sandstone in which, in some way not easy to understand, considerable numbers of adjacent quartz grains have united to form or have been absorbed into a new individual, leaving absolutely no trace of their former separate existence. The introduction of new silica, or the separation of silica from decomposing silicates in the rock itself, may well have been essential factors in the recrystallization. I shall make no attempt to explain the process further than to point out its probable analog}' with the ^^rocess by which the new microclines were formed in the Archean mica-schists. The close alignment of the clastic minerals inclosed in the large quartz areas, their frequent fracture, and their occasional separation, indicate that the time of crystallization probably followed a period of stress; while the very vague j^arallel elongation of the individuals of the background in the unstrained sections would seem to show that they crystallized under static conditions. Unquestionable proof of a period of stress later than the crys- tallization is given by the numerous slides, in which these grains are seen to have suffered fracture and distortion. The microscopical study of the quartzites thus supplies important evidence, not afforded by the outcrops, as to the erogenic history of the district. 406 THE CRYSTAL FALLS lEON-BEARIXG DISTRICT. SECTION V. THE RANDVlLIiE DOLOMITE. The Sturgeon quartzite is succeeded by a formation consisting, so far as is known, almost wholly of crystalline dolomitic rocks. Excellent exposures belonging to this formation are situated within a short distance of Randville station, on the Milwaukee and Northern Railway, and it may therefore convenienth' be named the Randville dolojnite. DISTRIBUTION, EXPOSURES, AND TOPOGRAPHY. Owing both to its great thickness and to its intermediate ^^osition in the series, the Randville dolomite in the Felch Mountain range coveis a larger share of the surface than any other member of the Algonkian suc- cession. The overlying formations are frequently interrujjted, because of the changes in direction of pitch of the secondary synclines in which they occur. Ill these gaps the dolomite covers the whole interior of the syncli- norium. Wliere the higher formations are present, they divide the dolomite into two or more parallel east and west belts, one of which lies south of the northern quartzite and the other north of the southern. Only in portions of sees. 35 and 36, T. 42 N., R. 29 W., wliere the rise in the axis of the main syncline has lifted it above the present surface of denudation, is the dolomite entirely absent from the main trough. Natural exposures of the dolomite are not so numerous as of the quartzite, but they are more evenly distributed. Moreover, owing to its proximity to the Groveland iron formation, the dolomite has been penetrated by many test pits and diamond-drill borings put down in search of ore, and these supply important information in the covered areas. From the western end of the map to sec. 34, T. 42 N., R. 29 W., the dolomite is for most of the way separated into two or more parallel belts. The southern belt is es^jecially well exposed in sees. 35 and 36, T. 42 N., R. 30 W., and in sec. 31, T. 42 N., R. 29 W., and for 2 miles to the northeast, beyond which it has been found only in test pits and drill holes. In the middle of sec. 35, T. 42 N., R. 29 W., the base of the formation is brought to the surface by the westerly pitch of the main fold and is well exposed along Sturgeon River. North of the strike fault, which, as alread's' described, has brought the RANDVILLE DOLOMITE IN FELGH MOUNTAIN RAISIGE. 407 Archean inica-sclnsts into contact with tlie dolomite and quartzite in the northern part of the same section, the Randville formation runs east in a single belt, which probably continuously widens as the throw of the fault diniini-shes. It has been found 'in several places in the north half of sec. 31, T. 42 N., R. 28 W., and near the east line of this section the appearance of the overlying mica-schists again divides it into two belts, which pass to the north and south of the Felcli Mountain syncline. The northern belt has been proved by test pits only, but the southern is well exposed natui'ally in the neighl)orhood of the Northwestern mine. Other exposures also occur south of the unconformable mica-schists and quartzite of the upper series, in the central ])ortion of sec. 33, T. 42 N., R. 29 W. The dolomite is relatively a weak rock, and generally occupies lower ground than either the quartzite below or the iron formation above it. The belt in contact with the southern belt of quartzite especially is valley making throughout most of its extent. The outcrops usually form low, steep-isided knolls elongated with the strike and of slight relief above the basement; these occasionally unite into linear ridges, as in sec. 35, T. 42 N., R. 30 W. The northern belt is one of low general relief, from which, how- ever, similar isolated knobs often protrude. The largest and most prominent of these is the peak in the northeast quarter of NW. ^ sec. 36, T. 42 N., R. 30 W., which rises 80 feet above its base, covering 8 or 10 acres. No actual contacts between the Sturgeon and Randville formations have been found, but from their close association and continuity, as well as from the structural characters, Avhen these are determinable, they seem everywhere to be strictly conformable. Near the quartzite the dolomite becomes distinctly more impure and contains a larger proportion of silicates and quartz. It is altogether probable that between them come transition beds, as indeed is shown by some of the drill records. In one of these " talckv mica-schists, micaceous limestone, altered actinolite-schist, and quartzite" are described as being interbedded near the junction. The determination of the thickness of the Randville formation is beset with the same difficulties as are encountered in the case of the quartzite, namely, the uncertainty as to the exact position of the contacts and the possibility of faults and subordinate folds within the formation itself The best sections give a wide range of values from a mininuim of about 500 408 THE CRYSTAL FALLS IRON-BEARING DISTRICT. feet near Felcli Mountain to a maximum of nearly 1,000 feet in the western part of the district. While the discrepancies may be partly due to lack of precision in the data, it is probable that the thickness of the formation is not uniform, but really increases from east to west. On the Fence River, 18 miles northwest of Randville, the thickness is probably about 1,500 feet. Accordingly, accepting- each of these determinations as approximately correct, 700 feet may be taken as a fair estimate of the average thickness of the Randville dolomite within the Felch Mountain range. PETROGRAPHICAL CHARACTERS. The outcrops of the Randville formation consist exclusively of dolo- mite, more or less pure, and always thoroughly crystalline. A few comparatively thin layers of schists, probably both micaceous and amphibolitic, and als(i of quartzite, are mentioned in certain drill records, to which I have had access as occumng interbedded with the dolomite; and wliile the lithological determinations are perhaps not entitled to much weight, thev at least prove the existence of i-ocks which are not dolomite within the formation. In the field, however, such interbedded layers do not outcro}), and they must constitute an extremely small part of the total thickness. From the results of my w(irk the Randville formation appears as a lithological unit. Macroscopically the dolomites are rather coarse-grained marbles, of various shades of color, of Avhicli jjinkish or bluish white are the most common. They always inclose, more or less abundantly, large flakes and aggregates of tremolite, which are particularly noticeable from their projec- tion above the weathered surface. Occasionally tremolite and other silicates are the most abundant, and sometimes, for small thicknesses, are essentially the only constituents. Quartz and chlorite are also often present, but in much smaller amounts. The weathered surface is usually dulled to a light brown or creamy yellow in a thin superficial skin, but is not deeply iron- stained, except when the silicates containing ferrous iron are present. The following partial analyses of three specimens from different parts of the range show that the carbonate is normal dolomite. The insoluble portion consists chiefly of tremolite. These analyses were made for me by Mr. G. B. Richardson, a graduate student in geology in Harvard University. KANDVILLE DOLOMITE IN PELCII MOUNTAIN RANGE. Analyses of Randville dolomite. 409 I. II. III. Insoluble in HCl 2.0 1.2 53.2 42.3 9.7 2.1 48.9 38. 29.1 2.2 39.3 27.7 FOiOi CaCO , MgCO , Total 98.7 98.7 98.3 The outcrops, while often entirely massive, usually possess decided structural features. These are indicated by color banding, by differences in texture, and by the banded arrangement of the components. Slight variations in the body color of the rock, proceeding from no distinguishable variation in composition, often occur in alternate parallel layers, which are persistent within the limits of observation. With the color banding often go variations in texture, which, however, are neither so regular nor nearly so persistent. The characteristic form taken by these is in thin layers, which as they continue open out into nodule.s. Such layers consist of closely packed crystalline grains, very much coarser than the body of the rock, which have grown normal to the boundaries. Adjacent layers are not strictly parallel and sometimes cross each other. They are believed to represent ancient fracture and slipping surfaces, which followed very closely the original bedding, in which the new carbonate individuals have had room for larger growth. The arrangement of the accessory minerals, especially the tremolite, also is usually a banded one. Layers rich in tremolite alternate with layers poor in tremolite, while within the layers the orientation of the tremolite individuals is usually at random. The structure brought out in these various ways is, on the whole, a parallel structure. If corresponds with the strike and dip in all the localities where these can be independently confirmed by the attitude of the adjacent for- mations, and it also has been tin-own into minor folds. 1 therefore regard the structm-e as having originated partly in chemical differences in the material originally deposited and partly in secondary growths in the open spaces and rubbing zones determined by relative movements along the sur- faces of easiest fracture at the time of the earliest folding, and for both reasons preserving in the subsequent metamorphism the true stratification of the formation. 410 THE CEYSTAL FALLS IRON-BE ARINO DISTRICT. Under the microscope the dolomites show uo featui-es of special inter- est. They are thoroughly crystalline rocks, chiefly composed of coarse o-rains of dolomite with which is associated a considerable number of acces- sory minerals. Of these the most important are tremolite, diopside, chlorite, muscovite, phlogopite, quartz, and rutile, while apatite, tourmaline, pyrite, and magnetite are rare. The dolomite is by tar the most abundant constituent in most of the slides, and furnishes the general background for the accessories. The shape of the grains in many sections is decidedly oval, and the long axes lie in the same direction, thus producing a foliation. Tremolite is abundant in some of the sections, and is entirely absent from none. It occurs in long-bladed individuals and aggregates, usually bounded by the prism, but one or both pinacoids are also sometimes present. It includes portions of the carbonate background. Diopside is rather rare; it occurs usually in small single individuals, with sharp crystal outlines. It is sometimes surrounded by tremolite, from which it is distinguished by its high obliquity of extinction and its almost rectangular cleavage. Partings parallel to both ])inacoids, as well as a transverse parting in prismatic sec- tions, are also observable. Quartz occurs in irregular grains completely interlocking with the dolomite, and in some cases with tremolite. In the slides examined it is in all cases a secondary as well as a rare constituent. In no case is there any indication that it is clastic. Chlorite is an abundant constituent of some of the slides, while from others it is entirely absent. Muscovite in little frayed plates is plentiful in some sections. Quite pos- sibly some of these may be original clastic particles. The most interesting mica, however, is phlogopite, which is very abundant in one locality near the base of the formation. It occurs in large, cleanly bounded plates, each of which is a multiple twin, and evidently a product of secondary crystallization. Some of these plates have been strongly bent, thus showing that the dolomite, like the quartzite, has been deformed since it crystallized. The thin sections therefore show that the rocks of this formation have experienced even more nearly complete reconstruction than is shown in the case of the quartzites, for here none of the constituents, except jjossibly some of the smaller micas, are present in their original form. Also the evi- dence for disturbance after crystallization is of similar character and equally MANSFIELD SCHISTS IN FELOH MOUNTAIN RANGE. 411 strong. Accordiiig-ly, a close agreement in the sequence and in the charac- ter of the i)rincipal events thus indicated in the history- of the two rocks may he recognized. These considerations make it quite certain that tlie recrystalhzation of the two formations was • essentially contemporaneous. From the character of the accessory minerals in the dolomite it is probable that the crystallization was not accompanied by the introduction of foreign material from outside, in notable quantities, but consisted in a mineralog- ical rearrangement of the elements present in the rock from the. beginning. SECTIOX VI. THE MANSFIELD SCHISTS. Above the Randville dolomite comes a formation composed chiefly of fine- to medium-gr;iined mica-schists. Owing to their exceedingly soft character and small thickness, these rocks are exposed naturally in only a few localities in the Felch Mountain area. A series of phyllites less meta- morphic but otherwise similar, and occupying the same stratigraphical position, immediately above the dolomite, outcrop characteristically at the Mansfield mine, and especially north of it, near the Michigamme River, in T. 43 N., R. 31 W. For these reasons it is convenient to name the forma- tion for the Mansfield locality. DISTRIBUTION, EXPOSURES, AND TOPOGRAPHY. The existence of the Mansfield formation in the Felch Mountain area is known mainly from test pits and the records of diamond-drill borings and early explorations. Fortunately, these are so widely distributed that the persistence of the formation is well proved. Many drill holes have passed tlu-ough it into the dolomite. Immediately above it comes the mag- netic Groveland formation, which even when covered betrays its presence to the compass needle. With the upper and lower limits thus determined, and with the large body of data supplied by the test pits and records, there is no difficulty in indicating its approximate boundaries for the greater part of the map. On the west, mica-schists belonging to the Mansfield formation have been proved b}" diamond drilling to occur between the dolomite and Grove- land formations in the south half of sec. 34, T. 42 N., R. 30 W. Farther east there is a line of outcrops in the eastern portion of section 35, and the 412 THE CRYSTAL FALLS IRON BEARING DISTRICT. schists have also been found in test pits on both sides of the western exten- sion of the Groveland syncline in sec. 36, T. 42 N., R. 30 W. In sec. 31, T. 42 N., R. 29 W. (the Groveland section), they have been penetrated in 10 drill holes, besides numerous test pits, giving altogether a cross section more than half a mile in length from north to south. In the nortliern half of sections 32 and 33 numerous test pits have exposed the Mansfield formation, proving that it borders on both sides the narrow syncline, the interior of which for a mile and a half is occupied by the magnetic Groveland jasper. Tlu'ough sees. 34, 35, and 36, T. 42 N., R. 29 W., and sec. 31, T. 42 N., R. 28 W., the mica-schists have not been discovered, probably both because they are but feebly represented and because but few test pits have been sunk tlu'ough the Cambrian blanket. In sees. 32 and 33, T. 42 N., R. 28 W., the mica-schists have been found in scattered test pits and borings on both sides of the interior jasper of the Felch Mountain spicline, and also on the south side of section 33. The thickness of the Mansfield formation is so small — not more than 200 feet — that it produces no very noticeable efi'ects on the general topog- ra]3hy, in spite of the ease with which it weathers. In the western portion of the district, througli sees. 34 and 35, T. 42 N., R. 30 W., with the dolo- mite it underlies a broad low-lying plain, which is bounded on the south by a ridge of the Sturgeon quartzite backed by the. Archean plateau. On the north, a broad ridge, through which diagonalljnfass the Archean gran- ites and gneisses, the quartzite, and the dolomite, defines this valley as far east as the middle of section 35 ; in the northern and central portions of this section it spreads out into a swampy lowland, diversified by glacial sand plains, expressive of the gradual widening of the trough and of the gen- erally horizontal attitude of the soft rocks of the interior. The most defi- nite topographical feature directly due to the Mansfield schists is the narrow steep-sided valley which runs east from this lowland for nearly 2 miles, on the south side of the Groveland syncline. The ancient stream valley filled with the Cambrian sandstone, already mentioned, follows along this narrow belt. PETROGRAPHICAL CHARACTERS. The hand specimens from the various test pits, the drill cores, and the few small outcrops indicate that the Mansfield formation is quite uniform in character throughout the Felch Mountain area. The great majority of MANSFIELD SCHISTS IN PELOH MOUNTAIN KANGE. 413 the speciinous ure of fine-grained mica-schists, the color of wliicli varies from light to dark, according as muscovite or biotite is the })redominant mica. Garnets, in some localities, are very abundant, especially near the contacts with intrusives. It appears from the records of explorations that thin seams of jaspery iron ore interlaminated with the schists have been encountered in occasional drill holes and test pits, but no specimens of such occurrences have been obtained. Their existence is of interest, as showino- the likeness in an imjiortant character of these more altered rocks with the slates occupying the same relative position in the Iron Mountain and Norway areas. The outcrops and specimens are frequently well banded in lighter and darker layers, the color banding in seme cases not coinciding with the schistosity. Just south of the GroA^eland mine, in a test pit which was sinking at the time of my visit, the color bands which mark the true strati- fication, as shown by the contact with the underlying dolomite, are closely crumpled and cut by the foliation of the rock, which is much the more dis- tinct of the two structures. Near the contact with the overlying Grroveland formation the mica- schists become both more siliceous and more ferruginous, and there is accordingly a distinct passage between the two formations. This does not necessarily signify a transitional character in the original sediments, but may be altogether due to the downward transportation of silica and iron from the upper rock. The mica-schists are generally very tender rocks, and the material on the dumps of test pits sunk in them is usually far gone in decomposi- tion after a few years' exposure to the weather. From even the freshest specimens the little flakes of mica often rub off on the fingers. Where penetrated by intrusions, however, as in sec. 36, T. 42 N., R. 30 W., and in sec. 31, T. 42 N., R. 28 W., they become very much harder. Under the microscope the rocks of this formation are seen to be in the main thoroughly crystalline, though very fine-grained, aggregates of biotite, muscovite, chlorite, quartz, and feldspar, with the iron ores, rutile, tourma- line, and apatite as the accessories. Garnets are abundant in some of the sections, and with these also occur actinolite, epidote, titanite, and an unde- termined colorless amphibole in stout single prisms. In the eight thin sec- tions which I have examined from this formation I have found no material 414 THE CRYSTAL FALLS IRON-BEARING DISTRICT. which is certainly original and fragmental, although almost every slide contains grains that may possibly be such. On the other hand, it is evident that the large majority of the individual grains have formed in place. The micas are in most cases the most abundant constituent; sometimes muscovite, though iisually biotite, predominates. The two micas are often intergrovvn. The biotite is usually very deeply colored, both brown and green, and, except in the thinnest slides, is almost opaque even in cleavage sections. The larger mica flakes do not exceed 0.5 mm. in length, and average not more than 0.25 mm. Quartz generally occurs in irregular grains, full of fluid inclusions, and inclosing the various accessories. It frequently appears in little triangles in the interspaces between adjacent flakes of mica. Rarely part of the peri- meter is rounded and embedded in a mica, thus suggesting a clastic origin. Feldspar is very abundant in some of the slides and entirely absent from others. Both microcline and plagioclase occur, and in forms similar to the quartz. Biotite sometimes penetrates in irregular shredded edges and filaments into the interior of the feldspars, and in such cases may be a metasomatic product, as described by Irving and Van Hise ^ in the mica- schists of the Gogebic district. But much of the feldspar, as shown by its form and freshness, has recrj^stallized. The alignment of these minerals is with the schistosit)^ of the rock, which they thus determine. When the schistosity cuts the lines of stratification, as it frequently does, the latter are but faintly marked in the thin section by very slight mineralogical diff'erences. Thus a dark baud, which may be very striking macroscopically, may be due merely to the predominance of deeply colored biotite; a light band, to the predominance of muscovite. Sometimes, however, in these bands a grain of quartz, or a stout flake of muscovite, lies out of the general orientation and with the direction of the band. Such grains are very possibly original. The schistose structure, as has already been stated, is determined by the general parallelism of the long axes of the constituent grains. Since the greater part, if not demonstrably all, of these grains have formed in this position, and have not been forced mechanically into it, the cases in which the schistosity cuts the bedding support the inference as to the time of the general recrystallization of the series grounded on the facts observed ' The Penokee-Gogebic iron-bearing district of Michigan and Wisconsin, by R. D. Irving and C. R. Van Hise : Mon. U. S. Geol. Survey, No. XIX, 1«92. GKOVELAND FORMATION IN FBLGH MOUNTAIN RANGE. 415 ill the lower t'onujition, luunely, that this time foHowed a period of great stresses. Also a period of still later stress has affected the recrystallized con- stituents of the schists, just as it has those of the quartzite and dolomite. It is shown 1)>' lines of fracture crossing the slides along which ferric oxide has infiltrated, and by occasional straining and bending of the quartz and mica. G-arnetiferous varieties of the schists are found in close proximity to basic igneous rocks, probably in every instance intrusives, and are evidently the result of contact metamorphism. With the garnets occur actinolite in felted mats and clusters, and abundant magnetite ami pyrite. A colorless amphibole in large single crystals bounded by the prism and clinopinacoid, and giving low extinctions, is often associated with the actinolite. SECTION VII. THE GROVELAND FORMATION". The ferruginous rocks which compose this formation are well exposed in the central portion of sec. 31, T. 42 N., R. 29 W., in the vicinity of the Groveland mine, and thus may properly be termed the Groveland formation. DISTRIBUTION, EXPOSURES, AND TOPOGRAPHY. The magnetite, which is always an abundant constituent of these rocks, has made it possible to trace them for long distances throughout the trough, by means of the disturbances effected in the compass needles. The same disturbances had led to the sinking of a great number of test pits on the part of former explorers for iron ore, and the material thrown out of these has served to check and substantiate the inferences from the magnetic attractions. Finally, in several localities excellent natural exposures of the iron-bearing rocks occur. So, altogether, the available data as to the surface distribution of the Groveland formation are fairly satisfactory. On the west the presence of the Groveland formation through sees. 34, 35, and 36, T. 42 N., R. 30 W., is shown by one principal and other minor lines of attraction, as well as by test pits and outcrops. The principal line of attraction begins in section 34, near the southwest corner, and runs to the northeast, in conformity with the strike of the northern belt of dolomite, finally ending in the northeastern portion of section 36. This line of attrac- tion is very vigorous and strongly marked. Two other lines, parallel with the principal line, but more feeble and much shorter, cross the boundary between sections 35 and 36, and on the northern of these ferruginous rocks 416 THE CRYSTAL FALLS IRON-BE ARING DISTRICT. outcrop in the western part of section 36. Near the center of section 36 another hne, marking the western end of the Groveland syncUne, begins and continues for a mile and a half east to the eastern portion of sec. 31, T. 42 N., R. 29 W. Along the western portion of this line are many test pits, and in section 31 the fine exposures of the Groveland hill. Four hundred paces north of the center of sec. 32, T. 42 N., R. 29 W., another line of attraction begins, and may be followed toward the east without interruption nearly to the east line of section 33 of the same town- ship. Along this line, which is comparatively feeble and crosses wet ground, there are but few test pits. In the eastern part of section 33, beyond the point at which the attractions cease, many pits have been sunk to and into the Mansfield formation, which is there somewhat ferruginous. From this point east for 4 miles the Groveland formation has not been recognized. In the northern part of sees. 32 and 33, T. 42 N., R. 28 W., the fer- ruginous rocks are again well exposed on Felch Mountain for nearly a mile along the strike, and may be identified for half a mile farther by the vigor- ous disturbances produced in the magnetic needles. In the southeastern quarter of section 33 the Groveland formation is again encountered in a small and much-disturbed area, in faulted contact with the Archean. The most conspicuous hills within the Algonkian belt owe their relief to the fact that they are underlain by the Groveland formation, but else- where this formation has left but little impress on the topography, perhaps because the local base-levels are cut nearly to the bottoms of the synclines in which it is preserved. The two lulls referred to — Felch Mountain, in sees. 32 and 33, T. 42 N., R. 28 W., and the Groveland hill, in sec. 31, T. 42 N., R. 29 W. — stand 100 feet or more above the average level of the surrounding Algonkian territory, and in both instances the infolded second- ary synclines are exceptionally deep and broad. The magnetic lines which indicate the other synclines pass through low ground, and the Ijelts of dis- turbance are much narrower than in the cases of the two principal hills. There seems to be, so far as the collected material warrants a judgment, no litholoarical difference between the rocks of the naiTOw and those of the broad and deep synclines, and accordingly the relief of the latter is believed to be caused by their depth below the adjacent base-levels and not by their more resistant character. GROVKLAND FORMATION IN FELCH MOUNTAIN RANGE. 417 PETROGRAPHICAL CHARACTERS. The rocks of tlii' (irntveliind formation have a general family likeness, which makes it very easy to distingnish them in tlie field from all the other members of the Algonkian series. Among- them two main mineralogical kinds may be recognized, the usnal one of which consists of quartz and the anhydrous oxides of iron, while the other, which is mucli rarer, is made up essentially of an iron amphibole, quite similar to the griinerite of the Marquette range, with quartz and the iron oxides as associates. As seen in the field, the rocks of the first kind are generally siliceous, heavy, and dark colored, the weight and color, which has a tinge of blue, being due to the presence of abundant crystalline ii-on oxides. A large part of the silica is easily recognized as crystalline quartz, in some instances, indeed, in the form of detrital grains. The visible iron oxides occur both as little spangles of specular hematite and also in irregular dark-blue masses and single grains, the latter often having the crystalline form of magnetite. Many, if not most, of these last, however, seem to be really martite, as they give a dark purple streak, and in fine powder are not attracted by a hand magnet. In the first kind there is much variety in external appearance, deter- mined by the varial>le proportions in which the chief constituents occur and by the different ways in which these constituents are arranged. Con- siderable areas, for example, consist maiidy of granular quartz merely darkened by the intimately mixed iron oxides, and in these, so far as the eye can judge, the rock is a ferruginous quartzite. Closely connected with such occurrences, or included most iiTegulaidy in them, are others in which the ferruginous constituents are so abundant and the quartz so subordinate that they would pass for lean iron ores. Between such rare extremes we find all intermediate proportions of mixtures of the quartz and the iron oxides. One form of arrangement of the constituent minerals is in narrow parallel bands, in which the quartz and the iron oxides alternately predom- inate. Such alternations are sometimes so frequent and regular as per- fectly to reproduce the lean "flag ores" of the Marquette range.^ Regular banding, however, is not common. Usually the light or dark bands are 'Geol. Survey Michigan, Vol. I, Part I, by T. B. Brooks, pp. 93-94. MON XXXVI 137 418 THE CRYSTAL FALLS IRON-BEARING DISTRICT. suddenly cut oft', as if by faulting, or tajjer to thin edges, or occur in sepa- rated pebble-like forms. Neighboring lenses and fragments of bands are most frequently i-oughly parallel with one another, but often they are jumbled together in the greatest confusion. They no doubt represent an original more continuous banding, which has suffered brecciation. Masses thus shattered are also traversed and cemented by numerous small A-eins filled chiefly with quartz, chalcedony, and specular hematite. The posi- tions in which the separated patches of the Groveland formation now survive, namely, in and near the bottoms of synclines, and therefore at the jioints where sharp turning and crowding together have taken place, suffi- ciently explain the extensive brecciation observed in these brittle beds. Very prevalent in all the varieties of the first kind of rock, in massive, banded, and brecciated alike, is the occurrence of some of the constituents in small roundish spots, which give to the whole formation a very detrital aspect. In the quartzitic phases, as well as in the most ferruginous bands, the eye recognizes, besides the little grains of clear quartz, which seem to be unquestionably detrital, numerous small dots of blue hematite and bright red dots of jasper. These are more abundant in some layers than in others, but seem never to be entirely absent, and are exceedingly characteristic of the formation wherever found. In a few localities the iron constituent is almost entirely in the form of little micaceous scales of specular hematite, which have a parallel arrange- ment. Hematite-schists, however, are not very common. The best exam- ples occur in the northern part of sec. 36, T. 42 N., R. 30 W., along the northern syncline. The second kind of rocks of the formation, the griinerite-schists, have been found in small thickness and in one locality only, namely, in the southern parts of sec. 33, T. 42 N., R. 28 W., where they underlie, in a series of small anticlines and synclines, banded siliceous beds composed of quartz and magnetite or martite. Under the microscope the essential constituents of the first or prevalent kind of rock of the Groveland formation are quartz, magnetite, martite, and hematite. With these, much smaller quantities of clilorite, epidote, and apatite are generally associated as accessories. Of rarer occurrence are calcite and probably siderite, sericite, tremolite, griinerite, pyrite, limonite, chalcedony, rutile, titanite, tourmaline, microcline, and plagioclase. b GKOVELAND FORMATION IN FELCH MOUNTAIN EANGE. 419 Quartz occurs in two ways — first as rouuded detrital particles, and secondly as grains which have crystallized in place. The deti'ital grains, which are easily recognized by their form, size, and freedom from inclusions of the ores, consist of single individuals, often surrounded with rims of later growth. They are also usually larger than the neighboring indigenous grains. While detrital quartz is not abundant and, indeed, is often entirely absent from the thin sections, its occurrence is of interest as conclusively establishing- the sedimentary origin of the iron-bearing formation! The secondary quartz grains are the most abundant constituents of the thin sections, and form the genei'al background for the other minerals. They always inclose separate crystals of the iron oxides, usually in gi'eat abundance, and often also chlorite and little prisms of apatite. These grains usuall}' have the shape of irregular polygons bounded by straight lines, frequentl} with reentrant angles, and adjacent grains completely inter- lock. In size the secondary quartz grains range from about 0.03 to 0.4 mm. in diameter. Grains of approximately the same size occur together in bands or in the rounded areas to be mentioned later. The iron ores include both magnetite, or martite, and crystalline hem- atite, the former being much the more abundant. The magnetite and martite can not be distinguished in thin section, as their color in reflected light and crystalline form are the same. They occur in irregvilar bands composed of aggregates of crystals, the edges of which interlock with the adjoining and inclosed areas of quartz, and show the triangular, rhombic, and square sec- tions of magiietite individuals. Magnetite also occurs in isolated, irregular aggregates interlocking with the secondary quartz grains, and of similar dimensions to these, but is especially abundant as single minute crystals interposed in the grains of secondary quartz, ranging in size from such as are barely recognizable under a No. 9 objective to octahedra 0.03-0.05 mm. in diameter. A single quartz grain ^ mm. in diameter may inclose a hun- dred or more such minute individuals. Hematite is much rarer than mag- netite, and seems to be found only in the secondary quartz grains or in veins. In the former it occurs in separate crystalline plates, of deep red color in transmitted light, under the same conditions as to number and size as the magnetic crystals. Throughout some sections, and in certain bands and rounded areas in other sections, it is more abundant as inclosures than magnetite. Such rounded areas formed of several quartz individuals, each 420 THE CRYSTAL FALLS IRON-BEARING DISTRICT. of which thus holds a great number of hematite plates, appear macroscopic- all)^ as the little jasper dots already described. Chlorite and apatite are also often embedded in the secondary quartz grains, the former in thin plates and the latter in small hexagonal prisms. Epidote is quite common in small irregular areas intercalated between the quartz gi-ains or in the magnetite bands. Many of the slides contain a small amount of rhombohedral carbonate, much if not all of which is calcite. It occurs chiefly in the quarts bands, in irreo-ular grains which interlock with the secondary quartz grains, and, like them, inclose little crystals of magnetite and hematite. Specimens the slides from which contain carbonates effervesce freely in scattered spots with cold dilute acid. Most of the carbonates are clear white under the microscope, and are evidently calcite. Sometimes, however, the carbonate areas have a very light-brown tint, and are partially surrounded with a limonite border and jjenetrated by brownish filaments along the cleavages. In such cases it is difficult to decide whether they are calcite stained with limonite, or siderite partially oxidized to limonite. However, if part of these areas are siderite it is nevertheless certain that the small magnetite and hematite crystals which they inclose have not been derived from them. These little crystals are inclosed in the carbonates just as they are in the adjoining grains of secondary quartz, while the alteration of the siderite, if it is siderite, is to limonite. Carbonates also occur with tremolite, quartz, chalcedony, epidote, and hematite in the numerous thread-like veins which traverse some of the thin sections. The feldspars have l^een found in onl}^ a few thin sections, as well- scattered but minute angular grains of microcline and plagioclase. ]\Iany slides, however, contain areas of matted sericite and quartz wliicli probably represent original grains of feldspar. Rutile and tourmaline are also occasionally inclosed with the iron ores in the grains of secondary quartz. Small roundish areas of titanite, prob- ably detrital, occur very sparingl}^ in a few of tlie thin sections. The most interesting features of the thin sections are certain very distinct structural arrangements of the quartz and iron ores. In almost every slide, in ordinary polarized light (with the analyzer out), the minute interpositions of the iron ores are seen not to he equally distributed through- out the background, Init to be concentrated in round or oval areas, never GROVELAND FORMATION" IN FELCII MOUNTAIN RANGE. 421 excuodiug ;i inilliiiK'tur in (liiunL-ter. Tliese oval forms are coiitiiUMl to the more siliceous bands, and are much more distinct in some of the slides than iu others. ( )ften the outlines are reenforced b}- rims of closel}' set mag- netite individuals, somewhat coarser than the dust-like crystals within. The long' diameters of adjacent ovals are 2)arallel to one another and to the band in which they lie, and ai'e often closely packed like pebbles. ()cca.sionall3' the little grains of iron ore within the ovals have a distinctly concentric arrangement. Between crossed nicols these areas are seen to have had in some instances a distinct influence on the crystallization of the secondary quartz. When they are large and closely packed, each oval includes a large number of interlocking quartz grains, and occasionally in such cases there is some difference in size between the quartz grains inside and those outside the ovals. In the triangular and quadrangular areas lying between the larger ovals, and bounded by curving segments of their perimeters, the secondary quartzes are frequently larger than those within, and are placed normal to the boundaries, precisely as if they had grown outward from the ovals into free spaces. Often, however, a single individual of secondary quartz lies partly within and partly without the oval. On the other hand, when the ovals are small, one or more may be completely or partially inclosed within a single quartz individual. The interlocking quartz grains within the large ovals show no indications of having formed in open spaces, even when the included iron ores have a tendency, as occasionally happens, to a concentric arrangement. The faulting and brecciation so plainly seen in many of the thin sections have also displaced and separated the oval areas. It seems perfectly clear to me that these forms represent a structure originally possessed by the rock from which the various phases of the iron formation have been derived, and which has been preserved through the subsequent metamorphism. From the facts described above, it is evident that the Groveland for- mation is made up of highly metamorphic rocks, which still, however, retain some original clastic material as well as certain original structural characters. With the exception of the rather rare clastic grains of quartz, titanite, feldspar, etc., the minerals which now chiefly compose these rocks — namely, quartz and the crystalline iron oxides — are not cla.stic, but have crystallized in place. It is a matter of great interest, therefore, to determine, if possible. 422 THE CRYSTAL FALLS IRON-BEARING DISTRICT. in what form these constituents were present in the original deposit. (.)n this question the microscopic structure seems to me to have a distinct bearing. Forms similar to the ovals in these rocks occur in the iron-bearing formations of other districts in the Lake Superior region. In the Gogebic district of Michigan and Wisconsin, R, D. Irving and C. R. Van Hise^ have supposed that such forms have resulted from pi-ocesses of solution and redeposition after the rock was formed, and are tlierefore concretionary. They regard that portion of the formation — which they have named fer- ruginous cherts — in which such forms occur, as an alteration product from an original dejiosit of cherty carbonate of iron. On tlie other hand, J. E. Spurr' lias shown that similar forms are exceedingly abundant throughout the iron-bearing formation of the Mesabi range of Minnesota, and are there original. In the least-altered stages Mr. Spurr has found that these oval and roundish areas are filled with a green substance, which chemically is a hydrous silicate of iron, in composition A^ery close to glauoonite, with which it is also optically identical. The oval and rounded forms, moreover, are those characteristic of glauconite in green sands of all geological ages. Starting with this original substance, which is very unstable when exposed to oxidizing and carbonated waters, Mr. Spurr has traced an interesting series of clianges, the final result of which along one line is the complete oxidation of the iron to hematite or magnetite and the separation of the silica as chalcedony and quartz. Tlu'oughout these changes the original form of the glauconite grains is jireserved in the new minerals. Without going into the details of these changes, and without accepting Mr. Spurr's conclusions in their entirety as to the steps involved, he has clearly shown, as I have satisfied myself from the study of the large number of Mesabi slides in my own collection, that the green glauconitic substance is the source of the iron and silica of the ferruginous cherts of the Mesabi range, and that the peculiar spotted structure of these cherts is inherited from the original forms of the glauconite grains. Between the ferruginous quartzites of the Groveland formation and the ferruginous cherts of the Mesabi range there is a very close resemblance, especially in structure. The essential difi"erence is that the former contain ' Loc. eit., pp. 2.54-257. ■ The iiou-ljcaring rocks nf the Mesabi rauge iu Jliuiiesota, by .1. E. Spurr: Bull. Geol. aud Nat. Hist. Surv. of Minn., No. X, 1894, 259 pp., 12 pis. UPPER nURONIAN OF FELCri MOUNTAIK RANGE. 423 little or IK. chalcedony, the silica being crystallized quartz, while the latter have a great deal of chalcedonic silica. Also the former contain small amounts of detrital material, which the latter generally lack, but the essen- tial difference between them is one of degree of crystallization only. If the silica of the Mesabi cherts had originally crystallized entirely as quartz, or if after passing through the stage of mixed chalcedony and quartz it had subsequently crystallized as quartz, there would be no essential difference between the iron formations of the two districts. There are, then, at least two possible forms in which the iron and silica of the Groveland formation may have been deposited originally, as indicated by the conclusions of observers who have studied the similar iron-bearing formations in other districts of the Lake Superior region in which these formations are less altered than here. Either of these forms — namely, a cherty iron carbonate, as on the Gogebic range, or a glauconitic greensand, as on the Mesabi range— could give i-ise, under the action of vigorously oxid- izing waters, to rocks of the mineralogical composition of those in question, and since no trace of either original form has been found in the Groveland formation the choice between them may perhaps be regarded as still open. My own opinion, based on the microscopic structure which, as I interpret it, shows that the Groveland formation was in the beginning largely made up of rounded particles having the same general form as the glauconite grains of the Mesabi range, is that the iron and silica were originally present largely in the form of glauconite. SECTION VIII. THE MICA-SCHISTS AND QUARTZITES OF THE UPPER IIURONIAN SERIES. Through the eastern part of sec. 32, T. 42 N., R. 28 W., and entirely across section 33, runs a belt of mica-schists and thin-bedded ferruginous quartzites which seem to have unconformable relations with the formations just described. These rocks are seen on the west in a cut in the North- western Railway in the SE. i of the NE. ^ of sec. 32. At the western end of this cut the strike is northwest and the dip northeast at an angle of about 35°. At the eastern end there is a decided bending in the strike to a more nearly east-and-west direction, and the bedding surfaces carry striations which dip east at an angle of 10°, all indicating that these outcrops prob- ably lie on the south limb of a gently eastward-pitching synclinal fold, and 424 THE CEYSTAL FALLS IRON-BEAEING DISTEIOT. near the axial plane. East from this point similar schists and quartzites form a ridge, low and flat-topped, which extends immediately south of the railway almost to the east line of section 33, and sinks gradually beneath the great swamp of the eastern portion of that section. The formation notice- ably disturbs the compass needles, and this fact, together with the rusty appearance of the outcrops, has probably led to the sinking of the numer- ous test pits by which the continuity is chiefly established. But low-lying natural exposures are not lacking. North of the center of the NE. ^ of sec. 33 similar schists have been found in two test pits. Also, parallel with the outcropping southern belt and a quarter of a mile or more farther north, a faintly marked zone of magnetic disturbances runs east and west through the swampy ground south of Felch Mountain and probaljly connects the last-mentioned occur- rences with the exposures of the railway cut. It therefore seems likely that the low ground through the middle of sections 32 and 33 is wholly occupied by an open syncline of these soft and easily disintegrating rocks. Between the exposed ^southern limb of this syncline and the southern Archean the lower Algonkian formations are found in the southeastern portion of section 33. Actual contacts are not visible, but there are note- worthy discordances in strike and dip, and especially clear proof of great disturbances in the lower rocks in which the upper have not shared. In the SE. i of the NE. 4 of the SW. i of sec. 33, about 200 feet thickness of the Randville dolomite, striking east and west and dipping north at about 70°, is exposed between the Sturgeon quartzite below and the mica-schists to the north. Between the dolomite and the schists is a covered interval of some 40 feet. The latter also strike about east and west, but dip north at 30° or less. Between a quarter and three-eighths of a mile east of this locality (the interval being without outcrops) the Mansfield and Grroveland formations lie against the Archean gneisses with a faulted contact. They have been thrown into a series of southeastward-pitching minor folds, and have been intruded by a mass of diabase and also by a pegmatite dike. The true strike of these formations at this locality is toward the northeast, and the dip, as shown both by the direction of pitch and the order of sujjer- position, is toward the southeast. Five hundred feet north of this disturbed area and directh' across the strike of the lower formations therein, the UPPER HURONIAX OF FP:LCH MOUNTAIN RANGE. 425 upper schists and (niartzites coiitiiiuc their .southeastward strike without deviation. These general relations indicate tliat the ferruginous mica-schists and quartzites are part of an upper series which overlies unconformably the Groveland and all the lower formations. This series lias not Ijeen found elsewhere in the Felch Mountain area. PETROGRAPHICAL CHARACTERS. The rocks of this formation, as seen in tlie outcrops, are principally soft and deeply iron-stained mica-schists in which occur frequent tliin beds of ferruginous and micaceous qnartzite. Under the microscope tlie schists are composed mainly of biotite, quartz, inuscovite, and magnetite. Chlorite, as an alteration product of the biotite, is frequently abundant, and garnets also occur in .some sections. These schists are much coarser in grain than those of the Mansfield forma- tion, and are wholly crystalline. No clastic material has been recognized in the thin sections. The quartzites also are thoroughly recomposed rocks, without recog- nizable clastic particles. Quartz is the most aljundant con.stituent, and with it muscovite, biotite, and magnetite are constantly associated. The micas and the magnetite are frequently inclosed in a background of large inter- locking quartz grains, which is very similar to the background of the Stur- geon quartzite. Such inclosures lie in general alignment throughout the thin sections, but, unlike many of the inclusions of the Sturgeon quartzite, they seem not to be clastic particles but to have crystallized in place. In one slide among the inclusions in the large quartzes of the background is a colorless isotropic substance, of low refraction, occurring in large polygonal areas, but without definite crystal form. It is usually stained with limonite, which has penetrated from the margins along straight lines, as if following cleavages. This interesting mineral, which is certainly not garnet, and probably not opal, deserves further investigation The rocks of the upper series, like those of the lower series, are greatly altered. From their mineralogical composition and structure it is evident that as originally deposited they consisted of beds of mud separated by thinner beds of sand. But as they now stand they have been as greatly changed from their original condition as tlie bedded rocks below. Also, 426 THE CRYSTAL FALLS IRON-BEARING DISTRICT. since the time of inetamorphi.sm they liave been subjected to stress, as is clearly shown by the optically strained condition of the secondary quartz grains and the bending- and twisting of the micas. From these facts we may reasonabl}^ infer that the general metamor- phism of both series was accomplished after the deposition of the upper series and before the latter was folded. Reconstruction so complete as that shown by the upper series is not believed to take place except at con- siderable depths below the surface, and hence the part of the upper series now visible must then have been deeply covered by overlying rocks, which were afterwards entirely swept away before the deposition of the Cambrian. In the earth movements which folded this mass of material and brought it up within the reach of denuding agents, we may recognize the causes which have strained and broken the secondary minerals of both series alike. SECTIO]y IX. TlIK IXTRITSIVES. The Alffonkian formations of the Felch Mountain area have been cut by later intrusives, among which both acid and basic rocks are represented. The latter have also been recognized in the Archean, in which, indeed, the freshest and least-altei-ed occurrences have been found. The acid rocks consist of fine- to medium-grained pink granites, occurring in narrow dikes. A number of these dikes have been found in the Stvu-geon formation, both in the area of tine exposure on the south side of sec. 3.5, T. 42 N., R. 30 W., and also in sees. 34 and 35, T. 42 N., R. 29 W. Two o-ranite dikes are also known in the highest member of the lower series, but none have been detected in the Randville or Mansfield forma- tions. One of these occurs on Felch Moinitain, the other, a very coarse pegmatite, is found cutting tlie Groveland formation in the southern part of sec. 33, T. 42 N., R. 28 W. Basic dikes and intrusive. sheets are found in many localities. Some are highly schistose and greatly altered, others are massive and but little changed. They probably belong to many eras of eruption. The least altered are diabases, in one occurrence of which, from the Archean, the aus'Ites are almost intact. CHAPTER IV. THE MICHIGAMME MOUNTAIN AND FENCE RIVER AREAS. By reference to the general map, PI. Ill, it will be seen that an oval- shaped Archean area, about 11 miles long from northwest to southeast and liaAing an extreme breadtli of nearly 4 miles, runs through portions of Ts 44, 45, and 46 N., Rs. 31 and 32 W. The country to he described in the present chapter includes that portion of this Archean mass (together with the younger rocks on its eastern border) which lies east of the line between Ranges 31 and 32 W., as well as the territory to the south in the prolongation of the axial line, as far as the south line of T. 43 N., R. 31 W. A gap about 6 miles broad not covered by oui' work intervenes between this arbitrary southern boundary and the western termination of the Felch Mountain work at Randville. In the northern portion of the area now under consideration (which lies along and is twice crossed by the Fence River) the geological structure is exceedingly simple, while in the southern portion, especially in the neigh- borhood of Michigamme Mountain, it is rather complex. The boundary between these two divisions falls in the neighborhood of the mouth of the Fence River in sec. 22, T. 44 N., R. 31 W. It is therefore convenient in what follows to refer to the northern portion as the Fence River area, and to the southern as the Micliigamme Mountain area. By referring to PI. Ill, the broad geological structure of the whole territory of which the above-mentioned Archean oval is the center is evi- dent at a glance. It is an anticlinal dome, the core of which is Archean, around which the younger Algoukian formations run in a series of concen- tric rings, on all sides dipping outward from the inner nucleus. In the Fence River area, on the eastern long side of the dome, the Algonkian formations have a constant eastward dip, and are free from important secondary folds. In the Michigamme Mountain area, how^ever, which lies in the prolongation of the main axis of the dome, these encircling forma- 428 THE CRYSTAL FALLS lEON-BEAEING DISTRICT. tious fall away gently to the south in a series of waves, produced by several concentric minor folds transverse to the main axis. Of these minor folds but one is at all distinct to the east of the general anticlinal axis, while to the west of this axis at least three are well made out within the Michigarame Mountain area. The much greater bi'eadth of the Algonkian formations on the west side of the dome than on the east is probably due to the persistence of these minor folds toward the northwest. The general character and aspect of the formations of the two areas and their succession is iii so many respects identical with the formations of the Felch Mountain range that no doubt can be entertained that they are really the same formations. Nevertheless certain differences mark these rocks with a distinct individuality. These differences will be considered in detail in the descriptions of the several formations. In general they may be summarized as involving a great reduction in thickness of the Sturgeon formation, with a corresponding increase in the Randville dolomite, the appearance of surface igneous rocks at the Mansfield horizon in the Fence River area, and a less uniform and complete metamorphism in the whole Algonkian series. SECTION I. THE ARCHEAN. The rocks of the Archean core are well exposed through the west- central sections of T. 44 N., R. 31 W., while farther north in T. 45 N., R. 31 W., outcrops are few and scattered. Much less attention was paid to this area than to the Felch Mountain Archean; our work, as a rule, stopped with the location of the boundary, and, therefore, the following brief state- ments as to its character embody observations along the southern and east- em margins only. The prevalent rock in the Archean is granite, varying from medium to coarse grain, and often carrying very large porphyritic Carlsbad twins of flesh-colored microcline. Banded gneisses and mica-gneisses and mica- schists, such as are so abundant in the Felch Mountain Archean, are rare but not entirely absent. While in many localities the granites are much crushed and even sheeted along adjacent parallel fractures, their original!}^ massive character is sufficiently evident. They have the composition and structure of typical igneous granites. The primary minerals are entirely without definite arrangement. AUCIIEAN OF MlCeiGAMME MOUNTAIN AREA. 429 111 tlR' Arcliean areas p-raiiites of two af;es Imv'e been found, the younger in tlie form of narrow dikes. Basic igneous rocks, also in dike form, are rather abundant. One of these under the microscope proves to l)e a but little altered diabase, in which the augite is almost intact. These acid and basic intrusions are i)rol)ably connected with tlie surface flows of like character which are so abundant at the ]\Ian.siield horizon along the Fence River. Of mucli interest is the occurrence of a small mass of quartz-porphyry in contact with the Archean, and below the lowest Algonkian sedimentary formation. The locality is in sec. 21, T. 44 N., R. 31 W., in the southeast quadrant of the Archean oval. The upper surface of contact of this sheet with the lowest sediments is covered, and hence it is not entirely certain whether it is intrusive or extrusive, and therefore whether it belong-s to Archean or Algonkian time. The general relations, however, appear to indicate that it is a surface flow wdiich suffered erosion before the deposition of the ]:)asal Algonkian member, and is tlierefore to be classed with the Archean. The exposure is 250 feet long by 100 broad. The rock consists of a very finely granular matrix of a warm gray color, through which are sprinkled quite uniformly little grains of blue quartz, and larger rounded grains of pink feldspar. Flakes of biotite are scattered through the ground- mass and coat the cleavage surfaces, which are developed in two distinct systems, intersecting at an angle of about 10°. Immediately below the porphyry is coarse porphyritic granite, sheeted in waving surfaces parallel to the contact, wdiich dips eastward about 40°. The lower portion of the porphyry contains a number of fragments of the underlying granite, one of which is over 4 feet in leuafth. Under the microscope the groundmass is a line-grained crystalline aggregate of quartz, greenish biotite, and a little feldspar. The quartz phenocrysts are beautifully corroded, and have the characteristic bipyrami- dal iovm, while the feldspars are extensively altered to biotite, sericite, and quartz. Biotite-gneisses related to this porphyry in external appearance occur among the Archean outcrops inclosed in the "B" line of magnetic attraction in sec. 7, T. 45 N., R. 30 W., and may be described here for comparison. They are dark-colored, fine-grained rocks, which weather to light pink. They are eminently schistose, and the cleavage surfaces are coated with 430 THE CRYSTAL FALLS IRON-BE ARIXG DISTRICT. biotite plates of medium size. Minute grains of blue quartz are occasionally distingiiisliable by the eye. Under the microscope these gneisses have a fine to medium grained groundmass composed of quartz, microcline, orthoclase, plagioclase, green biotite, and nuiscovite, and a little scattered epidote. Within it are large roundish areas of quartz and feldspar, sometimes single individuals, but more often consisting of several fragments. The gneissic foliation is pronounced and is caused by a general elongation of the constituent minerals in a common direction. The onl}" essential differences between these gneisses and the porphyries described above are this strong foliation and the coarser groundmass. SECTION II. THE STURGEOK FORMATION. The Sturgeon formation as a distinct member of the Algonkiaii series is hardly known in this area apart from the Randville formation. Neverthe- less, purely clastic sediments unmixed with the carbonates of calcium and magnesium were deposited and are now visible along one section between the Archean granites below and the dolomites above, and for these it is convenient to retain the name, although their total thickness is so small and their continuity so uncertain that they can not be shown on the geo- logical map. The general conditions of sedimentation here were such, perhaps in consequence of the low relief of the neighboring land, that lime- stones began to form a relatively short time after the submergence of the Archean surface, so that the two lower Algonkian formations probably by no means represent equal periods of time with the same formations in the Felch Mountain range. The time represented by both together is perhaps not g-reatly different in the two areas, but since in the entire absence of fossil evidence it is impossible to draw the line of equivalence, while at the same time the lithological break is a shai'p one, it seems desirable to carry over the Felch Mountain names, extending the Randville dolomite downward to the lower limit of limestone deposition, and retaining the name Sturgeon formation for the basal sediments which are free from carbonates. These basal sediments are found only in sec. 15, T. 44 N., R. 31 W., where they are exposed in low-lying outcrops in the banks and bed of the Fence River. Elsewhere throughout the 10 or 12 miles through which the Archean extends in this area no outcrops have been found in the flat and generally swampy belt which intervenes between it and the dolomite above. KANDVILLE DOLOMITE IN FENCE RIVER AREA. 431 The expofsuix's ret'cnxMl to consist of soft, light-weathering slates and gi'aywackes, with which are interbedded layers of coarser texture. They are very evenly banded in pale shades of yellow, red, and green, and the structure thus brought out dips eastward at an angle of 52°. Besides this a secondary cleavage is quite prominent, especially in the finer-grained beds, also dipping eastward, but at a considerably higher angle. At the eastern side the slates are overlain by the lowest marble beds, here extremely impure and highly charged with chlorite and quartz sand. The thickness of slates exposed is about 100 feet, and between the Archean and the most western outcrops there is room for about as much more. The total thickness, then, can not exceed 200 feet. A thin section of a specimen from one of the coarser layers shows it to be a graywacke, the most prominent constituent of which is qi;artz in small roundish and oval grains. These are embedded in a groundmass composed of chlorite in minute irregular plates, ferric oxide, and kaolin. The quartz grains while having generally clastic shapes are bounded by minutely rough edges which interlock with the fibrous minerals of the groundmass. Evidently much new quartz has been deposited round the original grains. SECTION III. THE BANDVILLE DOLOMITE. DISTRIBUTION AND EXPOSURES. In the Fence River area the dolomite, as already stated, lies on the east side of the Archean, and occupies a belt over half a mile in width, which extends from the mouth of the Fence River on the south for about 10 miles to the north and west, to our western boundary near the north- west corner of T. 45 N., R. 31 W. lu this distance it is twice crossed by the river, and on these natural sections and in their neighborhood the only known outcrops of the dolomite have been found. The northern river sec- tion passes through sees. 22 and 28, T. 45 N., R. 31 W., and discloses an excellent series of closely connected exposures for a distance of about 2,900 feet, measured at right angles to the strike. The southern section is 5 miles farther south, and is much less continuous, laying bare the extreme upper and lower portions only of the formation. Elsewhere through the dolomite belt the rock surface is concealed by swamps or glacial drift^ to which last it contributes but few scattered bowlders of noticeable size. 432 THE CEYSTAL FALLS IRON-BEARING DISTRICT. South of the Archeaii dome in the Michigamme Mountain area the dolomite tops the hiw arch in a broad crumpled sheet, in the minor syn- clines of which the higher formations are more and more implicated as we go south. This broad sheet, with its included tongues of phyllite, extends to the south line of T. 44 N., R. 31 W., beyond which it disappears beneath the hig-her formations, except in a single narrow belt which continues along the main axis for about a mile farther south. Exposures sufficient in num- ber to indicate several minor folds are found along the Michigamme River and scattered through sees. 28, 32, and 33, T. 44 N., R. 31 W., and sec. 4, T. 43 N., R. 31 W. FOLDING AND THICKNESS. In attitude the Randville formation in the Fence River division of the district is an eastward-dipping monocline, the inclination of which is gen- erally moderate. The rocks are usually heavily bedded and nearly always show distinct alternations in coarseness and color, so that structural obser- vations are made with much more cei'tainty than in the Felch Mountain range. The more conspicuous minerals secondarily developed here — coarse carbonates and tremolite — have formed chiefly in the old planes of bedding. Oblique structures are generally absent except in the close vicinity of the basic dikes which intersect the formation along the upper river section. The surfaces of contact with the dikes stand at high angles, and nearly parallel to these the neighboring dolomite has well-developed cleavages, along which new minerals have formed, intersecting the true bedding. It is evident that the stronger igneous rocks in these cases have lurnished resistant surfaces against which the dolomite has been kneaded in the general tilting of the series. The eastward-dipping monocline is a simple one, 5^et the observed angles of inclination are by no means uniform. Thus, along the upper river section the dip ranges from 25° to 60°, with 40° as the mean of about a dozen observations. The variable dips are so scattered through the cross section as to indicate no widespread roll in the formation as a whole, but rather a great number of minor undulation's pi'obably distributed through- out its thickness. Such undulations are visible in favorable localities, as, for example, on the north bank of the river in the NW. ^ of NW. 4» sec. 28, T. 45 N., R. 31 W., where fresh surfaces have been exposed in blasting FOLDING AND THICKNESS OF KANDVILLE DOLOMITE. 433 for tlie dam. The light-blue and pearly-white layers of tlie beautiful mar- ble here seen are thrown into a series of unsymmetrical folds. The western sides of the little anticlinals are short and overturned, while the eastern sides are long and gently inclined. Evidently, if the same system of sec- ondary folding holds throughout the entire thickness of the formation, sur- face observations would show everywhere eastward dips at A'ariable angles, dependent upon the portion of the fold which hapi^ened to constitute the particular outcrop, and gentle dips would be more abundant than steep dips. This would completely explain the observed variations. Similar variations and lack of regular sequence in the dips are found in the southern river section. Five good observations range between 20° and 58°, all eastward, but none of the exposures is sufficiently extensive to show minor folds. The mean of these observations is about 40°. The surface width of the dolomite zone on each section is a little less than 3,000 feet, assuming that a fair proportion of the covered zones on each side is underlain by the same formation. If the average observed dip is taken to represent the average dip of the rock, the thickness in each case would be a little over 1,900 feet. This is probably too great, and is certainly too great if the same kind of internal crumpling visible in parts of the upper river section is characteristic of the formation throughout. The average dip evidently would more nearly be represented by the dips of the long eastern limbs of the little anticlines. Assuming that these are less than the mean, we find the average of the dips below 40° to be 30° for each section. This gives a thickness of about 1,500 feet, which still is per- haps beyond the truth, but is probaljly much nearer it than the first value. It is interesting to compare this result with the thickness of 500-1,000 feet obtained on the two Felch Mountain sections. A part of the increase is probably due, as already explained, to the earlier beginning of limestone deposition in the Michigamme area. But an important part of it is pi'ob- ably not depositional at all, but is the result of plications. The whole series here is but gently tilted as compared with the walls of the Felch Mountain trough, and hence the strong horizontal pressures have acted in a direction but slightly inclined to the bedding. The result has been the secondary crumpling within the formation which must contribute in an important degree to its present apparent thickness. In the . scattered outcrops of the Michigamme Mountain area the MON XXXVI 28 434 THE CRYSTAL FALLS IRON-BEARING DISTRICT. dolomite strikes and dips toward all points of the compass. This irregularity is cansed by the gentle arching over the general northwest-southeast axis, combined with much sharper local folding about a series of axes which run more nearly east and west. The l)est-defined east-aud-west folds occur west of the main axis in sec. 32, T. 44 N., R. 31 W., in which three syn- clines and three anticlines are found along a north-and-south section 4,000 feet long. Tlie two southern syuelines are sufficiently deep to include the overlying Mansfield phyllites. The secondary folds die out toward the main nortli-and-south axis and bi'oadeu toward the west. East of the main axis Ijut one secondary* fold has been recognized, namely, the syncline which forms Michigamme Mountain. This is tlie deepest of the secondary folds, and the only one containing the Groveland formation. PETROGRAPHICAL CHARACTERS. The Randville formation in this area is richer in litliological varieties than in the Felch JMountain range. As originally deposited, a much larger proportion of sand and mud was mingled with the carlionates, and tlie prog- ress of subsequent metamorphism also has been less uniform. Depending upon the interaction of these two factors, we find, as the extremes of variation, on the one liand coarse saccharoidal marbles, sometimes very pure, but most often filled with secondary- silicates, and on the other hand fine-grained little-altered limestones, which occasionally are ho impure as to be rather calcareous or dolomitic sandstones and shales. The more impure varieties occur, as might be expected, near the contacts with the adjacent formations. On the Fence River, in sec. 16, T. 44 N., R. 31 W., the base of the dolomite rests on the Sturgeon formation. The rock is filled with grains of quartz and feldspar and scales of chlorite, and is so soft that it ma}^ be crushed between the fingers. In sec. 32, T. 44 N., R. 31 W., the toj) of the formation is in contact with the Mansfield slates, and between them is a com- plete sei'ies of transition beds. Near the junction the limestone becomes dark colored and contains thin bands in which the clayej' material greatly exceeds the carbonates. These are succeeded by alternating beds of slate and impure limestone in nearly equal volume, and it is only high up in the slate member that the calcareous bands completely disappear. Apart from these belts of extreme impurity at the base and top of the formation, the presence of scat- tered fragmeutal grains of quartz and feldspar is j-ather general tln-oughout. The prevalent colors are white, various shades of pink, both light and PBTKOGKAPUIOAL CUAUACTKKS OF KANDVILLE D(JLOMITE. 435 (U't-p l)hu', iuifl pale green. Where weathered, the u.sual colons are light brown or buff. The lighter-colored rocks in general are chai-acteristic of the Fence River area where metaniorphisni is more uniform and more intense, and the (hirker colors of tlie Michigamme Mountain area to which the less crystalline forms are wholly contined. Bands differently colored are nearly always present in the same outcrop. In the Michigamme jMountain area the torsional strains attendant upon the formation of folds in two directions have developed two systems of frac- ture in the dolomite. In these secondary quartz has formed, occasionally in large amount. Of nuich interest is the occurrence in close connection witli such vein quartz of occasional thin bauds of pegmatite, doubtless aris- ing from the action of deeply derived ' waters. In similar spaces coarse secondary carbonates, tremolite, and oxides of iron also have commonly formed. Over the small anticlinal axes and domes of this area the orieinal bands of the rock have often been shattered, and are now recognizable only in displaced fragments cemented together b}' the new minerals. In the Fence Ri^'er area the general secondary folding has been attended Avith differential movements along the bedding, which left narrow open spaces where the adjacent surfaces failed to fit in their final position of rest. These sjiaces are now indicated by coarsely crystalline carbonates and silicates arranged nor- mal to the original walls. Where the space was a wide one the outer walls are usually lined with coarse calcite, while the interior is filled with quartz. In the Michigamme Mountain area certain pink bands of the dolomite have a beautiful oolitic texture, which is most clearly brougjit out in Aveath- ering by the geometrical regularity of distribution of the harder shells or cores of the little rounded grains. The forms are not different from and are quite as distinct as those in the oolitic limestones of recent deposition. The chemical composition of the dolomites is illu.strated by the follow- ing partial analyses by Mr. R. J. Forsythe, of Harvard University: ' Analyses of dolomites frovi Michigamme Mountain area. I. II. III. 14.25 11.15 47.18 18. 48 9.34 . 12. 57 45.98 •19.22 Al2(Fe..)03 5.38 36.60 16.38 CaCOj MgCOs 436 THE CRYSTAL FALLS IRON-BEARING DISTRICT. The ratio of CaCOg: MgCOg is too great for normal dolomite, but approximates that for 2CaC03+MgC03. Under the microscope the chief differences in the various thin sections are in the degree of metamorphism and in the quantity and character of the foreign fragments. The least altered varieties are those highest in the series from the Michigamme Mountain area. These consist of a background of extremely fine-grained calcite, with a few rounded fragmental quartz grains, and scattered particles of chalcedony. Mixtures of small quartz particles, chalcedony, and calcite slightly coarser than the background occur in short vein-like gashes. The prevalent deep color of these rocks is due to the even sprinkling through the background of a black opaque pigment, which may be carbonaceous. Altogether the microscopic characters are those of a little-altered, slightly cherty limestone. The more crystalline varieties of the dolomite contain several secondary minerals, namely, tremolite, diopside, chlorite, muscovite, phlogopite, pyrite, and the oxides of iron. Of these, tremolite is very common and abundant, especially in the Fence River area, where the rarer pyroxene, diopside, also is found. Phlogopite comes in but two of the thin sections, while muscovite occurs in nearly all. The general habit of these siHcates is precisely the same as in the dolomite of the Felcli ]\Iountain range. They are developed pari passii with the passage of the unaltered dolomite into marble. The fragmental inclusions within the dolomite are of interest. These are little pebbles of quartz, feldspar, mica, titanite, magnetite, and augite; and are evidently derived mainly from preexisting granites or gneisses. Titanite and augite are very i-are; the others are represented in almost every slide. The quartz grains are seldom more than a millimeter in diam- eter and commonly are much smaller. While the general shape is oval or rounded in most cases, the perimeters are usually extremely irregular and interlock with the carbonate grains of the Ijackground, which indicates that they have been enlarged since deposition by the formation of new silica. This is very evident in the few instances in which the original smooth out- line, or part of it, is preserved by a film of difi"erent material inside the present perimeter. The feldspar pebbles include orthoclase, microcline, and plagioclase, nncrocline being the common species. They are usually much decomposed and iron stained. The feldspars are especially abundant in the slides from the Fence River area. PETIUKJRAI'HICAL GHAKAGTKRS OF UANDVILLE DOLOMITE. 437 Tho clastic |)cl)l)les give us striking' proof of the general and severe internal strains suffered by the dolomite, the etfects of which have healed over without a scar in the carbonate matrix. The pebbles are always Optically strained. Very often they are fractured and the parts separated, and sometimes they have been reduced to small fragments. In tliese cases the breaks have been completely healed by the flow or redeposition of the Si'roundmass in the interstices. These effects are foitnd in g-reater or less degree in every thin section. The oolitic varieties are very interesting under the microscope. They consist of little oval or round areas, averaging 2 mm. in diameter, packed together as closely as possjble. Each oval consists of a single or compound nucleus, surrounded by several thin and very even concentric layers. The nucleus in a few cases is a single roundish quartz individual, evidently a clastic grain. In most cases, however, it is composed of a great number of minute quartz grains, or of several coarse calcite grains, with films of iron oxide between. The arrangement of these separate quartz and calcite indi- viduals is such as to indicate that they have filled interior cavities. The surrounding thin layers are calcite in all cases. Sometimes two adjoining nuclei, each within its own rim of several layers, are together included within a common series of shells. In one such case the outside rim trav- ersed the edges of the rings surrounding one of the nuclei with a decided unconformity, as if the latter had been eroded before the deposition of the former. The oolitic structure, I believe, has not hitherto been noted in limestones of undoubted pre-Cambrian age. SECTION IV. THE MAIVSFIELD FORMATION. The typical locality of the Man.sfield formation is the Michigamme River valley in the vicinity of the Mansfield mine, which lies a mile west of the border of my field of work, and is described by Mr. Clements. The same formation, however, is present in the Michigamme Mountain area, where its relations to the adjacent formations are clearly defined. In the Fence River area rocks of very different character and derivation occur at the Mansfield horizon. These occur in typical development to the west, on the Hemlock River, and are hence called the Hemlock formation. 438 THE CRYSTAL FALLS IKON-BEARING DISTRICT. DISTRIBUTION, EXPOSURES, AND TOPOGRAPHY. The Mansfield rocks of the Michigamme Mountain area consist of phylHtes or mica-slates of various colors. They are found in the series of east-west synclines, which have already been described in connection with the Randville formation. The best exposures occur in sec. 32, T. 44 N., R. 31 W., between the center and the west quarter post, and still farther north along- the south bank of the ^lichigamme in the northwest quarter of the section. The}" are also found round the western edge of the Michi- gamme Mountain syncline in sec. 33, T. 44 N., R. 31 W., and in sec. 5, T. 43 N., R. 31 W., but here the exposvu-es are mainly in test pits. Test pits have likewise penetrated them in sec. 10, T. 43 N., R. 31, where they succeed the dolomites as the surface rock over the general arch. Their extent iu the covered portions of this area is probably considerable, but the structure is so complex and the outcrops so few as to forbid any but the most approximate outlining of their general boundaries. The geological position of the Mansfield rocks is free from doubt. In the princii)al syncline of section 32 they are seen to overlie the dolomites and to pass downward into them by a relatively slow gradation, while on the Ijorders of the Michigamme ^Mountain syncline they are jJi'oved to underlie the Groveland formation. The passage to the higher formation likewise is graded, though more rapidly, and is marked in certain bands by an increase in clastic quartz grains and by changes in the character of the matrix in which these are set. The portions of the surface underlain by the Mansfield formation are without special features, and are indistinguishable topographically in the gently rolling plain, the greater portion of which is formed in the dolomites. In section 32 the outcrops are miniature ridges elongated with the strike, the height of which, however, is less than the contour interval of the map. FOLDING AND THICKNESS. The folding of the Mansfield rocks, so far as it can be determined in this area, has already been described in the account of the preceding formation, which they overlie. The rocks are known only in the sec- ondary synclines winch lie transverse to the general direction of the main axis south of the Michigamme River. In the southern of these syn- clines, in sec. 32, T. 44 N., 11. 31 W., between the limestone rims on the FOLDING AND T[IICKNESS ()1< MANSFIELD FORMATION. 439 north ;int liberal estimate for ])ossible minor crumples, it is improbable that a less thickness than 300 to 400 feet occurs here. On the eastern side of the main axis the phyllites below the Groveland formation are very mvicli thinner than this, the thick- ness at the Interrange exploration, for example, being only about 100 feet; but there, as well as along the whole western edge of the Michigamme Mouulaiu syncline, the lower contact with the dolomite is ])robably faulted. It seems entirely safe, therefore, to place the average thickness of the Mansfield formation in the Michigamme Mountain area ;it not less than 400 feet. PETROGRAPHICAL CHARACTERS. The ]\Iansfiield formation consists almost entirely of ^'ery tine grained mica-slates or phyllites. The prevailing colors are dark green, black, and light olive-green. These are often mottled irregularly with red, due to the infiltration of iron oxides along the secondary cleavages. The cleavage surfaces have a dull luster, caused by the ])arallelism of the micaceous minerals, which are too minute, however, to be distinguished by the eye or lens. The phyllites are often finely banded in different colors and shades. Near the base of the formation bands of limestone and near the toj) thin bands of graywacke are interbedded, as has already been stated. Quartz and calcite lenses are not unusual in the minutely puckered portions of the formation. The secondary cleavage is the jjrominent structure, and, indeed, the only structure of the outcrops where the color and texture bandings do not appear. Its general direction is transverse to the main arch, or nearly east and west, and its dip is almost vertical. The north-south compression thus appears to have been the stronger, or to have been active somewhat later in point of time than the east-west compression. Under the microscope the phyllites are seen to be composed principally of fine leaves of muscovite and chlorite, often also with a little biotite, and with a variable and usually small amount of quartz, feldspar, and sometimes 440 THE CRYSTAL FALLS lEON-BEAEING DISTRICT. calcite. Magnetite, ilmeuite, and limonite are usually rather abundant. Pyrite also occurs in a few grains in nearly every slide. The differences in color depend mainly upon the relative proportions of the chlorite and nuis- covite, the former being characteristic of the dai'k-colored, the latter of the light-colored, rocks. The very dark-green or black varieties contain also an opaque and jn'obabl}' organic pigment in very minute particles. The quartz and feldspar grains are usually verj- small and irregularly shaped. The larger, however, of which a few occur in the slides froin the less com- pressed rocks, have well-rounded contours. In other cases extremely flattened and strung-out lenses composed of many small particles represent what were doubtless <:)riginally single clastic grains. Two varieties of cleavage are well illustrated in the thin sections, namelv, that caused by the parallelism of the component minerals, and " ausweichungs-clivage." The former is characteristic of the coarser- grained varieties, and the latter of the finer grained, where the direction of pressure has made a large angle with the bedding. In some cases the little leaves of muscovite outline parallel and equal folds, less than 0.2 mm. from crest to crest, each of which is ruptured, sometimes with slight displace- ments, sometimes with none, entirely across the slide. The structure is most distinct in the red phyllites, in which the fractures and the arrangement of the muscovite plates afe clearly outlined by the ferruginous stain. Each kind of cleavage in a different way tells the storj- of extreme pressure. SECTION V. THE HEMLOCK FOR3IATION. The Mansfield formation of the Michigamme Mountain area changes along the strike into rocks of an entirely different character, which, as already said, have been named the Hemlock formation. DISTRIBUTION, EXPOSURES, AND TOPOGRAPHY. * The Hemlock formation in the Fence River area consists of several varieties of schists which occupy a belt between 2,000 and 3,000 feet in width between the dolomite on the west and the Groveland formation on the east. The best exposures occur on the two river sections already referred to (p. 431), but outcrops are by no means lacking elsewhere.' At the northwest corner of the area, in sec' 6, T. 45 N., R. 31 W., the schists are found striking N. 60°-70° W., and dipping northeast about 40°. East THE UEMLOOK FORMATION. 441 of the center of sec. Hi, '1\ 46 N., li. 31, a few exposures occur, tlie sti'uc- ture of which strikes a few degrees west of north and dips eastward at angles of 45° to 50°. In sections 21 and 28, a mih- and a lialf south, numer- ous outcrops in siniihn- attitudes are found ah)ng the northern river .section. In sees. 3 and 4, T. 44 N., H. 31 W., a few scattered outcrops only have been found, but throufiliout section 10 they are very abundant. For the next 2 miles south tln-ough sees. 15 and 22, T. 44 N., K. 31 W., onlv a few small exposures have been discovered which have the same northerlj' strike and eastward dip. Thus for a distance of 1 1 miles along the strike exposures occur at comparatively short intervals, the longest gap being 3 miles. In general this belt is one of slight elevation above both the dolomite country on the west and the iron formation country on the east. The areas of best exposure are characterized by very rough topographical details, which are entirely lost in the generalized curves of the map. Abrupt strike ridges, separated by narrow ravines, succeed one another at short intervals. In the covered areas the surface, while retaining its general elevation, has been leveled off by the deposition of till in the hollows, and has the smoothly undulating contours characteristic of till-covered areas. FOLDING AND THICKNESS. No secondary folds have been detected within the Fence River area of the Hemlock formation, and on account of the metamorphism and cleavage structural observations are not possible from wdiich they might be inferred. The only clear evidence as to the attitude of the rocks was afforded by the contact at one locality between beds of amygdaloid and agglomerate. There the dip is eastward at an angle of 50°. The surface width of the formation varies between 2,000 and 3,000 feet. If 50° is taken as the dip, the thickness would be from 1,500 to 2,300 feet. If the average dip is assumed to be 40°, or the average of the observed dips of the underl^-ing dolomite, the thickness would be from 1,300 to 1,900 feet. Or if 30° be taken, the average of the lower dips of the dolomite, the thickness would be 1,000 to 1,500 feet. We may say, therefore, that the thickness north of the southern river section is probably not less than 1,000 nor jn-obably more than 2,300 feet. South of the southern river section the thickness diminishes rapidly. 442 THE CRYSTAL FALLS IRON-BEARII^G DISTRICT. PETROGRAPHICAL CHARACTERS. The exposures tlu-ougli section 10 and the northern ])art of sec. 15, T. 44 N., R. 31 W. — the southern river section — give us a nearly complete seqv;ence across the Hemlock formation, the principal gaps being on the extreme east and west, thus leaving the details of the relations with the dolomites below and the iron formation above undisclosed. In this section of 3,000 feet in length, the rocks are chiefly chloritic and epidotic schists, with which are associated schists bearing biotite, ilmenite, ottrelite, and araphibole, greenstone conglomerates or agglomerates, and amygdaloids. These rocks ai-e characterized by a generally fine and even grain, by a lack of sedimentary characters, and by a double structure. In most of the varieties minerals, which have formed quite independently of and later than these structures, are macroscopically conspicuous. The prevailing color is green, passing to dark purple and black in the varieties in which biotite, hornblende, and magnetite abound. The distinction made in the field between the several varieties of the schists is a rough one, indicating the predominating minerals rather than iniplving the absence of the others. In fact all the varieties are intimaxely related. The chloi-ite-schists are very fine-grained green rocks, usually from their color evidently very epidotic; they weather to greenish or pink- ish white. The cleavage surfaces are often plentifully sprinkled with little flakes of Ijiotite. Frequently also black needles of ihnenite, brilliant plates of ottrelite, and large clusters of actinolite run irregularly through them, quite independent of the cleavages. The biotite-schists are nuicli darker, and lack the green coloring Through them also the same metamorphic minerals are frequentl}' interlaced. By an increase in these minerals the passage to the other varieties in limited exposures is a very eas)^ one. Greenstone-conglomerates and amygdaloidal rocks occur in a few exposures. In the former, light green or gray aphanitic inclusions, of angular shapes, ranging from an inch to 2 or 3 feet in long diameter, are inclosed in a matrix of chlorite-schist or biotite-schist. The chlorite-schists often hold round or lens-shaped eyes of epidote, and epidote and quartz. That these are filled cavities can in most cases be shown only by the micro- scope, vet some of the larger amygdules have a banded structure evident to the naked eye. These rocks are of structural interest since they are the PETROGRAPHICAL CHARACTERS OF OEMLOC^K FORMATION. 443 only Hiciiihers of rlic urea wliicli ])ossess undoubted bedding-. The i)lane of contact, between an aniyg-daloid and a layer of greenstone-conglomerate in SE. \ sec, 10, T. 44 N., R. 31 W., dips eastwanl at an angle of SO^. Two well-marked systems of cleavage traverse all the rocks of the southern river se(;tion. The angle betv/een their strikes is always acute toward the north, varying from F)° to as high as 34° in different exposures, while the direction of the bisectrix is almost constant at N. 8°-10'^ W. The dip of both systems is toward the east at about the same angle, namely 50° to 60°. The two systems are usually both well developed, so that the outcrop edges break down by weathering along zigzag lines. The character of the cleavages varies from fine partings which divide the surface into rhombs, sometimes extremeh' i-egular in the more aphanitic rocks to a single perfect scliistosity capable of minute subdivision, along which the com])f)nent minerals are visibly aligned, in the more crystalline. Along the cleavages seams of quartz and calcite have frequently formed. Along the upper river section the rocks of the area are distinctly more crystalline, and are chiefly biotite-schists and biotite-hornblende-schists, the latter often veiy coarse. They are sometimes banded, but very irregularly, the lenticular character of tlie banding- sug-o-estinp- the rhondjic cleavag-es of the southern section. Some of the finer-grained biotite-schists contain round or elongated areas of quartz and epidote, which resemble amygdules. With these are associated considerable thicknesses of sericite-schists, full of little eyes of blue quartz; these are evidently metamorphic acid eruptives. The width of the northern section is about 2,000 feet. Under the microscope the Hemlock schists of the Fence River area have a general porphyritic habit. Two main divisions only are clearly distinguished. One of these is the fine-grained mica (sericite) schists, which are characterized Ijy the presence of muscovite as well as biotite in the microcrystalline groundmass, and true phenocrysts of feldspar and bipyramidal quartz, while the other embraces all the other varieties, which, diverse as they undoubtedly are, have yet certain important characters in common and are connectf^d by gradations. The sericite-schists are obviously metamorphosed acid lavas, and need not be described in detail here. The origin of the second division, however, is far more obscure. The least altered of these rocks possess an exceedingly fine grained microcrys- talline groundmass, made up of very pale chlorite and a colorless aggregate 444 THE CRYSTAL FALLS IRON-BEARIXG DISTRICT. with feeble double refraction, which seems to be quartz. Between crossed nicols the groundniass is almost isotropic, and it is by no means improbable that certain reddish patches here and there may really be glass. Little crys- tals of magnetite are abundantly scattered through the groundmass, and are often arranged in parallel curving lines, very suggestive of the flowage lines brought out on the surface of weathered rhyolites by the ferruginous stains. In man}- sections the groundmass includes minute lath-shaped plagioclase feldspars, much altered and with indistinct boundaries, which are often arranged in pa,i-allel lines. The groundmass also is generally sprinkled with little irregular grains of epidote and calcite. In this groundmass are included in variable combinations and propor- tions much larger crystals and grains of common hornblende, actinolite, biotite, ottrelite, calcite, ilmenite, epidote, and zoisite. Of these biotite, calcite, ilmenite, epidote, and zoisite are the most constant and abundant. Biotite is present in all or nearly all of the thin sections. It is always brown, and is characteristically developed in stubby individuals, very thick for their basal dimensions. These individuals are large and lie scattered through the slides. They frequently inclose portions of the groundmass. The mica cleavage most frequently stands across the cleavage of the rock. In many of the darker-colored schists, however, biotite plates intermediate in size between the large porphyritic individuals and the small chlorite plates of groundmass are present in large numbers, constituting a sort of secondary groundmass. These are generally aligned with the cleavage of the rock and are sometimes gathered in bands, but in color and stubby habit are similar to the phenocrysts. Ilmenite in brownish-black prismatic sections is a common constituent. It usually lies at random through the slide. It incloses the quartz and epidote grains of the groundmass. Epidote and zoisite are exceedingly abundant, often in well-formed crystals. Manj^ of the epidote and zoisite individuals contain darker colored inner nuclei, the nature of which is uncertain. In some cases the nuclei are irregular in shape and have the characteristic pleochroism of epidote, but are more strongly colored than the surrounding zones. In other cases they have sharp crystal boundaries, isomorphous with e])idote, are brown in color, and inclose grains of magnetite; these may be allanite. The nuclei are too small, however, for determination. Generally tliey do not extinguish exactly witli PETKOGRAPIIICAL CHAKACTERS OF HEMLOCK FORMATION. 445 the .suiToiindinii- zoiu's. It is prohahlo that luauy of these nuclei represent an early g'eneration of e[)idote, like the small irregular grains of the ground- mass, which were subsequently enlarged to porphyritic size. Inclosui'es of zoisite are not uncoiumon in tlie large epidote individuals. Large lenticular aggregates of epidote with calcite, chlorite, and biotite are found partially replacing feldspar individuals, which were no doubt original phenocrysts. Similar aggregates unmixed with the remains of feldspar are not infrequent, and may reasonably be attributed to the same source. Epidote with quartz is also the common tilling of the amygdaloidal cavities. Common hornblende, actinolite, and ottrelite are very common as porphyritic constituents of the schists. Hornblende occurs in very large well-formed single crystals and clusters placed at random through the gToundmass. It is characteristically associated with ottrelite and biotite, and often has formed somewhat later than the latter. It is always crowded with inclusions, which in the laminated varieties carry the structure through without reference to the position of the host. Ottrelite is abundant in some of the sections, and is distinguished by its characteristic pleochroism. It occurs in large individuals and multiple twins, and like the large horn- blendes and biotites is full of inclusions. The general characteristics of these schists then are, lir.st, a groundmass composed of chlorite, quartz, magnetite, epidote, and in some cases contain- ing plagioclase microlites, and secondly the presence in this groundmass of much larger porphyritic individuals of several secondary minerals. The varieties are determined by the varying ratio of the porphyritic constituents to the groundmass, by the nature of the predominant secondary minerals, and also by the diiferences in grain of the groundmass. This, while gen- erally extremely fine grained is much coarser, Ijut without minei'alogical change, on the northern river section where the schists are more distinctly crystalline. The cleavage of the schists is determined by the aiTangement of the minute particles of the groundmass, and not by the parallelism of the large secondary minerals. These last, further, are never faulted or broken, and in general are unstrained optically. They must have formed then after the compression and tilting of the series. The origin of these schists, I think, is not doubtful. As important points of evidence we have, first, the absence of rocks possessing any sedi- mentary characters throughout the whole section. Next we have the 446 THE CRYSTAL FALLS IRON-BEAKmU DISTRIC3T, undoubted presence of lavas in the series, shown ))y the sericite-schists, amyg'daloids, and greenstone conglomerates or agglomerates. Furthermore, the minerals which compose the schists are those which would result from the alteration, in connection with dynamic metamorphism, of igneous rocks of basic or intermediate chemical composition. Finally, the grain and character of the groundmass, and in some slides the presence therein of plagioclase microlites disposed in flow lines, point directly to an igneous orio-in and to consolidation at the surface. Mr. Clements has reached similar conclusions for the formation above the Randville dolomite' on the western side of tlie Archean dome. There metamorphism seems to have progressed less far than in the Fence River area, and among the more basic rocks he has recognized andesites and basalts. I conceive, then, that the Hemlock rocks of the Fence River area are a series of old lava flows, varying in composition from acid to basic, which first underwent dynaiBic disturbance, which developed the secondary cleav- ages, and afterwards, in a state of rest, the porphyritic minerals were formed. It is an interesting fact, for which I can suggest no explanation, that metamorphism is further advanced in tlie northern part of the area than in the southern, and the schists more distinctly crystalline. This is also true of the underlying dolomite. SECTION VI. THE GROVELAND FORMATIOIf.^ DISTRIBUTION, EXPOSURES, AND TOPOGRAPHY. The Groveland formation in this area, as in the Felch Mountain range, consists mainly of siliceous iron-bearing rocks, which hold much fragmental material, together with certain subordinate schists. While it is of wide extent throughout the area, its known outcrops are limited to three local- ities, namely: The vicinity of Michigamme Mountain, in sees. 33, T. 44 N., ' Volcanics of the Michigamme district of Michigan, by J. Morgan Clements: .lour. Geol., Vol. Ill, 1895, No. 7, p. 801. -This formation was originally named by me the Michigamme Jasper (Am. Jour. Sci., March, 1894). The name Michigamme was subsequently used for one of the Upper Marquette formations, in the Preliminary Report on the Marquette District, l.'ith Ann. Rept. U. S. Geol. Survey. I now aban- don the old name, although it is entitled to stand by the rules of priority, in order to avoid the con- fusion which would necessarily arise from its retention. THE GitOVELAND EOKMATION, 447 Iv. 31 W., and 3, '1\ 43 X., \\. 31 W.; the exposures and test pits at the Shohleis exploration in sec. 21, T. 45 N., R 31 W., and the test ])its at tlie Doane exploration in sec. K;, T. 45 N., K. 31 W. The last two localities are 1 mile apart, and tlie inoiv soutliern is 8 miles north of Michiffamme Mountain. In spite of the poverty of the formation in, outcrops, its distribution throug-hout the area has been well determined throug-h its magnetic jjroper- ties (following the methods described in Chapter II). Adjacent to the Fence River area of the Hemlock formation it gives rise to a strong mag- netic line which passes through the outcrops and test pits of the Sholdeis and Doane explorations. To the north this line was followed to the south- ern side of sec. 32, T. 46 N., R. 31 W., where it is said to connect with a magnetic line followed by Mr. Clements from the western to the northern side of the Archean dome. To the south it continues into the Micliii>-amme Mountain area to within a mile of the outcrops of Michigamme Moun- tain. There the magnetic line gives way to a broad zone of disturbances, feeble and difficult to interpret, but consequent I believe mainly upon the flattening of the foi'mation as it begins to pass over the general northwest- southeast anticlinal axis. This zone connects directly with the exposures of Michigamme Mountain which produce similar irregular distui-bances of the needles and which visibly constitute a thin crumpled sheet, on the whole but gently inclined. For the stretch of 13 miles just described the Groveland formation occupies a continuous belt on the east side of the main anticlinal axis. In the Fence River area it lies east of and upon the Hemlock formation, while in the Michigannne Mountain area it holds the same relations to tlie Mans- field formation. The eastern belt was not traced farther than a mile soixtheast of Michi- gamme Mountain. . In the central and southeastern portions of T. 43 N., R. 31 W., however', in the direct prolongation of the anticlinal axis, we found a broad belt of slight magnetic disturbance, along the western mar- gin of which lie 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 indefinitely, while the other maintains a general southerly course to the south line of the township. In section 26 large 448 THE CRYSTAL FALLS IRON-BEARING DISTRICT. angular bowlders evideiith' derived from tlie Grovelaiid formation are found in the zone of magnetic disturl^ance, but no outcrops have been discovered. There can be little doubt that these disturbances roughl}' outline the position of the Groveland formation in the axial region. Except in ^[ichigamme Mountain, the most elevated point of the dis- trict, the Groveland formation is not topographicall}' ^^I'ominent. In the Fence River area it produces a more subdued and somewhat lower-lying surface than the underlying formation, but the difference is slight and is of little moment in comparison with the confusing effects of glaciation. FOLDING AND THICKNESS. In the Fence River area there is no reason to suppose that the Grove- land formation contains within itself minor folds of any importance. Our knowledge of its attitude is supplied almost wholly by the magnetic obser- vations, and these indicate that it has a general eastward dip like the under- lying members of the succession. Here and there it may be divided into two or more parts by sheets of intrusive material, and also may be slightly criimpled, but on the whole it must be regarded as a single persistent sheet, having a general eastward dip. At Michigamme Mountain the Groveland formation caps tlie hill in a well-marked syncline, the axis of which runs northwest and southeast. The structure is distinctly shown by the attitude both of the ferruginous rocks and of the underlying ])hyllites. At the Interrange exploration half a mile south, a secondary embayment of the same syncline, but more open, is found. Tliese are the only folds of the Michigamme Mountain area sufficiently deep to include the iron-bearing rocks. The thickness of the formation can onlv be guessed at, as no complete section is exposed, and the data for determining its upper limit are decidedly shadowy. The mag- netic observations indicate a breadth of from 400 to 600 feet, and as in the Fence River area it is certainly nuich thinner than the two lower forma- tions, its thickness may be approximately 500 feet. PETROGRAPHICAL CHARACTERS. In general aspect the iron-bearing formation in this area is strikingly like that of the Felch Mountain range, and all the varieties there found are represented here also. It is therefore unnecessary to repeat the detailed descriptions already given. By way of broad comparison, however, it may PETKOGRAPHIOAL CHAHACTEKS OF GROVEL AND FORMATION. 449 he .said that tliu toriuatiou coutaius less iron than in tlie Felch .Mountain range, and consequent!}- the lighter-colored varieties are more abundant, that it contains niort' detrital material, and that in tlie ]\Iichigamme Moun- tain area the texture is generally closer and less granular. Moreover, in passing north from the Michigamme Mountain area to the Fence River area we find at the Sholdeis and Doane explorations that the lower portion of the formation is com])osed of ferruginous quartzite, which is succeeded higher up by actinolite-schists and grihierite-schists similar in all respects to the characteristic rocks of the Negaunee iron formation in the western Marquette district. In this change in character as the Marquette district is approached is found the lithological support for the view, first suggested by the distribution of the lines of magnetic attraction, that the Groveland formation is the southern equivalent of both the Ajibik quartzite and the Negaunee formation of the Marquette district. The passage to a more crystalline condition in going from south to north is also in accord with the like changes already noted in the lower formations. Under the microscope the close texture of the Groveland rocks of Michigamme Mountain is seen to be due to the minuteness of the quartz grains of the groundmass, and to the abundance therein of chalcedony. The coarse quartz grains are all detrital and are often beautifully enlarged. In many slides feldspar pebbles occur, and in many also sericite and chlorite are prominent in the groundmass. The iron oxides, including both mag- netite and hematite, in single crystals and also in agg-regates, are well dis- tributed, as in the Felch Mountain sections. A similar grouping of these in round j^ebble-like areas as in the Felch Mountain range is also beauti- fully shown. In one slide tlie matrix is a rhomljohedral carbonate, prob- ably calcite, in which are embedded quartz grains and the iron ores in single crystals and irregular aggregates. The most interesting features of the thin sections from Michigamme Mountain are the pressure eff'ects. In many slides the detrital quartz grains are strained to an extraordinary degree. In one case the stage was rotated 45° before the black wave of extinction completely traversed a little pebble 0.3 mm. in diameter. Almost every section is crossed in sev- eral directions by fractures healed by the deposition of coarse quartz and the iron oxides. MON XXXVI 29 450 THE CRYSTAL FALLS IRON-BEARING DISTRICT. In the Fence River area the lower portion of the formation consists of quartzite, more or less ferruginous and micaceous. It contains beautifully rounded and enlarged grains of quartz, and also less abundantly rolled grains of microcline. Muscovite, biotite, and epidote occur, with the gen- eral background of interlocking later quartz. The more ferruginous laj^ers have a groundmass almost exclusively of hematite, in which the clastic particles are set. The hematite is in parallel micaceous scales, which com- pletely cover the cleavage surfaces. Above these layers come crystalline actinolite-schists and griinerite-schists, the former with garnets and Ijoth carrying particles of the iron oxides. These rocks are not distinguishaljle in the field or in thin section from certain varieties of schists of common occurrence in the Negaunee formation of the Marqiiette district. 50 n I m !l CHAPTER V. THE NORTHEASTERN AREA AND THE RELATIONS BETWEEN THE LOWER MARQUETTE AND LOWER MENOMINEE SERIES. From the northerninost outcrops of the Fence River area to the north- ern end of the Repubhc trough the air-lhie distance is about 1 1 miles. This intervening territory, on one side of which we find the typical formations of the Menominee district and on the other the typical formations of the Marquette district, remains to be described in this chapter. It may con- veniently be referred to as the Northeastern area. As was shown in the report on the Marquette district, in the pro- ductive portion of the Marquette range between Negaunee on the east and Repulilic on the west, the lower Marquette series consists of two or three clearly marked formations, which, perhaps, may further be subdivided according to individual taste. ^ The lowest of these, the Ajibik- quartzite, which rests on the Archean complex, is fragmental in origin, and is prevail- ingly a white vitreous quartzite, which in one or two localities is conglom- eratic near the base. Often it is represented by a muscovite-schist as the result of the dynamic metamorphism of the original arkose. In the eastern part of the productive area of the Marquette district and along the northern side of the main fold, in the western part of the district, this formation is overlain by the Siamo slates.^ Elsewhere the slates are not present, or are not known. The next formation is the Negaunee iron formation,* which has already been referred to in Chapter II. This rock, which has many phases, as there noted, is clearly marked oif from the lower quartzite by its great richness in iron and by the fact that over the whole Mai-quette district it nowhere appears to contain fragmental material, except in the transitional zone between it and the lower formations. 'The Marquette iron-bearing series of Michigan, by C. E. Van Hise and W. S. Bayley, with a chapter on the Republic trough, by H. L. Smyth: Mou. U. S. Geol. Survey, Vol. XXVIII, 1897, p. 221. 2 Op. eit., pp. 528-529. ' Op. cit., pp. 313-315. ' Op. cit., pp. 328-407. 451 452 THE CRYSTAL PALLS IRON-BEAEING DISTRICT. Above these conformable formations comes the unconformably placed Upper Marquette series, tlie base of which rests now on one member, now on the other, or on the Archean. East and south of Negaunee, and extending thence to the shore of Lake Superior at Marquette, is a series of rocks which resemble lithologic- ally neither the Upper nor the Lower Marquette series in the productive area. It consists, in ascending order, of quartzite with basal conglomer- ates, dolomite, and slates, and thus bears a close resemblance lithologically and stratigraphically to the three lower members of the Menominee series. This series, named by Wadsworth the Mesnard series, has been regarded b»y him as belonging with the Upper Marquette series, or at least as over- lying the Lower Marquette formations just described. Maj. T. B. Brooks had earlier correlated the dolomite with the Lower Marquette quartzite, and had supposed that there was a gradual passage from one into the other along the strike. Mr. C. R. Van Hise has recently stated that its position is below the Ajibik quartzite. This series is found only in the eastern part , of the Marquette area, between Goose Lake and Lake Superior, a distance of about 6 miles. Elsewhere, over by far the greater part of the Marquette district, no member of it has been recognized. The geological structure of the Marquette railge presents the general features of an east-west striking complex syncline or synclinorium. The pre-Cambrian sedimentary rocks, with their associated intrusive and extru- sive igneous rocks, occupy the trough, in which there is much local com- plexity of structure. The trough is flanked on the north and south by the •older Archean crystallines. At the western end of the district the peculiar Republic^ trough branches from the main synclinorium, and runs southeast into the Archean rocks for 6 or 7 miles, having a nearly constant width of about one-half to three-quarters of a mile. In this trough the Algonkian rocks have been so closely compressed that they stand essentially on edge. The interior is occupied by the younger Upper Marquette quartzites dud schists, betAveen which and the underlying Archean walls the older Lower Marquette iron formation and quartzite here and there occur. The northwestern end of the Republic trough is about the western ' Op. cit., p. 525. THE NOETHEASTEEN AEEA. 453 rmiit of iniuing' develop lueiit, tliough not of exploration, on the wouth side of the Marquette synclinorium. Up to this point outcrops, producing mines, and old explorations are sufficiently abundant to permit the separate formations to be traced and mapj^ed with comparative ease, and to indicate at least the larger structural features. At this northwestern end of the Republic trough the Lower Marquette sei'ies makes an abrupt turn to the south, and may be followed for a mile or more by occasional outcrops and test pits. The Negaunee iron formation is persistently present beneath the U]3per Marquette .quartzite, and gives rise to a very strong and persistent line of magnetic atl action, which was followed in our work for about 12 miles to the south and southeast into the Northeastern area. For about 4 miles from the sharp turn at the mouth of the Republic trough it runs nearly due south; afterwards it turns somewhat to the east of south, and follows that course for about 6 miles, after which it turns more and more toward the east, and finally, where we left it, its course w^as only slightly south of east. That this magnetic line is caused by and marks the position of the Negaunee iron formation there can not be the slightest doiibt, for that rock outcrops in a few scattered localities, occurs abundantly in the i:lrift, aiul has been found in occasional test pits and drill holes throughout this distance. The underlying quartzite outcrops beneath the iron-bearing formation near the northern end of the line, but farther south it is entirely covered by the drift, so far as the territory has been examined. The overlying Upper Marquette rocks are also known to be present just west of the Negaunee formation as far south as sec. 19, T. 46 N., R. 30 W. The magnetic line which accompanies the Negaunee formation may be called the A line. Taking into account the connected Republic trough and its exposures of the Lower Marquette rocks, it is seen that the A line par- tially surrounds a dome of the Archean crystallines, ai^J that in going from the interior of this dome outw^ard across the A line we pass from older to younger I'ocks. The dip along the A line is, therefore, on the whole, toward the west, although the observed dips at the few localities where determinations have been made are either vertical or slightly inclined from the vertical toward the east. The southern part of the A line, as far as it has been traced, passes through sees. 5, 8, 9, 15, and 16 of T. 45 N., R. 30 W. In section 5 it is just 5 miles east of the Grroveland formation, which. 454 THE CRYSTAL FALLS lEON-BEAElNG DISTRICT. as was shown in earlier chapters, is a magnetic rock occupying a definite place in the Menominee succession, and is underlain by other typical Menominee formations, and finally by the Archean. Between the A line and the magnetic line caused by the Groveland formation, which may be called the C line, is a third magnetic line, which may be called the B line. This was traced parallel to the A line and less than half a mile away, from near the south end of the latter to the north end, and finally entirely round an elliptical area, closing again upon itself at the starting point, the perimeter of the ellipse being 25 miles in length. Throughout this entire distance not a single outcrop could be discovered along the B line. Within the inclosed area, however, in sees. 6 and 7, T. 45 N., R. 30 W., and in sec. 19, T. 46 N., R. 30 W., several exposures of granites and crystalline schists were found, which left no doubt that the greater part of the area inclosed by the B line is occupied by Archean rocks of the same general character as those partially inclosed by the A line on the east and entirely by the C line on the west. The area between the A and B lines as far south as sec. 19, T. 46 N, R. 30 W., has been proved to contain the basal member of the Upper Marquette series. The southwestern quadrant of the B-line ellipse is nearly parallel to the C line and only IJ miles away. The known facts with reference to the B line, then, are these: (1) It represents a magnetic rock; (2) this magnetic rock completely encircles an Archean core. It may further be inferred with practical certainty that this formation, which carries such constant magnetic properties for 25 miles, must be sedimentary. With regard to' its structure, the foregoing con- siderations would necessarily involve the conclusion that it dips away from the Archean core on all sides, and this conclusion is fortified by the unsymmetrical separation of the horizontal maxima on the magnetic cross sections. It follows, therefore, that on the eastern side of the oval, where the formation is parallel to the A line, it dips toward the east, and on the western side, where it is parallel to the C line, it dips toward the west. This conclusion is further supported by the dips within the ellipse in the outcropping Archean rocks that show structure. These all happen to lie east of the major axis, and all dip toward the east. East of the B line, and between it and the A line, is found the basal member of the Upper Marquette series. The rock which is manifest in the THE NORTHEASTERN AREA. 455 B line must, therefore, be older than anj- member of the Upper Marquette series. The Negaunee iron formation, represented in the A line, dips west, while the rock of the B line dips east. They are both older than the basal member of the Upper Marquette series, and are both j^'ounger than the Archean. Thej' are both strongly and persistently magnetic. For 8 or 1 miles they run parallel to each other less than half a mile apart. Their broad structural relations to the Archean basement of the region are pre- cisely similar. Therefore, although the rock that gives rise to the B line has never j-et been seen, it may be concluded with the utmost confidence that it is the Negaunee iron formation, and that the A and B lines represent this rock brought up in the two limbs of a narrow and probably deep synclinal fold. This conclusion carries the Negaunee iron formation 3 J miles farther to the west, and in the northeast part of T. 45 N., R. 31 W., leaves a gap of but 1^ miles between the Lower Marquette and the Menominee sei'ies. Here, between the B and C lines, is precisely the same situation as between the A and B. One magnetic rock, represented by the B line, dips west; the other, the Grroveland formation, represented by the C line, dips east. Between them no magnetic disturbances can be found. The area between them must have a synclinal structure, and if they are not one and the same formation each must undergo an extremely rapid and precisely similar change in lithological character (namely, the loss of magnetite) in a very short distance and be represented on the opposite side of the syn- clinal fold by a nonmagnetic formation. Each of these rocks is ^Dersistently magnetic in the direction of the strike for great distances. That each should independently lose its magnetite in the direction of the dip in this particular locality is very improbable. And, therefore, the grounds for the conclusion that the B and C lines represent one and the same formation are quite as firm as those upon which rests the conclusion that the A and B lines repre- sent the same formation. The greater portion of the Northeastern area is without outcrops, yet through the structural and lithological residts of the magnetic work we are able to bridge over the gap and to show with a high degree of j^robability that the Negaunee iron formation of the Marquette range is identical with the Groveland iron formation of the Felch Mountain range. Further, when we recall the differentiation of the Groveland formation in the 456 THE CRYSTAL FALLS lEON-BEARING DISTRICT. northern part of the Fence River area into ferruginous quartzite at the base and griinerite-schist in the upper portion, it would seem probable that the Groveland formation represents the underlying Ajibik quartzite as well as the Negaunee formation of the western part of the Marquette range. This conclusion has an important bearing on the interpretation of the early geological history of what is now the Uj)per Peninsula of Michigan. If the formations which constitute the whole of the Lower Marquette series over the 25 miles or more of the productive and best-known portion of the range are represented in the Menominee district and the intervening area by a single formation, and that the highest in the Felcji Mountain succes- sion — namely, the Groveland formation — the formations below the Grove- land formation are all older than the Marquette rocks and do not occur at all within the productive portion of the Marquette range. Why are these lower formations absent? To this question there seem to be two answers which are a priori possible. It is conceivable that the quartzite, dolomite, and slates of the south, or some of them, may have been deposited in a succession of unbroken sheets over the whole Marquette area, in continuity with the simi- lar Mesnard formations on the east, and that afterwards the main Marqtiette area was elevated above the sea and entirely stripped of these formations by long-continued denudation. Finally, when the time of deposition of the Groveland formation came round, this elevated area had again been reduced to sea level, and subsided below it, so that the Ajibik quartzite and the Negaunee iron formation, and their southern equivalent, the Groveland formation, were deposited in an unbroken sheet over the whole. If this hypothesis is correct, two consequences should follow from it: First, we ought to find some discordance between the Groveland formation or the Lower Marquette quartzite and the lower formations in the marginal areas between the Menominee and Marquette, and the Mesnard and Marquette areas, respectively, or at least a gradual cutting out of these lower forma- tions by the iron-bearing members and the lower qviartzite, and, secondly, we ought to find, in the lack of discordance, rocks present in the areas of continuous deposition which represent the time of denudation. With regard to the first of these consequences no verification is possible, at least in the territory between the Marquette and Fence River districts, from lack of outcrops. Throughout the Northeastern area, from the north- THE NORTHEASTERN AREA. 457 western end of the Republic trough in T. 47 N., R. 30 W., to the C hue in T. 45 N., R. 31 W., there are no exposures whatever of the Algonkian rocks which underhe the Groveland formation. Somewhere in this distance of about 11 miles the lower formations disappear, but whether by unconformity or overlap is an unanswerable question; nor (for the same reason) can it be definitely settled whether elsewhere farther to the south there is any discordance. That there is general parallelism between the Groveland formation and the lower rocks, and strict conformity in some places, is true. But this is not at all inconsistent with a period of erosion between them, if that erosion antedated the later and more severe orogenic disturbance. In the Mesnard area the observed relations have been interpreted by Van Hise to mean that the lower formations disappear by overlap. The facts at present known on the Felch Mountain side are capable of the same interpretation, but they are not sufficiently definite to exclude the possibility of a period of erosion below the iron-bearing formation. With regard to the second consequence — the deposition in the sub- merged areas of formations which would represent the erosion period in the elevated area — the evidence at hand is decidedly against the existence of such formations. The alternative hypothesis is that the lower quartzite, dolomite, and slate formations of the Menominee area were not deposited over the western Marquette area at all, but disappear toward the north and east by overlap, and this hypothesis is much more likely to be the true one. We can suppose, as I have already pointed out,^ that this part of the Upper Peninsula was a slowly subsiding area, the central portion of which, now occupied by the Marquette rocks, stood initially at a greater elevation above the encroaching sea than the rest. While the quartzite-dolomite-slate triad was going down in the Mesnard area on the east and the Menominee area on the south and west, the central Marquette area still remained above the sea. At last, when the Groveland foi'mation began to be deposited, the Marquette high land was finally submerged and covered, as the sea marched over it, first, with a sheet of arkose made up of its own disinte- grated ddbris, and, finally, with the same nonclastic sediments as chiefly compose the Groveland and Negaunee formations. ' Relations of the Lower Menominee and Lower Marquette series in Michigan, by H. L. Smyth : Am. Jour. Soi., Vol. XLVII, 1894, p. 222. CHAPTER VI. THE STURGEON RIVER TONGUE. By William Shirley Bayley. DESCRIPTION AND BOUNDARY OF AREA. The Sturgeon River area of Algonkian sediments, like the Felch Mountain area, is an east-west tongue of conglomerates, slates, and dolo- mites, very narrow at its eastern extremity and widening out toward the west until it finally plunges under drift deposits that separate it from the large Huronian area of the Crystal Falls district. The tongue occupies the western portion of T. 42 N., R. 27 W., the central and northern portions of T. 42 N., R. 28 W., T. 42 N., R. 29 W., and T. 42 N., R. 30 W., and the southern parts of T. 43 N., Rs. 28 W., 29 W., and 30 W. The best exposures of the rocks constituting the tongue are found in sees. 7, 8, 17, and 18, T. 42 N., R. 28 W., and in sees. 1 and 3, T. 42 N., R. 29 W., on or near the northwest branch of the east branch of the Sturgeon River; hence the name Sturgeon River tongue (PI. LI). On the south the sedimentary rocks are bounded by an area of granites, gneisses, hornblende-schists, and mica-schists, that are cut by granite and quartz veins, by dikes of diabases, and by other greenstones. This area separates the Sturgeon River tongue of sediments fi-om the Felch Mountain tongue lying from 2 to 3 miles farther south. The exact line of demarca- tion between the granite-schist complex and the sedimentary rocks is difficult to draw, because for the eastern 7 miles the latter are bordered by green- stones whose position in the granite-schist complex or in the sedimentary series can not be determined at present. The line as drawn on the map places the greenstones in the Archean. It begins near the south side of sec. 7, T. 42 N., R. 27 W., and runs a little south of west to the quarter post between sections 17 and 18, in the next town west, then northwest to near 458 51 U.S GEOLOGICAL SURVEY MONOGRAPH XXXVI PL LI, GEOLOOK^AL MAP OF STIHGEOX HIVJ-JU TONOUE SCALE i INCH - i MILE COT^TOUR INTKRVAL 2() FEET '^ Outci'ops williout observed strike or dip =r Outoroijs vvilli deternuned strike and dip TOutcrope with slaltnesH or 3(.'hiBtosity. ^ ■ Teel pilH bottomed in rock, \^nTI('AL SCAIiE or SE(-/riON 1 INCH -i:i2 460 THE CEYSTAL FALLS IRON-BEARING DISTRICT. and argillaceous slates in Ts. 42 and 43, Rs. 29, 30, etc., of mica-slates in Ts. 41 and 42, Rs. 29, 30, etc., and of "calciferous sandrock" near the south boundary of T. 42 N., Rs. 27 and 28 W. In a list of specimens gathered from these townships Burt^ mentions sienitic greenstone, trap, granite, granulite, and talco-micaceous slate. On the land plats made by these surveyors conglomerate is noted on the west line of sec. 8, T. 42 N., R. 28 W., and marble at the south corner between sections 3 and 4 in the same township. In 1851 Messrs. Foster and Whitney reported^ the existence of an arm of Azoic rocks about 18 miles in length and 10 in breadth, extending east- erly into Ts. 42 and 43 N., R. 28 W., and located its position on their map of the Upper Peninsula. Brooks,^ in his description of the northern iron belt of the Menominee district, refers to the existence of outcrops of hornblendic rocks, mica- schists, and gneisses, cut by trap dikes, which he regarded as equivalents of the various greenstone-schists exposed along the Menominee River. "Near the center of this hornblendic belt, in the north part of sees. 22, 23, and 24, T. 42 N., R. 29 W., a line of weak magnetic attraction was observed. This is regarded as an indication here of the existence of an iron-ore belt." The gneiss, granite, etc., north of the north quarter post of sec. 31, T. 42 N., R. 29 W., he declares to have the appearance of typical Laurentian rocks. "If future investigations prove them to be Laurentian, a very troublesome structural problem would be presented here, as we would have Laurentian rocks conformably overlying beds unmistakably Huronian."* The only distinct reference made by Brooks to the sedimentary beds of the district is in the following paragraph:^ A range of marble associated with quartzite, chloritic and talcose rock, and overlaid by a chloritic gneiss, with beds of chloritic schist and gneissoid conglom- erate, the whole dipping at a high angle to the south, passes about 5 miles north of ' Geological report of the survey of a district of township lines in the State of Michigan, in the year 1846, by Wm. A. Burt: Thirty-first Congress, first session, 1849-50, Senate Documents, Vol. Ill, No. 1, p. 84. ^ Report on the geology and topography of the Lake Superior land district, hy J. W. Foster and J. D. Whitney, Part II, The Iron Region: Thirty-second Congress, special session, 1851, Senate Documents Vol. Ill, No. 4, p. 14. 3 Iron-Bearing Rocks (economic), by T. B. Brooks: Geol. Survey of Michigan, Vol.1, 1869-1873, N.Y., 1873, p. 161. * Op. cit., p. 175. ' Op. cit., p. 176. LITBRATLTRB ON STUEGEON RIVER TONGUE. 461 the North belt (i. e., the Pelcb Mountain tongue). These may represent the north side of the trough or basin, of which this iron belt is the south outcrop. No iron has, however, been found, so far as I know, on this range. In Romiuger's first I'eport on the Menominee district only a single reference is made to this area. He declares that a series of test pits put down in the W. ^ sec. 26, T. 42 N., R. 29 W., and in the SW. i sec. 14, T. 42 N., R. 30 W., are in decomposed granite.^ A specimen of the conglomerate referred to by Brooks as overlying the marble in the belt 5 miles north of Felch Mountain is described and pictured by Van Hise^ in his paper on the Principles of North American pre-Cambrian Geology (see also PI. LIII). It is stated to be from the Felch Mountain district. The more exact location of the ledge from which it was obtained is near the northwest corner of sec. 17, T. 42 N., R. 28 W., in the Sturgeon River tongue. Thus the only distinct reference to a tongue of sediments north of the Felch Mountain range is that of Brooks, although the existence of sedi- mentary rocks in this portion of the Menominee district was reported by Hubbard and Burt. Brooks believed that the Sturgeon River rocks repre- sented the northern rim of a syncline whose southern i-im constitutes the I'dch Mountain range, although both he and Rominger discovered a granite-schist complex underlying the country between the two areas of fragmental rocks. RELATIONS BETWEEN" THE SEDIMENTARY ROCKS AND THE GRANITE- SCHIST COMPLEX. As has already been stated, the country between the Sturgeon River tongue -of sediments and the Felch Mountain tongue is underlain by a com- plex of granites and various schists, traversed by fresh and altered diabases and by granite and quartz veins. Brooks recognized these rocks as pre- senting a Laurentian aspect, although he felt constrained to call them Huronian, because of the supposed structural difficulties involved in any other view of their age. No contacts of this granite-schist complex with the bedded rocks of the Sturgeon River tongue have been discovered. Nevertheless, there can be little question as to the relative ages of the two series. As has been 1 Geol. Survey of Michigan, by C. Rominger, Vol. IV, 1881, pp. 198-199. 2 Sixteenth Ann. Kept. U. S. Geol. Survey, 1896, p. 801 and PI. CXV. 462 THE CRYSTAL FALLS IROISr-BEAEING DISTEIOT. stated, the granites and schists extend southward to the Felch Mountain fragmentals, and here they are unconformably beneath the latter. Moreover, since the Sturgeon River rocks and the lower members of the Felch Moun- tain series are identical in character, it is probable that they are of the same age, in which case the granites and schists that are older than the Felch Mountain rocks are older also than those of the Sturgeon River tongue. The relations of the sedimentary series to the granites on the north have not been determined, because no contacts are exposed. The granites, however, can be traced northward until they are found unconformably beneath the rocks of the Lower Marquette series at Republic, and these, so far as is known, are the oldest sediments in Upper Michigan. There can be little doubt, therefore, that the relations of the sediments to the northern granites are the same as those with the southern schist complex. The granites of the two areas surrounded by the sediments are prob- ably of the same age as the northeni and southern granites. The rocks of the area in sees. 7 and 8, T. 42 N., R. 28 W., and sec. 12, T. 42 N., R. 29 W., are demonstrably beneath the conglomerates, though their relations with the dolomites have not been determined. A well-marked contact between the granites and the conglomerates is exposed at the south base of a small hill of granite in the NE. | sec. 7, T. 42 N., R. 28 W} The conglomerate here is well bedded. Its strike is N. 60° W., and its dip almost vertical. It consists largely of pebbles and bowlders of granite identical with the granite composing the hill, and a matrix constituted entirely of granitic debris. The contact, though exposed for only a short distance, seems to be an erosive one. It is certainly not an igneous one. From a consideration of the facts as given above, there can be little doubt that the rocks of the granitic areas within the Sturgeon River tongue and of those bounding it on the northern and the southern sides are older than the sediments within the tongue, though this has not been proved for the granites with respect to the limestones. From the lithological similarity of the Sturgeon River fragmentals with those of the Felch Mountain district, and from the structural relations exisiting between the rocks of the two districts, it is practically certain that the Sturgeon River sediments are of the same age as the Felch Mountain 1 The exact location of the coutaot is 400 paces N., 280 W., of the southeast corner of section 7. BASEMENT COMPLEX OP STURGEON RIVER TONGUE. 463 ones — i. e., Meuomiuee (Huroiiiau) — while the granites and schists belong to the Basement Complex on which the Lower Algonkian beds throughout Michigan have been laid down. THE BASEMENT COMPLEX. • The Basement Complex rocks in the area studied comprise gneissoid granites, biotite-schists, and hornblende-schists, cut by dikes of greenstone and by veins of quartz and granite. The granites are best exposed in the NE. 4 sec. 7 and the NW. ^ sec. 8 and the NE. i sec. 7, T. 42 N., R. 28 W., where they occur as bare knolls of a fairly coarse pink rock, separated from one another by stretches of sand. The best exhibition of rocks with the typical aspect of the Basement Complex is along the west half of the east-west quarter line of sec. 19, T. 42 N., R. 28 W., and south of the center of this section. Here we find hornblende-schists and hornblende-gneisses cut by veins and dikes of red granite and by greenstones that are usually schistose. Near the west quarter post of the section is a high hill bare of vegetation. On this hill the rocks are especially well exposed. In addi- tion to the types already mentioned, there is present here a coarse white pegmatitic-looking granite that apparently cuts the hornblende-gneiss. All the members of the Basement Complex in this area are so similar to the corresponding members of this complex elsewhere in the Lake Supe- rior region that they demand but little description. They are described here only in sufficient detail to establish their character. THE GNEISSOID GRANITES. The gneissoid granites north of the fragmental tongue, and those of the two areas surrounded by the sedimentary rocks, are mediumly coarse aggregates of a dark-red feldspar, white quartz, and a dirty green chloritic substance. The red feldspar is in excess, sometimes to the exclusion of the other components, when the hand specimen resembles a dense red felsite. Almost all specimens are gneissoid. The constituents are usually lenticu- lar, but in a few specimens, particularly those taken from near the contacts with the sedimentary rocks, they are drawn out into long slender string-like masses, giving the specimens a streaked appearance. The microscopical features of all the granites are those common to these rocks elsewhere in the Basement Complex. They consist of clouded ortho- 464 THE CRYSTAL FALLS lEON-BEAEING DISTEICT. clase, some plagioclase and a little microcline, quartz in varying quantity, and more or less green chlorite that seems to have been derived from biotite. All the constituents present abundant evidence of the efFects of pressure. In the least-crushed rocks the quartz shows undulatory extinc- tions, and the feldspar grains granulation around their edges. As the criTshing action increased, the granulation increased, so that the most crushed granites now consist of large grains of feldspar and of quartz in an aggregate of broken fragments of orthoclase, quartz, plagioclase and micro- cline, and a few wisps of green chlorite. Movement in the crushed rock mass has drawn out the granulated aggregate between the large grains of feldspar into bands and lines, thus producing the schistose structure noted in the hand specimens and in the ledges. In the more highly schistose granites a considerable quantity of new microcline and a small quantity of new plagioclase have developed within the granulated aggregate, and in a few instances muscovite has been found in fairly large plates of pale-yellow color. This muscovite occurs on the contact between the larger quartz and orthoclase grains, but more particularly in the granulated matrix. The granites in the area between the Sturgeon River and the Felch Mountain tongues are not so abundant as those in the northern area of Base- ment Complex rocks, or in the areas surrounded by the sediments, but in their essential features they are identical with these. Occasionally the sur- face of a fresh fracture through these southern granites shows the outlines of porphyritic orthoclase crystals, but these crystals are riot sufficiently numerous to impart a porphyritic aspect to the rock. Some of the granite specimens examined from this district are so com- pletely granulated that they can with difficulty be distinguished in the hand specimens from the schistose arkoses near the base of the fragmental series. In thin section they differ from the latter in containing no rounded quartz grains and in the possession of very little niica. The feldspathic constitu- ents are nearly all decomposed, and very much of the quartz present in the granites is of secondary origin. Thus in all essential respects the gneissoid granites of this district are like those in the Marquette district elsewhere described.^ 'The Marquette iron-bearing district of Michigan, hy C. E. Van Hise and W. S. Bayley, with a ■chapter on the Republic trough, by H. L. Smyth : Mon. U. S. Geol. Survey, Vol. XXVIII, 1897, pp. 171-176. BASEMENT COMPLEX OP STURGEON EIVEE TONGUE. 465 THE AMPHIBOLE-SCHISTS. In additiou to gneissoid granites, the southern area of the Basement Complex contams a number of ledges of dai-k-colored schistose rocks. These, in some instances, are cut by dikes of granite similar to the granite already described. These dark schists may be classed as greenstone-schists and as horn- blende-schists. The former are heavy rocks, with dull greenish-gray luster and distinct schistose structure. They resemble closely in their micro- scopic as well as in their macroscopic features the schistose dike greenstones to be referred to later. They are doubtless altered or squeezed diabases or gabbros. The hornblende-schists are usually line-grained, bluish-black rocks, with a very even schistosity, closely resembling slaty cleavage. On the surfaces of cross fractures may be seen long slender prisms of glistening black hornblende aiTanged in as distinct lines as the lines of particles in an evenly bedded sedimentary rock. Often the cleavage surfaces are coated with thin layers of golden-yellow mica scales. In most specimens there may also be noticed a fine banding parallel to the foliation. In thin section these hornblende-scliists differ from the schistose dike diabases and from the greenstone-schists, referred to above, in the presence of large quantities of quartz, and of some biotite, and to some extent in structure. The greenstones owe their schistosity to the flattening of their components, while in the hornblende-schists this structure appears to be due largely to the crystallization of the hornblende in elongated prisms with their major axes parallel. The parallel arrangement of the amphibole in the latter rocks is thus much more pronounced than in the schistose greenstones. The hornblende-schists are comjDosed of hornblende, quartz, biotite, plagioclase, magnetite, and sphene. The hornblende is in long prisms of the usual yellowish green color. The mineral is compact, but it is full of inclusions of quartz grains similar to those constituting a large part of the matrix lying between the amphiboles. It was evidently formed in situ as an original crystallization, and not, like much of the hornblende of the schistose greenstones, by the alteration of augite or of some other component of a basic crystalline rock. The biotite MON XXXVI 30 466 THE CRYSTAL FALLS IRON BEARING DISTRICT. is in small dark greenisli-brown flakes interspersed between quartz and feldspar grains, wliich together constitute a matrix surrounding the other components. The quartz is unusually free from inclusions. It contains a few liquid inclosures and occasionally a few flakes of biotite and needles of hornblende. The plagioclase, where present, is in irregular grains with ragged outlines, as though a newly formed mineral. It apjjears to act the part of a cement surrounding the other minerals with which it is in contact. Small round grains of sphene and magnetite occur very abundantly scat- tered through the matrix. Often the magnetites are surrounded by borders of sphene ; hence it is probable that this mineral is a titaniferous variety and that the round grains of sphene are pseudomorphs after mag'netite grains that have been completely altered. In a few specimens large colorless areas with the outlines of porphy- ritic crystals are observed in the midst of the finer-grained groundmass of schist. Between crossed nicols these break up into a coarse-grained aggre- gate composed of the same minerals that constitute the rest of the rock, except that in it altered plagioclase is common and amphibole is rare. These probably represent phenocrysts of plagioclase which have suffered alteration into quartz and new plagioclase that may differ somewhat from the feldspar of the original crystal. The banding of some of the hornblende-schists has already been referred to. Under the microscope the only differences noted in the bands are the quantity of hornblende present in them and a variation in the coarseness of grain. The coarsest of the bands have the composition and structure of the schistose greenstones. They contain large quantities of plagioclase, both fresh and altered, and large grains of hornblende that are not in the definite prismatic form characteristic of this mineral in the main mass of the rocks. OEIGIN OF THE AMPHIBOLE-SCHISTS. From the gradations often observed between the hornblende-schists and the greenstone-schists, it is plain that the two rocks are genetically related. The latter, from their similarity to schistose dike greenstones in composition and structure, are believed to have been derived from massive diabases or gabbros. The hornblende-schists are in all probability derived .from similar basic rocks, though the presence in them of what appears to BASEMENT COMPLEX OF STURGEON EIVEK TONGUE. 467 have once been plagioclase pheiiociysts may indicate that the original rocks were in the form of hxvas. The principal difference between the hornblende-schists and tlie greenstone-schists seems to be in the nature of the amphibole in the two rocks and in the presence of quartz and newly formed plagioclase in the first named. The materials of the greenstone-schist were derived from the alteration of those of the original rock, as were also those of the hornblende- schist, but the former now consist mainly of the direct jjroducts of this alteration, whereas in the latter the substances now existing liaA^e been worked over and entirely recrystallized. THE BIOTITE-SCHISTS. Mica-schists are not common in the Sturgeon River tongue. They constitute by no means so large a part of the Basement Comj)lex in this dis- trict as they do in the other portions of the Lake Superior region that have been studied. Indeed, only a few ledges of this rock have been observed in the country between the Sturgeon River and the Felch Mountain sedi- mentary tongues, and most of these are along the southern edge of a g'reen- stone knob 300 to 400 paces north of the southeast corner of sec. 17, T. 42 N., R. 28 W. The mass of this knob is a dark hornblende-schist. On the south side of the top of the knob this rock is in contact with a A^ery evenly banded or streaked rock of a general dark-gray color. In the hand specimen it resem- bles very closely a fine-grained banded aug-en-gneiss. Near its contact with the hornblende-schist the rock is apparently porphyritic, with pheno- crysts of feldspar from 1 to 15 mm. in length, and an occasional one of quartz scattered through a matrix composed of narrow altei-nating bands of almost black and light-gray material. On cross fractures of the rock the phenocrysts are seen to • be drawn out in the direction of the bands. Cleavage takes place verj'- readily along the j^lanes of the banding, yielding- siTrfaces covered with tiny scales of black biotite. A little farther from the contact the light-colored bands are thicker and more distinct. At first glance they appear to be uniformly thick for long distances, but a more careful inspection shows that they wedge out rapidly and are replaced by other bands of the same character. The dark bands are not thicker than 468 THE CRYSTAL FALLS IRON-BEARmG DISTRICT. sheets of paper. They are the cross sections of the mica coatings on the cleavage planes. The inspection of this rock in the hand specimen and in the ledge leads to the same conclusion — that it is an intermediate or an acid lava, a porphyrite, or a porphyry that was squeezed until it became schistose and sheared until it became fissile. Under the microscope the feldspar phenocrysts, though much decom- posed and filled with inclusions of quartz, muscovite, and other decomposi- tion products, are well enough preserved to exhibit in some places twinning striations. The greater portion of the phenocrysts are untwinned. The twinned material borders the grains, fills in cracks between Carlsbad twins, and is irregidarly distributed through the untwinned material, occurring more particularly in those places where the decomposition of the original feldspar is most complete. The twinned feldspar is fresher than the untwinned variety. This fact and the manner of its distribution indicate a secondary origin for it. The quartz phenocrysts are rare. They present their usual characteristics. The groundmass in which the phenocrysts lie is a fine-grained aggTe- gate of biotite, quartz, and plagioclase. The biotite is a greenish-brown variety. It occurs in large plates arranged in parallel position and in small flakes occupying the same parallel position and lying between the quartz and the plagioclase grains. The banding noticed in the hand speci- men is due to the arrangement of the large biotite flakes in bands. These are separated from each other by bands of quartz and plagioclase that are free from the large biotites, though they contain innumerable small flakes of this mineral. Only when a porphyritic crystal lies in the way of tlie bands do these depart from their uniform directions. Here they bend around the phenocrysts, leaving on both sides of them little triangular areas in which the components are much finer grained than elsewhere in the rock. The light-colored components are quartz and plagioclase. These min- erals are in small grains that appear to be intercrystallized in the manner of the secondary aggregate that constitutes the fine-grained matrix of many greenstones, of the aporhyolites, and of other rocks that have sufi"ered intense metamorphism. The quartzes are nearly always crossed by strain shadows and the fresh clear plagioclase by interrupted and bent twinning bars. BASEMENT COMPLEX OF STURGEON RIVEE TONGUE. 469 Here and there in the midst of this fine-grained groundmass are noticed lenticuhxr and k)iig' narrow aggregates composed of grains of plagiochise that are much larger than the grains of this mineral occurring- in the sur- rounding matrix. They look as though they might be the crushed remains of what were originally plagioclase phenocrysts. Thus the microscopic study of these rocks tends to confirm the results of their field study. They were probably porphyritic lavas or intercalated flows that have sufl'ered alteration as the result of intense pressure and movement. Their present composition suggests that they were oi'Iginally quartz-porphyries or perhaps andesitic porphyrites. Whatever their original nature, their origin is different from that of the biotite-schists of the Mar- quette district.-' THE INTRUSIVE ROCKS. The intrusives in the schists and gneissoid granites of the Basement Complex are granites, identical with the gneissoid granites above described, and greenstones. The former cut only the schists. They are probably apophyses from the larger granite masses. The greenstones cut the schists and the granites. They are similar in all respects to the greenstones in the sedimentary series, and thus are the youngest rocks in the district, with the exception of the horizontal Paleozoic sandstones and limestones that cap some of the higher hills. The greenstones are all more or less altered diabases. In some the ophitic structure may be detected, but in most of them no traces of their original constituents nor of their structure remain. Nearly all are more or less schistose. The only evidence that the most schistose phases were once massive igneous rocks is in their composition and their occurrence in dike- like fissures. As the schistosity of these greenstones increases, the amount of their altei'ation also increases; there is a greater abundance of horn- blende present in them and a greater quantity of quartz, until in the most schistose phases the rocks are now typical hornblende-schists. One of the best examples of these greenstones occurs in the series of ledges extending in nearly a straight line for 6 miles from the southern portion of sec. 13, T. 42 N., R. 29 W., to the northeast corner of sec. 14, T. 42 N., R. 28 W. Except in its eastern ledges the rock constitutes bold, ' Mou. U. S. Geol. Survey, Vol. XXVIII, pp. 200-203. 470 THE CRYSTAL FALLS lEON-BEARING DISTRICT. rounded, bare knobs with ahnost perpendicular sides, usually situated in the midst of swamps. The main mass of the knobs is a rather fine-grained, slightly schistose, gray rock exhibiting the diabasic structure on weathered surfaces. On the south sides of the knobs the rock is much denser, and in most cases is much more highly schistose than the main rock mass. Under the microscope these rock's present the usual features of schistose dike greenstones. They consist almost exclusively of hornblende, plagio- clase, and quartz. The hornblende, which is the common yellowish-green variety, occurs in long plates and in columnar crystals, some of which are idiomorphic in cross section, and also in slender needles penetrating the quartz and feldspar. These two minerals form an aggregate between the larger hornblendes. The feldspar is mainly a calcium-soda plagioclase, though a small quantity of albite may also be present. It occurs as irreg- ular grains embedded in a mosaic composed of rounded grains of the same feldspar and of quartz, and appears to be a new crystallization subsequent to that of the greater portion of the plagioclase. At any rate, a single large grain often fills the interstices between numbers of the mosaic grains and extinguishes uniformly over large areas. The magnetite in the rock is titaniferous. It occurs in little crystals and in small irregular grains that are often surrounded by a granular zone of leucoxene. This rock may serve as a type of nearly all the other dike greenstones in the district under discussion. Some may be more schistose than this one, while a few ma}^ be more massive, but in general characteristics they are all similar. The more schistose rocks diff'er from the less schistose varieties simply in the possession of a greater amount of quartz and a greater quantity of what appears to be newly formed feldspar. Their greater schistosity is due to the more uniform elongation of their components. The fine-grained greenstones found on the edges of the coarser-grained ones, and occasionally as independent dikes, are weathered diabases of the normal type. COMPARISON OF THE STURGEON RIVER AND THE MARQUETTE CRYSTALLINE SERIES. The Basement Complex in this area is essentially like that in the Mar- quette district, except that the altered tuffs so abundant in the northern area are absent from that now under discussion. The biotite-schists of the two ALGONKIAN KOCKS OF STURGEON EIVBR TONGUE. 471 areas seem also to be different in origin, althoug-h this can not be stated with certainty, since the origin of the Marquette schists is not so clear as is that of the Sturgeon River schists. There is enough similarity between the crystalline series in the two areas to leave no doubt as to their practical identity. If the Marquette Basement Complex is Archean, the crystalline series underlying the conglomerates in the Sturgeon River tongue is also Archean. THE ALGONKIAX TROUGH. The sedimentary rocks comprised within the Sturgeon River Algonkian tongue may be separated into a conglomerate series and a dolomite series. The conglomerate series consists of schistose conglomerates, arkoses, qnartz- ites, slates, and certain sericitic schists that are squeezed arkoses. The dolomite series embraces crystalline dolomites, a few thin beds of quartz- ite, a few breccias and conglomerates, and some slates. It is possible that a third series, composed essentially of slates, also exists in the district, but if so it is not advisable to separate it from the dolomite series, since its exposures are very few in number, and the slates which comprise its main mass are so nearly like the slates belonging in the dolomite series that they can with difficulty be distinguished from these. Associated with the sedimentary rocks are great inasses of basic igneous ones. Some of these are unquestionably intrusive masses, as shown by their relations to the conglomerates, while others appear to be interleaved sheets. A very few, apparently bedded greenstones, on close examination seem to be composed of intermingled sedimentary and igneous material. These may be altered tuffs. Nearly all the sedimentary as well as the igneous rocks embraced in the trough are schistose, and thus are sharply distinguished from the brown Potsdam sandstones and the Silurian limestones that here and there lie approximately horizontal on their upturned edges. The squeezing of the pre-Potsdam formations has been so intense that both conglomerates and dolomites have been forced into closely appressed folds, which in the con- glomerates are for the most part apparently isoclinal. The strike of the latter rocks is nearly east and west, and their dip nearly perpendicular, except in one or two cases. The dolomites are less closely folded than the conglomerates. Their dips are mucli less steep, and their strike varies 472 THE CRYSTAL FALLS IROJS[-BEAEING DISTRICT. considerably, except in the narrow eastern portion of the tongue, where it is approximately parallel to that of the conglomerate, i. e., a few degrees north of east. RELATIONS BETWEEN THE CONGLOMERATE AND THE DOLOMITE SERIES AND CORRELATION WITH THE FELCH MOUNTAIN FRAGMENTALS. The relations of the conglomerates to the dolomites are best shown by the distribution of their respective outcrops, as members of the two series are nowhere in contact. In the central portion of the tongue the conglomerate outcrops are limited to the district between the central granites and the southern area of the Basement Complex. The dolomites, on the other hand, are limited to the country north of the central granite. Its outcrops are found scattered over the northern tier of sections in T. 42 N., Rs. 28 W. and 29 W., and the southern tier of sections in T. 43 N., Rs. 28 W. and 29 W. Between them and the granite to the north is a belt of country devoid of exposures. It is heavily drift covered, consisting of sand plains and sand hills, from beneath which no ledges of any kind protrude. This barren belt measures about a half mile in width, in sec. 2, T. 42 N., R. 28 W., gradually increasing in width till it reaches the center of sec. 1 in T. -tS N., R. 29 W., where it opens out into the large Pleistocene area whose southeast edge is shown on the map (PI. LI). In the eastern portion of the district the northern granites and the conglomei'ates approach each other, and the dolomite belt becomes very narrow, finally disappearing toward the east side of T. 42 N., R. 28 W. The relative distribution of the conglomerate and dolomite ledges, when considered with reference to the triangular outline of the area embraced between the northern and the southern granite-schist complexes, suggests that the two formations constitute a western-pitching syncline with the dolomite in the center and the conglomerates with their associated beds on the two flanks. The conglomei'ates comprising the southern flank are well exposed, but those of the northern flank are not seen. They are believed to underlie the glacial deposits in the barren strip of country bordering the northern granites. The conglomerates, according to this view, are older than the dolomites. Toward the center of the dolomite area, in the north half of sec. 6, T. 42 N., R. 28 W., and at a few places farther west, there are ferruginous ALGONKIAN EOGKS OF STURGEON ElVBR TONGUE. 473 beds in the dolomite series. If these represent the upper portion of the dolomite formation, as is the case with similar rocks in the Felch Mountain range, it is clear that as we approach the center of the Sturgeon River tongue the rock beds met with are younger than those on its borders. This is in line with the supposition that the Sturgeon River tongue is a westward- pitching syncline. The belief that the conglomerates are beneath the dolomites in the Sturgeon River area is further strengthened by the fact that the jirincipal conglomerate in the Felch Mountain range is benea.th a dolomite which is identical in character with the Sturgeon River dolomite. This conglomerate is regarded as the base of the Lower Menominee series in this district, with the dolomite above it, known as the Randville dolomite, immediately suc- ceeding it. If the conglomerates and dolomites in the two districts are the same, the Sturgeon River rocks are Lower Menominee. RELATIONS BETWEEN THE DOLOMITES AND CONGLOMERATES AND THE OVERLYING SANDSTONES. At several places the conglomerates- and dolomites are overlain hy well-defined Lake Superior sandstone. The sandstone usually caps hills, on the lower slopes of which ledges of the underlying rocks appear. The contacts between the overlying sandstone and the underlying rocks are rarely seen, but the fact that the former are always horizontal, while the latter are always very steeply inclined, leaves no doubt that there is a strong unconformity between them. THE CONGLOMERATE FORMATION. The conglomerate formation comprises very much squeezed granitic conglomerates, arkoses, sericite-schists, quartzites, a few beds of banded rocks believed to consist largely of tuffaceous material (see pp. 486-487), and occasional beds of slates. Nearly all the members of the series are schistose, the arkoses in some cases passing into very well characterized sericite-schists. Occasionally the arkoses show obscure traces of ripple marking, and more frequently very well defined cross bedding. All the rocks of this formation strike in a nearly uniform direction, N. 76°-84° E., and dip almost vertically. In one or two instances observed the dip is as low as 65°, but in most cases it varies between 85° N. and 85° S. The strike of the schistosity is approximately parallel to the strike 474 THE CEYSTAL PALLS IRON-BE ARIKG DISTRICT. of the bedding, as is also the direction of the elongation of the pebbles so abundant in the conglomeratic layers. From the slight changes iii dip observed in the beds, as well as the great width of the formation in some places, it is evident that folding must exist. It is probable that in the wider portions of the area occupied by these rocks there are present two or more folds, so closely appressed that the beds on the opposite limbs can not be correlated. Hence they appear as members of a consecutive series of conformable members with a nearly uniform dip throughout. In the narrower portions of the area it can not be told whether more than one fold is loresent or not. In any event, the folding is practically isoclinal. The ledges of the conglomerates and their associated beds occur in the southern portion of the Sturgeon River tongue throughout its entire extent. No exposures have been found north of the granite areas in the central and western portions of the tongue. IMPOKTANT EXPOSURES. The arkoses, the sericite-schists, and the conglomeratic phases of the series can be best studied at the dam of the Sturgeon River near the north- west corner of sec. 17, T. 42 N., R. 28 W. Here they form a continuous ledge of well-bedded layers striking N. 83° E., and dipping 85° S., which measures at least 250 yards in width and 400 yards in length. (See PI. LII.) The conglomerates are pink in color. They contain immense numbers of white quartz pebbles and bowlders, fewer and smaller ones of pink granite, and many fragments of red feldspar in a matrix composed of moderately coarse granite debris. All the fragments and pebbles in these rocks, as well as then- matrix, show plainly the eflPects of pressure (PL LIII.) The matrix of all specimens is more or less schistose, and the coarse sand grains embedded in it are in many cases elongated in the direction of the schistosity. Most of the pebbles and bowlders in the conglomerate are also flat and parallel to the schistose plane. How far these phenomena are due to mashing, to rota- tion into parallel positions during flattening, and to original sedimentation, respectively, can not be determined in most cases, since the schistosity of the rock and the elongation of the pebbles are both approximately parallel to the bedding — i e., the pebbles are nearlj^ in the positions assumed by unequidimensional pebbles in a well-bedded conglomerate. In a few us GEOLOGICAL SURVEY MONOGRAPH XXXVl PL III n \ ^ mW^ ^ '^ Mm>^^'f |;f %^^. vf f ^l\%^ % ,^ JULIUS BiEN aco.N.r MAP OF EXPOSURES IN SEC. 7 AND IN PORTIONS OF SECS. 8, 17, AND 18, T. 42 N.,P. 28 W., M ICH I GAN VICINITY OF DAM ON EAST BRANCH OF STURGEON RIVER I A. Arkose SI. Slate Cj. Quartzite INTRUSIVE C. Conglomerate G.T. Greenstonetiiffaceous ,,j G. Greenstone G . S . Greenstone Schist ALGONKIAN EOCKS OP STURGEON EIVER TONGUE. 475 instances the scliistosity may be seen to meet the bedding at a very acute angle. In this case the pebbles are usually arranged with their longer axes parallel to the schistosity, though there are always present a large number that lie parallel to the bedding planes. In the least schistose phases of the rocks the pebbles are nearly round and the matrix possesses a well-defined fragmental texture, but in those beds in which the schistosity is more pronounced the matrix is seiicitic and the pebbles are lenticular. The most completely schistose jihases resemble augen-gneisses. In these the matrix is an almost typical sericite-schist. The quartz pebbles have been crushed and flattened into long narrow stringers or plates of quartz, some of which are continuous for long distances (6 or 7 inches), while others are broken into separate parts, which when rounded on their edges yield quartz lerses like the "augen" of so many augen-gneisses.-' The nonconglomeratic beds interstratified with the conglomerates are usually more completely schistose than the latter. The least schistose beds are arkoses. These often show ripple marking and current bedding. As the schistosity increases, the quantity of sericite present also increases, until in the most highly schistose phases sericite-schists result. Some of the arkoses, as well as some of the finer-grained conglomerates, in addition to being schistose, are also foliated — i. e., they are built up of plates or leaves, along the planes between which they split very easily. When this is the case, the cleavage surfaces are covered by small scales of silvery mica. The foliation is so pronounced in many cases that the rocks are almost fissile. Besides these rocks there are present near the dam great ledges of coarse and fine grained greenstone (see PL LII), whose relations to the sedimentary beds at first glance appear to be those of interleaved flows. Upon close inspection some of these masses disclose intrusive features. Although they almost invariably follow the bedding of the fragmental rocks, some of the greenstones can be seen to cut across the layers in such a manner as to leave no doubt of their intrusive character. ' The best examples of these extremely schistose conglomerates are not found in the exposures referred to above, but they are well developed along the line between sees. 11 and 12, T. 42 N., K. 29 W. Here the width of the series is but one-fourth mile, whereas the total width of these rocks and tbeir associated greenstones near the dam measures a full mile. 476 THE CEYSTAL FALLS IRON-BEARING DISTRICT. On the old road leading to the dam the conglomerates and arkoses are intruded by an altered diabase in a most complex way. To the north of the road is the great mass of the greenstone, within which are considerable areas of the conglomerate. Within the belt of conglomerate, on the other hand, are several bands of the eruptive rock which roughly follow the bed- ding of the sedimentary one, but which cut across it in a minor way. At the contact of the main mass of greenstone and the conglomerate are numerous interlaminations of the two rocks, the greenstone having intruded the conglomerate along- its bedding planes. At one place a dozen' alternations of the two were noted within a foot. Moreover, for some distance from the greenstone the conglomerate appears to be impregnated with material from the intrusive, so that it has taken on a greenish tinge. This impregnation in one instance has gone on so far as to produce what is apparently a greenstone matrix containing separate pebbles from the con- glomerate, the groundmass of the conglomerate having apparently been absorbed. The greenstone adjacent to the conglomerate is traversed by narrow pegmatite veins in various directions, some of the lai'gest being not more than 2 inches in width. There is no evidence of a granitic intrusion, the pegmatites appearing clearly to be the result of an interaction between the basic igneous rock and the more acid fragmental one. At one place along the contact there is a belt of very coarse hornblendic material that is cut through and tlirough by the pegmatite veins. East and west of the dam for some distance are other ledges of con- glomerate. They, however, as a rule, present no features different from those exhibited by the great ledge described above. In all, especially in those occurring in sees. 9 and 10, T. 42 N., R. 28 W., the interbanding of conglomeratic and nonconglomeratic layers is beautifully shown. Near the north quarter post of sec. 11, T. 42 N., R. 28 W., the arkoses have a purple rather than a pink tinge. On cross fractures they are seen to be spangled with glistening black needles and plates of hornblende, which lie with their long axes in all azimuths. The little crystals appear to be more abundant in some layers than in others. The best exposures of quartzite are found near the north quarter post of sec. 11, T. 42 N., R. 28 W., and at 1,300 paces W., 150 N., of the south- east corner of sec. 7, T. 42 N., R. 29 W. The rocks are black. They occur in beds varying in thickness from a few inches to several feet. ALGONKIAI^ EOCKS OF STURGEON EIVER TONGUE. 477 PETKOGRAPHICAL DBSCEIPTIONS. As might naturiilly be expected, the least schistose of the arkoses and conglomerates exhil^it the fewest evidences of alteration in the thin section. In addition to the pebbles in the conglomerates, these rocks consist of ronnded and angular grains of quartz, microcline, orthoclase, and of various plagioclases, and a few of microperthite, embedded in a finer-grained aggregate of the same minerals, tiny flakes of gi-een biotite and of color- less mnscovite or sericite, a few plates of chlorite, particles and crystals of magnetite, and little nests and isolated grains of epidote, with occasionally some calcite. Many of the feldspar grains are altered into sericitic products, colored red by small particles of various iron oxides and by red earthy substances. The composition and microstructure of the schistose arkoses and of the schistose matrices of the conglomerates vary greatly in different speci- mens, being determined largely by the original composition of the different beds and the amount of squeezing to which they have been subjected. No attempt will be made here to describe in detail all the changes suffered by these rocks; a simple statement of the tendency of these changes will be given. The quartz pebbles in the moderately schistose conglomerates show plainly that they have been under great stresses. The smaller ones all exhibit undulatory extinction. The larger ones are sometimes peripherally granulated, and sometimes etched or con-oded on their edges, as though they had suffered partial solution. By this process small portions of the original particles have been separated from them, and the dissolved silica has been redeposited among the grains of the surrounding matrix as sec- ondary quartz. In their interiors many of the larger pebbles have been changed to a mosaic of differently oriented parts, which interlock so per- fectly that they appear to have crystallized together. The groundmass in which the pebbles lie is, in a few cases, a frag- mental aggregate of quartz and several feldspars, with the addition of seri- cite and other crystallized components. In most cases, and in all in which schistosity is marked, no fragmental structure is noticeable. The ground- mass is an interlocking mosaic of fairly large quartz grains that appear to have crystallized in situ, between which are smaller grains of the same 478 THE CRYSTAL FALLS lEON-BBARING DISTRICT. character, large and small spicules and plates of sericite, crystals of magne- tite, and a few needles of chlorite and other secondary substances. Between these, again, is often a cement of what seems to be secondary quartz. The schistosity of the specimens is due to the arrangement of the sericite in approximately parallel positions, and to the elongation of the quartz grains in the same direction. The pink color of the rocks is produced by red earthy substances in the feldspars and in their decomposition products. In the most schistose phases of the conglomerates the quartz pebbles have been mashed into plates, several of which join, end to end, forming sheets, which in the thin section appear as long narrow lines of variously oriented quartz grains, each of which is crossed by strain shadows. The larger quartz grains in their matrices are broken into parts, and these parts are differently oriented with respect to one another. Other grains seem to have entirely recrystallized, for they are now made up exclu- sively of the same kind of interlocking quartzes as are present in the fine portions of the groundmass in which the coarse quartz grains are embedded. In the groundmass of these rocks sericite is very abundant, and feldspar is rare. From the proportions of these minerals present it would appear that the former has been derived largely from the latter. Biotite is also present in many specimens as small green flakes, but this mineral is not widely spread. The conclusion from the study of the thin sections of the schistose conglomerates is that there has been a crystallization of new substances, principally quartz, sericite, biotite, and magnetite, from the materials of the original granitic sediments. Perhaps a portion of the crystaUization was the result of alteration of the original components before squeezing took place. The larger portion, however, was accomplished under the influence of pressure. The result of the mashing and recrystallization is a schist, which between crossed nicols has the aspect of a typical crystalline schist, but which in natural light exhibits its conglomeratic nature in the presence of the large quartz lenses, with the outlines of flattened pebbles, in a fine- grained groundmass. The pink arkoses differ from the conglomerates simply in the absence from them of the pebbles. The schistose varieties are similar in every respect to the schistose groundmass of the squeezed conglomerates. Both in the hand specimen and in the thin section the schistose arkoses exhibit striking resemblances to jnuscovitic gneisses. ALGONKIAN EOCKS OF STUKGEON ElVEK TONGUE. 479 The purple arkoses differ from the pink ones just described in contain- ing chlorite and hornblende, and in addition some apparently newly formed feldspar, notably a feldspar with the microcline twinning. As a rule, these rocks are more feldspathic than the matrices of the conglomerates, and they contain much less quartz. The larger grains of both quartz and feldspar are corroded as if partially dissolved. They have lost their smooth, rounded contours of sand grains, and now possess irregular jagged ones, which, however, are not due to secondary enlargements. The characteristic components of these rocks are the chlorite and the hornblende. The former mineral is present in plates intermingled with grains of epidote, while the hornblende is in dark-green or light-green plates, and in acicular or columnar crystals that are idiomorphic in cross section. The crystals are distributed indiscriminately through the rocks, with their longer axes lying in all azimuths. They were evidently formed after the squeezing that made the rock schistose. The plates, moreover, include within themselves such great numbers of the other components of the rock that their parts often appear to be independent. Under crossed nicols, however, many of these apparently independent plates are discov- ered to polarize together. No evidence is present in any of the sections as to the source of the material that gave rise to the hornblende. The fact, however, that all of the horublendic rocks are banded, that some layers are rich in amphibole while others are completely devoid of this mineral, sug- gests the notion that the hornblendic schistose arkoses consist partly of sedimentary and partly of tuffaceous materials. As we shall see later, this origin is ascribed with more confidence to some very peculiar rocks to be discussed later. Crushing effects are noticed in some of the hornblendic arkoses, but their present condition appears to be due more to chemical changes pro- duced in them than by mechanical action. The chemical changes were no doubt superinduced by the mashing, but this can only be inferred from the fact that they are more pronounced in the schistose phases of the rocks than in those phases in which the schistosity is poorly developed. THE DOLOMITE FORMATION. The dolomite formation comprises, as has been stated, both dolomitic limestones and calcareous slates, and occasionally quartzites, sandstones, 480 THE CEYSTAL FALLS IRON-BEAEING DISTRICT. and conglomeratic and brecciated beds. As a rule, exposures are small and scattered. Their distiibution lias already been described. All ledges observed may be seen by reference to the map (PI. LI). IMPORTANT EXPOSURES. Good exposures of the dolomites occur in the NW. J sec. 6, T. 42 N., R. 28 W. The ledge nearest the northwest corner of the section is a hard flesh-colored dolomitic marble, containing here and there little quartz grains. This is cut by joints', and is traversed by small chert bands. The bedding is more or less contorted, but its general strike is N. 45° E., and its dip is 45° NW. About one-fourth mile east of this ledge is a small, bare knoll, composed of interlaminated pink marbles, conglomerates, red sand- stones, and red slates, varying in thickness from a few inches to a foot or more. The conglomerate consists of marble pebbles and slate and chert fragments in a calcareous quartzitic matrix. The strikes and dips are uni- foi-m throughout the ledge, the former being nearly east and west and the latter 45° S. The difference in dip of the beds of these two exposures indicates plainly the presence in this place of a little westward-pitching anticline. Other prominent exposures of the dolomite series are in the NW. | sec. 1, T. 42 N., R. 29 W., and in the SE. 4 sec. 35, T. 43 N., R. 29 W. In the first-named locality is a high, bare knob, and a cluster of small ledges, in which dolomites, conglomerates, and slates are all well exposed. The dolomites, for the greater part, are massive pink marbles crossed by joint planes. In places the rocks take on a greenish-yellow tinge, and become schistose. At 1,500 paces N., 1,930 W., of the southeast corner of sec. 1, T. 42 N., R. 29 W., the dolomite forms a well-defined bed, striking N. 45° E. and dipping 70° SE. Above this, to the southeast, is a bed of coarse-grained granitic sandstone or quartzite, which in turn is over-lain by beds of gray quartzite alternating with thin slates and fine-grained conglomerates. Farther south is a ridge of well-bedded, fine-grained quartzite and bluish-gray slate, the individual layers being usually less than one-half inch in thickness. This rock grades into a gray schistose dolomite, and the whole quartzite-slate series strikes N. 75° E. and dips 63° S. The exposures in section 35 are almost pure marbles, in which no traces of bedding have been detected. ALGONKIAN BOOKS OF STURGEON EIVER TONGCTB. 481 I'ETliOGUAPHlCAL DESCUIPTION. Ill thill section tlie iniirl:)les appear as very close-gTaiued aggregates of calcite aud dolomite, usually untwinned, but occasionally twinned in the ordinary manner of these minerals. Here and there among the car- bonates are rounded quartz grains, but the greater portion of this min- eral appears to have crystallized in situ between the calcite and dolomite individuals. All the marbles are of the same general character. They diifer only in the quantity of silica present and in the presence or absence of the tiny dust grains producing the color. The schistose varieties owe their schistos- ity to the elongation of their components. The quartzites and slates interbedded with the marbles possess no unusual characters. They are similar to the corresjjonding rocks inter- stratified with the Marquette dolomites. The conglomerates interstratified with the dolomites, slates, and quartzites are of two kinds. One is com- posed of marble and slate fragments cemented by quartzite, and the other of small granite pebbles embedded in granite sand. The latter are evi- dently composed of the detritus of the granites underlying the dolomite series, while the marble-bearing conglomerates, or perhaps more ^jroperly breccias, are interformational beds conformable with the beds below them, and also with those above. They are similar in every respect to the inter- bedded breccias in the Kona dolomites on the Marquette range. SLATES AND SANDSTONES ON THE STURGEON RIVER. The rocks in the SW. ^ sec. 34, where the road to Sagola crosses the Sturgeon River, are placed in the dolomite formation, although they differ somewhat from that portion of the series described. These rocks are white calcareous sandstones, that look very much like the Potsdam sandstone where it overlies limestones, and a light-green slate, which near joint planes and other cracks has a light purple color. According to Dr. J. M. Clements, who visited the spot, the slate overlies the sandstone. "The river," he writes in his notebook, "gives a section through these rocks, and makes the strike seem to be N. 35° W., dip 50° N. It appears to me, how- ever, that the true strike is about N. 85° E., and dip 40° S." If these rocks belong to the marble series, they constitute its upper part. The slate closely resembles some of the slates in tlie Kona dolomite formation of the MON XXX.VI 31 482 THE CRYSTAL FALLS lEON-BBAEING DISTRICT. Marquette range. It is a very fine grained rock composed of very small splinters of quartz, flakes of sericite, and a few of chlorite. THE IGNEOUS ROCKS. The igneous rocks associated with the sedimentary beds in the Sturgeon River tongue are all greenstones in composition. Many of them are unques- tionably intrusive; a few ma}' be tuffaceous. The intrusive greenstones do not differ essentially from those cutting the Basement Complex. Some of them are in the form of small bosses. Others are clearly dikes, though for the most part these dikes follow the bedding of the sedimentary rocks. Still others may be intrusive sheets. The rocks regarded as possibly tufPaceous are distinctly banded. Some are made up of alternate bands of dark and light shades. The darker bands consist principally of a schistose greenstone, and the lighter ones principally of arkose or granitic sandstone. These rocks are well bedded, apparently constituting a definite portion of the conglomerate series near its lower horizon.^ THE INTRUSIVE GREENSTONES. The intrusive greenstones are usually fairly massive rocks, with a dark bluish-green color and a moderately tine grained texture. On their edges they often pass into schistose phases, presenting the structure and appear- ance of chlorite-schists. A very typical schist of this character occurs on the southern edge of the great greenstone mass 1,525 to 1,600 paces north and 300 to 400 west of the southeast corner of sec. 18, T. 42 N., R. 28 W. In the hand specimen the rock appears to be a well-characterized chlorite- schist, spangled with plates of a light-colored muscovite measuring 1.5 to 2 mm. in diameter. The intrusive character of some of the greenstones is clearly shown by the fact that they occur immediately on the strike of the conglomerate bands, and often cutting across them, as is the case at 300 paces east of the north- west corner of sec. 17, T. 42 N., R. 28 W. (see PL LII), and at 400 paces south, 100 west, of this same corner. PETROGRAPHICAL DESCRIPTION. The greenstones intrusive in the Algonkian sediments are not essen- tially different from those cutting the members of the Basement Complex. ' See Van Rise's Notebook 184, pp. 21-23. IGNEOUS BOOKS OF STUKGEON EIVER TONGUE. 483 They differ from the latter in containing, as a rule, less quartz and a very nmch greater abundance of epidote. The epidote is all secondary, as is also the quartz, so that the only noticeable difference between the two sets of greenstones is dependent upon differences in the nature of their alteration, which in turn are probably the results of differences in environment. Both sets of greenstones have been squeezed, but those in the Basement Complex are associated with crystalline schists, while those in the Algonkian series are associated with fragmental beds. In addition to hornblende, plagioclase, epidote, and a little quartz, almost all the later greenstones contain biotite, small crystals of magnetite, and irregular grains of ilmenite or of a titaniferous magnetite. Their structure is schistose through the arrangement of the larger hornblendes and biotites and the elongation of the feldspar grains in approximately parallel directions. As a rule, their thin sections present no unusual features. They all show dirty green hornblende plates, greenish-brown biotite flakes, magnetite crystals, etc., embedded in a mass of irregular grains of decom- posed plagioclase, the principal decomposition product of the feldspar being in almost all cases epidote. Often the proportion of epidote present is very great. It occurs as colorless crystals and grains scattered through the hornblende, and as light- yellow plates and grains embedded in the mass of altered plagioclase. In the rock at 500 paces east, 125 north, of the southwest corner of sec. 8, T. 42 N., R. 28 W. (PI. LII), the replacement of the plagioclase by epidote has pro- ceeded so far that no trace of the feldspar can be discovered. In the hand specimen the rock is seen to be a massive mixture of black ghstening horn- blende crystals in a yellowish-green groundmass possessing a sugary texture. In the thin section the hornblende is present as bluish-green plates that are often idiomorphic in cross section. The groundmass in which they lie is composed of epidote and quartz. The epidote is in large yellowish-green irregularly-outlined plates, including particles of magnetite and small rounded quartz grains. Most of the quartz is in isolated grains between the epidote plates and in little nests of interlocking grains. Small magnetite granules are scattered everywhere throughout the section, through all of the components indiscriminately The coarser greenstones show plainly in the hand specimen the ophitic structure, even where the rocks are schistose. In the section this structure 484 THE GEYSTAL FALLS lEON-BEARING DISTEICT. is often obscured by the abundance of decomposition products. Under low powers of the microscope, however, it can nearly always be detected. In a few of the finer-grained varieties, phenocrysts of plagioclase are occasion- ally met with. They are clouded by inclusions of biotite flakes and shreds of hornblende and by tiny ^^articles of a kaolinitic or sericitic mineral. From their composition and structure, it is clearly evident that the intrusive greenstones, whether massive or schistose, are altered phases of diabase or of diabase-porphyrite. The dark-green chlorite-schist referred to as occurring in the edge of one of the greenstone masses is a chloritic biotite-schist spangled with large flakes of a light-colored mica. The rock consists of biotite, chlorite, musco- vite, quartz, and rutile. The liiotite is in broad thin plates, arranged approximately parallel, and embedded in a mass of chlorite, the greater portion of which is a greenish-brown variety that looks as though it may have been derived from hornblende. A smaller ^sortion of the chlorite is in light-green plates, like the chlorite so frequently found in chloidte-schist. The quartz is in small rounded grains exhibiting strain shadows, scattered here and there through the chlorite and between the biotite plates. It is much more abundant in soiiie portions of the rock than in others, forming bands rich in quartz, between others in which very little of this mineral is present. The rutile is in large quantity. It constitutes large greenish- yellow grains. Some of these are rounded forms, others are prismatic crystals measuring 0.08 mm. to 0.12 mm. in length, while still others are clearly defined elbow twins. They occur everywhere throughout the slide, but are rare in the quartz. They are most abundant in the chlorite and in the large plates of light-colored mica that have been mentioned as character- istic features of the hand specimens. These have all the properties of mus- covite. They lie indiscriminately among the other com2Donents, irrespective of the schistosity of the rock, and contain very few inclusions, with the exception of the rutile grains. The lines of biotite, to the arrangement of which the rock owes its schistosity, do not bend around the muscovite as they do around the eyes in an augen-gneiss, but they continue their courses up to the edge of the muscovite grain, and there abruptly stop. From these facts it is clear that the muscovites have originated since the rock containing them was rendered schistose. As in the case of many other secondary minerals, it appears that these were produced from the components of the IGNEOUS ROCKa OF STURGEON RIVER TONGUE. 485 schist by a process which resuhed in the absorption of all of them except rutile. The process may have been connected with contact action, but no evidence in favor of this supposition has been obtained. There are a few other types of greenstone occasionally met with among the dike and other intrusive forms of the district, but they do not diifer in any marked degree from those described, except that some are quite schistose. One or two of these contain oval aggregates of epidote, plagioclase, and quartz, that may represent inclusions of foreign rocks. They are now, however, so much altered that it is difficult to determine their character with any degree of certainty. The rock of one or two other exposures in the area underlain mainly by the conglomerates deserves mention before the banded greenstones are discussed. The rock referred to is a heavy, lustrous, black schist that resembles in many respects a hornblende-schist. On fresh fractures across the schistosity parallel lines, darker than the main mass of the rock, may be easily detected. These are the edges of cleavage planes, whose surfaces are coated with brassy yellow mica plates. In thin section these rocks differ very little from the schistose greenstones referred to above. They consist of a heterogeneous schistose mass of green hornblende, cloudy plagioclase, quartz, epidote, chlorite, and magnetite. Biotite flakes are met with occasionally, but they are by no means common, except on the cleav- age surfaces. Rocks of this class have not only been made schistose by squeezing, but they have also suffered shearing along what are now the cleavage planes. They are almost identical in microscopic and macroscopic features with the hornblende-schists in the Basement Complex. THE BANDED GREENSTONES. Distincth' banded rocks, composed partly of basic material with the composition of greenstone, form a well-defined hillock in sec. 17, T. 42 N., R. 28 W., about 250 paces north of the west quarter post of this section, and a group of outcrops on the east bank of the Sturgeon River, imme- diately west of this point. The rocks in question are banded in mediumly coarse-grained dark bands, containing large quantities of green hornblende, and in fine-grained lighter ones, that resemble in the hand specimen bluish-black quartzites or cherts. In some bands there are large lenticules of white quartz, that show 486 THE CRYSTAL FALLS IRON-BEARING DISTRICT. plainly on weathered surfaces, like the flattened pebbles in a squeezed con- glomerate or the drawn-out parts of quai'tzose layers in a mashed bedded rock. These bands, though not very well defined, run continuously for long distances, and strike and dip conformably with the conglomerate beds exposed 200 paces to the north. PETROGRAPIIICAL DESCRIPTION. In the thin section the lighter-colored layers of these rocks are seen to be composed of very irregularly outlined and rounded quartz grains, cemented by a mass of finer quartzes and small grains of zoisite, little clumps of chlorite, some decomposed feldspar, and particles of magnetite. Occasionally a plate of yellowish epidote occurs in the midst of this aggre- gate, and scattered here and there through it are large plates of green horn- blende with the cellular structure so common to secondary minerals. These hornblendes lie irregularly in the slide, and include grains of all the other components. The quartz grains are small and are independently oriented, but frequently little groups of them, with the outlines of sand grains, are met with. There is Httle evidence of schistosity in these layers, but they exhibit a banding produced by the alternation of coarser and finer constit- uents. In the darker layers the proportion of hornblende is much greater than it is in the hghter ones. Indeed, some bands consist almost exclu- sively of large cellular plates and radial aggregates of plates of this mineral, only the small interstitial spaces between the large amphiboles being filled with an aggregate of quartz-zoisite, small hornblende needles, and magnet- ite. In some sections biotite is also present. It occurs most abundantly in the quartz-zoisite aggregate, filling the interstitial spaces between the amphiboles, but is present also as inclusions in this latter mineral. Some of the biotite in the hornblende appears to grade into its host, and certain portions of the amphibole possesses the brown color of the mica, with the optical properties of the hornblende. The large amphiboles are evidently the youngest components in the rocks, though they were plainly produced before the schistosity. In those layers in which the schistosity is strongly marked this structure is produced mainly by the parallel arrangement of the biotite and the small amphibole needles and plates in the quartzose aggregate. The larger cellular hornblendes lie across the schistose planes, and when they do so, the lines of biotite and of small amphiboles pass IGNEOUS ROCKS OF STUEGEON EIVER TONGUE. 487 around tlieiu exactly as tliey would do were the large hornblendes present before the rock was squeezed. Sometimes the amphibole masses that form. so large a proportion of the schistose bands are single crystals, sometimes they are fragments of crystals, and at other times they are groups of radiating crystals. The magnetite is very much more abundant in the hornblendes than in the surrounding quartz aggregate, sometimes being confined exclusively to this mineral, as though it were one of the products (the hornblende being the other) resulting from the decomposition of some original constituent, probably augite. Little particles of hematite, on the otlier hand, are abundantly disseminated through the quartzose aggregate, and are practically absent from the hornblende. Much of it appears to have been derived from magnetite. The evidence derived from the microscopic study of sections of these banded rocks, so far as it relates to their crigin, is disappointing. The quartzose layers are, in all pi'obability, sedimentary. The hornblendic layers, however, differ from these so much in composition that their material must have had a different source. It is possible that the quartzose layers represent sediments derived from the granitic portions of the Basement Comjjlex, while the hornblendic layers represent sediments derived from the basic portions of the Basement Complex; or, it may be that the acid layers have the origin ascribed to them, while the basic ones are mixed sediments and basic tuffs. The sections of the dark layers of these rocks resemble so strongly the sections of the basic laj^ers in the Clarksburg, series of mixed tuffs and sediments in the Marquette district that the writer is inclined to regard the rocks as composed partly of tuffaceous material. On the other hand, the banded rocks occur so close to the boundary between the sedimentary area and the Basement Complex, which near this boundary is composed mainly of basic schists, that it would seem but natural that they should contain large qiiantities of basic material derived from these schists. The original structure of the layers has been so completely destroyed by mashing that it can not give any evidence as to the nature of the beds. We are therefore compelled to rely entirely upon their com- position to aid us in discovering their origin. This indicates simply that much of their material was derived either from volcanic ashes or from the debris washed from the basic portions of the Basement Complex. INDEX A. Page. Aa structure, sketch of 120, 121, 122, 123. 124 (See Ellipsoidal structure.) Aci Castello, basalt from 121 Aci Trezza, basalt from 121 Acid intrusives, described 45-49, 190-198, 426 age of 49 distribution of 190 in A rchean 45-49 in Felch Mountain Kange 426 in Hemlock formation 77 in "Upper Huronian 164 Acid lavas, of Hemlock formation, described 80-94 banding of 91,92,93,94 micropegmatitic texture in 89 pressure effects in 87, 88 schistose, described 87-94 Acid py roclastics, described 94, 95 Acid volcanics of Hemlock formation, described 80-95 Actiuolite from hornblende 215 of adinole .' 209 of cblorite-schist 442 of metabasalt 105, 133 of picrite-porphyry 214, 215 of py roclastics = 147 of slate 205,209 of spilosite 206 orientation of 105 (See Amx^hibole.) Actinolite- schist from gray wacke 57 of pyroclastics 147 Adaraello tonalite 230 Adams, F. D., on analysis of slates and granites 58 on pyroxene zone about olivine 256 Adinole, analyses of 208 compared with analyses of clay slate and spilo- site 210 in Mansfield formation 64 described 208-209 Ajibik quartzite 451 correlation of xxv, xxvi relations to Groveland formation xxi,449,456 Albitefrom feldspar 151,201 of metabasalt 99 of spilosite, plate of 302 Algonkian, contact "veith Archean, effect on topog- raphy 386 deposition of 456, 457 distribution of 331,427 folding of 427,428 of Felch Mountain Range, distribution of 384 succession in 384-385 structure of 384-385 intrusives in, described 426 Page. Algonkian of Marquette district, distribution of 452-453 of Sturgeon River tongue, described 458-437 comparison with Algonkian of Felch Moun- tain tongue 462 folding of 471-t72 igneous rocks of 482-487 pressure effects iu 471 relations to Archean 461-463 relations to Lower Marquette 462 relations to Archean xvil, 331, 399, 427, 458 relations to drainage 334, 335 relations to Paleozoic formations 331, 383 relations to quartz porphyry 439 (See Huronian, Upper Huronian, Lower Hu- ronian.) Allanite in acid lavas 89 included in epidote-zoisite 444 Allen, Andrews, referred to 22 Alteration of audesine ,. 224 of basic volcanics 152 of biotite 43, 393 of bronzite, plate of 306 of calcite 146 of diabase 469 of ellipsoidal basalt 292 of feldspar 42, 170, 171, 192, 201. 224, 228, 478 plate of 288 of glauconite 422 of grit 168-169 of hornblende. 234,235,237 of metabasalt 1 17 described 126-135 of picrite-porphyry 213 of slate 14 of tutfs 141 Amasa (town) 12,143,162 Amasa area, ore deposits of 177 succession in 177-178 (See Hemlock mine.) American Black Slate Co., analysis of slate from 61 Araphibole of greenstone 486, 487 of metabasalt 127 of mica-schist - 415 of picrite-porphyry 214 orientation of 127,214,405,486 (See Actinolite, Hornblende, TremoUte.) Amphibole-peridotite, described 253-254 crystallization of 257 gradation to olivine-gabbro 254-260 gradation to wehrlite 254-260 Amphibole-schist of Sturgeon River Archean, de- scribed 465-467 Amphiboliie, analysis of 397 489 490 INDEX. Page. Arapliibolite of Arcliean, described 395-397 from basic volcaiiics 152 intruding tnica-scbist 392 Aniygdaloidal texture in metabasalt. . . . 102, 113, 122, 123. 442 described 124^126 plate of 280,282,284,290 cause of distribution of, in basic lavas 95 in eruptive breccia 136 in Hemlock scbist - ■^^S in pyroelastics 138, 14C Amygdules of calcite, plate of 282 of clilorite, plate of 284 of feldspar, plate of 284 of quartz, plate of 284 order of deposition of 125 pressure eftects in 126, 128 Analyses of adinoles 208 of iimpbibolite ^ 397 of clay slate 59,61 of clay slate, spilosite, and adinole, comparison of 21 of dolomite 409,435 of ;:;neiss 391 of granite 389 of bornblende 242 of hornblendegabbro 263-264 of iron ore, by Brooks, referred to 19 of iron ore 181 of Mansfield mine 69 of Mansfield slate 59,61 comparison witb contact products 209-211 of metabasalt 103,106.107 of mica-diorite 231,263-264 of niica-scbist ^9^ ofnorite 245,263,264 ofperidotite 259,263,264 of picrite-porphyry 219 of spilosite 207 Auatase included in bornblende 236 of gabbro and norite - 236 of metabasalt 129 {See Octabedrite.) ■ Andesine, alteration of 224 altering tomuscovite 224 altering to epidote-zoisite 224 of diorite 224 of gabbro and norite 233 of metabasalt 104 twinnina; of 104 Andesite, of Hemlock formation 107 Anticline determined by magnetic observations- 366, 372, 373 described - 370-371 Antoine dolomite of Menominee district corre- lated XXV, XXVI Apatite included in biotite 217,234,239 in feld spar 234 in quartz 194,419,420 in spbene 244 of acid lavas 91 of biotite -granite 192 of gabbro and uorite 239 of metabasalt 99 of periodite 252 of picrite-porphyry 216 of outlining feld spra- crystals 239 Apbanitic texture in metabasalt 98, 99, 212 Apobasalt, use of term 96,98 Page. 274 87 276 276 276 Aporbyolite "witb perlitic parting, plate of Aporhyolite-porpby ry described breccia, plate of perlitic parting of. plate of pressure effects in, plate of use of term 80 Aragon mine 69 Arcbean, described xviri, 38-49, 385-397, 428-430, 463^71 acid dikes in, described 45-49 age of 39 basic dikes in, described 46-49 biotite-granite of, described 40-43 iUkes in, described 45-49 distribution of 38, 39, 331, 427 ellipse 26, 38, 333, 427 erosion of xxiv, 39 folding of xxiii intrusives in xviu, 38, 45-49 metamorpbism of x viii, xxiii of Crystal Falls district correlated witb Arcbean of Marquette district xxv, xxvi of Felcb Mountain range, described 385-397 distribution of 385 topography of 386, 387 of Marquette district correlated T,7itb Archean of Menominee district xxv, xxvi correlated witb Archean of Sturgeon River tongue 470-471 distribution of 452-453 of Micbigamme Mountain and Fence River areas, described 428-430 of Sturgeon River tongue, described 463-471 comparison with Archean of Marquette dis- trict 470-471 relations to Algonkian 459, 460, 461-463 origin of 39 relations to Algonkian xvii, 331, 378, 379, 386, 399, 427, 458, 460. 461-463 relations to Cambrian 26, 331 relations to drainage 334,335 relations to Groveland formation 416,424 relations to Huronian. (See Relations to Algon- kian.) relations to Lo^er Huronian xix, 55 relations to Mansfield formation 424 relation.s to overlying formations 39 relations to Paleozoic rocks 26, 331 relations to quartz-porphyry 429 relations to Randville dolomite xix, 51. 53, 55, 407 relations to Silurian rocks 26 relations to Sturgeon quartzite xis, 398, 401 relations to Upper Huronian xxii topography of 38,39,333 Arenaceous slate group of Rominger 164 Argentine Republic, gabbros of 255, 256 Arkose of Sturgeon River conglomerate formation described 478-479 pressure etfects in 479 Armenia mine, description of ore body at 183 location of, 178; op 186 table of shipments from, op 186 Ash beds described 142-143 Azoic system 16, 17, 375 of Sturgeon Rivertongne 460 Augite altering to hornblende 100 altering to pilite 211 INDEX. 491 Augito altering to uralito , lOi, 2U1, 212 cU'avajroof 237 I'ryatallizatiou of 257 of amphiboU'-peridotito 255 of gabhroautl norito 237 of horubleiide-gabbro 243 of inetabaaalt 100,104,211 of iiietiidolerite 200,201 of peridotite 250 of tuffs 138 iu dolomite 436 included iu hornblende 237, 250, 255, 257 iuc'lnding biotito 250 including hornblende 250,255 including hyperathene 255 zonal intergrowth with hornblende, plate of 320 zonal intergrowth with olivine and hornblende. 255-260 Augite-andesite of Heinloct formation 108 Ausweichungs-clivage. (See Pyroxene) 440 B. Bad Water village 374 Balsam village 143 Banding of acid lavas 87,88,01,92 described 93,94 of acid pyroclastics 95 of Archean granite 49 of ash bed 143 of biotite and chlorite 225 of biotite-granite 43-44 of biotite-schist 467,468 of Bone Lake crystalline schist 149 of chert of Mansfield mine 69 of gneiss 390 of granite 45,49 of greenstone 485-487 of Groveland formation 417-416 of hornblende-schist 465, 40G of Mansfield schist 413 of mica-gneiss 197 of quartzite 401 of Handville dolomite 409 of rhyolite-porphyry ' 81 of spilosite-desraosite, plate of ,. 306 of Sturgeon formation 431 of tuff 138 of volcanic conglomerate 144 Barrois, Charles, referred to 44 Basalt. (See Metabasalt.) Bascora, F., on devjtritied lavas 80, 87, 96 Base level. (See Peneplain.) Basement Complex of Stnrgeon Kiver tongue de- scribed (see Archean) 463-471 Basic intrusives described 198-212 age of 48, 49 correlation of 189 effect on topography, sketch of 46 in A rchean 49 in Felch Mountain range 446 in Hemlock formation 77 described 204 in Mansfield slate described 203-204 in other intrusives described 204 in Upper Huronian 164,211 described 204 metamorphism of Mansfield slate described 204-211 Page. Basic intrusives, schistosity of 47-48 transfer of material to slate 211 Basic lavas described 95-135 Basic volcanics, alteration of 152 described 95-148 Ba83ett,V.H., indebtedness to 106,264 Bayley, "W. S., indebtedness to 12 on Clarksburg formation 132 on metamorphism of basic volcanics 152 on schistose pyroclastics 148 referred to xv. 22 Eebb, E. C., indebtedness to xvi. 12 Becke, F., on granodiorite 231 on relations of pyroxene and amphibole 258 on tonalite 230-231 on tonalite-porphyrite 229 referred to 1 05 Bedding of Upper Huronian 167 Benson, Vt., analysis slate from 61 Bessemer ore of Mansfield mine 68 of Mansfield slate 27 Biotite, alteration of 43, 393 altering to calcite 192,225 to chlorite- 43, 47, 192, 193, 215, 216, 225, 228, 248, 403, 425 to epidote 43 to epidote-zoisite 225 to rutile 43,192,202,225 to sagcnite 192,193 to sphene 192,225 banding with chlorite 225 crystallization of 258 from feldspar 42,82,89,90,92,150,170,192,201 included in augito 250 included iu feldspar 151, 171 included in hornblende 251, 260, 261 included in muscovite 298 included in quartz 47,171,394,403,466 included in plagioclase 193,484 including apatite 217,234,239 including epidote 225,226 including ilmenite 216 including iron oxide 252 including sagenite 403 including sphene 225, 404 including zircon 234 iutergrowu wifh muscovite 393, 414 of acid lavas 89,91 of amphibolite 296 of amphibole-peridotite 254, 258 of amphibole-schist 466 of araygdules 124 of basic dikes 47 of biotite-granite 42, 43, 192, 193 of biotite-schist 468,484 of Bone Lake crystalline schist 150 of diorite of gabbro and norite of granite , of graywacke of greenstone of Hemlock schist of metadolerite of metamorphosed Mansfield slate of mica-diorite, plate of of mica-schist of muscovite-biotite-gneiss, plate of 225 234 198 170 4S6 444 202 205 308 392. 393 298 492 INDEX. Biotite of mnscovite-'biotite- granite 193 of peridoiite 252,257,261 of picrite-porphyry 215 of pyroclastica 147 of sedimentaries developed by intrusion 195 of sedimentary inclusions in granite 197 of spilosite 206 ; orientation of 393,425,468,486 , paralle] growth, -witb, muscovite 170 | penetrating feldspar 4=14 | pressure effects in 43, 248 j relations of orientation to hornblende 486 . relations of orientation to nmscovite 484 ^ replacing feldspar - 193 ' Biotite-gneiss of Arcliean 429 Biotite-granite described 40-43, 191-193 banding of 43,44 plate of 308 micropegmatitic structure in, described 192-193 schistosity of 44 Biotite-scbist of Algonkian of Sturgeon Kiver tongue 484 of Arcbean of Sturgeon Kiver tongue 467-469 of Hemlock formation 442 Birkinbine, John, on ore shipments 186 Blaney mine. {See Hope mine.) Blocklava. (SeeEllipsoidalstructure, Aastructure ) Bog iron ore of Upper Huronian 182 Bone Lake described 35 referred to 15^ Bone Lake crystalline schists described 148-152 Bonney, T. G-.,on alteration of olivine 218 on ellipsoidal structure 118-119 Botryoidal ore 180 Brackett, 11. N., on ultrabasic intrusives 220 Brauuer, J. C , on ultrabasic intrusives 220 Breccia, eruptive, of Hemlock formation, described. 135-136 Tolcanic, use of terra 137 Breeciation of Groveland formation 418 of metabasalt 11''^ Brittany granite compared with Crystal Falls granite 44 Brogger, W. C, on diorite and gabbro families 242 on monzonito group 232 on rock analyses 105 on use of term diorite 222 Bronzite altering to serpentine 238 plate of... 306 altering to talc 238 plate of 306 included in hornblende 250 plate of 306 including ilmenite 238 including rutile 238 in zonal intergrowth with hornblende, plate of. . 318 of broDzite-norite 244 plate of 318 alteration of, plate of 306 of bronzite-norite-porphy ry 246 of gabbro and norite 238 of peridotite 250 Bronzite-norite described 244-247 plate of 318 analysis of 245 crystallization of minerals 2(32 intruding hornblende-gabbro 243, 249, 265 Bronzite -norite -porphyry 246 plate of 320 Paga Bronzite-norite-porpbyry altering to serpentine 246 intruding hornblende-gabbro 249 Brooks, A. H., referred to 22 Brooks, T. B., on composition of iron ore 181 on correlation of Menominee rocks 19 on correlation of Upper Huronian 164 on Felch Mountain range 376, 377. 378, 379 on iron-bearing rocks of Michigan 16-17 on magnetic observations 24, 337 on Menominee district 19 on Mesuard series 452 on Paint Kiver district 11 on Sturgeon Kiver tongue 460-461 on Upper Huronian 172, 173 referred to xv, 21 Brooks, Kominger, and Pumpelly, map of Upper Pen- insula of Michigan 18 Brown, E F., on analysis of iron ore 69 Brule Kiver described 31 referred to 13, 15, 161 Building stones of Hemlock formation described. . . 153, 154 Burt, William A., map of part of Upper Peninsula . 15 on Crystal Falls rocks 13 on Felch Mountain range 375 on Sturgeon Kiver tongue 460-161 referred to - 16,21 C. Calciferous limestone, relations to Potsdam 383 Calcification of chlorite 132 of feldspar 131 of metabasalt 117,130,132,133,134 Calcite, alteration of 146 altering to limonite - 153 developed by dynamic action 432 from biotite 192,225 from feldspar 82,90,131,132 from hornblende 203, 21 5 of acid lavas 89,93 of amygdules 124 plate of 282 of basic dikes 47 of biotite-granite 192 of ellipsoidal metabasalt, plate of 292 of Groveland formation 420 j of marble 481 of metabasalt 100,101,117,127.128,129,132 j plate of 290 I of metadolerite 203 j of peridotite 252 ; of py roclastics 146, 147 of tuff 142 orientation of 132 pseudomorphs after feldspar 132 replaced by iron carbonate 133 replacing chlorite 132 replacing feldspar 131 Caledonia mine. (See Mansfield mine.) Calumet and Heckla mine 399 Cambrian sandstone 29 deposition of xxiv erosion of xxiv relations to Algonkian 331,473 relations to Arcbean xxiv, 26, 331 relations to Ke weenawan 162 relations to intrusives 188 INDEX. 493 Page. Camliriaii sniulstonc, relatioiiH to Upper Hurouiaii xxiv, 155.161,162 Carbon of Mansflold slate 60 Carbonate ilevcloped by dynamic action 432 Carbonation of metabasalt 117, 130, 132, 133, 134 Carboniferous clays, analyses of 59 CataL-lastic structure in feldspar 44 in granite 194 in quartz 41, 44 Cellular texture in hornblende 486 Channing, J. Part, figure by 65 on Mansfield mine 66 referred to 22 Cbanning (town) : 175 Chert fragments in conglomerate 64 of MansfieUl formation described 62 of Upper Huronian 166 pressure efl"ect3 in 177 relations to ore deposits 182 road material 154 Cherty carbonate from organic matter 184 in conglomerate 64 Chester, F.B., (?) on gabbros 247 Chicago, Milwaukee, and St. Paul Eailway 95, 143, 175 Chicago and Northwestern Railway 156, 175, 423 Chlorite altering to epidote 132 banding with biotite 225 calcification of 132 from biotite 43, 47, 192, 193, 215, 216, 225, 248, 403, 425 from feldspar 82,99,111,131,201 from hornblende 100,203,214,215,237 included in quartz 403, 419, 420 including epidote 248 including ilmenite 146, 216 including iron ore 218 including sagenite 403 including titanite 404 of adinole 208 ! of altered slate 209 I of amygdules 124,125 i plateof 280,284 \ of arkose 479 of basic dikes 47 , of biotite-granite 192, 193 . of biotite-schist 484 ; of Bone Late schist 151 of dolomite 410 i of graywacke 170 of metabasalt 98, 99, 101, 117, 118, 127, 128, 129, 131, 132, 133, 134 of mica-schist 196 of phy Uite 440 of picrite-porphyry 213, 217, 218 of pyroclastics 145 of spilosite 206 plate of 302, 304 of spilosite-desmosite, plate of 306 of slate 205, 209 of tuff 141^ 142 of volcanic conglomerate 143,144, 145 orientation of us, 127, 133, 146 pseudomorphs after biotite 217, 228 after garnet 403 replaced by calcite 132 Chlorite-schiat from basic volcanica 152 from gray wacte 57 Page. Cblorite-scbist intrusive in .^Mgonkian of Sturgeon River 432 of Hemlock formation 442 of Upper Huronian ig6, 174 Cbloritization of metabasalt 117 Chrustscbofr, C. von, on relations of pyroxene and amjihibole 258 Claire mine, location of 178, op. 186 table of shipments from op. 186 Claire Mining Company. {See Claire mine.) Clarke, F. "W., analysis by 61 Clarksburg volcanics 132,165 comparison with banded greenstones of Sturgeon Kiver tongue 487 replacing Michigamme and Ishpeming forma- tions XXVI Clastic volcanic, plate of 284 Clay slates, analyses of 59, 61 analyses of, compared with altered clay slates .. 209,211 composition of 58,60 from granite 58 of Mansfield formation described 57-62 origin of 5g relations to ferruginous chert 63 relations to ore bodies 63 relations to siderite-slate 63 Cleavage of .7cid lavas 88-89 of augite 250 of biotite-schist 467 of Hemlock schist , 443,445 of hornblende ? ^51 of phyllite 439, 440 of pyroxene 237 Clements, J. Morgan, on ellipsoidal structure, referred to US on Hemlock volcanics 446 on Sturgeon River sandstone 481 on volcanics of Crystal Falls district 20 referred to xx vi, 437, 447 Cole, G. A. J., on ellipsoidal structure 118, 119 Columbia mine, description of ore bodies 182 location of 179, op. 186 table of shipments from op. 1 86 referred to 161 Commonwealth 385 iron ores at xxvi Compass, dial, use of 24, 341 , 342, 344 Concentration of ore in synclinal troughs (see Iron ore deposits, origin of) 183, 184 Conglomerate altering to sericite-schist 475 basalt of Upper Huronian 163 of Hemlock formation 76, 152,153 * intruded by diabase 476 of Mansfield ore deposit 63-64, 68 intruded by greenstone 475-470 of Sturgeon River xviir, xxtv, 461,462, 481 described 472-479 of Upper Huronian xxn, 166 pressure effects in 474, xviii volcanic, described 143-145 plate of 284 use of term 136 Copper, absence of, in Huronian volcanics 125 Corrigan, McKinney Sc Co. (*S'ee Crystal Falls mine.) Cortlandt series, comparison with Crystal Falls in- trusives 222 494 INDEX. Page. CoutcMchiiig 380 Credner, H., on Felch Mountain range 376 on Menominee district 377 on oritrin of iron ore 71 Cretaceous subaideDce xxiv Cross, Whitman, on metadolerite 97 Crystal Falls area, ore deposits of 178 comparison with output of Menominee mines... 186 comparison with output of region 186 discovery of ore deposits 175 Crystal Falls district, drainage of 31-36 elevations in 30, 31, 332 folding of 26 geographical limits of 25 physiography 13,29-37,329-335 relations to Marquette district 11,25,329 relations to Menominee district 11,25,329 structure and stratigraphy of 25-29 Crystal Falls mine, analysis of ore from 181 location of 178, op. 186 table of shipments of op. 186 referred to 161 Crystal Falls series, correlation with Marquette series xxv, xxvi correlation with Menominee series xxv, xxvi metamorphisra of xxiv, xxv, sxvi Crystal Falls syncline xxni, 26, 178 ' described 158-161 Cry stall ine schist of Bone Lake described 148-152 of Upper Huronian* 166, 167, 171, 172 Crystallization of minerals of basic rock described . . 257-259 of minerals of intrusive series 262 Culver, G.E., referred to 22 Current bedding in quartzite 53 D. Dakyns. J. R.,on plutonic rocks 222 Dalmer, K.. on ellipsoidal structure 118, 119 Dana. J. D., ou ellipsoidal structure... 120,121,124 sketch by 120 on metadolerite 96 on origin of volcanics 78 referred to 95 Darton, N. H., ou ultrabasic intrusive^ 219-220 referred to 95 Dathe, E., on ellipsoidal structure 118, 119 Deer River 38, 75,88, 333 described 31,32-35,334 development of 32-35 topography of valley 29 Delphic mine, location of op. 186 table of shipments from op. 186 Desmosite gradation to spilosite, plate of 306 of Mansfield foi-mation 64 described 207 De Soto Mining Co. {se(? Mansfield mine) --- 65 Devitrification of aporhyolite 87 of metabasalt 102,103,126 of tuffs 138 Dewitt, N. Y.,picrite-porphyry at 219 Diabase, alteration of 469 altering to greenstone 466-484 altering to borublende-schist 466 intruding conglomerate 476 intrudiu^ granite 429 intruding Felch Mountain series 426 Page. Diabase intruding Sturgeon River series 469 (^SeeMetadiabase.) Dial compass, use of 24,344 described 341-342 Diamond-drill work at Hemlock mine, figure of 177 in8ec.20, T.45N.,R.33 W 176 Differentiation of magma 265 Dike, associated with ore deposits at Paint River mine 183 in Monitor mine 183 in Paint River mine 183 acid, in Archean, described 46-49 basic, in Archean, described 46-49 effect on topography, sketch of 46 (See Basic dikes, Acid dikes.) Diller, J. S.,on ultrabasic intrusives 220 referred to 95 Diopside of gabbro and norite 238, 212 Diorite, comparison with granodiorite 231 crystallization of minerals of 262 intruded by diorite-porphyry 265 intruded by granite 194 intruded by hornblende-gabbro 265 of intrusive series described 222-232 use of term 22.223 Diorite-porphyry intruding diorite 265 intruding hornblende-gabbro 265 Diorite-schist associated with ore deposits 183 Diorite. See Motadiorite. Dip needle, use of 24. 344 described 342-343 Doane exploration 447, 449 Dolerite, contact with granite 192 dikes associated with ore deposits 18.1 grading into ba.salt 200 including sedimentary rocks 20:J intruded by granite 194, 204 intruding Archean 48 intruding Hemlock formation 77 intruding pyroclastics 147 intrusive, endomorphic effects of 211 metamorpbism of Mansfield slate, described 204, 211 relations to intrusives of other districts 189 relations to picrite' porphyry 212 use of term 96 {See Metadolerite, Basic intrusives.) Dolomite described 408-411, 431-437, 479-482 analyses of 409, 435 containing foreign minerals 436 metamorphisra of 432 of Michigamme Mountain and Fence River areas, described 431-437 of Randville formation, described 408-411 of Sturgeon River tongue, described 479-482 distribution of 1 - 472 relations to conglomerate -* 471 relations to Felch Mountain fragmentals . . . 472-473 relations to Lake Superior sandstone 473 orientation of 410 (See Randville dolomite.) Drainage of Crystal Falls district 31-36,334-335 Drift. (See Pleistocene.) Dunn Iron Mining Company. {See Dunn Mine.) Dunn mine, analysis of ore from 181 depth of 185 description of ore bodies 182 INDEX. 495 I'a-r. Dunn luiiif, liK-aliuii of 179, op. 180 table of shipment H from op. 186 Dynamic iictiou. {Sec Pressure effects.) Economic ])iodiict8 of Hemlock formation de- scribed 153-154 Elevations of Crystal FiiUs district 30,31,332 Elldaleii, S-i\-eaeu LUlletlinta of 92 Ellipsoidal structure m metabasalt described 112-124 figure of 112,113,114 plate o( 116,292,298 Ells, R. "\Y., on ellipsoidal structure 118. 119 Enlargement of feldspar 144 of quartz 57,85,404-405 Enstatite of gabbro and uorite 238 Epibasalt, use of term 98 Epidiorite, use of term...^ 97,222 Epidolerite. use of term 97 Epidoto from biotite 225 /rom cblorite 132 from feldspar 82,92,111.169,170,201,248,483 from hornblende 100,203 included in biotite 225,220 248 151 47 444 47 445 47 - 443 151 442 226 included ni chlorite included in feldspar included in bornblende ' included in ilraenite included in quartz including zoisite of basic dikes of biotite-scbist of Bone Lake schist of chlorite-schist of diorite of greenstone 483, 485, 486 of metabasalt 101, 102, 117, 118, 127, 134 of py roclastics 147 of spilosite 206 plate of 302 of tuff 141 of variolite Ill of volcanic conglomerate 143 Epidote-scbist from basic volcanic 152 of Hemlock formation 442 Epidote-zoisite — from feldspar 42,99,127,192,224 from bornblende 237 including allanite 444 including spbene 225 of metabasalt 99, 131, 132 of araygdules 124 zonal structure in 101 Epjdotizatiou of metabasalt 117 Eriksen, E. T ., referred to 22 Erosion of Archean xxiv, 39 of Cambrian u xxiv of Huronian sxi, xsii, xxiv of Pleistocene xsiv, 29 Eruptive breccia formed by intrusion 195 of Hemlock formation described 3 35, 136 Escanaba Piver described 335 Eutaxitic structure in metabasalt 103 Exploration for iron ore 12, 73, 330 in Mansfield slate 73 of Mansfield ore deposit 67 Page. Fairchild, C.N. .indebtedness to 331 Fairbanks, H. "W., on ellipsoidal structure 118 on gabbros 240 on ultrabasic rocks 247 Fairbanks mine. (See Lincoln mine.) False bedding in TTjipcr Huronian 168 Felch Mountain range described 374-426 elevation of 1 331 geographical position of 374 literature on 375-383 relations to Sturgeon River rocks 462, 472, 473 Feldspar, alteration of 42, 127, 170, 1 92, 201, 224, 478 plate of 288 altering to albite 151, 201 altering to biotite 42, 82, 89, 90, 92, 150, 170, 192, 201 altering to calcite 82,90, 131, 132 altering to cblorite 82, 99, 111, 131, 201 altering to epidote 92, 111, 161, 170, 201, 248. 483 altering to epidote-zoisite 42, 99, 127, 192, 224 altering to feldspar Ill, 131, 151, 169, 170, 171, 201, 248 altering to iron oxide 42, 92 altering to mica 169, 170 altering to muscovite . . . 42, 82, 90, 92, 192, 201, 224, 228, 248 plate of 286 altering to paragonite 41, 42 altering to quartz 42, 99, 52, 111, 131, 151, 169, 170, 171, 201, 248 altering to sericite 52, 89, 99, 101, 111, 127, 131, 477 altering to zoisite 201 plate of 286 cataclastic structure in 44 enlargement of 144 included in dolomite _ 436 included in bornblende 202 plate of 312 included in microcline 294 included in uralite 202 including apatite 234 including biotite 151, 171 including epidote 151 including bornbleiide 151 including iron oxide _ 151, 234 including muscovite 468 including quartz 191.468 including rutile 234 of acid lavas 90 of ampbibole-peridotite 255 of amygdulea 124 plate of . 284 of basalt, plate of 282 of biotite-granite 41,42, 191, 193 of biotite-scbist 467, 468 of Bone Lake schist 150, 151 of bronzite-norite-ijorpbyry 246 of diorite 223-225 of dolomite 436 of gabbro 234, 241, 242, 248 of granite 198, 388. 464 of graywacke 56, 169 of greenstone 470 of Groveland formation 420 of metabasalt 99, 102, 104, 127, 129, 211 of metadolerite 200, 201 of micadiorite 227 of mica-schist 414 496 INDEX. Page. Feldspar of muacovite-biotite-gneiss, plate of 298 of peridotite - 252,257,260,261 of phyllite *« of pjroclastics 146, 147 of quartz-niica-diorite-porphyry, plate of 310 cf rhyolite-porphyry 81, 82, 84, 85 pressureeffects, plate of of sedimentary inclusions in granite of slate, altered ' *•■ of spilosite -• 206 145 255 Ill 278 197 205 414 57 of tuff of ultrabaaic roots of variolite of TOlcanic conglomerate 144,145 of volcanic sand, plate of 296 orientation of - 84,90,171 outlined by apatite '39 penetrated by biotite - - - • by tourmaline penetrating hornblende, plate of 322 perthitic 41,42 pressure eflects in 42, 90, 92, 169, 254, 388, 464 plate of 276 replaced by biotite 1^3 replaced by calcite 131 twinning of 41,42,89,191,201,224,233,468 zonal jntergrowtb witb borublende 256 plate of 322 zonal intergrowtb witb pyroxene, plate of 322 zonal structure in 197,233 Fence River described 31,35,334,335 referred to XIX, 38, 39, 50, 51, 152, 329, 333, 408, 427, 430 Fence Eiver area described 427-450 168 62 63 22 417 Page. Folding of Upper Huronian, relations to intrusivea. 189-190 time of 161,188 Foliation of araphibole 395 of gneiss 430 of granite 387 of Sturgeon formation 402 Forbes, indebtedness to 331 Ford River, described 334, 335 Forsythe, E. J., analyses by 435 Foster, J. "W., and Whitney, J. D., on Felcb Mountain range 375 on Crystal Falls district - 16 map of Lake Superior land district 17 referred to 21 Fouque, F., on ellipsoidal basalt 120, 121, 124 Fox, H., on ellipsoidal structure 118 Frankenstein, wekrlite from 219 Friction breccia of Hemlock formation 136 Fumar(dc action on sedimentaries 57 Fundamental Complex. (,Sce Arcbean.) Futterer, Karl, on pressure eflects in quartz 90 on pressure effects in feldspar 91 Ferruginous cbert of Mansfield formation described. relations to clay slate Finlay, J. E., referred to Flag ore of Groveland formation Floodwood road Florence, iron ores at xxVi 175 87 278 45 244 discovery of ore deposits at Plow structure in aporhyolite-porphyry plate of in granite in bornblende-gabbro inmetabasalt 99,102,105 plate of 280 in tufl's 138 f^&m Banding.) Folding determined by magnetic observations, de- scribed 306-372 of Algonkian xxill, 163,427,428 of Sturgeon Paver 471,472,474 of Arcbean - ^-^'" of Crystal Falls district 26 of graywacke, caused by intrusion 195 of Groveland formation 448 of Hemlock formation 441 of Mansfield formation 438, 439 of Marquette district 26, 452 of Menominee district •. 26 of Eandville dolomite 432-434 of Sturgeon quartzite - 399-400 of Sturgeon Eiver series 471,472,474 of Upper Huronian described 158-162 sketch of 179 Gabbro described altering to greenstone-schist crystallization of minerals of intruded by granite intruding bornblende-gabbro pressure effects in relations to peridotite Garnet of actinolite-scbist of graywacke of Mansfield schist of mica-schist pseudomorphs after chlorite Geikie, A., on ellipsoidal structure on Tertiary basalt Geographical limits of Crystal Falls district of Felcb Mountain range Glass in metabasalt in tuff {See Devitrification.) Glauconite, alteration of of Groveland formation of Mesabi iron-bearing formation Glidden exploration Globular basalt. (&m Ellipsoidal structure in meta- basalt.) Gneiss, analyses of developed by intrusion of Arcbean xviii described recrystallizatiou of Gneissoid biotite-granite described Gneissoid granite, included in granite-porphyry, sketch of of Sturgeon Eiver tongue described pressure effects in ■ Granite altering to clay slate altering to pbyllite analyses of gradation into quartzite banding of composition of contact with dolerite 233-249 466 . 262 194 265 247, 248 261 450 167 413 415 403 114, 118 75 25 274 99 138 422 XX 422 161, 163 391 198 43 390,391 390, 391 43,44 45 463, 464 44 58 58 389 51,52 45 58 192 INDEX. 497 Page. Granite, contact -with sodimeiUaries 194-198,300 pliito of 298 ioUationof 287 iiicliuUug sedimentary fragments 195 iutnuied by diabase 429 intrudi;ig iron-bearing formation 370,381,426 intruding dolerite 204 intruding Folcli Mountain range 426 intruding gneiss 381 intruding granite 429 intruding Sturgeon formation 426 intruding Sturgeon Kiver series 459 iloribiban (Brittany) compared witli Crystal Falls granite - 44 of Arcbean xviii,38,45,49,428,429 described 40-13,387-389 of Micbigarame River 16 of Sturgeon River 459,464 of Sweden 44 porpbyritic 40,45 jjressure effects in 44, 194,464 relations to quartzite 376 relations to otber intrusives 194 (See Biotite-granite, Muscovite-biotite-granite, G-rauitite, Gneissoid granite.) Granite-porphyry, sketch of 45 Granitite 226 described 40-43 Granitic texture in diorite 223 Granodiorite, comparison with diorite of Crystal Falls 231 Granopbyric texture 85 Granular texture in hornblende-gabbro 240, 244 plate of - 316 in perlodotite 250 Granulation of feldspar 42,169 plate of 276,316 of quartz 51,90,169 Graywacke altering to actinolite-scbist 57 altering to chlorite-schist 57, 166 altering to miea-scbist 57, 166 brecciated by intrusion 195 intruded by granite 194 plate of 298 magnetitic of Upper Huronian 176 of Mansfield slate 56,57 of Sturgeon formation 431 of Upper Huronian 166, 167, 168, 169-174, 176 of sec.l9,T.46X.,R.32 W 27 pressure etfects in 170 recrystallizatiou of 195, 198 plate of 298 Great slate formation of Menominee district sxv, xxvi Great "Western mine, description of ore bodies at . . . 182 depth of 185 fispuring of rocks 185 location of 178, Op. 186 table of shipments from Op. 186 Green Bay, streams tributary to 31. 334 Greenstone 14 from diabase 484 from diabase-porpbj'rite 484 in conglomerate of Hemlock formation 442 intrusive in Algonkian of Sturgeon River tongue 482-485 intrusive in Arcbean 38 intrusive in conglomerate .' 475, 476 MON XXXVI — 32 Greenstone of Mansfield mine... topograjihy of Greenstone-schist from diabase . from gabbro Pago. 63 333 466 466 of Arcbean of Sturgeon River 465, 467 Gregory, G. W.,on ellipsoidal structure 118, 119 Grifdth Sc Nathaniel quarries, analysis of slate from. 61 Grit altering to chlorite ]68, 169 altering to muscovite 108, 169 Groveland formation described sx, 415-423, 446-450 correlation of (see Relations) xxv, xxvi distribution of 415,416,446-448 folding of 448 intruded by granite 426 magnetic observations in 447, 448 origin of 423 pressure effects in 449 quartzites, resemblance to Mesabi chert 422-423 relations to Arebean 416, 424 relations to Ajibik quartzite xxr, 449, 456 relations to Hemlock formation xx, xxi, xxvi, 447 relations to Mansfield formation 411,413,438,447 relations to Negaunee formation xx, 449, 455 relations to Eandville formation xx relations to Upper Huronian xxii, 425 thickness of xvii, 448 topogranby of 415, 416, 446-448 Groveland mine 413, 415 Grijnerite- schist of Groveland formation 418 Giimbel, , on epidiorite 97 H. Halleflinta, of Elfdalen, Sweden Hampton Village, N. T., analysis of slate from . Harriman.F. J., referred to Hartz Mountain, analysis of spilosite from Hatch, F., on limburgite . 92 61 22 207 221 Hawaii, volcanics of 75, 120 Ha-wes. G. "W., on metadolerite 96 Hedstrtim, H., on micropoikilitic texture 83-84 Hematite deposits of Upper Huronian described... 180,182 of Groveland formation 417, 418, 419, 420, 450 of Mansfield ore deposits described 69-70 of Paint River district 18 Hematite from basic volcanics 152 from siderite 168 included in biotite 252 included in quartz 419 of am y gdules 1 25 of Bone Lake schists 151 of clay slate 57 of quartzite 450 orientation of 418 Hemlock formation described xx, 73-154, 440-446 acid volcanics of, described 80-95 classification of 79, 80 correlation of xxv, xxvi distribution of 27,73,74,440-441 economic products from 153, 154 folding of 441 metamorphism of xxiv, 446 of Micbigamme Mountain and Fence River areas described 440-446 origin of 27,78 pressure effects in xxiv, 75 pyroclastics of, described 135-148 498 INDEX. Page. Hemlock formation, pryoclastics of, plate of 140 relations to lutmsiyes 77,204 relations to Mansfield formation xs, xsi, xxvi, 64, 76 relations to Groveland formation xx, xxi, xxvi, 447 ralations to Randville dolomite xx, 75 relations to Upper Hurouian 77 rtiyolite-porphyry of, described 81-87 sedimentaries of, described 152, 153 thickness of ' XVII , 27, 74, 75, 441 topograpby of 73, 74, 440, 441 Hemlock mine 157, 1 77 analysis of ore from 181 diamond drill work at, figure of 177 exploitation of ore deposits at - 175 location of Op. 186 table of shipments from Op. 186 Hemlock Kiver described 31 Hesse-Darmstadt, wehrlite from 219 Hillebrand, "W. F., analvsis by 61 Hobbs, Wm. H., on luetamorpbism of schists 130 HoUister mine, location of 178, Op. 186 Holy oke formation 20 Hoosau schists of western Massachusetts 394 Hope mine, location of 178, Op. 186 table of shipments from Op. 186 | Horizontal needle, deflection of, figure of 345 j Hornblende, alteration of 234-235 altering to actinolite 215 altering to calcite 203,215 altering to chlorite 100, 203, 214, 215, 237 altering to epidote 100,203,237 altering to magnetite 215 crystallization of 257,258,259 from augite 100,201 grading into hornblende 241,245,251 plate of 312 included in augite 255 included in feldspar 151, 484 included in quartz 466 including anatase 236 including augite 250,255,257 including biotite 251,260,261 including brouzite 250 plate of 306 including epidote 47 including feldspar 235,241 plate of , 312 including ilmenite 236,246 including iron oxide 47,215, 251,487 including malacolite 238 including' olivine 251, 257, 260 including pyroxene 237, 241, 245, 251, 260 plate of - 3 12 including quartz 47 including rutile 236,239,248 including spinel 236 inclusions in 240, 245, 248, 260 plate of 318 intergrown with augite, plate of 320 intergro wn with augite and olivine ' 255-260 intergrown with bronzite, plate of 318 intergrown with feldspar 256 plate of 322 intergrown with pyroxene 251, 260 plate.of 320 intergrown with spinel 256 Page. Hornblende of amphibolite 396 of amphibole-peridotite 253, 258 of arkose 479 of basic dike 47 of biotite-granite 192 of Bone Lake schists 150 of bronzite-norite-porphyry 246 of diorite 226 ofgabbro 234,237,248 of greenstone 470,483,486 of hornblende-gabbro 240, 241, 243, 245 plate of 312 of hornblende-gneiss 174, 465 of Hemlock schist 445 of metabasalt 100 of metadolerite 200,202,203 ofperidotile 251,252,257,260 of picrite-porphyry 215 of tuff 141,142,145 of volcanic conglomerate 145 of volcanic sand, plate of 296 orientation of 174,214,215,395,396,465 penetrated by feldspar, plate of 322 pbenocrysts in Bone Lake schists 150 pressure effects in 237 relations of orientation to biotite 486 zonal structure in 2-26, 234, 235, 236, 256 {See Hornblende intergrowth.) Hornblende-gabbro described 240-244 plate of 312,314,316,318 analyses of 242,263-264 crystallization of minerals of 262 intruded by bronzite-norite 243, 249,265 intruded by diorite 265 intruded by gabbro 265 intruded by peridotite 246, 260, 265 Hornblende-gneiss of Arcbean described 395-397 intrusive in Upper Huronian 167, 173, 174, 175 pressure effects in 175 Hornblende-schists from diabase 466 from gabbro 466 intrusive in Sturgeon River Algonkian 485 of Arcbean of Sturgeon River 465-467 of Upper Huronian 173 Hornblende-slate described 14 distribution of 13-14 Hubbard, Bela, on Sturgeon River tongue 459,461 Hulst, N. P., on composition of ore deposits 69 Huron Mining Co. {See Columbia mine.) Huronian rocks, correlation of xsv, xxvi distribution of, plate of 160 erosion of xxiv folding of XXII, 25, 163 metaraorphism of xxiii relations to Arcbean 39, 378, 379, 380, 410 relations to Cambrian xxi v, 28 succession xxv, xxvi unconformity within xxvii (See Algonkian, Upper Huronian, Lower Huronian.) Hutcbings.W.Maynard, analyses by 59 on composition of clay 60 on spilosites 206 on transfer of material from basic intrusive 211 Hyalopilitic texture in metabasalt 98, 99, 104 Hyperstbene included in augite 255 Hypidiomorphic texture in gabbro 233 INDEX. 499 Page. 75 105 84 265 95 Iceland, volcanica of Iddiug-s. J. P., on niagmatic differentiation ou micropoikilitic texture on origin of igneous rocks referred to , lUinoia Steel Co. {See Tonngstown mine.) Ilmenito altering to rutile * 202, 239 altering to sphene 239 included in biotite 1 216 included in bronzite 238 included in chlorite 216 included in horubleude 236, 246 including epidote ' 444 ini-luding quartz 444 of basic dikes 47 of Bone Lake schist 151 of chlorite schist ' 442 of gabbro and uorite 236,238,239 of Hemlock schist 444 of metadolerite 202 of picrite-porphy ry 216 of sedimentary inclusions in granite 197 Instruments, magnetic, use of, described 341-344 Interrange exploration 439, 448 Intersertal texture in tuetabasalt 98, 99, 104 Intruaives described 187-265, 469-470, 482-485 age of 188,189 correlation of 189 effect on magnetic observations 371,372 in Archean xviii, 38,45-49 in Felch Moantain range 426 in Hemlock formation 77 in Huronian xxm, 190 in Mansfield slate 63,64,204^211,413 in Sturgeon Kiver tongue, described 469-470, 482-485 in Upper Huronian 164, 174, 175, 190 influence on topography 54, 333 metamorphism of Mansfield slate, described- .204^211, 413 relations to Cambrian 188 use of term 187 IntrusiTe series, related, described 221-265 composition of 263-265 crystallization of minerals of 262 relative age of rocks of 265 textures of 262 Intrusives, unrelated, described 190-221 Iron-bearing formation described sx, 415-423, 446-450 correlation of 20, 330 intruded by granite 381 magnetic observations in 338, 339 near Crystal Falls, succession in 179,180 of Felcb Mountain range 378 described 415-423 ' distribution 377 Eominger on 381-383 of Penokee district, located by magnetic work . - 24 of Upper Huronian 168 relations to marble 376 relations to volcanics of Penokee series , sxi (See Groveland formation, Iron-ore deposits.) Iron carbonate. {See Siderite.) Iron County H Iron Mountain 385 slates of, correlation witb Mansfield schist 413 Iron ore, analyses of : 69 by Brooks, re I'erred to 19 Page. Iron ore, composition of. isi from siderite 70. 71, 184 from eruptive rock 334 origin of 70, 71, 184 Credner on 71 Irving on 70,71,13.0 Spurr on 422 Van Hise on 39, 70, 71, 130, 168 Iron ore deposits associated with intruaives 183 comparison -with ore deposits of adjacent dis- tricts 180-181 concentration of, at Mansfield mine 72 conditions favorable for, "Van Hise on 72 discovered at Crystal JFalls 175 discovered at Florence 175 explorations for 12,330 of Amasa area 175, 177 of Armenia mine, description of 183 of Crystal Falls area 178,186 production of ore from, table of op. 186 of Crystal Falls district, compared with Me- nominee iron ores 19 table of production from op. 186 of Columbia mine described 182 of Dunn mine described 182 of Great Western mine described 182 of Felch Mountain range, Burt on 375 of Hemlock mine 177 of Mansfield formation 27,62 described 65-73 of Mansfield mine 63 described 67-70 of Paint Kiver district. Brooks on 18 of Paint Hiver mine ]83 of Pesbakumme Falls 375 of sec. 20, T. 45 K., K. 33 "W 176 of S6c.34,T.46N.,R.33'W 176 of Ux^per Huronian 28, 166 described 175-186 distribution of, described 176-180 history of 175 methods of mining in, described 184-185 origin of, described 183-184 prospecting in 185 relations to adjacent rocks, described 182-183 size of, described 184 sketch of occurrence of 182 table of shipments from op. 186 relations to claj' slate 63 search in Bone Lake schists 151 shipments from Crystal Falls area, table of op. 186 Iron oxide altering to sphene 212 from feldspar 42, 92 included in chlorite 218 included in feldspar 151, 234 included in hornblende 47, 215, 251 included in quartz 419 included in siderite 133 of acid lavas 89, 91 of amygdules 124 of biotite-granite 192 of Bone Lake schists 151 of conglomerate 64 of dolomite 53 of gabbro and norite 239 of Groveland formation 417, 419, 449 500 INDEX. Page. 1,99,117,127,131 202 Iron oxide of metabaaalt of raetadolerite of musco^ite-biotite-gneisB, plate of of peridotite of picrite-porphyry 314,216, of variolites veins in Groveland formation Iron pyrites of basic dikes of peridotite "^^-^ Iron Star Co. {See Great "Western mine.) Iron Star mine. {See Great "Western mine.) Ishpeming formation, correlation of xxv, XXVI Irving, K, D., in concretionary structure in ferrugi- nous cliert *22 on Felcli Mountain range 379-380 on magnetic mapping ■ - 24, 336 on mica-schist ^^^ on origin of iron ore 70,71, 130 Ivrea, Italy, norite from, analysis of 244, 245 Italy, noritea from 235 Jacob's statf, use of, described 342 Janesville.N.Y., analysis of slate from 261 Johnston-Lavis, H. J., on resorption by intrusives. . 227 Judd.J.AV., on basalts 104 Julien, A. A., on lithology of Crystal Falls district . 21 on Upper Huronian 173-174 K:ayser,E., analyses by 207-208 22 394 221 220 219 95 162 Page. Lake Superior Survey, reconnaissance by 329 topography by 22 Lamont mine, location of 178, Op. 186 table of shipments from Op. 186 [See Monitor mine.) Lamont Mining Company. {See Lamont Mine.) Land Survey, United States 23, 343 Lane, A. C, referred to 21 Lang, H. Otto, referred to 105 Larsson, Per, on composition of iron ore 69 Laurentian. {See Arcbean.) Lavas, basic, described 95-135 Lawson, A. C, on ellipsoidal structure 118-119 on Laurentian and Coutcbiching 380 Lean ore in Upper Huronian 182 Lee Peck mine, location of 178, Op. 186 table of shipments from Op. 186 Kelley, F. T., referred to Kemp, F. J., analysis of mica-schist ou limburgite on ultrabasic intrusives on picrite-porphyry referred to Keweenawan series relations to Cambrian sandstone . Kona dolomite, correlation of xxv, xxvi Krakatao, tuff from 142 L. La Platte River, olivine gabbro from 255 Labradorite of gabbro find norite 233 of bornblende-gabbro 241 of metabasalt 104,105 twinning of 104 Lakes of Crystal Falls district, origin of 32 Bone, described 35 referred to 156 Huron, Huronian rocks of north shore of xvii Michigan, streams tributary to 31 Superior, streams tributary to 31 Light, described 34 Liver, described - 34 Mary 161 Squaw 335 SuuDog 335 Tobin 164 Trout 335 Lake Superior land diatrict, map of, by Foster and Whitney 17 Lake Superior sandstone described 28 unconformity with Huronian 28 {See Potsdam sandstone.) Lake .Superior Survey 183 Leith, C. K. , indebtedness to Leucoxene of greenstone of metabasalt of metadolerite surrounding magnetite Lewis, H. C, on saxonite-porpbyry ou alteration of olivine Lewis, indebtedness to Life in Mansfield slate, presence of carbon. Light Lake described 12 470 100 202 470 220 218 331 60 34 Limburgite, porpby ritic intrusive, described 212-221 Lime rock of Mansfield mine 63, 67, 69, 70 Limestone of Felch Mountain range, Rominger on.. 383 of hemlock formation 153 {See Dolomite, Randville dolomite.) Limonite deposits of Upper Huronian 180 . from calcite 153 from siderite 1 68 of Groveland formation 420 Lincoln mine 179 analysis of ore from 181 location of 178, Op. 186 table of shipments from Op. 186 Lincoln Mining Co. {See Lincoln mine.) Lindgren, "Waldemar, on granodiorite 231 Liver Lake described 34 Long Portage 174 Long Portage series 173 Lessen, K., on altered clay slates 206,209 Lower Huronian series described xvii, xviil, 50-154, 398-423, 430-450, 471-481 erosion of XXI, xxii magnetic rocks of, described 338-341 relations to Arcbean xix, 55 relations to intrusives 187, 190 relations to Upper Huronian xvii, XXir, 158, 160, 161, 162, 163, 176, 424 auccessiou in xxv, xx vi, 50 thickness of xvii Lower Marquette aeries, distribution of 453-455 relations to Crystal Falls series xxv, xxvi relations to Menominee series xxv, xxvi, 451-457 relations to Sturgeon River series 462 ore deposits, comparison with Crystal Falls ore deposits Lower Menominee, relations to conglomerate of Sturgeon River tongue relations to Lower Marquette 451-457 Luster mottling in raetadolerite 100 181 473 INDEX. 501 M. Page. Magnotio instruuieuts, «3e of, described 341-344 Ma.i;netic ubaervations 24,77 described 336-372 efiect of iutniaivos on, described 371-372 iu Grnveland format ion 377, 415, 416, 447, 448 in Negaunee forniiition 453,455 in melabasalt 134 in Sturgeon Hi ver tongue 460 in Upper Hurouian 175, 176 described 156-157 enfolding 366-373 Magnetic rocks described 338-341 depth of, determination of 354-356 Magnetism iu magnetic rocks, distribution of 339-341 466 152 215 217 168 394 252 436 487 394 466 151 Magnetite altering to sphene from basic volcanics from hornblende from olivine from siderite included in microcliue included in biotite included in dolomite included in hornblende included in quartz of amphibole-schist of Bone Lake schist of graywacke' 170,176 of greenstone 470, 487 of iron-bearing formation 338, 417, 419, 421 of metabasalt 100 of metadolerite 202 of peridotite 261 of picrite-porphyry 218 of pyroclastics 147 of sedimentary inclusions in granite 197 of volcanic conglomerate 143 surrounded by leucosene 466, 470 surrounding amygdaloidal cavities 102 titaniferous 47 Malacolite included in hornblende 238 of gabbro and norite 238 Manhattan mine, location of op. 186 table of shipments from op. 186 Manganese ore of Upx>er Huronian 181, 182 Mansfield (to-mi) 12,54,130,178,203,204 Mansfield formation described xx, 54^73, 411-415, 437-440 adinole of, described 208-209 Bessemer ore in 27 clay slate, described 57-62 analysis of 59, 61 compared with analyses of contact product. 209-211 compared with clay 60 compared with New York slate 61 compared with Vermont slate 61 chert of described 62 correlation of xxv, xx\a, 413 desmosite of 207 distribution of 27, 54, 438 exploration in 73 folding of 438, 439 gray wacke of, described 56-57 iron ores of 27, 62 described 65-73 metamorphosed by intrusives 204^211, 413 of Felch Mountain range described 411-415 Page. Mansfield formation of Michigamme Mountain and Fence lii ver areas described 437-440 phyllite of, described 57-G2 possible continuation of 55 relations to adjacent formations gg 64 relations to Archean 424 relations to Groveland formation 411, 413, 438, 447 relations to Hemlock formation xx, xxi, xxvi, 64, 76 relations to intrusives 63, 64, 203, 204 relations to Randville formation xx, 55, 4 U , 434, 438 relations to slate of Michigamme Mountain 56 relations to Upper Huronian xxii siderite-slate of, described 62 spilosites of, descri bed 206-207 structure of g^ thickness of xvn, 64, 65, 412, 438, 439 toijography of 54^433 Mansfield mine 27 54 411 described 65-70 analyses of ore from 59 cross section of g3 concentration of ore at 72 figure of cracks in Q5 location of op. 186 table of shipments from op. 186 Mansfield ore deposit, exploration of 67 concentration of 70 origin of 70 relations to surrounding beds 68 Marble of Randville formation 434, 435 described 408-311 of Sturgeon River dolomite 480, 481 relations to iron-ore formation 376 relations to quartzite 376 {See Dolomite, Randville dolomite.) Marquette district, correlation of formations of XXA', XXVI, 189, 330, 451-457, 470-471 folding of 26, 452 Clarksburg formation of iy2 iron-beai'ing formation of, connection with Crys- tal Falls iron-bearing formation.. 330 Martin-Garcia, olivine-gabbro from 255 Martite of Groveland formation 417-419 Mary Lake jei Mastodon Iron Co. {See Mastodon mine.) Mastodon mine, analysis of ore from 181 caving system at 185 location of 179, op. 186 method of mining at 184,185 table of shipments from op. 186 Matrix of ellipsoidal metabasalt 122, 132,133,135 described 114-118 plate of 116, 284, 298 Matthews, E. B., indebtedness to 331 referred to 22 Maurer, E. R., referred to 22 McCutcheon'a Lake 91 McKim, J. A., referred to 22 McNair, F. W., referred to 22 Melaphyres 98 Menominee district, correlation of formations of xxv, XXVI, 330 folding of 26 relations to Crystal Falls district 11, 25 structure of 378 iron ore of, compared with Crystal Falls iron ore '. 19,181,186 502 INDEX. Page. Menominee Eiver 175, 335, 374, 375, 460 described 32, 334 ilenomiDce series of Sturgeon Pdver 462 Merriam, "W". N., indebtedness to 12 on dike at Glidden exxjloration 183 magnetic lines traced by 12 referred to xv,xxvi,22 sketcli of folding of Upper Huronian 179 Merrill, G.P., on ultrabasic intrusives 220 Mesnard quartzite of Marquette district 452 correlation of xx v, xxvi Mesabi cbert, resemblance to Groveland quartzite. - 422, 423 Mesabi iron-bearing formation , origin of 422 Mesqua-cum-a-cum-sepe (riTer) 15 Metabasalt described 98-135 alteration of 126-135 aniygdaloidal, plate of 280,282,284,290 analyses of 103,106,107 carbonation of 132, 133; 134 conglomerate of Upper Huronian 163 devitrification of 102,103,126 ellipsoidal, described 112-124 figure of 112,113,114 plate of 116,292 matrix of, described 114-118 plate of 284,.298 eut-axitic texture in 103 flo'w structure in, plate of 280 fragments in volcanic sand, plate of 296 grading into dolerite 2O0 intrusives described 211-212 perlitic parting in, plate of 294 road material IS'i eilicification of 133, 134 spberulitic 98, 108 texture of, plate of 280, 286, 288, 290, 294 use of term 96,97,98 variolitic, described 108-111 plate of 110 wea^tbering of 134-135 Metadiabase, use of term 96, 97 Metadolerite, described 199-204 analysis of 203 distribution of ■ 199 relations to Hemlock volcanics 204 relations to Mansfield slates 203, 204 relations to other intrusives 204 relations to Upper Huronian 204 use of term 96, 97 Metamorpbism of Arcbean x viii of Crystal Falls series xxni, xxiv, xxv, xxvi of Hemlock formation xxiv,44G of quartzite 425-426 of Upper Huronian 28,425-426 {See Alteration.) Method of location of ledges 23 Methods of mining 184, 185 Methods of work of Lake Superior Division 22, 23, 24 Metropolitan mine, granite dike at 381 Mica, from feldspar 169,170 included in microcline 394 of acid lavas 89, 91 of biotite-granite 43-192 of granite 198,388,464 of Hemlock schist 444 of bornblende-gabbro 243 Page. Mica of bornblende-schist 465, 485 of Mansfield slate 205 oi" mica-schist 414 of quartzite 425 orientation of 43, 44, 89, 91, 198, 387, 388 Mica-diorite 226 analysis of 231,263,264 comparison with monzonite 232 opbitio texture in, plate of 308 Mica-gneiss, developed by intrusion 195, 197 from basic volcanics 152 of graywacke 170 of Upper Huronian 166,167,171,172,173 pressure efi'ects in 171 Miea-scbist described 392-395, 423-426 analyses of 394 developed by intrusion 195, 196 from basic volcanics 152 intruded by amphibolite 392 intruded by pegmatite 392 of Arcbean, described 392-395 of Pelch Mountain range, described 423, 426 of graywacke 170 of Mansfield schist 413 of Upper Huronian 166. 167, 171, 172, 173, 174, 425 pressure efi'ects in 171 relations to Kaudville dolomite 392 relations to Sturgeon quartzite 392 Mica-slate described 15 distribution of 14, 15 of Mansfield formation 439 Mica-titanite incl uded in d olomite 436 Michel-Levy on method of feldspar measurement. . . 104, 224 on ophitic texture 200 on texture of rhyolite-porphyry 81, 85, 105 Michigamme dam 55 formation, correlation of xxv, xxvi, 28, 165 replaced by Clarksburg volcanics xxvi Michigamme jasper, magnetic observations in 339 Michigamme Mountain xx, 56, 446, 447, 448 analyses of dolomites from 435 elevation of 333 slate of - 56 Michigamme Mountain area described 427-450 Michigamme Piver 16, 21, 54, 65, 156, 161, 163, 164, 166. 167,173, 174, 187, 190, 194, 203, 204, 205, 240, 241, 243, 245, 249, 253, 329, 330, 331, 333, 411, 432, 438 described 31,334,335 course of 54, 55 changing channel of 56 Michigan Exploring Co. (See Michigan mine.) Michigan mine 157 location of op. 186 table of shipments from — op. 186 Microcline, including feldspar 394 including magnetite 394 including mica ^ 394 including quartz 393, 403 of biotite-granite 41,42 of diorite 225 of mica-schist 394,414 of quartz-mica-diorite 228 orientation of 294 Microgranitic texture 83 described 86,87 INDEX. 503 Page. Microgranitic toxturo in quartz-niica-diorite 228 in quartz-mioa-diorite-porphyry, plate ol' 310 Microlitic feldspar 99 Micro-ophitic texture in metabasalt .'-. 104,212 Micropegiuatitic texture in acid lavas 89 in biotite-granite 191,192, 193 indiorite 224 in gneisses 390 in metadolerite 201 in quartz-niica-diorite 227 in tonalite 230 {See Pegmatitic texture.) Micropoikilitic texture in rhyolite-porpbyry de- scribed 83-86 plate of 270,272 in feldspar 171 origin of 85 (See Poikilitic texture.) Milwaukee and Nortliern Ewy 406 Mines, Aragon 09 Armenia, description of ore body at 183 location of 178, op. 186 table of shipments from op. 186 Blaney. {See Hope mine.) Caledonia, {See Mansfield mine.) Calumet and Hecla 399 Claire, location of 178, op. 180 table of shipments from op. 186 Columbia, descrijjtion of ore bodies 182 location of 179, op. 186 table of shipments from op. 1S6 referred to 161 Crystal Falls, analysis of ore from 181 location of 178, op. 186 table of shipments from op. 186 referred to 161 Delphic, location of op. 186 table of shipments from op. 186 Dunn, analysis of ore from 181 depth of 185 description of ore bodies 182 location of 179, op. 186 table of shipments from op. 186 Fairbanks. {See Lincoln mine.) Great Western, dessription of ore bodies at 182 deptli of 185 fissuring of rocks 185 location of 178, op. 186 table of shipments from op. 186 Groveland 413,415 Hemlock 157, 177 analysis of ore from ■... 181 diamond drill woik at, figure of 177 exploitation of ore deposits at 175 location of op. 186 table of shipments from op. 186 Hollister, location of 178, op. 186 Hope, location of 178, op. 186 table of shipments from j op. 186 Iron Star. {See Great "Western mine.) Lament, location of 178, op. 186 table of shipments from - op. 186 {See Monitor mine,) Lee Peck, location of 178, op. 186 table of shipments from op. 186 Lincoln 179 Page. Mines, Lincoln, analysis of ore, from 181 location of 178,op.l86 table of shipments from op. 186 Manhattan, location of op. 186 table of shipments from op. 186 Mansfield 27, 54, 411 described 65-70 analyses of ore from 69 Cross section of G3 concentration of ore at 72 figure of cracks in 65 location of op. 186 table of shipments from op. 186 Mastodon, analysis of ore from 181 caving system at 185 location of 179, op. 186 method of mining at 184, 185 table of shipments from op. 186 Metropolitan, granite dike at 381 Michigan 157 location of op. 186 table of shipments from op. 186 Monitor, dike in {see Lamont mine) 183 !North western 399, 407 Paint Kiver 183 dike in 183 location of 178, op. 186 ore deposits of 183 table of shipments from op. 186 Pewabic 69 Sheldon &. Schafer. {See Columbia mine.) Smith. (See Armenia mine.) ITnion. {See Columbia mine.) "Wauneta. {See Hope mine.) Toungstown 182, 186 location of 178, op. 186 table of shipments from op. 186 Mining, depth of 185 Mining methods 184-185 Miarolitic texture in tonalite 229 Mixed ore in Upper Hurouian 182 Monitor mine, dike in {see Lamont mine) 183 Morbihan, Brittany, granite, compared witb Crystal Falls granite 44 Muscovite, development of 484 from feldspar 42, 82, 90, 92, 192, 201, 224, 228, 248 plate of 286 from schistose x>y roclastic 146 from staurolite 196 included in feldspar 468 included in quartz 171, 394, 403 including biotite 298 iutergrown with biotite 393, 414 of acid lavas 89 of basalt, plate of 290 of basic dikes 47 > of biotite-granite 192 of biotite-schist 484 of Bone Lake schist 151 of granite 464 of gray wacke 170 of metabasalt 127,129 of mica-schist : 392, 393 of muscovite-biotite-granite 193 plate of 298 of pbyllito 440 504 INDEX. Page. Mascovite of sedimentarieBmetamorpliosed by intru- sion 195,197 of spilosite, ijlate of 30^ parallel growth with biotite 170 relations of orientation to biotite - 484 Muscovite-biotite-gneiss, plate of 298 of Sturgeon forn?ation 401 Muscovite-biotite-granite ^^^ described 193,194 Navitic texture in raetabasalt 98, 99, 104 Needle, dip, use of 344 described 342-343 Keedle, borizontal, deflection of, figure of 345 Kegative crystals in quartz 41 Negaunee iron-bearing formation 451 correlation of XX, xxv, xxvi, 449, 455 distribution of ■ 453-453 magnetic observations in 338, 339 relations to Groveland formation xx, 449, 455 Ket Kiver described 31 New Haven ^^ Korite, analyses of 244,245,263-264 described 233-249 Nortb iron range. (See Felcli Mountain range.) Northeastern area described 451-457 Northwestern mine --• 399,407 Norway Portage 167,173,174,191,226 Norway Rapids -40 Norway slates, correlation with Mansfield schist 413 0. Octahedrite, included in hornblende 236 of gabbro and norite 236 (See Anatase.) Oligoclase in biotite-granite 41 , 42 Olivine, altering to magnetite 217 altering to pilite 211,217 altering to serpentin > 213, 217, 251, 253, 255 altering to tremolite 218 crystallization of 257 included in hornblende 251, 257, 260 included in pyroxene 255,256 including spinel 252, 255 of amphibole-peridotite 255 of gabbro and norite 238,239 of metabasalt 104,211 of meladolerite 201 of peridotite 251 of picrite-porphyry 213, 217 pseudomorphs of 203, 213, 214 zonal intergrowth -with augite and hornblende.. 255-260 plate of 320,322 zonal structure in 258, 259 01i\'ine- gabbro, gradation to amphibole-peridotite . . 254-260 Oolitic testurein dolomite 435,437 Ophitic texture in diorite 223 in gabbro 233 in greenstone - 483 in hornblende-gabbro 240, 244 in metabasalt 212 in metadolerite 48, 200 in tonalite 230 Organic matter, origin of cherty carbonate 184 Page. Orientation of actinolite 105 of amphibole 127,214,486 of biotite 393,425,468,486 of calcite 132 of chlorite 118,127,133,146 of dolomite 410 of feldspar 84,90,171,394,396 of hematite 418 of hornblende 174, 214, 215, 395, 396, 465 of mica 43,44,51,89,91,198,387-388,478 of phenocrysts of granite-porphyry 45 of quartz 51,84,118,132,133,171,402,404,475 Ornamental stones of Hemlock formation described. 153, 154 Orthoclase, alteration of 224 of acid lavas 89 of biotite-granite 191 of diorite 224,225 of granite 464 of m asco vite-biotite-granite 3 94 of rhyolite-porphyry 182 Ottrelite of chlorite-schist 442 of Hemlock schist 445 Oval structure in Groveland formation described... 420-423 Pahoehoe structure in basalt (see Ellipsoidal struc- ture) 119,120,121,123,124 Paint Eiver 11,21,161,164,167,174,179,187,190,194 described 31 Paint River district 18 iron ore of 18 named by Erooks 11 Paint River Falls 18 Paint River Iron Company. {See Paint River mine.) Paint River mine 183 dike in 183 location of 178, op. 186 ore deposits of 183 table of shipments from op. 186 Paint rock associated with ore deposits 183 of Mansfield ore deposit 57, 68 Palagonite tuffs, use of term 138 Paleozoic rocks, deposition of xsiv relations to lower formations 26, 28, 376, 383 (See Potsdam sandstone.) Paragonite from feldspar 41, 42 Parallel texture of hornblende-gabbro 244 plate of 314 Patton, H. S., on hornblende 251 on hornblende-picrite 254,255 referred to 21 Peevie Falls 167 Pegmatite in Randville formation 435 intruding mica-schist ., 392 veins in greenstone 476 Pegmatitic texture in gneisses ■ 390 Peneplain of Crystal Falls district, relations to Michigan and "Wisconsin peneplain ] 3, 31 Penokee district, correlation of series xxi, 189 iron formation of, located by magnetic work 24 ferruginous carbonates of 130 Peridotite described 249-262 age of 262 analyses of 259,263,264 crystallization of minerals 262 distribution of 249,250 INDEX. 505 Page. , Poridotite, ruliitious of 249,250 relations to g.ibbro 246,200,261,265 (See Welirlite.) Peril tie purtiug iu aporljyoiito 87 I'l^itoof 274,276 iu Ij.tsalt tuH'. plate of 294 Pertliite in biotite-granlte 41,42 Pesbakiiiiiine Falls, iron ore at 375 Pesliakauinio JRirer 14, 15 Pewabie mine 69 Phenoorysts of granite-porpbyry, orientation of 45 Phillips, H.F, indebtedness to 331 referred to 22 Phyllite, of Mansfield formation 411,439 descri bed 57-62 Pblogopite of dolomite 410 Pbysiograpby of district 13,29-37,329-335 Pichler, A., on alteration of staurolite 196 Pickings, Matber & Co. {See Hemlock mine.) Picotite of picrite-porpbyry 217 of olivine 252 Picrite-porpbyry described 212-221 Pilite from augite 211 from olivine 211, 217 of metabasalt 211 of metadolcrite 201 of picrite-porpbyry 218 pseudpmorpbs after olivine 202 pseudomorpbs after pyroxene 218 Pilotasitic texture in metabasalt 98, 99, 104, 212 Pine. {See Timber.) Pinnite, including rutile 410 of dolomite 41Q PlagiocJase included in hornblende 235, 241 including biotite I93 434 including hornblende 484 of acid lavas 89 of ampbibolite 296 of aiupbibole-schist 4gg of biotite-granite 41 42191 of biotite-scbist 4q8 459 of diorite 223 ''24 of gabbro and norito 233 234 of greenstone 483' 484 of bcrnblende-gabbro 240 241 of metabasalt ' jq-^ of mica-diorite 227 23'' plate of gQg of mica-schist ^^^ of muscovite-biotite-granite _ 193 of pyroclastics -^^g of rhyolite-porphyry g2 of tuffs jgg orientation of gog pressure effects in 234 zonal structure in 193-194 Plagioclase-basalt of Hemlock formation 108 Poikilitio texture in amphibole-peridotite 253 in contact of granite and sedimentary, plate of. . 300 in granite jgg in hornblende 235 '>5i in hornblendegabbro '241 plate of 019 in metadolcrite 200 in pyroxene '_' 253 iu quartz-mioa-diorite-porphyry, plate of 310 inperidotite .' 250 {See Micropoikilitic texture.) Point Bonita, California. 40,45 45 Page. ^ . ^,, . 99,108,113 Point Jlornto, California, gabbro from 240 Polar magnetic picrite-porpbyry 212,219 Porphyritic granite sketch of I'orphyritic limburgite described 212-221 Porphyritic texture in gabbro 233 241 plate of in graywacke iu granite in metabasalt in picrite-porpbyry - . . in quartz-mica-diorite in schistose pyroclastics 312 167 45 127,129 217 228 146 383 376 87,1 479 48 43, 248 177 {See Ehyolite-porphyry, ' Granite-porphyry, Aporhyolite-porpbyry.) Potsdam sandstone, relations to Calciferous lime- stone relations to pre-Paleozoic rocks, xxiv, 26, 28, 331, 376, 473 relations to Silurian Pressure effects in acid lavas in Algoukian of Sturgeon Paver tongue 471 in amygdules 120,128 iu aporhyolite-porpbyry, plate of 276 iu arkose iu basic dikes in biotite in chert iu conglomerate ^-^ iu feldspar 42, 90, 92, 169, 234, 248, 388, 464 plate of 276^278 in gabbro 217' "48 '°g™"e ''.'.'44, 194! 464 iu graywacke ■, ^a iu Groveland formation 440 iu Hemlock formation ^c in hornblende ^gy i n hornblende-gabbro, plate of gig in hornblende-gneiss i^c iu mica-gneiss -.t^-, in mica-schist ny, iu quartz 41,51,53,57,81,82,91,93, 133, 169, 225, 388, 402, 404, 426, 464, 468, 477, 478, 48-( Pl'i*e«f 268, 276^278 in quartzite jo in Eandville formation 435-437 in rhyolite-porphyry, plate' of 276 278 in sedimentaries, caused by intrusion 195 Pleistocene deposits, XXIV 29,33 75 332 333 erosion of XXIV ' ' 29 relations to Huronian 155 Prospecting for iron ore jg= Platauia, Gaetano, on ellipsoidal structure 121 122 Pseudo-conglomerate, eruptive igg Pseudomorpbs after olivine 213 214 251 253 after pyroxene 214 218 of calcite, after feldspar 130 of chlorite, after biotite 217 228 of pilite, serpentine, and magnetite, after pyrox- ene 218 of serpentine, after olivine 251 253 Pumpelly , Raphael, on amygdules 125 on Felch Mountain series 380 on luster mottling 200 on Menominee iron region 377 on Michigamme Eiver 335 506 INDEX. Page. Purapelly, RapliEel, ou pseudo-amygd ules 203 with Broolis and Rominger, map of Upper Pe- ninsula of Michigan .'. 18 Pyroclastics, acid, described 94, 95, 135-148 of Hemlock formation described 135-148 plate of 140 intruded by dolerite 147 road material 154 Pyroxene altering touralite 48 crystallization of 258,259 included in hornblende 237, 241, 245, 251, 260 plate of 313 including olivine 255,256 including serpentine 245 lutergrown with hornblende 251 of amphibole-peridotite 258 of bronzite-norite-porphy ry 246 of gabbro 237,238,241,245 of metadolerite 201 of peridotite 250,260 of picrite-porphyry 213,218 of tuffs 138 pseudomorphs 213,214.218 zonal iutergi owth with feldspar, plate of 322 zonal intergrowth with hornblende 260 plate of * 320 zonal intergrowth with olivine, plate of 320, 322 Quartz and magnetite, oval structure in Groveland formation 420-423 augen of conglomerate 475 cataclastic structure in 41 , 44 enlargement of 57,85, 404, 405 from feldspar . . 42, 52, 99, 111, 131, 151, 169, 170, 171, 201, 248 from granite 52 globular texture in 86 granulation of 51,90 included in feldspar 40,191,394,403,468 included in hornblende 47 included in ilmenite 444 including apatite 194,419,420 including biotite 47,171,394,403,466 including chlorite 403, 419, 420 including epidote 47 including hornblende 466 including iron oxide 394, 419 including muscovite 171, 394, 403 including rutile 394, 420 including sericite 51 including tourmaline 420 inclusions in. . 41, 82, 191, 261, 388, 393, 396, 403, 404, 425, 466 lenticules in greenstone 485, 486 micropoikilitic zones about, plate of 270 of acid lavas 89,91,93 of amphibolite 296,466 of amygdules 125,126 plate of 280,284 of basic dikes 47 of biotite-schist 443,468,484 of Bone Lake schists 151 of conglomerate 477, 478 of diorite 225 of dolomite 410,436 of granite 191,193,194,198,388,464 plate of 298 Page. Quartz and magnetite of gray wacke 56, 57, 169, 431 of greenstone 470, 483, 486 of Groveland formation 419, 421, 449, 450 of Hemlock schist 444 of marble 481 of metabasalt 117, 118, 127, 129, 132, 133 of metadolerite 201 of mica-schiat 393,414 of peridotite 261 of phyllite '. 440 of pyroclastics 147 of q uartz-mica-diorite-porphy ry , plate of 310 of quartzite 52, 402, 404, 425, 450 of rhyolite-porphyry 81, 83, 84, 85, 86 pressure effects in, plate of 278 of rock intermediate between granite and quartz- ite 51 of sedimentary inclusion in granite 197 of sericite-schist 443 of tuff 142 orientation of 51, 84, 118, 132, 133, 171, 402, 404, 475 penetrated by tourmaline 57 pbenocrysts, aureoled, of rhyolite -porphyry, plate of 272 pressure effects in 41, 51, 53, 57,81, 82, 90, 91, 93, 133, 159, 225, 278,388, 402, 404, 426, 464, 468, 477, 484 plate of 268, 276 rhomboehdral parting in veins lu greenstone Groveland formation in ore deposits in Eandville formation Quartz-biotite-granite Quartz-diorite Quartz-mica-diorite Quartz-mica-diorite-poTphyry, plate of . Quartz-porphyry Quartz rock . 82 14 449 185 435 40 230 227,230 310 429 134 of Mansfield ore deposit 63, 67, 69, 70 Quartz-schist of Eandville formation 52 of Sturgeon formation 401 of Upper Huronian 173 Quartzite, apparent gradation into granite 39, 51, 52 correlation of xxv, xxvi current bedding in 53 metamorphismof 425,426 of Pelch Mountain Eange described 398-405, 423^26 of Kaudville formation 51-53 of Upper Huronian xxii, 425 pressure effects in 52 relations to dolomite 51, 376 relations to granite 376 relations to ore, deposits 182 {See Sturgeon quartzite.) E. Eailways ; Chicago, Milwaukee and St. Paul 95, 143, 175 Chicago and Northwestern 156,175,423 Mil waukee and Northern 406 Eaisin, C, on alteration of olivine 218 on ellipsoidal structure 118 Eandville dolomite described . . xix, 27, 50-53, 406-411, 431-437 analyses of 409, 435 correlation of xxv, sxvi distribution of 26, 50. 406-408, 431, 432 INDEX. 507 Page. Kanavillodolmiti'. I'ohliugof 432-431 iifFcluh llouutaiu tongue described 406-411 ofMichigammo Mouutainaud Fence River areas described 431-437 jiressuro eflfecta in 435, 437 quartzite of 51-53 relations to Archeau xix, 51, 53, 55, 407 relations to Groveland formation xx relations to Hemlock formation xx, 75 relations to intrusivea 434 relations to Mansfield foi'mation xx, 55, 411, 438 relations to mica-schists 392 relations to Sturgeon quartzite.. xix, 51,407, 430,4:^1,434 relations to Upper Huronian 424 relations to underlying formations 53 thickness of xvii, xix. 26, 53, 407, 408, 433 topography of 50, 406-408 Kandville station 406 Range. {See Town.) Ransome, F. L., on ellipsoidal basalt . 113, 114.118, 119, 122, 123 on feldspar sheaves 99 on ultrabasic iDtrusives 220 on variolites 108 Recomposed granite grading into quartzite 39 of Marquette district 52 Recryatallization of gneisses 390, 391 ofgraywacke 195,198 plate of .• 298 of hornblende, plate of 316 of mica-gneiss 171 of quartz 404 of Sturgeon river conglomerate xxiv Reibungsbreccia of Hemlock formation 136 of Upper Huronian , 161,166,177 Republic (town) .■ 333, 334 Republic trough 26, 452, 453 Resorption rim about olivine 258 Reyar, E., referred to 142 Rhombohfedral parting in quartz 82 Rhyolite-porphyry 80, 190 described 81-87 banding of 81 plate of 278 microgranitic texture in described 86, 87 micropoikilitic texture in described 83-86 plate of 270,272 of Hemlock formation 80 described 81-87 phenocrysts of, plate of 268 pressure effects in, plate of 276, 278 Richardson, G. B., analysis by 408, 409 Ridgway, J. L., indebtedness to 13. xvi Rieserferner, tonalite from 230, 231 Ripple marks in arkose 473 Road material of Hemlock formation 154 Roberts, C. T.. indebtedness to 66, 182 Romberg, J., on augite of gabbro 255 on olivine-gabbro j 256 on pyroxene zone about olivine 256 Rominger, C-, on correlation of Menominee rocks ... 19 on correlation of dolerite dikes 189 on correlation of Upper Huronian 164 on Felch Mountain range 374,379,381 on Menominee district 19 on iron-bearing formation near Crystal Falls 179, 180 ' on progressive metamorphism 52 Page. Komiuger, C.on Sturgeon River tongue 461 on Upper Peninsula of Michigan 20 referred to xv with Brooks and Pumpelly, map of Upper Penin- sula of Michigan 18 Kosenbusch, H., analysis by 166 on enstatite-porphyrite 247 on peridoti to 249 on picrite-porphyry 220,221 on spilosite 206 referred to 193, 105 Roth, J., on transfer of material from basic intru- sives to slate 211 on weathering of granite 58 Rothpletz, A.,on ellipsoidal structure 118,119 Rutile from biotite 43, 192, 202, 225 from ilraenite 202, 239 from titanic iron 170, 192, 193, 230 included in bronzite 238 included in feldspar 234 included in hornblende 236, 239,248 included in pinnite 410 included in quartz 394, 420 of acid lavas 93 of biotite-granite 192,193 of biotite-schist 484 of Bone Lake schist 151 of gabbro 236, 239 of graywacke 56, 170 of nietabasalt 129 of metadolerite 202 of metamorphosed Mansfield slate 205 twinning of 205, 236 S. Saalband in quartz-mica diorite * 228 Sagenite from biotite 192, 193 included in biotite 403 included in chlorite 403 of biotite-granite 192, 193 Sand, volcanic, plate of 296 Sandstone and slates of Sturgeon River dolomite formation described 481,482 Sandstone, Lake Superior. (5'ee Potsdam sandstone.) Sanford, S., indebtedness to 331 Santorin, basalt from 1200 Schafer, R. "W"., on basic rocks 247 Schistose acid lavas described 87-94 dikes in Archean 46-48 pyroclastics described 145-148 Schistosity of acid lavas 91 of basic dikes 47,48 of gabbro 2i8 plate of 316 of granite 44, 194 of greenstone 470 of hornblende-gneiss 175 of hornblende-schist 465 of metabasalt 127 of mica-schist 114 of pyroclastics 147 of Sturgeon River conglomerate 474, 475 Schists. (See Crystalline schists.) Schlieren in metabasalt 98 Scoriae of Hemlock formation 70 Section. (See Town.) 508 INDEX. Page. Sedimentariea, contact with, granite, described 194-198 plate of 300 included in dolerite 203 included in granite 195 of Hemlock formation described 152. 153 of Upper Huronian described 165-174 volcanic, described ^ 136-145 plate of '. 296 Sericite from feldspar 52, 89, 99, lOi, 111, 127. 131, 477 included in quartz 51 of acid lavas 91 of nietabasalt 99,101,127 of Kandville quartzite 52 of rock intermediate between quartzite and granite 51 orientatiou of 51, 478 Sericite-scbist from conglomerate 475 from graywacke 57 of Hemlock formation 443 Serpentine from bronzite 238 plate of 306 from bronzite-norite-porpbyry 246 from olivine 213,217,251,255 includ/sd in pyroxene 245 intergro wn witb augite 253 of picrite-porpbyry 213,214,218 pseudomorphs 214 pseudomorpba after olivine 253 pseudomorphs after pyroxene 218 Sbeldon &. Scbafer mine. (See Columbia mine.) Sboldeis exploration 447,449 Siamo slate 451 correlation of sxv, xxvi Siderite altering to hematite 168 altering to limonite 168 altering to magnetite 168 including iron oxide 133 interbanded witb clay slate 168 of gray wacke 170 of Groveland formation xx, 420 of metabasalt 117, 133 of siderite slate 62 origin of 168 replaced by silica 134 replacing calcite 133 source of iron ore 70, 71 , 184 Siderite-slate of Mansfield formation described 62 of Upper Huronian 108 relations to claj^ slate 63 Sideritization of metabasalt 117 Sidnaw (town) 1'5 Silicification of metabasalt 130, 133, 134 of siderite 134 Silurian rocks, unconformity with, underlying rocks . 26, 376 Slates described 1-1, 153, 169-174, 481, 482 alteration of 14, 166 of Mansfield formation described 204-211 altering to chlorite-schist 166 altering to mica-gneiss 166 argillaceous 14 fragments in Hemlock tufls 76 from Benson, Yt., analysis of 61 from Hampton village, N. Y., analysis of 61 from Janesville, N. Y., analysis of 61 from SoutbPoultney,Yt., analysis of 61 hornblende, described 13, 14 Page. Slates of Hemlock formation 153 of Mansfield formation, described 57-02 of SturgeonRiverdolomiteformation, described. 481,482 of Upper Huronian 166 described 169-174 relations to ore deposits 68, 182 transfer of material from basic intrusives 211 (See Mansfield formation.) Smyth, H. L., on peneplain 31 on Randville dolomite 27, 53 referred to xv. xvl 22, 26, 50, 74, 152 Smith, G. 0., on Archean granite 38 on ellipsoidal structure 118, 119 Smith mine. {See Armenia mine.) Soapstone associated \\ ith ore deposits 68, 183 of Mansfield formation 57 Soil of Crystal Falls district 36,37 Solar compass, use of 24 Sorby, H. C, on enlargement of quartz grains 404 South iron range. (See Menominee range.) South Mastodon mine. (See Manhattan mine.) South Mountain, Pa 124 South Poultney, Yt., analysis of slate from 161 Sphenefrom biotite 192,225 from ilraenite — 239 from titanic iron 144, 146, 170, 193, 212, 239 included in biotite 225 incl uded in epidote-zoiaite 225 including apatite 244 of acid lavas 91 of basic dikes 47 of biotite-granite 192, 193 of gabbro 239 of gray wacke 170 of metabasalt " 100 of metadolerite 202 of tuff 141 of volcanic conglomerate 144 pseudomorphs after magnetite 466 surrounding magnetite 466 Spherulitic texture in metabasalt (^. 114 pp. 1 pi. Price" 10 cents. 21. The Lignites of the Great Sioux Reservation; a Report on the Region between the Grand and Moreau Rivers, Dakota, by Bailey Willis. 1885. 8"^. .16 pp. 5 pi. Price 5 cents. 22. On New Cretaceous Fossils from California, by Charles A. White. 1885. 8°. 25 np. 5 pi. Price 5 cents. 23. Observations on the Junction between the Eastern Sandstone and the Keweenaw Series on Keweenaw Point, Lake Superior, by R. D. Irving and T. C. Chamberlin. 1885. 8^. 124 pp. 17 pi. Price 15 cents. 24. List of Marine Mollusca, comprising the Quaternary Fossils and Recent Forms from American Localities between Cape Hatteras and Cape Roque, including the Bermudas, by William Healey Dall. 1885. 8*^. 336 pp. Price 25 cents. 25. The Present Technical Condition of the Steel Industry of the United States, by Phineas Barnes. 1885. 8^^. 85 pp. Price 10 cents. 26. Copper Smelting, by Henry M. Howe. 1885. 8°. 107 pp. Price 10 cents. 27. Report of Work done in the Division of Chemistry and Physics, mainly during the Fiscal Year 1884-'85. 1886. 8°. 80 pp. Price 10 cents. 28. The Gabbros and Associated Hornblende Rooks occurring in the Neighborhood of Baltimore, Maryland, by George Huntington Williams. 1886. 8°. 78 pp. 4 pi. Price 10 cents. IV ADVERTISEMENT. 29. On the Fresli-Water Invertebrates of the North American Jurassic, by Charles A. White. 1886. 8*^. 41 pp. 4 pi. Price 5 cents. 30. Second Contribution to the Studies on the Cambrian Faunas of North America, by Charles Doolittle Walcott. 1886. 8°. 369 pp. 33 pi. Price 25 cents. 31. Systematic Review of our Present Knowledge of Fossil Insects, including Myriapods and Arachnids, by Samuel Hubbard Scudder. 1886. 8°. 128 pp. Price 15 cents. 32. Lists and Analyses of the Mineral Springs of the United States; a Preliminary Study, by Albert CPeale. 1886. 8°. 235 pp. Price 20 cents. 33 NotesontheGeologyofNorthemCalifornia, by J. S. Diller. 1886. 8°. 23 pp. Price 5 cents. 34 On the Relation of the Laramie Molluscan Fauna to that of the Succeeding Fresh-Water Eocene and Other Groups, by Charles A. White. 1886. 8°. 54 pp. 5 pi. Price 10 cents. 35. Physical Properties of the Iron-Carburets, by Carl Barus and Vincent Strouhal. 1886. 8°. 62 pp. Price 10 cents. 36. SubsideneeofFineSolidParticlesinLiquids,byCarlBarus. 1886. 8*^. 58pp. Price 10 cents. 37. Types of the Laramie Flora, liy Lester F. Ward. 1887. 8°. 354 pp. 57 pi. Price 25 cents. 38. PeridotiteofElliottCounty,Keutucky,byJ.S.Diller. 1887. 8='. 31pp. Ipl. PriceScents. 39. The Upper Beaches and Deltas of the Glacial Lake Agassiz, by Warren Uphani. 1887. 8"^. 84 ijp. 1 pi. Price 10 cents. 40. Changes in River Courses in Washington Territory due to Glaciation, by Bailey Willis. 1887. 8°. 10 pp. 4 pi. Price 5 cents. 41. On the Fossil Faunas of the Upper Devonian — the Genesee Section, New York, bv Henry S. Williams. 1887. 8°. 121pp. 4 pi. Price 15 cents. 42. Report of Work done in the Division of Chemistry and Physics, mainly during the Fiscal Year 1885-'86. F W. Clarke, Chief Chemist. 1887. 8^. 152 pp. Ipl. Price 15 cents. 43. Tertiary and Cretaceous Strata of the Tuscaloosa, Tombigbee, and AKabama Rivers, by Eugene A. Smith and Lawrence C. .Johnson. 1887. 8°. 189 pp. 21 pi. Price 15 cents. 44. Bibliography of North American Geology for 1886, by Nelson H. Darton. 1887. 8°. 35 pp. Price 5 cents. 45. ThePresentConditionofKnowledgeoftheGeology of Texas, by Robert T.Hill. 1887. 8°. 94 pp. Price 10 cents. 46. Nature and Origin of Deposits of Phosphate of Lime, by R. A. F. Penrose, jr., with an Intro- duction by N. S. Shaler. 1888. 8°. 143 pp. Price 15 cents. 47. Analyses of Waters of the Yellowstone National Park, with an Account of the Methods of Analysis emjiloyed, by Frank Austin Gooch and James Edward Whitfield. 1888. 8°. 84 pp. Price 10 cents. 48. On the Form and Position of the Sea Level, by Robert Simpson Woodward. 1888. 8°. 88 pi). Price 10 cents. 49. Latitudes aud Longitudes of Certain Points in Missouri, Kansas, and New Mexico, by Robert Simpson Woodward. 1889. 8°. 133 pp. Price 15 cents. 50. Formulas and Tables to Facilitate the Construction and Use of Maps, by Robert Simpson Woodward. 1889. 8°. 124 pp. Price 15 cents. 51. On Invertebrate Fossils from the Pacific Coast, by Charles Abiathar White. 1889. 8°. 102 151>. 14 pi. Price 15 cents. 52. Subaerial Decay of Rocks and Origin of the Red Color of Certain Formations, by Israel Cook Russell. 1889. 8°. 65 pp. 5 pi. Price 10 cents. 53. The Geology of Nantucket, by Nathaniel Southgate Shaler. 1889. 8°. 55 pp. 10 pi. Price 10 cents. 54. On the Thermo-Electric Measurement of High Temperatures, by Carl Barus. 1889. 8°. 313 pp., incl. 1 pi. 11 pi. Price 25 cents. 55. Report of Work done in the Division of Chemistry and Physics, mainly during the Fiscal Year 1886-'87. Frank Wigglesworth Clarke, Chief Chemist. 1889. 8°. 96 pp. Price 10 cents. 56. Fossil Wood and Lignite of the Potomac Formation, by Frank Hall Knowlton. 1889. 8^. 72 pp. 7 pi. Price 10 cents. . 57. A Geological Reconnoissance in Southwestern Kansas, by Robert Hay. 1890. 8°. 49 pp. 2 pi. Price 5 cents. 58. The Glacial Boundary in Western Pennsylvania, Ohio, Kentucky, Indiana, and Illinois, by George Frederick Wright, with an Introduction by Thomas Chrowder Chamberlin. 1890. 8°. 112 pp., incl. 1 pi. 8 pi. Price 15 cents. 59. The Gabln-os and Associated Rocks in Delaware, by Frederick D. Chester. 1890. 8°. 45 pp. 1 pi. Price 10 cents. 60. Report of AVork done in the Division of Chemistry and Physics, mainly during the Fiscal Year 1887-'88. F. W. Clarke, Chief Chemist. 1890. 8-^. 174 pp. Price 15 cents. 61. Contributions to the Mineralogy of the Pacific Coast, by William Harlow Melville and Wal- demar Lindgren. 1890. 8*^. 40 pp. 3 pi. Price 5 cents. 62. The Greenstone Schist Areas of the Menominee aud Marquette Regions of Michigan, a Con- tribution to the Subject of Dynamic Metamorphism in Eruptive Rocks, by George Huntington Williams, with an Introduction by Roland Duer Irving. 1890. 8'^. 241 pp. 16 pi. Price 30 cents. 63. A Bibliography of Paleozoic Crustacea from 1698 to 1889, including a List of North Amer- ican Species and a Systematic Arrangement of Genera, by Anthony AV. Vogdes. 1890. 8°. 177 pp. Price 15 cents. 64. A Report of Work done in the Division of Chemistry aud Physics, mainly during the Fiscal Year 1888-'89. F. W. Clarke, Chief Chemist. 1890. 8-^. 60 pp. Price 10 cents. ADVERTISEMENT. y T , ',"?■ »-V'''^'*'''''V'.!!;^' "i^^^' l:'it"n"i"'"« C(.al FioUl of Pennsylvania, Ohio, and West Virginia, bv Israel C. White. IWIl. y^. 212 p].. U jil. Prico 20 cents. ' i^mia, oy 66. Oxi a Ui(mi) of Volcanic Kuuk.s irom the Tewau Mountains, New Mexico, and on the Occur- rence of Primary Quartz in Certain Basalts, by Joseph Passon Iddings. 1890. 8°. 34 pp. Price 5 T-T .^'^'t^'^^" ■'^'^''1.^?^°^'' "ft^e Traps of the Newark System in the New Jersey Region, by Nelson Horatio Darton. 18fi0. 8^. 82 pp. Price 10 cents. h , y x^uibuu 68. Earthquakes in California in 1889, by James Edward Keeler. 1890. 8^. 25 pp. Price 5 S'^ 101 'i^ '^"price^S^tlnt^'^"^"*'"^ Biography of Fossil Insects, by Samuel Howard Scudder. 1890. ''\ 4 Repf I't oil Astronomical Work of 1889 and 1890, by Robert Simpson Woodward. 1890 8° 79 pj). Price 10 cents. : '^^;,^?^'^^ ^l "^« ,^^'^°"^'S,J°^??^ Insects of the World, including Myriapods and Arachnids, by Samuel Hubbard Scudder. 1891. 8°. 744 pp. Price 50 cents. =' -^ -^ ""> "y 72. Altitudes lietween Lake Superior and the Rocky Mountains, by Warren Unham 1891 8° J29 pp. Price 20 cents. ■ •.- . If -r?" y,'.^^'"sity of Solids by Carl Barns. 1891. 8-\ sii, 139 pp. 6 pi. Price 15 cents. 74. The Minerals of North Carolina, by Frederick Augustus Genth. 1891. 8° 119 nn Price 15 cents. ^ ^' iflQi '^t ^T'vT^ °^ ^^°-*^' American Geology for 1887 to 1889, inclusiTe, by Nelson Horatio Darton. 1891. 8°. 173 xip- Price 15 cents. ni, • f 'ip' ^ Dictionary of Altitude^ in the United States (Second Edition), compiled by Henry Gannett, Chief lopographer. 1891. 8^. 393 pp. Price 25 cents. /> i J o- f", 77. The Texan Permian and its Mesozoic Types of Fossils, by Charles A. White. 1891 8° 51 pp. 4 pi. Price 10 cents. " . . ^j. J?-^ A Report of Work done in the Division of Chemistry and Physics, mainly durine the Fiscal learl889-'90. F. AV. Clarke, Chief Chemist. 1891. 8". ISi pp. Price 15 cents ^ 79. A Late Volcanic Eruption in Northern California and its Peculiar Lava, by J. S. Diller o-rn Correlation Papers— Devonian and Carboniferous, by Henry Shaler Williams. 1891. 8° J79 i>p. Price 20 cents. 81. Correlation Papers— Cambrian, by Charles Doolittle Walcott. 1891. 8°' 547 ni) 3 nl Price 2o cents. ^^ " ^ ' 82. Correlation Papers— Cretaceous, by Charles A. White. 1891. 8°. 273 pp. 3 pi. Price 20 83. Correlation Papers— Eocene, by William Bullock Clark. 1891. 8°. 173 pp. 2 pi. Price lo cents. . 84. Correlation Papers— Neocene, by W. H. Dall and G. D. Harris. 1892. 8°. 349 pp 3 pi Price 25 cents. 1 1 • " i'*- 85. Correlation Papers— The Newark System, by Israel Cook Russell. 1892. 8°. 344 up 13 pi Price 2o cents. ^^ "i-i. 86. Correlation Papers— Archean and Algonkian, by C.E. Van Plise. 1892. 8". 549 pp 12 pi Price 2o cents. ^^' ^ n\, i^^Q^ Synopsis of American Fossil. Brachiopoda, including Bibliography and Synonymy, by Charles Schuchert. 1897. 8". 464 pp. Price 30 cents. s, i^.i y ju^y, "y a T ^-.TlieCretaceousForamimferaof New Jersey, by Rufus Mather Bagg, Jr. 1898. 8° 89 pp b pi. Price 10 cents. ^^" 89. Some Lava Flows of the Western Slope of the Sierra Nevada, California, by F. Leslie Kansome. 1898. 8^. 74 pp. 11 pi. Price 15 cents. Year 1 sqn "^pf ''^?''' w^pT" 1 "" 'of ^ IV^^^ ^V'^\ll?^ Chemistry and Physics, mainly diu-ing the Fiscal Year 1890- 91. F. W. Clarke, Chief Chemist. 1892. 8^. 77 pp. Price 10 cents. Price 10 i ''^^ American Geology for 1890, by Nelson Horatio Darton. 1891. 8°. 88 pp. 92. The Compressibility of Liquids, by Carl Barns. 1892. 8°. 96 pp. 29 pi. Price 10 cents. „f n 1 fO'^e Insects of Special Interest from Florissant, Colorado, and Other Points in the Tertiaries of Colorado and Utah, by Samue Hubbard Scudder. 1892. 8°. 35 pp. 3 pi. Price 5 cents. 94. The Mechanism of Solid Viscosity, by Carl Barns. 1892. 8°. 138 pp. Price 15 cents. PriceScehtT ^^^ ™ ™''''''^^^^''^°'^^^^'^'^'^''^'^^'^'''^^™^^^*°'^'^°^^^^ ^^^^- ^°- ^^PP- 07' r!^l?''y/'^''™''-'^^®l™°'^?"^™^''®°^^^'l'^"^«'^'yC:arl Barns. 1892. 8>^. 100pp. Price 10 cents. 97. IheMesozoicEchmodermataoftheUnitedStates, by W.B.Clark. 1893. 8°. 207 pp. 50pl. 1BO.. ^lo ^^i°on "^ ttie Outlying Carboniferous Basins of Southwestern Missouri, by David White. IbOo. 8*^. 139 pp. o pi. Price 15 cents. 99. Record of North American Geology for 1891, by Nelson Horatio Darton. 1892. 8°. 73 pp i^rice 10 cents. . ^ ^ ' Di,-i- ■^S°' I^iVl^^S^P^y and Index of the Publications of the U. S. Geological Survey, 1879-1892, by Phihp Crevelmg Warman. 1893. 8°. 495 pp. Price 25 .cents. "-^ . oy 101. Insect Fauna of the Rhode Island Coal Field, by Samuel Hubbard Scudder. 1893 8° ^( pp. 2 pi. Price o cents. -Rv„ 1 ^^^■■-.■^ Catalogue and Bibliography of North American Mesozoic Invertebrata, by Cornelius Breckinridge Boyle. 1892. 8°. 315 pp. Price 25 cents. VI ADVERTISEMENT. 103. Higli Temperature Work in Igneous Fusion and Elmllition, chiefly in Eelattou to Pressure, by Carl Barns. 1893. 8°. 57 pp. 9 pi. Price 10 cents. 104. Glaciation of the Yellowstone Valley north of the Park, hy Walter Harvey Weed. 1893. 8°. 41 pp. 4 pi. Price 5 cents. 105. The Laramie and the Overlying Livingstone Formation in Montana, by Walter Harvey Weed, with Report on Flora, hy Frank Hall Kno-witon. 1893. 8^. 68 pp. 6 pi. Price 10 cents. 106. The Colorado Formation and its Invertebrate Fauna, by T. W. Stanton. 1893. 8==. 288 pp. 45 pi. Price 20 cents. 107. The Trap Dikes of the Lake Champlain Eegion, by .Tames Fnrman Kemp and Vernon Freeman Marsters. 1893. S'^. 62 pp. 4 pi. Price 10 cents. 108. A Geological Reconnoissauce in Central Washington, by Israel Cook Russell. 1893. 8". 108 pp. 12 pi. Price 15 cents. 109. The Eruptive and Sedimentarv Rocks on Pigeon Point, Minnesota, and their Contact Phe- nomena, by William Shirley Bay ley. 1893. 8^'. 121 pp. 16 pi. Price 15 cents. 110. The Paleozoic Section in the Vicinity of Three Forks, Montana, by Albert Charles Peale. 893. 8°. 56 pp. 6 pi. Price 10 cents. 111. Geology of the Big Stone Gap Coal Fields of Virginia and Kentucky, by Marius E. Camp- bell. 1893. 8^.''l06pp. 6pL Price 15 cents. 112. Earthquakes in California in 1892, by Charles D. Perrine. 1893. 8°. 57 pp. Price 10 cents. 113. A Report of Work done in the Division of Chemistrv diiring the Fiscal Years 1891-'92 and 1892-'93. F. W. Clarke, Chief Chemist. 1893. 8". 115 pp. Price 15 cents. 114. Earthquakes in California in 1893, by Charles D. Perrine. 1894. 8^. 23 pp. Price 5 cents. 115. A Geographic Dietioiiary of Ehode Island, by Henry Gannett. 1894. 8°. 31 pp. Price 5 cents. - 116, A Geographic Dictionary of Massachusetts, by Henry Gannett. 1894. 8^. 126 pp. Price 15 cents. 117. A Geographic Dictionary of Connecticut, by Henry Gannett. 1894. 8°. 67 pp. Price 10 cents. 118. A Geographic Dictionary of New .lersey, by Henry Gannett. 1894. 8^. 131 pp. Price 15 cents. 119. A Geological Reconnoissauce in Northwest Wyoming, by George Homans Eldridge. 1894. 8°. 72 pp. Price 10 cents. 120. The Devonian System of Eastern Penuyslvania and New York, by Charles S. Prosser. 1894. 8°. 81 pp. 2 pi. Price 10 cents. 121. A Bibliography of North American Paleontology, by Charles Eolliu Keyes. 1894. 8°. 251 pp. Price 20 cents. 122. Results of Primary Triangulatiou, by Henry Gannett. 1894. 8^. 412 pp. 17 pi. Price 25 cents. 123. A Dictionary of Geographic Positions, by Henry Gannett. 1895. 8°. 183 pp. 1 pi. Price 15 cents. 124. Revision of North American Fossil Cockroaches, by Samuel Hubbard Scndder. 1895. 8^. 176 pp. 12 pi. Price 15 cents. 125. The Constitution of the Silicates, by Frank Wigglesworth Clarke. 1895. 8'^. 109 pp. Price 15 cents. 126. A Mineralogical Lexicon of Franklin, Hampshire, and Hampden counties, Massachusetts, by Benjamin Kendall Emerson. 1895. 8'-. 180 pp. 1 pi. Price 15 cents. 127. Catalogue and Index of Contributions to North American Geology, 1732-1891, by Nelson Horatio Darton. 1896. 8-. 1045 pp. Price 60 cents. 128. The Bear Eiver Formation and its Characteristic Fauna, by Charles A. White. 1895. S°. 108 pp. 11 pi. Price 15 cents. 129. Earthquakes in California in 1894, by Charles D. Perrine. 1895. 8°. 25 pp. Price 5 cents. 130. Bibliography and Index of North American Geology, Paleontology, Petrology, and Miner- alogy for 1892 and 1893, by Fred Boughton Weeks. 1896. 8°. 210 pp. Price 20 cents. 131. Report of Progress of the'Division of Hydrography for the Calendar Y^ears 1893 and 1894, by Frederick Haynes Newell, Topographer in Charge. 1895. 8°. 126 pp. Price 15 cents. 132. The Disseminated Lead Ores of Southeastern Missouri, by Arthur Winslow. 1896. 8°. 31 pp. Price 5 cents. 133. Contributions to the Cretaceous Paleontology of the Pacific Coast: The Fauna of the Knoxville Beds, bvT.W. Stanton. 1895. 8^. 132 pp. 20 ph Price 15 cents. 134. The Cambrian Eocks of Pennsylvania, by Charles Doolittle Walcott. 1896. 8°. 43 pp. 15 pi. Price 5 cents. 135. Bibliography and Index of North American Geology, Paleontology, Petrology, and Miner- alogy for the Year 1894, by F. B. Weeks. 1896. 8°. 141 pp. Price 15 cents. 136. Volcanic Eocks of South Mountain, Pennsylvania, by Florence Bascom. 1896. 8°. 124 pp. 28 pi. Price 15 cents. 137. The Geology of the Fort Eiley Military Eeservation and Vicinity, Kansas, by Robert Hay. 1896. 8°. 35 pp. 8 p'l. Price 5 cents. 138. Artesian-Well Prospects in the A-tlantic Coastal Plain Eegion, by N. H. Darton. 1896. 8°. 228 pp. 19 pi. Price 20 cents. 139. Geology of the Castle Mountain Mining District, Montana, by W. H. Weed and L. V. Pirs- sou. 1896. 8-^. 164 pp. 17 pi. Price 15 cents. 140. Report of Progress of the Division of Hydrography for the Calendar Year 1895, by Frederick Haynes Newell, Hydrographer in Charge. 1896. 8*^. 356 pp. Price 25 cents. ADVERTISEMENT. yil 141. The Eocene Peiiosits of the Middle Atlantic Slope iu Deluwaro, Maryland and Virsriui-i by William Bullock Clark. 1896. 8'^. 167 pp. 40 pi. Price 15 cents. 112. A Brief Contrilnition to the Geolony and Paleontology of Northwestern Louisiana bv T. Wayhind Vanghan. 1896. 8'^. 6.5 jip. 4 pi. Price 10 cents. ' 113. A Bibliography of Clays and the Ceramic Arts, by John C. Branner. 1896. 8°. 114 pp Price 15 cents. '■ ^ ' 144. The Moraines of the Missouri Coteau and their Attendant Deposits, by James Edward Todd 1896. 8"^. 71 pp. 21 pi. Price 10 cents. 145. The Potomac Formation in A^irginia, by W. M. Fontaine. 1896. 8^. 149 pp. 2 pi Price 15 cents. ' 146. Bibliography and Index of North American Geology, Paleontology, Petrology, and Miner- alogy for the Year 1895, by F. B. Weeks. 1896. 8°. 130 pp. Price 15 cents. 147. Earthquakes iu California iu 1895, by Charles D. Perrine, Assistant Astronomer iu Charo-g of Earthquake Obseryations at the Lick Observatory. 1896. 8"-'. 23 pp. Price 5 cents. ° 148. Analyses of Rocks, with a Chapter on Analytical Methods, Laljoratory of the Uuited States Geological Suryey, 1880 to 1896, by F. W. Clarke and W. F. Hillebraud. 1897. 8'-\ 306 pp Price 20 cents. 149. Bibliography and Index of North American Geology, Paleontology, Petrology, and Miner- alogy for the Year 1896, by Fred Boughton Weeks. 1897. 8°. 152 pp. Price 15 cents 150. The Educational Series of Rock Specimens collected and distrilmted by the United States Geological Survey, by Joseph Silas Diller. 1898. 8°. 398 pp. 47 pi. Price 25 cents. 151. The Lower Cretaceous Grypha^as of the Texas Region, by R. T. Hill and T. Wayland Vaiighan. 1898. 8°. 139 pp. 25 pi. Price 15 cents. 152. A Catalogue of the Cretaceous and Tertiary Plants of North America, by F H Knowlton 1898. 8°. 247 pp. Price 20 cents. ^o^o ^P' ■*■ Bibliographic Index of North American Carboniferous Invertebrates, by Stuart Weller 1898. 8°. 653 pp. Price 35 cents. 154. A Gazetteer of Kansas, by Henry Gannett. 1898. 8°. 246 pp. 6 pi. Price 20 cents 1.55. Earthquakes in California iu 1896 and 1897, by Charles D. Perrine, Assistant Astronomer m Charge ol Earthquake Obseryations at the Lick Observatory. 1898. 8^. 47 pp. Price 5 cents 156. Bibliography and Index of North American Geology, Paleontology, Petrology, and Miner- alogy for the Year 1897, by Fred Boughton Weeks. 1898. 8". 130 pp. Price 15 cents'! 160. A Dictionary of Altitudes in the United States (Third Edition), compiled bv Henry Gannett. 1899. 8^. 775 pp. Price 40 cents. '' ^T, x?-^' Earthquakes iu California in 1898, by Charles D. Perrine, Assistant Astronomer in Charo-e ot Earthquake Observations at the Lick Observatory. 1899. 8°. 31 pp. 1 pi. Price 5 cents In preparation: 157. The Gneisses, Gabbro-Schists, and Associated Rocks of Southeastern Minnesota, by C. W. 158. The Moraines of southeastern South Dakota and their Attendant Deposits, by J. E Todd 159. The Geology of Eastern Berkshire County, Massachusetts, by B. K. Emerson. WATER-SUPPLY AND IRRIGATION PAPERS. By act of Congress approved June 11, 1896, the following provision was made • "Provided, That hereafter the reports of the Geological Survey in relation to the o-auo-ino- of streams and to the methods of utilizing the water resources may be prin'ed in octavo form^'no'tto exceed one hundred pages in length and live thousand copies in number ; one thousand copies of which shall be lor the official use of the Geological Survey, oue thousand five hundred copies shall be deliv- ered to the Senate, and two thousand five hundred copies shall be delivered to the House of Renre- sentatives, for distribution." ' Under this law the following papers have been issued: 1. Pumping Water for Irrigation, by Herbert M. V'ilson. 1896. 8°. 57 pp. 9 pi 2. Irrigation near Phoenix, Arizona, by Arthur P. Davis. 1897. 8°. 97 pp 31 pi 3. Sewage Irrigation, by George W. Rafter. 1897. 8^. 100 pp. 4 pi. 4. AReconnoissanceinSoutheasternWashington, by Israel Cook Russeil. 1897. 8^^ 96 pp 7 pi 5. Irrigation Practice on the Great Plains, by Elias Branson Cowgill. 1897. 8° 39 pp 1'' pi' 6. Underground Waters of Southwestern Kansas, by Erasmus Haworth. 1897. 8^ 65 pp 12 pi' 7. Seepage Waters of Northern Utah, by Samuel Fortier. 1897. 8°. 50 pp 3 pi 8. Windmills for Irrigation, by Edward Charles Murphy. 1897. 8°. 49 pp. 8 pi 9. Irrigation near Greeley, Colorado, by David Boyd. 1897. 8°. 90 pp 21 pi 10. Irrigation in Mesilla Valley, New Mexico, by F. C. Barker. 1898. 8°. 51 pp. 11 pi 11. River Heights for 1896, by Arthur P. Davis. 1897. 8°. 100 pp. 12. Water Resources of Southeastern Nebraska, by Nelson H. Darton. 1898. 8° 55 pp 21 pi 1?" J'^"S'^*i°ii Systems iu Texas, by William Ferguson Hutson. 1898. 8''. 67 pp. 10 pi. 14. New Tests of Certain Pumps and Water-Lifts used in Irrigation, by Ozni P. Hood. 1889. 8^ al pp. 1 pi. 15. Operations at River Stations, 1897, Part I. 1898. 8'^. 100 pp. 16. Operations at River Stations, 1897, Part II. 1898. 8°. 101-200 pp 17. Irrigation near Bakersfield, California, by C. E. Grunsky. 1898. 8'=. 96 pp. 16 pi 18. Irrigation near Fresuo, California, by C. E. Grunsky. 1898. 8°. 94 pp. 14 pi 19. Irrigation near Merced, California, by C. E. Grunsky. 1899. 8°. 59 pp. 11 pL 20. Experiments with Windmills, by T. O. Perry. 1899. 8°. 97 pp 12 pi Hall. VlII ADVERTISEMENT. 2 pi. 7 pi. 21. Wells of Northern Indiana, by Frank Leverett. 1899. 8°. 82 pp. 22. Sewage Irrigation, Part II, by George AV. Eaiter. 1899. 8°. 100 pp 23. Water-Right Problems of Bighorn Mountains, by Elwood Mead. 1899. 24. Water Resources of the State of New York, Part I, bv George W. 99 pp. 13 pi. > > J ^ 25. Water Resources of the State of New York, Part II, by George W. Rafter. 101-200 pp. 12 pi. 26. Wells of Southern Indiana (Continuation of No. 21), by Frank Leverett. 1899. 27. Operations at River Stations, 1898, Part I. 1899. 8°. 100 pp. 28. Operations at River Stations, 1898, Part II. 1899. 8°. 101-200 pp. In preparation: 29. Wells and Windmills in Nebraska, by Edwin H. Barbour. 30. Water Resources of the Lower Peninsula of Michigan, by Alfred C. Lane. 62 pp. 7 pi. Rafter. 1899. S 1899. 8°. 8°. 64 pp. TOPOGRAPHIC MAP OF THE UNITED STATES. When, in 1882, the Geological Survey was directed by law to make a geologic map of the United States there was in existence no suitable topographic map to serve as a base for the geologic map. The prepnration of such a topographic map was therefore immediately begun. About one-fifth of the area of the country, excluding Alaska, has now been thus mapped. The map is published in atlas sheets, each sheet representing a small quadrangular district, as explained under the next head- ing. The separate sheets are sold at 5 cents each when fewer than 100 copies are purchased, but when they are ordered in lots of 100 or more copies, whether of the same sheet or of different sheets, the price is 2 cents each. The mapped areas are widely scattered, nearly every State being represented. About 900 sheets have been engraved and printed; they are tabulated by States in the Survey's "List of Publications," a pamphlet which may be had on application. The map sheets represent a great variety of topographic features, and with the aid of descriptive text they can be used to illustrate topographic forms. Tliis has led to the projection of au educational series of topographic folios, for use wherever geography is taught in high schools, academies, and colleges. Of this series the first folio has been issued, viz : 1. Physiographic types, by Henry Gannett, 1898, folio, consisting of the following sheets and 4 pages of descriptive text: Fargo (N. Dak. -Minn.), a region in youth; Charleston (W.Va.),a region in maturity; Caldwell (Ivans.), aregion in old age; Palmyra (Va!), a rejuvenated region; Mount Shasta, (Gal.), a young volcanic mountain; Eagle (Wis.), moraines; Sun Prairie (Wis.), drumlins; Donald- souville (La.), river flood plains; Boothbay (Me.), a fiord coast; Atlantic City (N. J.), a barrier-beach coast. GEOLOGIC ATLAS OF THE UNITED STATES. The Geologic Atlas of the United States is the final form of publication of the topographic and geologic maps. The atlas is issued in parts, progressively as the surveys are extended, and is designed ultimately to cover the entire country. Under the plan adopted the entire area of the country is divided into small rectangular districts (designated quadrangles), bounded by certain meridians and parallels. The unit of survey is also the unit of publication, and the maps and descriptions of each rectangular district are issued as a folio of the Geologic Atlas. Each folio contains topographic, geologic, economic, and structural maps, together with textual descriptions and explanations, and is designated by the name of a principal town or of a prominent natural feature within the district. Two forms of issue have been adopted, a "library edition" and a "field edition." In both the sheets are bound between heavy paper covers, but the library copies are permanently bound, while the sheets and covers of the field copies are only temporarily wired together. Under the law a copy of each folio is sent to certain public libraries and educational institu- tions. The remainder are sold at 25 cents each, except such as contain au unusual amount of matter, which are priced accordingly. Prepayment is obligatory. The folios ready for distribution are listed below. No. Name of slieet. State. Limiting meridians. Limiting parallels. Area, in square miles. Price, in cents. ^ Montana /Georgia 110°-lllo j. 85°-85o 30' 120° 30'-121o 840 30'-85o 1210-121° 30' 850-85° 30' 1050-105° 30' 85° 30'-80° 106° 45'-107o 15' I 77° 30'-78° 450-46° 34° 30'-35o 38° 30'-39° 35° 30'-36° 38° 30'-59o 350-350 30' 380 30'-39o 350-35° 30- 380 45'-39o 390-390 30' 3,354 980 932 969 932 975 932 975 465 925 \Tennessee California Tennessee California Tennessee Colorado Tennessee Colorado Virginia mest Virginia.. Maryland 3 25 4 25 •i P, 7 R Pikes Peak (oat of stock) 25 25 9 in ADthracite-Crested Butte 50 ADVERTISEMENT. IX No. 11 12 13 U 15 16 17 18 30 Name of sheet. State. Limiting meridians. Limiting parallels. Area, in Price, square in miles. Jackson ... Estillville . Fredericksburg Staunton Laason Peak... Knoxville Marys ville. . Smartsville . Stevenson . Cleveland Pikeville ilcMinn ville. Noiiiini Three Forks. Loudon Pocahontas .. MoiTistown.. Piedmont. Nevada Citv. fXevada City. <^ Grass Valley. iBanuer Hill . rGallatin..] /Tellow.stoue .Na. JCauyon...[ California Virginia Kentucky .Tennessee fMaryland \Virgiuia /Virginia \*U^est Virginia. California Teunessue North Carolina California California {Alabama Georgia Tennessee Teuue.-^see Tennessee Tennessee fMaryland \ Virginia Montana Tennessee /Viiginia OVest Virginia . Tennessee Virginia Maryland ["West Virginia. California 121° 00' 121° 01' 120° 57' t tional Park. Pyramid Peak - Franklin Brice ville Euckbannon... Gadsden Pueblo Do\\Tiieville ... Butte Special.. Truckee "Wartburg Sonera Nueces Bid^y6llBar ... Tazewell I Shoshone. [Lake Wyoming Boise Kichmond London Teumile District Special. Roseburg Hoi yoke California /Virginia \West. Virginia . Tennessee West Virginia . Alabama Colorado .... California Montana California Tennessee California 'i'exas California /Virginia \West Virginia. Idaho Kentucky Kentucky Cohii'ado ( )regon /Hassacbuaetta \Connecticut ... 112° 29' 120° 30'-121o 82° 30'-83° 770-77° 30' 79°-79° 30' 1210-122° 830 30'-81° 121° 30'-122o 1210-121° 30' 85° 30'-86° 84° 30'-85° 8o°-850 30' 85° 30'-8S° 76° 30'-77° lll°-112o 84°-Sl° 30' 81°-S1° 30' 83°-S3o 30' 79°-79° 30' 25"-121o 03' 45" 35"-121o 05' 04" 05"-121° 00' 25" 120O-120O 30' 79°-79° 30' 84°-84° 30' 800-80° 30' 86°-860 30' 104° 30'-105° 120° 30'-121° 30"-112° 30' 42" 120°-120° 30' 840 30'-S5o 1200-1200 30' 100°-100o 30' 121°-121° 30' 81° 30'-82o 1160-1160 30' 840-840 30' 840-84° 30' 106° 8'-106° 16' 1230-1230 30' 72° 30'-73o 38°-38o 30' 360 30'-37o 38°-38° 30' 360-38° 30' 400-41° 350 30'-36° 390-39° 30' 390-390 30' 350-350 30' 350 30'-36o 350 30'-36o 380-380 30' 450-460 350 30'-36° 370-370 30' 360-360 30' 390-390 30' 39° 13' 50"-39° 17' 16" 39° 10' 22"-390 13' 50" 39° 13' 50"-39o 17' 16" 380 30'-39° 36°-36o 30' 38° 30'-39° 340-340 30' 380-38° 30' 39° 30'-40° 450 59' 28"-46° 02' 54" 390-390 30' 360-360 30' 37° 30'-38° 290 30'-30° 390 30'-40o 370-370 30' 430 30'-44° 370 30'-38° 370-370 30' 390 22' 30"-39° 30' 30" 43°-43° 30' 420-12° 30' 957 938 938 3,634 925 925 925 980 975 969 969 3,354 969 11.65 12.09 11. C5 3,412 963 932 986 938 919 22.80 925 903 944 1,035 918 950 864 944 950 55 871 25 25 25 25 25 25 25 25 50 25 25 25 25 25 25 STATISTICAL PAPERS. Mineral Resources of tlie United States [1882], by Albert Williams, jr. 1883. 8^. xvii, 813 pp. Price 50 cents. Mineral Resources of tbe United States, 1883 and 1884, by Albert Williams, jr. 1885. 8°. xiv, 1016 pp. Price 60 cents. Mineral Resources of tlie United States, 1885. Division of Mining Statistics and Technology. 1886. 8^^. vii, 576 pp. Price 40 cents. Mineral Resources of the United States, 1886, by David T. Day. 1887. 8°. viii, 813 pp. Price (50 cents. Mineral Resources of the United States, 1887, by David T. Day. 1888. 8°. vii, 832 pp. Price 50 cents. Mineral Resources of the United States, 1888, by David T. Day. 1890. 8°. vii, 652 pp. Price Mineral Resources of the United States, 1889 and 1890, by David T. Day. 1892. 8^. viii, 671pp. Mineral Resources of the United States, 1891, by David T. Day. 1893. 8°. vii, 630 pp. Price 50 cents. X ADVERTISEMENT. Mineral Resources of the United States, 1892, by David T. Day. 1893. 8'^. vii, 850 pp. Price 50 cents. Mineral Resources of the United States, 1893, by David T. Day. 1894. 8^. viii, 810 pp. Price 50 cents. On March 2, 1895, the following provision was included in an act of Congress : "Provided, That hereafter the report of the mineral resources of the United States shall be issued as a part of the report of the Director of the Geological Survey." In compliance with this legislation the following reports have been published: Mineral Resources of the United States, 1894, David T. Day, Chief of Division. 1895. 8'^. xv, 646 pp., 23 pi. ; xis, 735 pp., 6 pi. Being Parts III and IV of the Sixteenth Annual Report. Mineral Resources of the United States, 1895, David T. Day, Chief of Division. 1896. 9P. xxiii, 542 pp., 8 pi. and maps ; iii, 543-1058 pp., 9-13 pi. Being Part III (in 2 vols.) of the Seventeenth Annual Report. Mineral Resources of the United States, 1896, David T. Day, Chief of Division. 1897. 8°. xii, 642 pp., 1 pi. ; 643-1400 pp. Being Part V (in 2 vols. ) of the Nineteenth Annual Eeport. Mineral Resources of the United States, 1897, David T. Day, Chief of Division. 1898. 8°. viii, 651 pp., 11 pi. ; viii, 706 pp. Being Part VI (in 2 vols.) of the Nineteenth Annual Eeport. The money received from the sale of the Survey publications is deposited in the Treasury, and the Secretary of that Department declines to receive bank checks, drafts, or postage stamps ; all remit- tances, therefore, must be by money order, made payable to the Director of the United States Geological Survey, or in currency' — the exact amount. Correspondence relating to the publications of the Survey should be addressed to The Director, United States Geological Survey, Washington, D. C, June, 1S99. Washington, D. C. [Take this leaf out and paste the separated titles upon three of your cata- logue cards. The hrst and second titles need no addition ; over the third write that subject under "which you would place tho book in your library.] LIBRARY CATALOGUE SLIPS. United States. Department of the interior. {TJ. S. geological survey.) Department of the interior | — | Monographs | of the | United States geological survey | Volume XXXVI | [Seal of the depart- ment] I Washington | government prilitiug office | 1899 Second title: United States geological survey | Charles D. Walcott, director | — | The | Crystal Falls iron-bearing district of Michigan | by | .J. Morgan Clements and Henry Lloyd Smyth | ■with I a chapter on the Sturgeon river tongue | by | William Shirley Bay ley | and | an introduction | by | Charles Richard Van Hise I [Vignette] | Washington | government printing office | 1899 A°. sssvi, 512 pp. 53 i3l. Clements (J. M.), Smyth (H. L.), Bayley (W. S.), and Van Hise (C. R.) United States geological survey | Charles D. Walcott, di- rector I — I The I Crystal Falls iron-bearing district of Michigan | by I J. Morgan Clements and Henry Lloyd Smith | with | a chap- ter on the Sturgeon river tongue | by | William Shirley Bayley | and I an introduction | by | Charles Richard Van Hise | [Vig- nette] I Washington | government piinting office | 1899 4^. xsxvi. 512 Jip. 53 pi. [U^'iTED States. Department of the interior. {U. S. geological survey.) Monograph XXXVl.] United States geological survey | Charles D. Walcott, di- rector I — I The I Crystal Falls iron-bearing district of Michigan | by I J. Morgan Clements and Henry Lloyd Smyth | with | a chap- ter on the Sturgeon river tongue | by | William Shirley Bayley | and I an introduction | by | Charles Richard Van Hise | [Vig- nette] I Washington | government printing office | 1899 4°. xxxvi, 512 pp. 53 pi. [TJxiTED States. Department of the interior. {U. S. geological survey.) Monograph XXSVI.] U S GEOUODICAL SURVEV Topograprir b, G E h,Di H L Smiih.ana from «u Su'xyeilm 1090-90 H TOI'OC.liAPIIK'AI. MAP ( ItYSTAl. r.VI.I.S liisTUUT, MM iii(;ax MAHQUHTTK lUHTHR'T. MICIUCAN