BERKELEY LIBRARY UNIVERSITY OF PLATE I. The Captain of the Canyon, Monument Canyon, Arizona. GEOLOGY PHYSICAL AND HISTORICAL BY HERDMAN FITZGERALD CLELAND, Pn.D, PROFESSOR OF GEOLOGY IN WILLIAMS COLLEGE WILLIAMSTOWN, MASSACHUSETTS , ' *.-*,-, AMERICAN BOOK COMPANY NEW YORK CINCINNATI CHICAGO COPYRIGHT, 1916 BY H. F. CLELAND ALL RIGHTS RESERVED CLELAND'S GEOLOGY W. P. I , - - : -. EARTH SCIENCES LIBRARY TO MY MOST HELPFUL CRITIC AND INDISPENSABLE AID MY WIFE 341808 PREFACE IN the preparation of this volume an attempt has been made to present an outline of the essentials of modern geology. By avoid- ing all details not necessary to an understanding of the fundamental principles of the science, it is hoped that this work will prove inter- esting to the student, although not less accurate because interesting. In the section on physical geology the human relation has been emphasized whenever possible, while in the historical section the history of life from the evolutionist's point of view has been taken up in broad outline. Much that may prove excellent in this work is due to the help of a number of eminent geologists, whose suggestions and criticisms have added many interesting points and have assisted in the elimi- nation of errors. The writer wishes especially to express his debt to Dr. W. D. Matthew, of the American Museum of Natural History, City of New York; to Professor Joseph Barrell, of Yale University; and to Professor N. M. Fenneman, of the University of Cincinnati, upon whom he has freely called for suggestions and criticisms and from whom much valuable assistance has been received. One or more chapters have also been read and helpfully criticized by the following geologists and educators, and their generous aid is acknowledged with keen appreciation : Messrs. H. E. Gregory, of Yale University; C. K. Schwartz, of Johns Hopkins University; J. B. Woodworth, of Harvard University ; J. S. Grasty, of the Uni- versity of Virginia ; Sumner W. Cushing, of the Salem (Massachu- setts) Normal School; J. W. Gidley and C. W. Gilmore, of the United States National Museum ; T. W. Stanton and F. H. Knowl- ton, of the United States Geological Survey ; Charles Schuchert and G. G. MacCurdy, of Yale University; T. D. A. Cockerell, of Boulder, Colorado; L. Hussakof, of the American Museum of Natural His- tory ; E. C. Case, of the University of Michigan ; O. P. Hay, of the Carnegie Institution of Washington; Sidney Powers, Cambridge, Massachusetts ; and C. L. Dake, of the Missouri School of Mines. 5 6 PREFACE For suggestions as to the most characteristic species of the vari- ous periods, credit is due to Professor G. D. Harris, for the Tertiary ; Messrs. T. W. Stanton and F. H. Knowlton, for the Mesozoic; and Drs. R. Ruedemann and E. M. Kindle, and Mr. L. Burling, for the Paleozoic. The numerous block diagrams which illustrate the text were made in wash rather than in line because the former are not only more attractive in appearance, but because being more realistic, they are more readily understood by the student. The writer is greatly indebted to Professors H. F. Osborn, W. B. Scott, F. A. Lucas, and S. W. Williston, for permission to use photo- graphs of restorations of extinct animals, made by them or under their direction, and to Prof. William Bullock Clark and Dr. John M. Clarke for the loan of a number of original drawings of the Geologi- cal Surveys of which they are directors. CONTENTS INTRODUCTION PAGE Astronomic or Cosmic Geology. Structural Geology. Dynamical Geology. Industrial Geology. Historical Geology. Length of Geological Time. Present Status of Geology. Fundamental Terms. Rocks. Sedimentary Rocks. Igneous Rocks. Metamorphic Rocks. Divisional Planes . . 21 PART I. PHYSICAL GEOLOGY CHAPTER I WEATHERING MECHANICAL AGENCIES . . . . . . . . , . .. V . 27 Frost. Talus. Rock Glaciers. Creep of Soils. Changes in Daily Temperature. Mechanical Action of Animals and Plants. Rain. Wind. Lightning. CHEMICAL AGENCIES ,. .35 Solution. Oxidation. Hydration. Carbonation. Organisms. Compari- son of Effects of Chemical and Mechanical Weathering. RESULTS OF WEATHERING 38 Spheroidal Weathering. Differential Weathering. Widening of Valleys. Rock Mantle and Soil. Kinds of Soil. Removal of Soil. CHAPTER II WORK OF THE WIND WIND AND SAND . . . . . . ... . . . . . . 44 Wind without Sand. Wind with Sand. Sand Dunes. Shape and Origin of Dunes. Migration of Sand Dunes. Beneficial Effect of Dunes. Ma- terial of Dunes. Height of Dunes. Eolian Sandstone. Dust. Loess. CHAPTER HI THE WORK OF GROUND WATER WORK OF GROUND WATER . . . - . . . . . .. ; . . 56 Quantity of Ground Water. The Water Table. Wells. Movement of Ground Water. Depth of Ground Water. Artesian Wells. Chemical Work of Ground Water. Solution. Replacement and Deposition. Belts of Weathering and Cementation. Desert Limestone. Mechanical Work of Ground Water. SPRINGS . . . . . . 62 Origin of Springs. Constant and Intermittent Springs. Mineral Matter in Spring Water. Mineral Springs. Temperature of Springs. Ther- mal Springs. Geysers. 7 8 CONTENTS PAGE STRIKING EFFECTS OF GROUND WATER 69 Swallow Holes. Caverns. Natural Bridges. Cave Deposits. Karst. Landslides. CONCRETIONS 75 Composition of Concretions. Time of Formation. Oolitic Limestone. Geodes. CHAPTER IV THE WORK OF STREAMS FACTORS IN STREAM EROSION 81 Material Carried by Streams. How the Sediment is Moved. Factors Deter- mining the Velocity of Streams. Water Wear. Solution. Vertical Erosion (Corrasion). Weathering and Vertical Erosion. Base Level of Erosion. Effect of Load. Factors Affecting the Rate of Erosion. Scour and Fill. Lateral Erosion. FEATURES DUE TO STREAM EROSION 89 Falls and Rapids. Exceptions Falls not the Result of Erosion. Pot- holes. Canyons. Instances of Rapid Erosion. Effect of Deforesta- tion on Rivers. Growth of Valleys. Valleys Formed in Ways Other than by Stream Erosion. The Direction of Valleys. Basins and Divides. Elevations Due to Unequal Hardness. Outliers. Rock Terraces. Stream Piracy. THE EROSION CYCLE 109 Youth. Maturity. Old Age. Effect of Elevation and Depression on Streams. PENEPLANATION . . .114 The Peneplain of Southern New England. The Appalachian Peneplain. The Laurentian Peneplain. Rate of the Denudation of Continents. How the Load of Streams is Measured. DEPOSITION IIQ Causes of Deposition. Flood Plains. Meanders. Oxbow Lakes. Natural Levees. Alluvial Cones and Fans. Piedmont or Alluvial Plains. Alluvial Terraces. Discontinuity of Terraces. Characteristics of River Deposits. DELTAS I30 Growth of Deltas. Structure of Deltas. DEPOSITION IN LAKES BY STREAMS AND BY OTHER AGENTS 133 Mechanical Deposits. Chemical Deposits. Organic Deposits Diatoms. Marl. Peat. Playas. Salt Lakes. Alkaline Lakes. Origin of Rock Salt. Extinct Lakes. CHAPTER V THE WORK OF GLACIERS GENERAL CONSIDERATIONS I4I Distribution and Size of Glaciers. Position of the Snow Line. Formation of Ice in Snow Fields. CONTENTS 9 PAGE MOUNTAIN GLACIERS 143 Formation. Cirques. Origin of Cirques. Development of Cirques. Fate of Cirques. Ablation. SURFACE OF MOUNTAIN GLACIERS . . ' . . . 147 Irregularities Due to Tension. Irregularities Due to Streams and Ice Tables. MOVEMENT OF GLACIERS . . . . 150 Rate of Movement. Differential Movement of Glaciers. Factors In- fluencing the Rate of Movement. Lower Limit of Glaciers. TRANSPORTATION OF MOLWTAIN GLACIERS I54 Surface Moraines. Subglacial Material. Englacial Material. EROSION BY MOUNTAIN GLACIERS 157 Plucking and Abrasion. Effect on the Material Carried. Factors In- fluencing the Rate of Erosion. DEPOSITS OF MOUNTAIN GLACIERS 159 Terminal Moraines. Submarginal Moraines. Ground Moraine. The Work of Glacial Streams. LANDSCAPE MODIFIED BY GLACIAL ACTION 163 Characteristics of Glaciated Valleys. Mature Glaciated Valleys. De- struction of Features of Glaciated Valleys. Fiords. PIEDMONT GLACIERS 167 CONTINENTAL ICE SHEETS * _. 168 Greenland. The Antarctic Continent. ANCIENT GLACIATION . 171 DEPOSITION ..'..-.' 171 Bowlders. Unstratified Drift. Moraines. Terminal Moraines. Moraines of the Last Great Ice Sheet in North America. Ground Moraine. Drumlins. Stratified Drift. Outwash Plains. Terraces. Deltas. Eskers. Kames. Relation between Stratified and Unstratified Drift. EROSION BY CONTINENTAL GLACIERS 182 Effect on the Underlying Rock. Modification in the Shape of the Hills. Effect of Glaciation on Drainage. Lakes and Ponds. Rivers. ICEBERGS . . . 188 Formation of Icebergs. Size and Work of Icebergs. GLACIAL MOVEMENT . . . . .189 Viscosity Theory. Expansion and Contraction. Regelation. Melting and Pressure. Growth of Granules. CHAPTER VI THE OCEAN AND ITS WORK GENERAL CHARACTER OF THE OCEAN 194 Topography of the Ocean Floor. Irregularities of the Ocean Floor. Com- position of Ocean Water. Temperature of the Ocean. Distribution of Marine Life. Age of the Ocean. I0 CONTENTS PAGE MOVEMENT OF THE WATER -. 198 Wave Motion. The Breaking of Waves. Force of Storm Waves. Height of Storm Waves. Tides. Tidal Currents. Tidal Bores. Earthquake Waves. Ocean Currents. MARINE EROSION 202 Factors in Marine Erosion. Shore Ice. Ice in Lakes. RESULTS OF MARINE EROSION 205 Effect of Erosion on Different Materials. Influence of Joints and Other Planes on Erosion. Coves and Headlands. Sea Caves and Blow- holes. Arches. Stacks. Marine Terraces. Striking Examples of Marine Erosion. Sea-captured Streams. Raised Beaches. Ancient Plains of Marine Denudation. The New England Marine Plain. TRANSPORTATION . ' . . . .217 Littoral or Shore Currents. Tidal Currents. FEATURES RESULTING FROM TRANSPORTATION 218 Beaches. Bayhead Beaches. Bars and Spits. Sand Reefs or Barrier Beaches. Tied Islands. Examples of the Constructive Work of the Sea. SHORES . 224 Smooth Shores. Cuestas. Rough Shores. Examples of Irregular Coasts. Proofs of Elevation and Depression. The Stability of the Atlantic Coast of North America. Cycle of Shore Erosion. DEPOSITION IN SEAS AND LAKES 233 Source and Extent of Land-derived Sediments. Stratification. Cross or False Bedding. LITTORAL DEPOSITS 235 Extent. Character of Littoral Deposits. Distinguishing Characteristics of Littoral Deposits. SHOAL-WATER DEPOSITS . . ... . . 237 Extent and Character of Deposits. Limestone. Lens-shaped Sediments. Dovetailing of Sediments. Basal Conglomerates. Subsidence Neces- sary for Great Accumulations. DEEP-SEA DEPOSITS . . .. . 241 Blue Mud. Globigerina Ooze. Radiolarian Ooze. Red Clay. CORAL REEFS AND ISLANDS ...... . i . . . . . .243 Coral-reef Problem. Subsidence Theory of Darwin. Submarine Bank Theory of Murray and Others. Change in Sea Level Due to Glaciation, or the Glacial-control Theory. CONSOLIDATION OF SEDIMENTS ... 248 Cementation. Effect of Pressure. Effect of Heat. CLASSIFICATION OF SEDIMENTARY ROCKS . . ..... . . . .249 Limestones. Sandstones. Shales. Deposits in Lakes and Deserts. Influence of Sedimentary Rocks upon Topography. CONTENTS H CHAPTER VII THE STRUCTURE OF THE EARTH PAGE STRUCTURAL FEATURES OF ROCKS . . . . . , > . . . . 252 Dip and Strike. Effect of Dip and Strike upon Outcrop. FOLDS . . . . . . . . . 254 Effect of Folding on Competent and Incompetent Strata. How the Structure of a Region is Determined. Origin of Folds. Warping. Zones of Flow and Fracture. JOINTS .258 Origin of Joints. Effect of Joints on Topography. FAULTS ../..'. . . . 261 Normal or Gravity Faults. Examples of Normal Faults. Reverse or Thrust Faults. Examples of Thrust Faults. Vertical and Hori- zontal Faults. Influence of Faults on Topography. Minor Features of a Fault Fracture. Detection of Faults. Origin of Faults. Rapidity of Fault Movements. CONFORMITY AND UNCONFORMITY . . . . ........ .270 Importance of Unconformities. Overlap. CONSTITUTION OF THE EARTH'S INTERIOR 272 Zone of Variable Temperature. The Interior Heat of the Earth. THEORIES OF THE PHYSICAL STATE OF THE EARTH'S INTERIOR 273 Internal Fluidity Theory. Solid Interior. Gaseous Center. Radioactivity and a Solid Center. Subcrust Theory. Summary. CHAPTER VIII EARTHQUAKES EARTHQUAKES AND ATTENDING FEATURES 275 The San Francisco Earthquake. Distribution of Earthquakes. Summary of the Causes of Earthquakes. Displacements. Depth of the Plane or Point of Origin. Earthquake Waves. Amplitude of Vibration. Vorticose and Twisting Movements. Duration. Frequency. Areas Affected by Certain Earthquakes. Instruments for Determining and Measuring Earthquakes. EFFECTS OF EARTHQUAKES . . . .. .287 Faults and Fissures. Changes in Level. Landslides. Earthquake Topog- raphy. Sounds. Loss of Life. Effect on Underground Water. Gases. Construction of Buildings in Earthquake Regions. Effect of Earth- quakes on the Sea. Evidence that a Region has been Free from Severe Earthquakes. CHAPTER IX VOLCANOES AND IGNEOUS INTRUSIONS VOLCANOES , ..... . 294 How Volcanoes Begin. New Volcanoes. Classification of Volcanoes. 12 CONTENTS PAGE MATERIALS ERUPTED 295 Gases. Fragmental Materials. Lava. Lava Streams. Effect of Compo- sition on Fluidity. Temperature. Surface of Lava Flows. Velocity of Lava Flows. Nature of Lavas. TYPES OF VOLCANOES ' .. - . / . . . 302 The Explosive or Fesuvian Type: Vesuvius. Krakatao. Katmai. Mt. Pelee. Bandai-san. The Quiet or Hawaiian Type: Crater of Kilauea. Eruptions. Lava Streams. Origin of Calderas. Steep Lava Cones : Volcanoes of the Chimborazo Type. Fissure Eruptions: Recent Icelandic Lava Sheets. CHARACTERISTICS OF VOLCANIC CONES . . . . . . ... . 311 Profiles of Volcanoes. Shape of Craters. Erosion of Volcanic Cones. Necks and Plugs. Age of Volcanoes in the United States. DISTRIBUTION AND NUMBER OF VOLCANOES 3I 8 Number of Volcanoes. Distribution. Cause of Distribution. Ancient Volcanoes. IMPORTANCE OF VOLCANISM TO MAN 321 Beneficial Effects. Harmful Effects. Volcanoes and Climate. SUBORDINATE VOLCANIC PHENOMENA 322 Mud Volcanoes. Solfataras. INTRUSIVE OR PLUTONIC ROCKS . .324 Injected Masses: Dikes. Sills. Laccoliths. Subjacent Masses: Stocks. Batholiths. Some Effects of Intrusions. IGNEOUS ROCKS 32Q Subdivisions Depending upon Chemical Composition. Subdivisions De- pending upon Texture. CLASSIFICATION OF IGNEOUS ROCKS . , 70 OO^ Coarse-grained Igneous Rocks: Granite. Syenite. Diorite. Gabbro. Peridotite. Compact or Fine-grained Igneous Rocks: Felsites. Basalts. Glassy Rocks: Obsidian or Volcanic Glass. Pitchstone. FRAGMENTAL VOLCANIC ROCKS . ... . 332 Tuff. Volcanic Breccia. Columnar Structure of Lava. AGE OF IGNEOUS ROCKS . . ... . ." . ... 334 THEORIES OF VOLCANISM . ."./,.' ... . - 334 Theory Based upon the Assumption that the Interior is Molten. Theories Based upon the Assumption that the Earth is Solid: Heat by Fric- tion. Formation of Lava Reservoirs by Relief of Pressure. Liquid- thread Theory. Abyssal Injection Hypothesis. RESUME' OF PRESENT KNOWLEDGE OF VOLCANISM .... 337 Origin of Volcanic Gases. Cause of the Ascension of Lava. Cause of Periodicity. Influences of the Atmosphere, etc. CONTENTS I3 CHAPTER X METAMORPHISM PAGB KINDS OF METAMORPHISM . . . . -".' . . . . . .341 Contact Metamorphism. Regional Metamorphism. CLASSIFICATION OF METAMORPHIC ROCKS . . . .' . , . '-. ' . 344 Quartzite. Marble. Slate. Schist. Gneiss. SUMMARY OF CAUSES OF METAMORPHISM , ... 347 Heat. Moisture. Pressure. ARRANGEMENT OF MINERALS . . . . . . . . . f t 348 Crystallization. Granulation. Relation of Cleavage to Pressure. From Igneous, through Sedimentary, to Metamorphic Rocks. Weathering of Metamorphic Rocks. Economic Importance. CHAPTER XI MOUNTAINS AND PLATEAUS CLASSIFICATION OF MOUNTAINS 352 Mountains of Accumulation. Residual Mountains. Fault or Block Moun- tains. Laccolith Mountains. Domed Mountains. Complexly Folded Mountains. ORIGIN AND DEVELOPMENT OF FOLDED MOUNTAINS . 358 Geosynclines. Lateral Pressure. Experiments in Mountain Building. Rate of Folding. To What the Topographic Features of Folded Moun- tains are Due. Cycle of Erosion of Mountains. THEORIES OF MOUNTAIN BUILDING 364 Cause of Lateral Pressure. The Elevation of Plateaus and Mountains. The Theory of Isostasy. The Distribution of Mountains. Perma- nence of Continents and Ocean Basins. Age of Mountains. CHAPTER XII ORE DEPOSITS CHARACTERISTICS OF DEPOSITS . .... . . . . . . . 370 Ores in Ready-made Cavities. Fissure Deposits. Form and Extent of Veins. Source of Vein Material. Cause of Precipitation. Replace- ment Deposits. Weathering and Concentration of Ores. Magmatic Segregation. Placer Gold Deposits. Sedimentary Iron Deposits. PART II. HISTORICAL GEOLOGY CHAPTER XIII HISTORICAL GEOLOGY FOSSILS . . . . . . . . . . ... .... 377 The Original Substance may be Preserved. Replacement. Casts and Molds. Footprints, Trails, etc. Entombment of Plants and Animals. Imperfection of the Record. I4 CONTENTS PAGE GEOLOGICAL CHRONOLOGY 381 Order of Superposition. Chronology Determined by Fossils. Use of Fossils in Determining Physical Conditions. Difficulties in Corre- lating Strata. DIVISIONS OF GEOLOGICAL TIME . . .- . . 383 CHAPTER XIV THE EARTH BEFORE THE CAMBRIAN THEORIES OF THE EARTH'S ORIGIN 385 Nebular Hypothesis. Planetesimal Hypothesis. Nebular and Planetesimal Theories Contrasted. PRE-CAMBRIAN ERAS 388 THE ARCHEOZOIC ERA 389 Distribution of the Archaeozoic Rocks. Characteristics of Archaeozoic Rocks. Thickness. Causes of Metamorphism and Deformation. Conditions during the Archaeozoic Era. Duration. Bearing upon the Theories of the Earth's Origin. THE PROTEROZOIC ERA 393 Archaeozoic and Proterozoic Contrasted. The Proterozoic in Different Regions. Iron and Copper Deposits. Life of the Proterozoic Era. Duration. Climate. Life before Fossils. CHAPTER XV THE CAMBRIAN PERIOD THE PALEOZOIC ERA J- . . 401 THE CAMBRIAN PERIOD .-. . . 402 Divisions of the Cambrian. Location of Cambrian Rocks. Physical Geography of Ancient Periods. Basal Unconformity. Physical Geography of the Cambrian. Character of the Cambrian Rocks. Pres- ent Condition of the Sediments. Volcanism. Close of the Cambrian. Other Continents. LIFE OF THE CAMBRIAN . 40 8 PLANTS ...... . ...... . . 409 ANIMALS . ... . . . .410 Crustacea: Trilobites. Other Crustaceans. Mollusc a: Gastropods. Molluscoidea: Brachiopods. Echinodermata : Cystoids. Worms. Coelenterata: Corals. Graptolites. Jellyfish. Sponges. Protozoa. SUMMARY 4l6 Evolution during the Cambrian. Climate and Duration. CONTENTS 15 CHAPTER XVI THE ORDOVICIAN PERIOD PAGE ORDOVICIAN PHYSICAL GEOGRAPHY . . . .- . 418 Close of the Ordovician. Cincinnati Anticline. Volcanism. Ordovician of Other Continents. PETROLEUM AND NATURAL GAS . . . . . . . ... . 424 Conditions Favoring the Accumulation of Oil and Gas. Origin of Oil and Gas. Life of Oil Wells and Fields. LIFE OF THE ORDOVICIAN . . ';* 427 Protozoa. Ccelenterata: Sponges. Graptolites. Stromatopora. Corals. Echinodermata : Cystoids. Crinoids. Blastoids, Starfish, Brittle Stars, and Sea Urchins. Molluscoidea: Brachiopods. Bryozoa. Mollusca: Pelecypods. Gastropods. Cephalopods. Crustacea: Trilobites. Other Arthropods. Fishes. PLANTS '.. V "... '. V '. .".. . . . ...... 436 Seaweeds. SUMMARY 436 Progress and Character of Ordovician Life. Climate and Duration of the Ordovician. CHAPTER XVII THE SILURIAN PERIOD SILURIAN PHYS CAL GEOGRAPHY 439 Geography of the Silurian. Character and Thickness of the Sediments. Clinton Iron Ore. Deserts. Origin of Rock Salt. Igneous Rocks. Other Continents. LIFE OF THE SILURIAN 444 Ccelenterata: Corals. Other Coelenterates. Echinodermata: Crinoids. Cystoids. Molluscoidea: Brachiopods. Bryozoa. Mollusca: Gastropods. Pelecypods. Cephalopods. Arthropoda: Trilobites. Eurypterids. Scorpions. Fishes. SUMMARY 450 Life on the Land. Migration. Climate and Duration. Close of the Silurian. CHAPTER XVIII THE DEVONIAN PERIOD DEVONIAN PHYSICAL GEOGRAPHY 452 Subdivisions of the Devonian. Geography. The Devonian in New York. Continent of Appalachia. Igneous Rocks. Devonian Oil and Gas. Devonian of Other Continents. LIFE OF THE DEVONIAN . . \ ...- :.,..'.. . ... . 456 Ccelenterata: Corals. CLELAND GEOL. 2 CONTENTS Echinodermata: Crinoids. Blastoids. Molluscoidea and Mollusca: Brachiopods. Bryozoans. Pelecypods. Gas- tropods. Cephalopods. Arthropoda: Trilobites. Barnacles. Eurypterids. Insects. Fishes: Ostracoderms. Sharks. Lungfish. Ganoids. Teleosts or Bony Fish. Comparison of Devonian and Modern Fish. Why the Verte- brate Type was " Fit." PLANTS . . . . . . . . . " . . . . .467 SUMMARY . . . . ..-.-.. . . 467 Migration and Evolution. Climate and Duration. CHAPTER XIX THE CARBONIFEROUS PERIODS MISSISSIPPIAN OR LOWER CARBONIFEROUS . . ...... . . .469 Close of the Mississippian. Other Continents. PENNSYLVANIAN OR UPPER CARBONIFEROUS 472 COAL FIELDS OF NORTH AMERICA 473 Productive Coal Fields : Eastern Canadian and New England Fields. Appalachian Field. Michigan Coal Field. The Indiana-Illinois Field. The lowa-Missouri-Texas Field. SUMMARY OF THE PENNSYLVANIAN . . . 475 Iron and Oil. Duration. Other Continents. PERMIAN 476 Permian Glaciation. Permian Deserts. Igneous Activity. Appalachian Deformation. Age of the Deformation. Other Continents. INVERTEBRATES OF THE CARBONIFEROUS . . 480 Protozoans. Coelenterates and Echinoderms. Molluscoids. Mollusks. Arthropods. Insects. VERTEBRATES OF THE CARBONIFEROUS . . .. . , . . . 485 Fishes. Amphibians. Origin of Amphibians. Rise of Amphibians. Reptiles. Rise of Reptiles. CARBONIFEROUS PLANTS 491 Ancestral Ferns and Seed Ferns. Club Mosses (Lycopods). Sphen- ophylls. Horsetails (Calamites). Cordaites and Other Gymnosperms. Conditions under which the Coal Plants Grew. COAL . . . . . ; . ; ; . . : -. . . . . 499 Mode of Occurrence. Origin of Coal. Necessary Conditions for Coal Formation. How Vegetable Tissue Accumulated. How it was Kept from Decay. How it was Changed to Coal and What Varieties Resulted. Conditions Favoring Coal Formation in the Pennsylvanian. Extent and Structure of Coal Beds. Climate during the Deposition of Coal. PROBLEMS OF THE PERMIAN . . 504 SUMMARY OF THE PALEOZOIC ERA . . . . . . . . . - . . . 505 The Building of the Continents. Evolution and Extinction of Life. Cli- mate. CONTENTS 17 CHAPTER XX THE MESOZOIC ERA: THE AGE OF REPTILES PAGE SUBDIVISIONS OF THE MESOZOIC . . . . .- . . . . . . . 508 PHYSICAL GEOGRAPHY DURING THE MESOZOIC . . .. .. . ,. . . .508 TRIASSIC ..... . 508 Atlantic and Gulf Coasts. Western Interior. Pacific Coast. Triassic in Other Continents. JURASSIC 512 Atlantic and Gulf Coasts. Western Interior. Mountain Forming in the West. Jurassic of Other Continents. LOWER CRETACEOUS (COMANCHEAN) . 514 Atlantic and Gulf Coasts. Western Interior. Pacific Coast. Lower Creta- ceous of Other Continents. UPPER CRETACEOUS (CRETACEOUS) . . . . '. . .,, ....... . . SI 6 Atlantic and Gulf Coasts. Pacific Coast. Western Interior. Upper Cretaceous of Other Continents. The Cretaceous Peneplain. Moun- tain-making Movements at the Close of the Mesozoic. Duration of the Mesozoic. LIFE OF THE MESOZOIC 521 Comparison of the Life of the Paleozoic and the Mesozoic. Plan of Study. INVERTEBRATES . ... 523 Chalk. Sponges. Corals. Crinoids. Sea Urchins (Echinoids). Starfish. Brachiopods. Pelecypods. Gastropods. Cephalopods. Ammonites. Naked Cephalopods (Belemnites). Crustaceans. Insects. FISHES AND AMPHIBIANS . 533 REPTILES 536 Reptiles with Mammalian Characters. DINOSAURS 539 Carnivorous Dinosaurs. Unarmored Quadrupedal Dinosaurs (Sauropoda). Unarmored Bipedal Herbivorous Dinosaurs (Unarmored Predentata). Armored Dinosaurs (Armored Predentata). Summary of Dinosaurs. Migration and Extinction of Dinosaurs. Size as a Factor in Ex- tinction. CROCODILES . ....". 552 MARINE REPTILES 552 Ichthyosaurus. Plesiosaurus. Mosasaurus-(Sea Lizards). Turtles. FLYING REPTILES (PTEROSAURS) . 558 TOOTHED BIRDS .560 Archaeopteryx. Hesperornis. Ichthyornis. MAMMALS . . ... '. '. .' . 563 PLANTS '.' '." . 565 Horsetails. Cycads. Ferns. Gymnosperms. Angiosperms. CLIMATE . . . . . '. . . . 569 Triassic. Jurassic. Cretaceous. COAL . . -. . . .. . .571 Triassic. Cretaceous. !8 CONTENTS CHAPTER XXI CENOZOIC ERA: THE AGE OF MAMMALS. TERTIARY PERIOD COMPARISON OF THE LIFE AT THE CLOSE OF THE MESOZOIC AND THE BEGINNING OF THE CENOZOIC . 57* Subdivisions of the Cenozoic Era. PHYSICAL GEOGRAPHY OF THE TERTIARY 574 EOCENE ' .... 574 Atlantic and Gulf Coasts. Pacific Coast. Western Interior. Eocene of Other Continents. OLIGOCENE 577 Atlantic and Gulf Coasts. Western Interior. Pacific Coast. Oligocene of Other Continents. MIOCENE 579 Atlantic and Gulf Coasts. Economic Products of the Miocene. Western Interior. Pacific Coast. Mountain Building. Basis for Separation into Periods. Igneous Activity. Miocene of Other Continents. PLIOCENE 585 Atlantic and Gulf Coasts. Western Interior. Pacific Coast. Pliocene Elevation. High Plains and Bad Lands. Pliocene of Other Continents. LIFE OF THE TERTIARY 590 Rise of Mammals. Archaic Mammals of Ancient Ancestry. Amblypoda. Ancestors of the Carnivores. Marine Mammals. Zeuglodon. An- cestors of Existing Whales. Ancestors of the Hoofed Mammals (Ungulates). Divergence of the Even and Odd-toed Hoofed Mammals (Ungulates). FACTORS IN THE EVOLUTION OF MAMMALS 600 Mammalian Teeth. Feet. Limits to Evolution. ODD-TOED MAMMALS (PERISSIDACTYLS) 603 Titanotheres. Rhinoceroses. Tapirs. Horses. Summary of the Evolution of the Horse. Probable Cause of the Evolution of the Horse. Cause of the Extinction of the Horse in North America. Elephants. Sum- mary of the Evolution of the Elephant. EVEN-TOED MAMMALS (ARTIODACTYLS) . . ... * . . .615 Camels. Deer. Cattle, Sheep, and Goats. Swine and Related Animals. A Climbing Ungulate. INSECTIVORES . . . . . 620 RODENTS (GNAWING ANIMALS) . . . . . . . 620 EDENTATES 621 TRUE CARNIVORES . . . . .622 PRIMATES (MONKEYS, APES, LEMURS) 622 BlR s . . . . | . '. '. . 623 REPTILES AND AMPHIBIANS ..' 624 FlSHES 625 CONTENTS I 9 PAGE INVERTEBRATES . ... .,., .. 626 Insects. Horseflies, Tsetse Flies, and Ants. VEGETATION 630 Grasses. Daemonhelix. Geological History of Sequoias. Diatoms. Ex- ceptional Preservation of Plants. Plant Localities in North America. CLIMATE 634 Difficulty in Determining Tertiary Climates. Eocene. Oligocene. Mio- cene. Pliocene. EFFECTS OF ISOLATION AND MIGRATION 636 Eocene Invasion. Oligocene Invasion. Miocene African Invasion. Plio- cene South American Invasion and Intermigration between the Old and New Worlds. Duration of the Tertiary. CHAPTER XXII QUATERNARY CHANGES AT THE CLOSE OF THE TERTIARY 643 Elevation. Glaciation. Changes in Life. DISTRIBUTION OF THE ICE SHEETS 645 Other Continents. North America. DEVELOPMENT OF THE ICE SHEETS . 647 Thickness of Ice Sheets at Center. GLACIAL AND INTERGLACIAL STAGES 648 Characteristics of Former Drift Sheets. HISTORY OF THE GREAT LAKES 651 Preglacial Drainage. Origin of the Basins. Great Lakes Stages. The Champlain Subsidence. OTHER PLEISTOCENE LAKES 656 Lake Agassiz. Lake Bascom. Great Basin Lakes. LOESS . % . 657 DURATION .-.?.-; 658 CAUSES OF GLACIATION ''.-''--.. . . . . 660 Elevation. Astronomical. Atmospheric Hypothesis. EFFECTS OF GLACIATION 662 LIFE OF THE PLEISTOCENE 663 Interglacial Deposits. North and South Migrations during Glacial and Interglacial Times. Deposits beyond the Ice Sheets or Protected from Them. Deposits on the Last Drift. Vegetation. Mammoths and Mastodons. Edentates. Pleistocene Carnivores. Horses, Camels, etc. Birds. PREHISTORIC MAN 674 Eolithic. Paleolithic Man. Neolithic Man. Man in North America. Birthplace of Man. Effect of the Advent of Man. FUTURE HABITABILITY OF THE EARTH 683 APPENDIX COMMON MINERALS . 685 INTRODUCTION GEOLOGY is the science that treats of the earth and its inhabitants as revealed in the rocks, and therefore deals with its constitution and structure, with the operation of the forces which led to its present condition, and with the occurrence and evolution of its life. In the search for this knowledge it calls to its aid astronomy, chemistry, physics, and biology. Geology is, in fact, a composite science, making use of the physical sciences in unrolling the complicated history and structure of this planet. Because of the breadth of its scope, geology has been divided into a number of branches which are, however, in such a large measure interdependent, that a general knowledge of all is often essential to a thorough understanding of any one. Astronomic or Cosmic Geology. Since the earth is one of the planets of the solar system, all theories of its origin must at the same time consider the origin of the other planets, and vice versa. Con- sequently, astronomy and geology are dependent upon one another in all attempts at determining the genesis of the earth. Structural Geology is a study of the materials of which the earth is built and their arrangement, and is especially concerned with the interpretation of the structures produced in the rocks by earth movements. The branch of geology which investigates minerals is mineralogy, and that which deals with rocks is petrology (Greek, petros, rock, and logos, discourse). Both mineralogy and petrology are closely allied to chemistry and optical physics. Dynamical Geology is a study of the agencies that have produced geological changes, together with their laws and modes of operation. Among the most important forces considered are water in motion, wind, glaciers, igneous activity, and earth movements resulting from strains. This branch of the subject is closely related to physio- graphic geology, since the latter deals with the evolution of the topography of the earth's surface and with the forces which have produced it. A study of physiographic geology is necessary to an interpretation of land surfaces. 21 22 INTRODUCTION Industrial Geology includes mining and economic geology and is the commercial application of geological principles. All of the above subjects are included in Part I of this volume, under the head of Physical Geology. Historical Geology includes paleontology (Greek, palaios, ancient, ontos, living being, and logos, discourse), or the study of the life of the past as shown by its fossil remains ; or, in other words, fossil botany and zoology. It also embraces p ale '0 geography (the geography of pre- historic lands), which is concerned with the boundaries of the lands and seas of the epochs and periods of the past, and with the evolution of the continents. It also includes stratigraphy, or the arrangement and succession of the strata, as indicated mainly by fossils. His- torical geology calls to its aid all other branches of geology, in order that the topography of the ancient land surfaces and the boundaries of the lands and seas may be known, and that the climates to which the earth was subjected may be determined. Such an exhaustive study is necessary, since the causes of the rapid extinction of certain forms of life, and of the sudden appearance and evolution of others, cannot be known with certainty until the environment under which they lived is learned. In general, it may be said that Historical Geology deals with the evolution of the continents and of the life of the past. Length of Geological Time. Without an appreciation of the vastness of geological time (p. 417) as compared with the brief span of a man's life, the work accomplished by the various geological agents cannot be understood. This conception of the length of geological time can, perhaps, best be grasped by a comparison: " Let a year be represented by a foot ; the average length of a human life is then measured by the breadth of a dwelling house, and human history is limited approximately to a mile; but the duration of geologic time is comparable to the circumference of the globe." Present Status of Geology. Much of the science of geology is definitely known and has been learned as a result of the accurate observation and careful reasoning of many geologists. It should, however, be borne in mind that many of the theories are subject to change, as will be pointed out from time to time in the following chapters. This is due to the fact that geology deals with many problems concerning which our knowledge is as yet incomplete, notwithstanding careful observations and deductions. For example, before it was known that the crust of the earth is heated, to some INTRODUCTION 23 extent, as a result of the radioactivity of certain minerals, a correct theory of the earth's interior was impossible. The true theory has probably not yet been found, but every advance in knowledge brings the solution nearer. The modification of geological theories from time to time should not be a source of annoyance to the student, but should rather serve to stimulate him to reason for himself. Fundamental Terms. There are a few terms with which the student must become familiar before a discussion of the subjects taken up in the following chapters can be understood. Of these terms only very elementary definitions will be given in this place, since more complete explanations will be taken up later. Rocks. With the exception of a comparatively thin layer of soil, which varies greatly in thickness and is entirely absent in some places, the earth is composed of rock which extends from the surface downward for many miles (the lithosphere), and probably through the central core (the centrosphere). In general, the rocks of the earth's crust can be classified according to their origin as of three kinds : (i) sedimentary, (2) igneous, and (3) metamorphic. Sedimentary Rocks. If one examines the sediment deposited by a muddy rivulet in a temporary pool of water, he will find that it consists of sand or clay, and that it is in layers. This deposition FIG. i. Niagara limestone, showing well-bedded layers with two sets of strong joints at right angles to each other. (U, S, Geol. Surv.) 24 INTRODUCTION represents on a minute scale what is occurring in the lakes and oceans of the earth. In the pool the deposit may be only a fraction of an inch thick, while off the shores of the ocean it may be many thou- sands of feet in depth. When, as has often occurred in the past, such an enormously thick deposit has been raised above the sea and streams have cut deep valleys into it, it is seen to be made up of layers, and the rocks composing it are consequently called stratified (p. 233) or layered rocks (Fig. i). The planes which separate the layers from one another are called bedding planes or planes of stratification. If the rock is made of hardened mud it is called shale, and is usually divided into many thin layers or lamina, the laminae being separated by planes of bedding. If the rock is composed of sand whose grains are cemented together by lime or other substances, it is called a sandstone. Sandstone is also stratified, but the bedding planes are usually farther apart than in shale. Sedimentary rocks are not always in the horizontal position in which they were deposited, but are often folded and tilted. Igneous Rocks. Although few persons have seen lava flowing from a volcano, many have seen the molten slag or glass of blast furnaces, which bears a resemblance to lava after hardening, and is, in fact, not unlike the lava of some volcanoes, both in composition and structure. Lava (p. 298) is an igneous rock (Latin, ignis, fire) ; that is, a rock which has been in a molten condition. The majority of igneous rocks, however, are not glassy, but are composed of dis- tinct grains or crystals. This crystalline structure, as we shall learn (p. 302), is brought about when molten rock cools so slowly that time is given for crystals to form. An igneous rock is, therefore, one which solidified from a state of fusion ; it is either glassy or grained (crystalline). It is apparent, therefore, that igneous rocks differ from sedimentary in a number of particulars ; the former are either glassy or crystalline and are devoid of stratification planes, while sedimentary rocks are seldom crystalline and are arranged in layers. Granite is a typical crystalline, igneous rock (Fig. 5, p. 29) and is composed, usually, of three minerals, the most conspicuous of which is feldspar. These feldspar grains or crystals are opaque, and white, pink, or gray in color. Mica, when present in granite, is usually easily recognizable by its glistening leaves which split into elastic scales. The third conspicuous mineral of granite is quartz. It is usually colorless, has the appearance of broken glass, and is harder INTRODUCTION than steel. A crystalline igneous rock is, therefore, made up of a number of minerals differing in color, in hardness, and in chemical composition. The importance of this character will be seen when the effect of the weather upon rocks is studied. Metamorphic Rocks are those which have been more or less pro- foundly changed from their original condition by heat and pressure, and are usually crystalline in texture. Most metamorphic rocks possess a cleavage which causes them to break easily in one direction. They are derived both from igneous and from sedimentary rocks. Metamorphic rocks have parting planes like sedimen- tary rocks, and a crystalline structure like igneous rocks. Divisional Planes. All rocks are more or less broken by planes which separate them into blocks. An ex- amination of a sandstone or limestone quarry will show that, in addition to the bed- ding planes, the rock is broken by two or more sets of fissures which run at right angles to the bedding. These are called joints (Fig. i). Joints occur also in igneous rocks, some often being ap- proximately horizontal and others tending towards the vertical. When beds are displaced along a joint or other crack so that the strata on the opposite sides of it do not match, the beds are said to be faulted (p. 261) (Fig. 2). FIG. 2. A fault. The thin-bedded band was once continuous. (U. S. Geol. Surv.) PLATE II. Pyramid Lake, Nevada. A lake resting on the floor of a steep-sided cirque. Lakes of this origin are common in the high mountains of the United States and Canada, where glaciers formerly existed. 26 PART I. PHYSICAL GEOLOGY CHAPTER I WEATHERING NOTHING endures. The most indestructible rock will, in time, dis- integrate ; the mountain peaks will crumble away, and the rough places will be made smooth. The forces which produce these results are calleH the agents of weathering. They vary in their effectiveness in different places and at different times in the same place, but under all conditions and at all times some agent is at work, reducing the exposed rock to soil. The rate at which rock weathers depends largely upon two factors: (i) the composition and structure of the rock, and (2) the physical conditions to which it is exposed. A sandstone (p. 36) in which the grains are held loosely together will disintegrate rapidly, while another, in which the cementing material is insoluble and abundant, may have a long life. The effect of dif- ferent physical conditions is obvious. In the dry regions of Mexico and Arizona churches and houses built of sun-dried brick (adobe) have lasted for several centuries ; houses made of a similar material would, in New England, crumble to a mound of clay in the course of a few years. MECHANICAL AGENCIES i. Frost. The property possessed by water of expanding upon freezing is of great importance in the disintegration of rocks in regions where the temperature falls below the freezing point, since upon freezing it expands one tenth and exerts the enormous pressure of 150 tons to the square foot. This force is well illustrated in the bursting of water pipes in which water has frozen. It is stated that in Finland freezing water is sometimes used instead of powder, and blocks of stone of 400 tons' weight are broken out in this way. All rocks, even the most dense, contain pores and fissures in which water may accumulate. Certain sandstones when weighed, then 27 PHYSICAL GEOLOGY FIG. 3. Sandstone ''set on edge," showing the scaling of the laminae by frost. soaked in water for twenty-four hours and again weighed, are found to have gained one eighth in weight. This fact shows not only that pores are present, but also that they have become filled with water. If such a rock, with its pores full of water, is frozen re- peatedly, its grains will be forced apart until it finally falls to pieces. This test is used in labo- ratories to determine the desirability of building stone in temperate regions. The complete disintegration of the rock is not attained by one freezing and thawing, but if the process is repeated many times, the rock may be completely re- duced to sand, as the water penetrates farther into the rock each time it thaws. It is readily seen that this wedge work of ice is usually more im- portant in moist, temperate regions in early and late winter. The obelisk which now stands in Central Park, City of New York, stood for many centuries in Egypt without apparent injury (although undoubtedly weakened by the extremes of daily temperature to which it had been so long subjected). But after one year's exposure to the moist, changeable climate of its new home, the hiero- glyphics near its base became almost illegible, and it was found necessary to coat the FIG. 4. Much-fractured limestone with strong bedding planes, illustrating the con- ditions favorable for the wedge work of frost and roots. The beds are bent into a low anticline. WEATHERING 29 monument with paraffine dissolved in turpentine to prevent further decay. Few, if any, tombstones in New England which are over one hundred years old show a polish. A marble slab at North Adams, Massachusetts, for example, upon which an inscription was chiseled in 1865 was practically illegible in 1905 and had to be recut. Such rapid weathering of marble as this is, however, unusual, Tombstones of red sandstone which were erected with the bedding FIG. 5. Granite broken by two sets of joints. Animas River Canyon, Colorado. (U. S. Geol. Surv.) planes l on edge have suffered severely, because the rain water has soaked down the more porous layers, and upon freezing has forced off sheets of the rock (Fig. 3). Exposed rock surfaces in polar regions are pulverized by frost, producing sand, and fine, dusty material which is shifted by the winds. Talus. 2 The most conspicuous effect of frost is seen in rock masses which are much broken by cracks and joints (p. 258) (Figs. 4, 5). 1 When rock is arranged in layers it is said to be bedded, and the planes separating the layers are called bedding planes. For a discussion of such rocks (stratified rocks) the student is re- ferred to page 233. 2 Talus slopes produced in arid regions by changes in daily temperatures are also important. 30 PHYSICAL GEOLOGY In such cases, blocks are forced from cliffs, building up slopes of loose fragments at their base, called talus. The formation of talus slopes (Fig. 6) can best be studied in the early spring, when fragments of the rock, loosened as the ice in the cracks melts, fall from the cliffs. These fragments are carried down by their weight until the declivity is too feeble for them to roll farther. When they come to rest they accumulate to form a slope, usually steep, whose angle is called " the angle of repose." The slope of talus (Fig. 14, p. 39) varies from 26 to 43 degrees, the angle depending upon the size of the fragments and upon their shape. If the fragments are angular, the slope will be FIG. 6. Diagram showing the formation of a talus slope and the destruction of a cliff. The successive faces of the cliff A, B, C, D and of the talus A, B y C, D are indicated by dotted lines. steeper than if they are rounded, since with the former an early lodg- ment is more likely. The largest blocks accumulate at the foot of the talus slope, the size diminishing regularly to the top. If talus accu- mulates under water the slope will be steeper, since the fragments are, to some extent, buoyed up by the water. It is evident that the alternate freezing and thawing of water in the pores and joints of a rock may bring about its complete disintegration unless it is protected by the soil of its own making. It should be remembered, however, that this agent seldom acts alone, but usually serves as an aid to the chemical agencies of weathering (p. 35), by breaking up the rock into small fragments and thus furnishing a larger surface upon which the latter may work. Rock Glaciers. A striking example of the above is shown in the formation under favorable conditions of rock glaciers or " stone rivers " (Fig. 7), many hundred feet in length in regions of severe' cold, when great masses of talus are slowly moved some distance down a valley, producing the appearance of a glacier. This is accomplished by the alternate freezing and thawing of the water in the interstices of the talus. WEATHERING Creep of Soils. Another important result of frost action is the creep of soils on slopes. If soil contains water, each freezing slightly raises the fragments at right angles to the surface of the hill, and each thawing permits gravity to pull them down hill. If the process is often repeated, the soil moves slowly down the slope. In the course of many years, many tons of earth may be thus carried to a lower level. 2. Changes in Daily Temperature. In regions where the air is dry and clear the radiation of heat is rapid and the range in daily temperature is wide, often varying 80 F., while in the Sahara Desert a change of 131 F. within a few FIG. 7. Rock glacier, McCarthy Creek, Alaska. The talus forming the rock glacier is derived from the high cliffs (cirque). (U. S. Geol. Surv.) hours has been re- corded. In such re- gions, the naked rocks are heated to a high temperature during the day and are cooled rapidly at night. Since rocks are not good con- ductors of heat, the side of the rock exposed to the sun's rays is often raised to a temperature of 120 F. or more during the day, while a short distance beneath the surface the rock is still cool. The result is that the outside shell is expanded, while the interior is still con- tracted. Strains are thus produced which tend to break off fragments of the rock, dark-colored rocks being particularly affected, since they absorb more heat. In the late afternoon and night, on the other hand, when the temperature falls, the interior which had been grad- CLELAND GEOL. 3 PHYSICAL GEOLOGY ually acquiring heat during the day is still warm when the surface is cool and contracted. The contracted exterior is then too small for the still expanded interior and the surface of the rocks tends to shell off (Fig. 8), forming onion-like, concentric layers, the process being known as exfoliation. It is stated that in certain parts of Africa the rock temperature rises to a height of 137 F. during the day and falls so rapidly at night as to throw off, by contraction, masses as much as 200 pounds in weight. Slabs of granite 8 to 10 inches thick and 10 feet long are known to have been broken off by changes in daily temperature. In the western part of the United States, where the climate is too dry to afford much scope for the operation of frost, cliffs are slowly disintegrated by these changes in tem- perature, producing talus slopes of large size. The alternate heat- ing and cooling of a rock causes its disin- tegration in still an- other way. When a rock is composed of minerals differing in color and composition, it is especially liable to disintegration by changes in daily temperature. Since dark-colored minerals absorb heat more rapidly than light-colored ones and also radiate it more quickly, rocks containing both expand and contract at different rates, with the result that the grains are gradually loosened until the surface is reduced to sand. The fact that the coefficients of expansion of the various minerals differ widely also aids in the dis- integration of the rock. Igneous rock, 1 composed of minerals of dif- ferent kinds, is therefore more easily disintegrated by this process than rocks made up of one mineral. Changes in daily temperature are especially effective in high alti- FIG. 8. Exfoliated granite. (U. S. Geol. Surv.) 1 Igneous rocks are those which have been formed from molten masses by cooling. See page 329 for a discussion of these rocks. WEATHERING 33 tudes, and mountain peaks often owe their jagged shapes, to some degree, to this action, although more largely to that of frost. In regions of deficient rainfall, talus accumulates at the foot of cliffs, the fragments forming the slope having been broken off by temperature changes. Mountains in desert regions are sometimes almost buried beneath rock fragments and sand, broken from their sides by changes in daily temperature. When the heated rocks of arid regions are wet by a sudden down- pour, they cool quickly and are broken asunder. In western Texas blocks 25 feet in diameter are reported to have been rent into several pieces in this way. (Hobbs.) 3. Mechanical Action of Animals and Plants. If one observes a cliff upon which vegetation is abundant, he will see that not only the large but also the small cracks of the rock are filled with roots and rootlets. As these roots and rootlets grow larger they tend to push the blocks of rock apart. The root of the garden pea, for instance, has a wedging force equal to 200 or 300 pounds a square inch. Abundant examples of this wedging process can be found in fertile regions, and are also often seen in cities, where the pavements are frequently broken and tilted by the enlarging roots of trees. Plants, earthworms, and burrowing animals open channels through which water from the surface can reach deep down into the soil. Moreover the organic matter carried into the tunnels by the animals is, upon its decay, a source of organic acids which actively attack the rocks and thus hasten the decomposition of those otherwise pro- tected by soil. It is thus seen that the mechanical disintegration of rocks is accomplished both by the agents of the weather and by organisms. 4. Rain. The mechanical effect of rain consists in (i) the impact of the raindrops upon the surface, which in the aggregate has a con- siderable effect, as, for example, in gravel deposits where the larger bowlders protect the gravel underneath from the impact of the rain, while that which is not so protected is removed. In the process of time, columns a score or more feet in height may result (Fig. 9). On a small scale, this same result can be seen in almost any soft material after a rain. (2) The mechanical work of rain is seen also in the softening of clay soils, which, on slopes, causes them to creep; (3) in the washing and later deposition of dust from the atmosphere, and (4) in the dissolving of some of the atmospheric gases which may later be used in weathering the rocks. Dew and hoarfrost, being 34 PHYSICAL GEOLOGY condensed from the lower layers of the air, absorb and furnish to the soil more gases and inorganic matter than rain. (5) Rain water is also effective in causing certain rocks to swell. In excavating the Panama Canal it was found that certain rocks which, when first uncovered, had to be blasted before they could be removed, be- came so soft after a few months' exposure to the tropical rains that they could be excavated with the steam shovels. The slides which have oc- curred in the Culebra Cut were due both to the softening of the rock in this way and to gravity FIG. 9. The work of rain water in sculpturing which tends to cause the TiroT k f Carth C ntaining b wlders ' near Bogen ' rock to move toward the excavation. The soften- ing action is taken advantage of in extracting diamonds from the inclosing rock in the South African mines. 5. Wind. The mechanical work of the wind carrying sand is very effective in wearing away rock, especially in arid regions, and will be discussed on another page (p. 45). 6. Lightning. When lightning strikes the earth it sometimes fractures large masses of rock. When it strikes sand, drops or bubbles of glass and irregular tubes or rods are sometimes formed by the partial fusion of the soil. These fulgurites, as they are called, are seldom more than a few inches long, but are sometimes several feet in length and two and a half inches in diameter. The entire summit of Little Ararat in western Asia, where electrical storms are extremely common, is said to be drilled by lightning. " A piece of rock about a foot long may be obtained, perforated all over with irreg- ular tubes, having an average diameter of three centimeters. Each of these is lined with a blackish-green glass." (A. Geikie.) This is, however, unusual, and the total effect of lightning is inconsiderable. WEATHERING 35 CHEMICAL AGENCIES The chemically active gases carbon dioxide. Un- less they are dissolved in water, however, their effect in the weathering process is unimportant, but in the presence of both moisture and heat they accomplish a great part of the work of chemical disinte- gration. It is evident, therefore, that the chemical decomposi- tion of rocks must vary greatly in effec- tiveness in different places and at different times in of the atmosphere are oxygen and FIG. ii. Joints in limestone, widened by solution. (Photo. H. L. Fairchild.) FIG. 10. Limestone bowlder channeled by water containing carbon dioxide. the same place, and we find that it is most active in moist, tropical regions, less rapid in temperate regions, and least im- portant in the frigid zones and in arid regions. i. Solution. Pure water is a poor sol- vent, but when it contains a consider- able quantity of car- bon dioxide its sol- vent power becomes greatly increased, so that limestone, gyp- sum, and other easily soluble rocks are slowly taken up by 36 PHYSICAL GEOLOGY it and carried away. Water obtains its supply of carbon dioxide from the air, from the decay of plants and animals, and from subterranean sources (p. 296). It has been estimated that the surfaces of certain limestones in England have been lowered at rates varying from one inch in 24 years to the same amount in 500 years. Although solu- ,---' ,---> r --J X -J - jr _i r -l... tion is most conspicu- ".'."]._" .'-~i~' '."!" . 1 ."."_V.r" --=--- --T--J.-A ously exhibited in limestone regions, B where the rock is "i ' i,,i C often furrowed by the FIG. 12. The formation of the residual soil B from rivulets (Fig. lo) the limestone A is shown. The soil was derived from which flow over the the limestone by the removal of the soluble portions and /. i i the concentration of the insoluble. Large areas of Ken- surtace > and the joints tucky and Virginia owe their fertility to this process. and other cracks are widened by its action (Fig. n), it is also effective on feldspar and even on quartz. Sandstones with calcareous cements are disintegrated by the solu- tion of the cement, causing the rock to fall to pieces and form sand. In regions of impure limestone the insoluble residue, such as clay and flint nodules (p. 77), will be left, covering the unweathered rock (Fig. 12). The depth of this cover often gives a basis for estimat- ing the thickness of limestone which has been dissolved and carried away. Many caves are formed by solution (p. 70). 2. Oxidation. Oxygen is effective only on rocks which contain minerals capable of taking up further oxygen and thus forming new compounds. The most important of these are iron compounds, and to them the red and yellow coloring, so conspicuous in rocks, is due. If oxygen alone is added to the iron molecule, a red color (Fe2Os) results; if moisture is present, however, the brown or yellow rust (hydroxide), common in moist regions, is formed. One noticeable result of oxidation is an increase in volume ; this being the case, the newly formed and bulky minerals crowd the grains of the rock apart and tend to produce disintegration. Complex silicates, such as feld- spar, mica, and hornblende (p. 690), are attacked by oxygen and carbon dioxide, and reduced to simpler and more stable compounds. 3. Hydration. The union of water with chemical compounds is known as hydration^ and is very important in weathering. An im- portant effect is the increase of the volume of the mineral acted upon. The operation of hydration and oxidation is well illustrated in the WEATHERING 37 weathering of iron pyrite (FeS 2 ) (p. 686) which often occurs dis- seminated through rocks. The first, and usually most conspicuous effect is the appearance of a yellow stain on the rock. If the pyrite is abundant, hydration may cause the rock to fall to pieces as a result of the increase of volume and of the formation of sulphuric acid. Building stones which are uniform in color when first quarried some- times become discolored, after an exposure of a year or more, by blotches of brown stain. Upon examination, it is usually found that the stain was formed from the weathering of small crystals of pyrite. From these blotches the stain spreads, sometimes covering an area of 100 or more square inches. 4. Carbonation. By the union of carbon dioxide, derived from the air and soil, with the calcium, magnesium, or iron of complex silicates, soluble compounds are formed which upon being carried away in solution cause the rock to crumble. This is an important cause of the disintegration of granite, although oxidation and hydra- tion are also effective in the same process. If organic acids derived from decaying vegetable matter are present in water, they tend to decolorize red and yellow rocks. Such decolorization can often be seen where water trickles over cliffs. For example, the red cliffs of the Vermilion River in northern Ohio are bleached wherever rivulets trickle over them. This is accomplished by the union of the carbon dioxide with the oxides of iron which gave the red and yellow color to the rock. 5. Organisms. Although not agents of the weather, the chemical action of plants and animals should be considered in a dis- cussion of rock disintegration. Certain bacteria are found in great numbers on the surface of bare rock. They live not only in low, moist regions, but even on mountain peaks, where they have been found coating the surfaces and crevices of the rocks. They draw their nourishment from the nitrogen and other compounds brought down in snow and rain. Rocks are attacked by the nitric acid which these bacteria form from the ammonia of the air and water. The chemical action of their excretions makes them an important though incon- spicuous agent of disintegration. Other organisms, such as lichens, mosses, and flowering plants, contribute to the decomposition of rocks. The roots of trees not only pry the rocks apart (p. 33), but they also act chemically by producing carbon dioxide and organic acids, which dissolve the lime and transform the silicates into car- bonates and other products. 38 PHYSICAL GEOLOGY Comparison of Effects of Chemical and Mechanical Weathering. Chemical decomposition of rocks is slow and long continued as compared with mechanical disintegration, which is a rapid process. By the former the rocks are broken up into fine particles, and by the latter into larger and smaller fragments. Chemical action is not only long continued, but is also more universal than mechanical action, being important under all climates, except in desert regions and on mountains where mechanical disintegration is so rapid that sufficient time is not permitted for conspicuous chemical action. Chemical decomposition tends to smooth surfaces, while mechanical disinte- gration tends to roughen them. Where the mechanical predominates, the slopes are stronger and tend to forrn cliff's. RESULTS OF WEATHERING Some of the most conspicuous features of scenery are produced by weathering. These features are seldom due to a single agent, but more often to two or more acting in conjunction. FIG. 13. Pinnacle Peak, Canadian Rockies. The ragged outlines are due largely to frost work. (Photo. M. H. Smith.) The rough and jagged peaks so characteristic of high mountains have been sculptured largely (i) by frost and (2) by changes in daily temperature (p. 31) (Fig. 13). The debris derived from such peaks WEATHERING 39 may accumulate in the valleys and on the sides of the moun- tains to great depths, in some cases more than a thousand feet. Where the supply of talus is too great for the stream in the valley to remove, the stream is dammed and a lake is formed (Fig. 14). The form of the crests and cliffs of high mountains is determined to a large degree by cracks which have a uniform direction (joints), as. when the water which fills FIG. 14. Drawing showing a stream so dammed by talus as to form a lake. them freezes, the rock is broken Note the angle of the talus slope, off along these planes. In tropi- cal regions the fractures of the rock are also important, since they permit the access of water, and chemical decomposition is therefore accom- plished at greater depths at these points. It is often possible to state from the shape of the topography in any one locality what the nature of the underlying rock is, but a general rule is impossible, since the same rock is differ- ently affected by the weather under different climates. For example, granite rocks which in a cold climate may be broken into jagged crests, may, in moist, tropical regions, be reduced to rounded forms through the chemical agencies. Spheroidal Weathering. Spheroidal weather- ing results from chemical action and should be distinguished from similar shapes which are pro- FIG ic Dia rams ^uced by exfoliation due to changes in daily showing the effect of temperature (p. 31). When water percolates weathering upon rock through the joints (p. 2J8) and horizontal planes &&(*$ ^to which all rocks are more or less divided, The corners and edges it attacks with its dissolved gases all the rock are most affected, and surfaces with which it comes in contact ; but b h e:o m b e CtS S phe n roidaT nce the corners and edges of the blocks formed (Modified after Hobbs.) by these joints and planes have a greater 4 o PHYSICAL GEOLOGY FIG. 16. Granite weathering under tropi- cal conditions. Rhodes' Grave, southern Rhodesia. The bowlders are residual frag- ments of a sheet of granite that once overlay the hill. (Photo. G. A. J. Cole.) surface exposed, they are more vigorously acted upon. Such places, too, encounter water from two or more directions and are more likely to be affected by the strongest solutions. The greater weathering of the edges, and especially of the corners, causes them to dis- integrate more rapidly, leav- ing a spheroidal core of un- weathered rock, embedded in less compact, weathered rock. When the rock is exposed to the action of wind and water, the unaltered, spheroidal core is exposed (Figs. 15, 16). Differential Weathering. When rocks are not uniform in character but are softer or more soluble in some places than in others, an uneven surface may be developed (Figs. 17, 18, 19) ; in deserts by the action of the wind and in moist regions by solution. Columns of rock which have been iso- lated in any way show the effect of differential weathering. In arid re- gions the lower parts of the columns (Fig. 20) are worn away more rapidly than the upper parts, be- cause the drifting sand is more abundant and effec- tive near the ground, and the bases grow smaller and smaller until the { - IT r I, . fie. 17. A bowlder showing differential monuments finally topple weat hering. the projecting portions are relatively over. insoluble silica, while the .main portion is limestone. WEATHERING FIG. 1 8. Differential weathering. The limestone has been dissolved, leaving the quartz veins projecting. Widening of Val- leys. Valleys are widened by the work of streams (p. 81), but a large part of the width of their upper portions is due to the work of the weather, which first disintegrates the rocks, after which rain, hillside creep (p. 31), and other agents bring the weathered material within reach of the stream which carries it away. Rock Mantle and Soil. We have seen that everywhere on the earth's surface the rock is being broken to pieces by one or more of the agents of weathering. This results in the accumulation of a mantle of rock waste which in the process of time would cover the lands to a great depth if it were not removed. The thickness of the mantle rock varies greatly. In tropi- cal regions the solid rock may not be encountered even at a depth of 150 feet, and in Washington, D.C., granite can be excavated with pick and shovel at a depth of 80 feet. In the Valley of Vir- ginia and in the Blue Grass regions of Kentucky a thick layer of soil, representing the insoluble portion of many feet of limestone, covers the underlying rock. Under normal conditions (Fig. 21) FIG. 19. The more rapid weathering of t fo e SQ {\ { s thickest on the a weak bed of limestone in the cliff has , i KOC^C nf formed the shelf. Helderberg Mountains, 'rests and * the baSCS * near Albany, New York. hills, and thinnest on the 4 2 PHYSICAL GEOLOGY FIG. 20. An erosion pillar, shaped largely by the work of wind-blown sand. Near Adamana, Arizona. slopes, since loose material has a tendency to creep down hill (p. 31). In regions which have been covered by glaciers (p. 168) the soil has often been removed from the hilltops by them. The same agencies that cause the disintegration of the rock break up the mantle rock to finer and finer par- ticles and form soil. Soil grades into the coarser sub- soil which has not yet been completely disintegrated. This subsoil is gradually brought to the surface by earthworms where the soil is clay, and by ants where it is sandy. In the aggregate, the work of these animals is important. It has been esti- mated that in England earth- worms bring 17 to 1 8 tons of material an acre to the surface each year, and that in Massachusetts ants bring up one fourth inch of earth. Leaves and other organic matter which are carried into the soil and subsoil by earthworms form organic acids which hasten the chemical disintegration of the rock. Roots of plants and overturned trees also help to mingle soil and sub- soil. The fertility of soil is greatly increased by the organic mat- ter, either animal or vegetable, which it contains, but its char- acter depends largely FIG. 21. Section showing the thickness of mantle . i i r rock on different parts of a hill. (Modified after Cham- upon the rock from berlin>) which it was derived. Kinds of Soil. Mantle rock and soil are moved by hillside creep (p. 31), by rain (p. 33), by avalanches, by landslides (p. 73), by slumping (p. 73), etc. ; all of which combine to remove it from the uplands and carry it to the valleys. There are two kinds of soil, (i) residual soil, that derived from the rock which it covers, such as that which overlies large areas where the country has not been affected WEATHERING 43 by glaciation, and (2) transported soil. Transported soils may be further classified as (a) alluvial, those which have been carried and deposited by streams, and which vary greatly in composition from the finest clay to coarse gravel ; (b) glacial soils which in any place may vary greatly, both in the character and the size of their constituents (p. 663) ; (c) soils of sand and clay deposited by the winds, such as the fertile loess of China (p. 53) and of the western part of the United States, and (d) talus soils of mountain regions. Removal of Soil. When, by deforestation, overgrazing by animals, or other causes, the vegetation which prevented the washing of the soil is removed, the soil may be carried away rapidly, and a fertile region may become almost a desert. This appears to have been true of portions of China and Greece. When the fertile soil is once removed, it is difficult for plants ever again to gain a foothold, and the region may be permanently desolated. It is stated that a single lumberman may in fifty years deprive the human race of soil that required tens of thousands of years to form. REFERENCES FOR WEATHERING BUCKLEY, E. R., Building and Ornamental Stones : Bull. Wis. Geol. and Nat. Hist. Surv. No. 4, 1899, pp. 11-34. DANA, J. D., Manual of Geology, pp. 118-129; I5 8 ~I59- DE MARTONNE, E., Geographie Physique, 1909, pp. 404-411. GEIKIE, A., Textbook of Geology, 4th ed., Vol. i, pp. 447-465. HAUG, E., Traite'de Geologic, 1911, pp. 371-401. MERRILL, G. P., Rocks, Rock- Weathering and Soils, pp. 173-285. SHALER, N. S., Aspects of the Earth, pp. 300-339. CHAPTER II WORK OF THE WIND THE conditions essential for the effective work of the wind are aridity and a scarcity of vegetation. Since such conditions prevail over more than one fifth of the land surface of the world, the work accomplished by this agent is of great importance. WIND AND SAND Wind without Sand. Wind is much less effective without sand than with it, but is nevertheless important. In semiarid regions and in those which are suffering from a long period of drought, cultivated fields may be excavated disastrously. In Wisconsin there are extensive regions of light lands which almost every year suffer from the drifting action of the wind. In these regions winds dry up the soil and sometimes sweep away the crops of grain, even after they are four inches high, uncovering the roots by the removal of one to three inches of surface soil. (King.) During the drought of 1894 * n Nebraska, the finely pulverized soil of the cultivated fields was blown out over extensive areas to a depth of two or more inches and was piled up in small dunes near fences and buildings. Blow-outs, as the pits excavated by the wind are called, are often the indirect result of the close grazing of a light soil, or are developed in land which is covered with a sparse vegetation. Blow-outs may be excavated to a depth ^ often feet or more, and at certain seasons of the year may be occupied by temporary lakes. The work of the wind in removing loose sand is termed deflation. Besides these more important effects of the wind, rocks are dis- lodged from cliffs by its force, as in the Orkney and Shetland Islands, where it is common to find pieces of flagstone or slate weighing several pounds, which have been detached from the precipices and blown upon the moors above during high gales. (A. Geikie.) Trees are blown down; water is thrown into waves; and birds, insects, and seeds are carried about. 44 WORK OF THE WIND 45 Wind with Sand. As soon as the wind picks up pieces of the mantle rock it has tools with which to work, and it becomes a geological agent whose effect in desert regions is not easily over- stated. It is from the man- tle of rock waste formed by the agents of the weather, and from the sediment car- ried to the deserts by the mountain streams that the sheets of sand which cover the deserts are made. The work of wind laden with sand is well illustrated in the artificial sand blast by which granite is polished and glass is etched in desired patterns. Telegraph poles in arid regions are often cut off near the base by wind-blown sand. Pebbles worn by the wind (Fig. 22) usually have a character- istic, brazil-nut shape (dreikanter) , the faces meeting in ridges. FIG. 22. Pebbles faceted by the abrasion of wind-blown sand (dreikanter) . FIG. 23. Surface eroded by wind-blown sand. The small table is the remnant of a once extensive bed. (De Martonne.) (See Fig 24.) This shape is explained as follows : the planing of the exposed surface of the pebble by the wind-blown sand continues until the pebble stands on a narrow base. It is then overturned by a slightly stronger gust 46 PHYSICAL GEOLOGY of wind and a new surface is exposed which is, in turn, planed by the blown sand. The pebble may be turned over by the undermining of the sand on one side, when the pebble falls into the depression thus made, and the wind is permitted to plane another surface. Pebbles of this sort are sometimes found in ancient rocks and afford evidence of the physical conditions of the time in which they were deposited. Wind scour wears away the softer strata of a desert much more rapidly than the harder, producing wide plains above which the harder rocks stand as isolated hills. In areas composed of horizontal strata the soft rocks are removed, and the region is lowered to a harder stratum, which may, in turn, be cut up and the whole region reduced to a still lower level. During this gradual lowering, the deserts are ft: FIG. 24. Diagram showing a table such as that appearing in Figure 23, formed by the wearing away of the beds A and B, by wind-blown sand. eroded, leaving extensive, flat-topped elevations capped by harder rock (Figs. 23, 24), which are later cut up into conical hills and finally destroyed. In desert regions where plains predominate, we sometimes find mountains whose bases are covered for 1000 to 2000 feet with sand, rising above the plains. (McMahon.) In places we find also bare rock exposed in the basins, showing that the wind is able to exca- vate the rock to low levels. In fact, it seems probable that so long as the ocean is held back from a desert, eolian excavation may go below sea level. The limit to eolian excavation is the level of under- ground water (p. 56), for when that is encountered, wind erosion is ineffective, since the sand is then held by the moisture and further removal prevented. Sand Dunes. When wind meets an obstacle, its velocity is lessened and, if it carries sand, some of its burden is dropped. The mounds of sand which are thus piled up by the wind are called dunes. Sand dunes are most abundant (i) in deserts, being as a rule more numerous in low-lying areas ; (2) on sandy coasts where the prevailing winds are on shore; and (3) near the beds of rivers whose volume varies, leaving broad areas of sand exposed during the dry season. Dunes of this origin are common in Nebraska, Kansas, Mexico, and many other regions. Sand dunes occur in the above-mentioned WORK OF THE WIND 47 regions because there only the wind finds sufficiently thick accumula- tions of dry sand and sufficiently extended flat surfaces for effective work. There also the winds blow in the same direction a sufficient length of time. Frequent changes in the direction of winds are as unfavorable to the development of dunes as vegetation or a rough topography. If the direction of the wind is constant, typical dunes will have a gentle slope on the windward side and a steep slope on the leeward side FIG. 25. Sand dunes. The direction of the wind was from the right to the left. (Photo. D. T. MacDougal.) (Fig. 25). This difference between the windward and leeward slopes is due to the fact that the sand is pushed up the former by the wind and dropped over the crest, where it comes to rest at a steeper angle. On the other hand, if the prevailing direction of the wind changes from season to season, the difference between the angles of the slopes will be less marked. The slope is generally steepest on high dunes, but is never greater than 10 on the windward and 30 on the opposite side. Since the winds vary greatly in velocity from time to time, the size of the sand particles carried up the dunes differs and usually produces distinct layers, or stratification. The inclination or dip (p. 252) of the stratification also varies widely in direction and steep- ness (Fig. 26), since it depends upon the direction and the force of the wind. The formation of this cross-bedding, as the layers in one CLELAND GEOL. 4 4 8 PHYSICAL GEOLOGY deposit which vary in direction are called, is clear when the conditions of their formation are considered. If the direction and force of the wind remain constant, the sand will be carried up the gentle slope and will fall over the steep slope, forming lay- ers of uniform inclination. If, however, as is nearly always the case, the force and direc- tion of the wind vary, the sand will be laid down on different sides of the dune at different times. In either case cross-bedding will be produced, but it will be more irregular in the second case than in the first. Shape and Origin of Dunes. When winds are moderate and the supply of sand is small, crescent-shaped dunes (Fig. 27 A) are formed ; if the wind is moderate but the supply of sand great, dune ridges are often developed which are at right angles to the direction of the wind (Fig. 27 B) ; while in regions where the prevailing winds are strong and the sands abundant, long ridges parallel to the direction of the wind usually result (Fig. 27 C). The crescent shape of dunes results when wind, blowing over a sandy stretch, heaps the sand into small piles. The grains of sand which are subsequently carried to the piles are deviated to the right and left, until crescents are formed. When the direction of the wind changes, the points of the crescents dis- appear and then form on the lee side in a new direction. Any obstacle, such as a bush, a rock, a fence, or even a mere rough- ness of the land surface, may cause the beginning of a dune. Dunes are also sometimes formed when the sand is wet by slow springs. These moist heaps serve to anchor additional particles of sand until a mound some feet in height is formed, which may afford lodgment for shrubs. The presence of obstacles is, however, not always essential to the formation of dunes. This can be observed on a small scale on a smooth asphalt street, where the dust is seen to be collected into small FIG. 26. A quarry in eolian limestone, Bermuda Islands. The cross-bedding was formed by the shifting winds which carried the sand. WORK OF THE WIND 49 ridges, perpendicular to the direction of the wind. The waves of the sea have the same origin as these ridges, but the molecules of the water are not carried along by ^ the wind as are the grains of sand, but after completing their orbits (p. 199), return to their original positions. Migration of Sand Dunes. Dunes are constantly migrating unless the sand of which they are composed is prevented from blow- ing by grass or other vegetation (Fig. 28). The forward move- ment is accomplished by the force of the wind, which shifts the sand of the dunes to the leeward. This can be seen on a windy day, when the crests of the dunes ap- pear to smoke. The rate at which dunes move varies, de- pending largely upon the velocity of the wind and the height of the dune, small dunes migrating the faster. In Denmark the rate is from three to twenty feet a year ; in France, on the Bay of Biscay, the sands have advanced at a rate estimated from 15 to 105 feet a year, burying in their progress forests, farms, vineyards, villages, and churches (Fig. 29). Some of these, after being buried for years, have been again uncovered by the further advance of the dunes. The church of Lege, taken down at the end of the seventeenth century and rebuilt two and a half miles inland, had again to be removed 160 years afterwards, showing an advance of the sands at a rate of 81 feet a year. (Wheeler.) On the south side of Lake Michigan forests FIG. 27. Diagrams showing the form of sand dunes. In A the wind blows from the upper left-hand corner. The supply of sand and strength of wind are moder- ate. In B the direction of the wind is from the upper right-hand corner. The supply of sand is large and the winds moderate. Under these conditions dunes transverse to the wind are formed. When the supply of sand (C) is large and the winds strong, dune ridges parallel to the direction of the wind are formed. 50 PHYSICAL GEOLOGY which were covered by sand dunes have been uncovered as the dunes moved on. There are hundreds, perhaps thousands, of square miles of buried towns and cities in Central Asia. One of the difficulties in connection with the maintenance of the Suez Canal is the sand which is constantly being blown into it from FIG. 28. Dunes held by mesquite bushes. (Photo. D. T. MacDougal.) the desert, necessitating frequent dredging. Many communities in the past which depended for their existence upon irrigation have been obliged to abandon their homes, because an unstable government failed to keep the irrigation canals free from drifting sand. The drifting of sand has often affected the drainage of the land ; FIG. 29. A sand dune covering a cabin. The origin of the sand is the seabeach. (Bermuda Islands.) . the Grand Calumet River (Fig. 30) formerly emptied into Lake Michigan in Indiana, but its mouth became so filled with drifting sand that the course of the stream was reversed and it now empties into the lake at Chicago, twenty-four miles distant. Large lakes have been formed in consequence of the damming of rivers by dunes, where they emptied into the sea. One such in France, Lake Cazaux, has a width of nearly seven miles and a depth of 130 feet. The ripples which mark the surfaces of sand dunes shift their posi- tions gradually and, in general, are affected as are the dunes. The movement of sand dunes can sometimes be prevented by plant- WORK OF THE WIND 5 ! ing grasses, shrubs, and trees on the gentle slopes in order that they may hold the sand with their roots. This has been done successfully in San Francisco, in Provincetown (Massachusetts), and elsewhere. Beneficial Effect of Dunes. Dunes are not, however, always a detriment to man. A writer states that the people of Holland and Denmark " deal as carefully with their dunes as if dealing with eggs, and talk of their fringe of sand hills as if it were a border set with pearls. They regard these as their best defense against the sea." (Kahl.) As this implies, Holland depends to a large degree for its FIG. 30. Map of the Grand Calumet River. The river formerly entered Lake Michigan at the east, but was cut off by sand dunes and now enters the lake at Chicago. protection from the sea upon sand dunes, which are from one to three miles wide and from 40 to 50 feet high. The sand of certain dunes in England (Padstow), which consists largely of shell fragments, is used to some extent for a fertilizer. Material of Dunes. The material of sand dunes varies but is usually quartz sand. However, in the Bermudas, Bahamas, and por- tions of England, dunes are composed of shell sand (CaCOs). In the Bermudas these sands are cemented by the rain water which dissolves the calcium carbonate and later redeposits it, thus forming stratified eolian rock. When the shallow, alkaline lakes of portions of New Mexico (Otero Basin) dry up, they leave on their beds thin sheets of various salts, chiefly gypsum. These soon curl up into leaves which, when blown together, are broken into gypsum and salt sands. The winds carry the light gypsum out to the plains, where it gathers in a 52 PHYSICAL GEOLOGY great series of white dunes, 60 to 100 feet in height, covering an area 15 miles by 40 miles in extent. Dunes are formed also of fine clay, as well as of disintegrated granite sand. Height of Dunes. The height of dunes in regions where the direc- tion of the wind is fairly constant is seldom more than 300 feet, but in such deserts as the Sahara, where the wind varies from season to season, the height may reach 1500 feet. In the latter case, the dunes do not migrate, and their greater height is due to the piling up of the sand from different directions. Eolian Sandstone. Extensive strata of sandstone of very ancient date are known to have been formed of wind-blown sand. Rocks of this origin can often be distinguished from the sandstones laid down on the ocean bottom. The following differences assist in recog- nizing the source of the original deposits, (i) The former consist chiefly of quartz, the softer minerals having been worn to dust and carried away, while in the latter the softer and harder minerals are more likely to occur together. (2) Since water-laid sands are carried in suspension, they are subjected to less wear than eolian sands, as the water between the particles acts as a cushion, and the grains are consequently less worn and more angular than the sand grains which have been buffeted by the winds. (3) The stratification of eolian sand (Fig. 26, p. 48) usually exhibits cross or false bedding (p. 47), i.e., it is not horizontal but varies greatly in inclination and direction within short distances. (4) Wind-blown sands may also be dis- tinguished from marine sandstones by the character of the fossils, if such exist. Dust. As sand grains are borne to and fro by the wind, striking against each other or against rock surfaces, the softer grains are reduced to dust, and even the harder ones may finally reach a similar state. The dust thus formed is carried by air currents, often to great distances. In a single storm in 1901 it is estimated that 1,960,420 tons of dust were carried from the Sahara desert to Europe, reaching Italy on the second day of the storm, and Germany and Denmark on the fifth day. It is probable that every square mile of the earth's surface has dust upon it from every other square mile. Even the snows of mountain glaciers and those of the Arctic and Antarctic regions con- tain dust, carried to them from lands hundreds of miles away. Loess. One striking result of the transportation of dust by winds is that regions to the leeward of deserts are constantly receiving dust which settles gradually upon them. Such a deposit of fine dust is called loess. The fine dust is carried by the WORK OF THE WIND 53 wind to the edge of the dry region, where it is precipitated by rain or falls slowly by its own weight. Here some of it is held by the grasses of the high plains, whose roots have left, upon their decay, the vertical columns characteristic of loess. " But if FIG. 31. Loess deposits, Shan-si, China. The canyon-like depression was excavated by the wind as the loess was loosened by traffic. The two levels on the right are old roads. The ability of loess to stand in almost vertical walls is shown. (Carnegie Institution.) the desert dust has ceased to be the plaything of the wind, it has not ended its jour- ney. From now on rills take charge of it and continue the work of which the wind is no longer capable." (De Martonne.) In this way loess is spread over a large terri- tory. In China there are extensive areas which have been built up by the accumulation 54 PHYSICAL GEOLOGY of such dust, in some regions to a depth of 1000 to 2000 feet. The fertility of the soil of these regions is remarkable. Although cultivated for many thousands of years without artificial fertilizer, it still retains its fertility. This is due largely to the con- stant supply of new dust from the desert. It is stated that the limit of the loess practically marks the extreme limit of the extension of Chinese agriculture and com- merce. (Richthofen.) Large areas in the United States (p. 657) and in Argentina are also covered with loess, and in all such regions grass and grains flourish, although trees are usually few. The principal deposits of loess in the United States were de- rived from the fine material of glacial deposits which were caught up by the winds during the dry phases of the interglacial periods (p. 657). Loess has the property of maintaining a vertical face when cut through artificially or by streams. In China the roads of the loess region are often in nearly vertical, walled canyons (Fig. 31), many feet below the surface, having been deepened by the blowing out of the dust of the traveled road. On either side of these roads cave houses have been excavated and furnish homes for many thousands of people. Dust is obtained by the winds from sources other than desert sand, such as fine volcanic ash, solid particles of smoke, pollen of flowers, and spores of plants. The amount of material thrown into the air during volcanic eruptions is enormous. The volcano Krakatao in the East Indies, for example, in 1883 threw volcanic dust to a height of several miles, which in fifteen days had encircled the globe. So abundant was the dust in the air that for many months after the eruption the sunsets were remarkably brilliant. In Kansas and Nebraska there are deposits of volcanic dust, locally 30 feet thick, which had their source in ancient volcanoes hundreds of miles away. REFERENCES FOR THE WORK OF THE WIND BEADNELL, H. J. L., Sand Dunes of the Libyan Desert: Geog. Jour., Vol. 35, 1910, PP- 379-395- COBB, C., Where the Wind Does the Work : Nat. Geog. Mag., Vol. 17, 1906, pp. 310- 317. DAVIS, W. M., The Geographical Cycle in an Arid Climate : Jour. Geol., Vol. 13, 1905, PP- 381-407- DE M ARTONNE, E., Geographic Physique, 1909, pp. 649-672. FREE, E. E., The Movement of Soil Material by Wind: Bull. U. S. Bureau of Soils, No. 68, 1911. GEIKIE, J., Earth Sculpture, 1898, pp. 250-265. HAUG, E., Traite de Geologie, 1911, pp. 387-403. HOBBS, W. H., Earth Features and their Meaning, 1912, pp. 197-222. HUNTINGTON, ELLSWORTH, The Pulse of Asia, 1907. KEYES, C. R., Relation of Present Profiles and Geologic Structures in Desert Ranges: Bull. Geol. Soc. America, Vol. 21, 1910, pp. 543-563. (Dr. Keyes holds extreme views on wind erosion.) WORK OF THE WIND 55 MACDOUGAL, D. T., The Desert Basins of the Colorado Delta : Bull. Am. Geog. Soc., Vol. 39, 1907, pp. 705-729- WALTHER, J., Das Gesetz der Wustenbildung, 1900. TOPOGRAPHIC MAP SHEETS, U. S. GEOLOGICAL SURVEY, ILLUSTRATING WIND WORK Moses Lake, Washington. Camp Clark, Nebraska. Lamed, Kansas. Norfolk, Virginia North Carolina. Kinsley, Kansas. Sandy Hook, New Jersey New York. Pratt, Kansas. Toleston, Indiana. St. Paul, Nebraska. Yuma, California Arizona. CHAPTER III THE WORK OF GROUND WATER TAKING the world as a whole, about 78 per cent, of the rainfall either soaks into the ground or is evaporated, the remainder the run-off being carried directly into streams and rivers. The amount of the precipitation which is retained in the soil depends upon (i) the climate, (2) the slope of the ground, (3) the porosity of the soil and rock, and (4) the amount and character of the vegetation. In moist climates the run-off may amount to as much as one half of the rainfall, while in arid regions, on account of the excessive evapo- ration and the dryness of the soil, there may be no run-off. That portion of the rainfall which sinks into the soil is called ground water. Once beneath the surface, it continues its descent through the pores and cracks of the rock until it may reach great depths. Quantity of Ground Water. All rocks are more or less porous, even granites con- tain some water; for example, chalk may hold two gallons of water a cubic foot, and sandstones may hold 20 to 30 per cent, of their weight. The total amount of water in the rocks is therefore very large, and it is probable that, if the ground water were squeezed from the rocks, there would be enough to cover the earth with a sheet of fresh water one hundred or more feet deep. Locally, the quantity of underground water may be much greater; as, for example, in Wisconsin and Minnesota, where the underlying sandstone alone contains enough water to form a layer 50 to 100 feet deep. The ground water of any region is not always derived from the local rainfall, but may have had a long subterranean course, as is true of the underground water of the Great Plains, the source of which is in the mountains, many miles distant. The Water Table. The level beneath which the rock is saturated with water is called the water table or the level of underground water. 1 This varies greatly in different regions. In humid portions of North America it is from one to forty feet below the surface ; in limestone regions, where the drainage is largely subterranean, such as in por- tions of Kentucky and Tennessee, it may be two to three hundred 1 " In deep mines in various parts of the world water is found only in the upper levels, within 2500 feet or less of the surface, while below that the mines are dry or even dusty." (Scott.) 56 THE WORK OF GROUND WATER 57 feet deep ; while in the Colorado Plateau, where the surface is cut by deep canyons, it is sometimes 3500 feet beneath the surface; or it may be entirely absent, except where water-bearing strata conduct water from other areas. In the Navaho Reservation in Arizona, for example, no water except artesian (p. 59) is encountered below a depth of 100 feet. (H. E. Gregory.) The water table varies with the slope of the land, being farther from the surface on the hills than in the valleys (Fig. 32). The greater depth beneath the surface of the hills is due to gravity, which between rains and during dry seasons tends to pull the water downward to the level of the valleys, but is unable entirely to do so because of capillarity and friction of the water with the grains of rock. As a result, the water table is never flat in a hilly region, although it is more nearly so after a prolonged drought. It necessarily follows that the depth of the water table in any place will depend largely upon (i) the slope of the land, (2) the porosity of the rock, and (3) the frequency and character of the precipitation, slow, soaking rains accomplishing more than sudden and brief downpours. In forested areas it is found that the water table is lower than under similar conditions of moisture, rock, and topography ,in other regions, because of the great quantity of water abstracted by the roots of the trees and lost by evaporation through the leaves. The headwaters of streams, however, should be kept forested, since much of the water of excessive rains is retained in the thick layer of forest mold, from which it slowly drains away and thus tends to prevent great floods. Wells. When wells are sunk, it is necessary that they penetrate to a permeable rock or to a much fractured one (Fig. 33) below FIG. 32. Diagram showing the water table or level of underground water A A A A and the effect upon natural and artificial depressions. the water table (Fig. 32), for otherwise they do not afford a perennial supply of water. The value of wells, both for drink- ing purposes and for irrigation, is inestimable. It is stated that in India more land is irrigated from wells than from streams, and PHYSICAL GEOLOGY FIG. 33. Diagram showing the source of well and spring water in fractured rock. (Modified after H. E. Gregory.) in southern California one half of the irrigation water and the greater part of the city supplies are drawn from the sands and gravels that underlie the val- leys. It is estimated that 75 per cent, of the population of the United States depends for its water supply directly upon underground water. Movement of Ground Water. Underground water seldom moves in definite channels, except in lime- stone regions, but percolates slowly through the pores and crevices of the rocks. Even in coarse sand- stone the rate of movement may be only one fifth of a mile a year, although in regions of soluble lime- stone it may flow several miles a day in tunnels. In such regions the direction of the underground flow may be opposite to that of the surface streams, since it is determined by the dip of the rock. Much of the ground water eventually reaches the surface again unless it enters into chemical combination with minerals of the rocks with which it comes in contact. A large amount is taken up by plants and passes into the atmosphere by evaporation ; some of it is drawn out in wells ; some seeps out, or is discharged in springs, either in river valleys or in lakes and seas. Large springs of fresh water come to the surface of the Mediterranean, the Gulf of Mexico, and other seas at short distances from the shore, and in certain places fresh water is obtained from springs on the ocean bottom by diving. The total quantity of mineral matter dissolved by the ground water is enormous. The greater part of the 4,975,000,000 tons of mineral matter carried to the ocean each year was obtained by the streams from the ground water which escaped through springs and seepage. Depth of Ground Water. We have seen that the rocks of the earth's surface are much broken by cracks of various kinds. This condition holds true of rocks below the earth's surface, down to a depth where the weight above them is greater than their strength to resist pressure. This outer zone is called the zone of fracture. The depth of this zone varies with the strength of the rock. In the case of soft rocks, such as shales, no crack may be found at a depth of 2000 feet, while in the strongest rocks some cracks may possibly \ . THE WORK OF GROUND WATER 59 exist " at a depth of at least eleven miles." (Adams.) At depths greater than eleven miles it does not seem possible that a crevice can open, and if a fracture should occur, the parts would actually weld together. It is evident from the above that water will not descend a greater distance than eleven or twelve miles under the most favor- able conditions, and usually far less than that. The temperature of the rocks, and therefore of underground water, increases i F. for each 60 to 100 feet of descent, a fact which accounts for the warmth of deep wells and springs coming from great depths (p. 273). Artesian Wells. Strictly speaking, an artesian well is one in which the water rises above the surface of the ground as a fountain, but the term is now, unfortunately, frequently employed for any deep well from which water is obtained, whether it flows to the surface or not. This change in usage is doubtless due to the fact that often FIG. 34. Block diagram showing the conditions favorable for artesian water. The porous beds ( dotted ) receive water from the rain which falls on their outcrops, and from the streams which lose somewhat in volume as they flow over them. Three water-bearing beds ( aquifers ) are shown, from two of which water can be obtained on the barrier island which is separated from the mainland by a salt-water lagoon. artesian wells, after flowing for some months or years, cease to do so and must be pumped because of the excessive withdrawal of water from the artesian basin. This was true of the first artesian well at Artois, France (from which the name " artesian" was derived). Many wells in the San Bernardino valley, California, which flowed strongly when first drilled, are now pumped. The conditions favoring artesian water (Fig. 34) are (i) a porous bed capable of absorbing and trans- mitting large quantities of water ; (2) relatively impervious beds above and below; (3) exposure of the porous stratum where it may absorb water supplied either by rain or by streams flowing over it; (4) an inclination of the water-bearing stratum so that gravity may force the water down; (5) a lack of easy escape of the water at lower points; and (6) a sufficient supply of water to maintain the " artesian head." The artesian water of South Dakota (Fig. 35), for example, is derived from a saturated sandstone bed which receives its water in the Black 6o PHYSICAL GEOLOGY Hills from the rain that falls on it and the streams that flow over it. It is covered by clays and shales as it extends eastward, and when borings are made at elevations lower than its source in the Black Hills, the water rises and supplies wells even 350 miles from this source. FIG. 35. Diagram showing the conditions favorable for artesian water, from the Rocky Mountains to eastern Nebraska. The Dakota sandstone under the imper- vious Pierre clay carries water from the Rocky Mountains and supplies artesian wells on the plains. ( U. S. Geol. Surv. ) Artesian wells vary in depth, some being 4000 feet deep, while others may be less than 100 feet. Artesian water, both for drinking pur- poses and for irrigation, is of great importance. It varies greatly in composition, some wells affording excellent water, while others may be so charged with salts as to be useless for drinking or irrigation. Springs corresponding to artesian wells are formed if the impervious bed overlying the porous bed is broken by a fissure or fault (p. 25). These springs may be of great volume. Chemical Work of Ground Water. (i) Solution. Pure water has little power to dissolve the minerals of which rocks are composed, but rain water is seldom pure since it receives carbon dioxide from the air, and, in passing through the soil, takes up this and other acids formed by the decay of organic matter. It may be heated in its downward course and is subjected to great pressure. Thus equipped, its solvent power is greatly increased, and in its descent through the rocks it carries away the more soluble minerals and the cement of many of the rocks, rendering them more porous and causing their decay. At or near the surface, water is an active agent in causing the disintegration of the rocks, both by the mechanical work of the frost and by its chemical action. (2) Replacement and (3) Deposition. When ground water contains much mineral matter, a slight change in temperature or pressure, or a mingling with other waters of a slightly different composition, may cause the dissolved material to be deposited. This results in replace- ment, and deposition in cavities. (2) Replacement results when in its descent ground water dissolves and carries away one mineral, deposit- THE WORK OF GROUND WATER 61 ing another in its place. Shells, bones, and trees are petrified by the replacement, molecule by molecule, of the original substance by mineral matter. (3) Deposi- tion occurs when minerals are taken from the rock in one place and later deposited elsewhere. In this way many metallic and other veins (Fig. 36) are formed (p. 371), and loose sands and clays are cemented into hard rocks. Besides this more impor- tant work, concretions (p. 75) and geodes (p. 78) are formed, and in regions of thick limestone cave deposits are built up (p. 70). Belts of Weathering and Cem- entation. The belt of weathering extends from the surface of the ground to the level of underground water and is of variable thick- ness. In this belt the greatest chemical decomposition of rocks occurs. This work consists mainly in hydration, oxidation, absorp- tion of carbon dioxide, and solution, and it is here that minerals with complex molecules are broken down into simpler compounds. This belt is, therefore, that portion of the earth's crust which is being prepared for its ultimate disintegration into soil. Great porosity, low temperature, and low pressure characterize this zone. The belt of cementation is beneath the level of underground water. In this belt, as the name implies, deposition rather than solution plays the leading part. The consolidation of sands and clays into hard rock is brought about here, both by the deposition of minerals obtained by solution from the belt of weathering and also by the pressure of the overlying rocks. The rocks of this deeper zone are more or less porous and fractured, and the temperature is compara- tively low. As the surface of the land is lowered by erosion, the belt of weathering invades the belt of cementation, and the minerals which were deposited in the pores and cracks of the latter may again be dissolved out. \ 62 PHYSICAL GEOLOGY Desert Limestone. In arid regions the underground water may, by capillarity, bring to the surface large quantities of lime which, upon evaporation, is deposited as desert limestone. About Valencia, Venezuela, for example, the underlying rock is almost entirely hidden by thick layers of this deposit, and extensive areas of New Mexico, Arizona, and other states are covered by this limy incrustation. Mechanical Work of Ground Water. The mechanical work of underground water is important in producing landslides (p. 73), but aside from this the effect is usually slight, since its movement is, for the most part, extremely slow. An interesting result of the drying out of underground water was observed in England at the end of a prolonged drought in the summer of 1911. It was found that the foundations of hundreds of houses which rested on clay began to settle after the return of the rains. In ordinary summers the clay is quite moist at a depth of 2.5 to 3 feet below the surface, but during the summer mentioned it was often dry at depths of 5 to 6 feet. The dry clay became powdery, and when the autumn rains began the water found its way into the fissures and washed out the clay, causing sliding and lateral movements. SPRINGS > The rain water which sinks into the soil and rocks through joints, fissures, and pores usually issues once more to the surface through seepage and springs (Fig. 37). Origin of Springs. (i) Springs commonly owe their existence to the presence of a stratum of pervious material overlying an impervious FIG. 37. Thousand Springs, Snake River, Idaho. (U. S. Geol. Surv.) THE WORK OF GROUND WATER one. The water penetrat- ing the pervious or frac- tured stratum (Fig. 38, A, B y C) is prevented from moving downward through the impervious layer whose slope it fol- lows until it emerges at the contact of the two layers. (2) A second class of springs rise through cracks or fissures (Fig. 40). These are often of great volume and .may have a tem- perature higher than the springs of the first type. (3) When the surface of a limestone region is lowered by streams (Fig. 39), an underground stream is often encoun- tered and gives rise to the springs of great vol- ume which are so fre- quent in such districts. Silver Spring in Florida forms a navigable stream valleys at, or above, the lowest part of the valley. FIG. 38. Diagrams showing the origin of springs. In A the porous stratum is indicated by dots, the saturated zone being shaded. B is an impervious stratum. A spring (sp) appears at the left, and during wet seasons, when the water table is high, a spring will flow also from the right of the hill. In B the impervious stratum is horizontal, and springs will flow from both sides of the hill. If the surface of the saturated zone (shaded) becomes so low that it does not reach the surface, the springs will cease to flow. C shows a porous stratum B overlain by an impervious stratum. In such a case the water is derived from the surface at B and appears as a spring (sp) at the left. from its source. Springs may flow into thalweg, which is a line following the FIG. 39. Large springs often issue from the base of limestone cliffs. Such springs are frequently contaminated, since their water enters through wide joints and sinks without being filtered by soil. CLELAND GEOL. $ PHYSICAL GEOLOGY FIG. 40. A fissure spring. The oases of deserts often owe their existence to springs. The oases of Kerid in the northern Sahara desert contain about 6000 acres, which support nearly 1,000,000 date palms. They lie at the foot of an escarpment which forms the northern boundary of the desert. From the base of this escarpment or cliff, numerous springs gush forth and furnish a constant supply of water for irrigation. The water of the springs falls as rain in the highlands many miles distant. After flowing as streams for a short distance, the water disappears in the sand. It then follows underground courses until the escarp- ment is reached. Constant and Intermittent Springs. Whether springs are constant or intermittent depends upon a number of factors : if the rainfall is not uniformly distributed throughout the year, if the region is not forested, or if the porous rock is too limited to hold a sufficient supjsly of water, an intermittent spring may result. In such a hill^s that shown in Fig. 38 A the glacial deposit (till) and sand allow the water to be absorbed in large amounts and to sink to the impervious stratum along which ground water flows to S^>. When the water stands at A) a spring may flow which will cease when the water is below that level. In unusu- ally dry seasons all may dis- appear. It is seldom, per- haps never, that a siphon operates to form an inter- FIG. 41. Diagram illustrating the possi- bility of the occurrence of a siphon spring in nature. If the vertical joints B and C do not reach the surface, the water filling the joints A y Ay A will continue to flow as a spring (Sp) until the joints are emptied, because the not begin to flow until the joints are filled above BC. (Modified after De Martonne.) mittent spring, but in Such a weight of the water in the arm CD is greater case as that shown in Fig. 41 than in B. When once emptied the water will it will be seen that the water will not flow until it has reached BC, after which it will discharge until the reservoir is empty. This is due to the fact that the weight of the water in the arm CD is greater than that of the arm B. Mineral Matter in Spring Water. Since springs are derived from underground water which has been in close contact with various rocks, THE WORK OF GROUND WATER they usually contain a much greater quantity of dissolved minerals than do streams. Silver Spring in Florida is carrying to the sea in solution 340 pounds of mineral matter a minute, or 600 tons a day, and it is estimated that, in central Florida, a little more than 400 tons of rock a square mile are annually carried away in solution. This would be equivalent to a lowering of the surface of the central peninsular section of Florida by solution alone at a rate of one foot in five or six thousand years. Falls Creek, Oklahoma (Fig. 71), receives water from springs which contain much lime carbonate. In the immediate vicinity of the springs, however, no deposits are formed, as there is a sufficient amount of carbon dioxide present in the water to hold the lime in solution, but by the time the stream has flowed a quarter of a mile large quantities of carbon dioxide have been given off, and travertine is de- posited in the bed of the stream in the form of dams which vary in height from a few inches to 15 feet, and are being built up faster than the stream can cut them away. The great lime- stone deposits at FIG. 42. Block diagram showing the formation of a travertine terrace and natural bridge. Water containing much lime carbonate emerges from springs in the lime- stone at the right. Travertine has been rapidly de- Tivoll in Italy, from posited, forming the terrace and natural bridge, which was quarried much of the stone used in the construction of the Coliseum and St. Peter's at Rome and the interior of the Pennsylvania railroad station in the City of New York, were laid down by springs. The quantity and rapidity of the deposition of limestone under excep- tionally favorable conditions is well shown in the great travertine natural bridge at Pine, Arizona, more than 125 feet high, which, together with a terrace of 25 acres, was formed by such a deposit (Fig. 42). Springs containing lime carbonate or gypsum in solution are called " hard," since, in washing, the fatty acids of the soap unite with the dissolved minerals to form the insoluble " curd." By abstracting carbon dioxide from the water in which they grow 66 PHYSICAL GEOLOGY algae may cause lime to deposit. In this way beds of so-called " petrified moss," more than ten feet thick, have been formed. In the Yellowstone National Park the deposits about the geysers were built up both by the evaporation of the water and by algae (p. 65). A reduction in temperature and pressure may also cause minerals in solution to be deposited, as may also the mingling of waters carrying in solution substances of different composition (p. 372). Mineral Springs. Mineral springs contain various salts or gases. Such springs are often called " medicinal " because of their supposed curative properties. The total value of mineral waters is large, amounting to $5,631,391, in 1913, in the United States alone. Temperature of Springs. The temperature of springs is usually much lower than that of the air in summer, being about 47 F. in Connecticut; and the water is often described as being " icy cold/' The temperature of such springs in middle latitudes is fairly constant if they come from a depth greater than 50 or 60 feet, since at this depth the water is not affected by daily or seasonal changes and has, consequently, about the average temperature of the region. Thermal Springs. The temperature of many so-called hot springs varies from lukewarm to boiling, (i) The heat is sometimes due to the presence of deep fissures through which the surface water has percolated until it has reached great depths, where its temperature has been raised by the interior heat of the earth. After being thus heated, the water is forced by hydrostatic pressure to a point on the surface which is lower than the point of ingress. The depth from which come springs like those of Bath, England, which have a tem- perature of 120 F., may be approximately told from well borings, such as that of a well at Berlin, Germany, the water of which has a tem- perature of 110.5 F. at a depth of 3390 feet. Springs located along fissures in the earth's crust occur in Virginia, Arkansas, Colorado, Nevada, and South Dakota, and are often the seat of popular health resorts. (2) The water of some springs is heated by chemical action. (3) Water in volcanic regions may be heated at comparatively shallow depths by the presence of uncooled lava. Of this class there are more than 3000 in the Yellowstone National Park, some of which deposit limestone (travertine) and others silica (geyserite). Hot springs may bring about a considerable change in the character of the rocks in the regions in which they occur, by causing the disintegration of some and THE WORK OF GROUND WATER by adding new material to others. This is due to the fact that hot water is a more powerful solvent than cold. Geysers. Geysers are springs which intermittently erupt col- umns of hot water and steam (Fig. 43). They occur in regions of comparatively recent volcanic activity, where the lava is hot at a relatively shallow depth. They are well developed in but three lo- calities in the world, and the total area occupied by them is probably less than ten square miles. The most notable geysers occur in Ice- land, New Zealand, and the United States, although smaller ones are to be seen in Mexico, Tibet, the Azores, and the island of Formosa. Some of them throw water to a great height. The Monarch Geyser in New Zealand became active in 1903 and is said to have thrown mud and stones to a height of 1000 feet. Such a height, however, is unique. In the Yellowstone National Park an eruption throwing water 300 feet vertically is rare. The quantity of water flowing from geysers varies greatly : in the smaller ones it may be only a few gallons FIG. 43. Lone Star Geyser, Yellowstone National Park. an hour, while in others, as in Old Faithful in the Yellowstone National Park, the discharge may be as great as 750,000 gallons an hour, a quantity sufficient to supply a city of 150,000 inhabitants. The water of geysers is rain water which has percolated through porous lava, and under normal condi- tions would be discharged as springs. Consequently, if the climate of the Yellowstone National Park should become arid, the geysers would disappear. This water, heated by its passage through the lavas, dissolves soda and potash, becoming alkaline and thus capable of dissolving silica from the silicates of the lavas. Accordingly, the waters erupted by geysers contain much mineral matter in solu- tion, the chief of which is silica. This silica is deposited about the openings of the springs as siliceous sinter, or geyserite, forming a 68 PHYSICAL GEOLOGY Observed mound both by evaporation and also through the action of minute plants (algae) which are capable of living in hot water and of secreting silica. It is stated that by evaporation alone a geyser can produce a maximum thickness of geyserite of one twentieth of an inch a year, while the increase from algae deposition under favorable conditions may be as much as eight inches during the same period. A geyser usually originates as a spring in a fissure, the opening of which is gradually built up by the deposition of siliceous sinter until a considerable mound or terrace is formed. As long as the tube through which the water reaches the surface is short or the circulation of the water unimpeded, a siliceous spring will flow. When, as a result of the building up of the mound or for other reasons, the tube becomes so long that the water can- not circulate with rapidity (Fig. 44), the water at some distance below the top of the tube will increase in tem- perature more rapidly than that at the surface. Eventually water at a depth of a number of feet will reach its boiling point with the resultant formation of bubbles of steam which, in turn, will cause the .water to spill FIG. 44. Cross section of a geyser, over the edge of the opening. This showing the boiling temperature at over fl O w promotes boiling by reduc- the right and the recorded tempera- . . . ture at the left. (After Campbell.) ng the pressure upon the water deep in the tube. As a consequence a large quantity of water, which was not quite at the boiling point because of the weight of the overlying column of water, will instantly burst into steam and will eject the overlying water from the tube, sometimes to a great height. Usually the eruptions are not regular, but in Old Faithful an eruption can be predicted at intervals of about sixty minutes. When a quantity of soap or lye is thrown into a geyser, the viscosity of the water is increased and its circula- tion correspondingly lessened. In this way an eruption may be hastened. As the lavas cool, the geysers must necessarily disappear. However, the loss of heat is very slow, as is shown by the fact that, although careful records have been kept since the Yellowstone basin was discovered, the Yellowstone geysers have shown little sign of change since they were first studied. The eruptions of Old Faithful, for example, continue to be regular. THE WORK OF GROUND WATER 69 STRIKING EFFECTS OF GROUND WATER Swallow Holes. In limestone regions it is not unusual to find many funnel-shaped depressions in the surface of the ground into which water may flow. These are called " sink " or " swallow holes " and may be very conspicuous features of the landscape (Fig. 45). They are formed either (i) through direct solution by surface waters along joints, in which case they are usually more or less circular in outline; or (2) by the falling in of the roof of a cavern, when they are often irregular in outline. Those formed in the first way are much more common than the latter, but are usually smaller. An example of sink holes formed by the falling in of a cavern roof occurred in the city of Staunton, Virginia, in 1910, when four "cave-ins" occurred within three weeks, the largest of which was 60 by 90 feet. During the formation of this largest one three trees and portions of a dwelling house were engulfed. (3) In regions underlain by salt, local sinkings result from the solution and removal of the salt by underground water. After a swallow hole is formed, more or less of the material imme- diately around the hole will be carried in by surface wash. More- over, a large amount of water entering through the sink may cause a rapid solution of the limestone in its immediate vicinity, resulting in the formation of large basins locally called " prairies " or " coves." In the United States these are well developed in Kentucky and Florida. If the bottoms of swallow holes become choked, small lakes or pools come into existence. A striking example is shown in the history of Alachua Lake, 1 Florida (Fig. 46). Previous to 1871 the waters of 1 Florida Geol. Surv., Third Annual Report, 1910, pp. 62-67. FIG. 45. Small swallow or sink holes in the Juras, Switzerland. 70 PHYSICAL GEOLOGY the principal stream of this region emptied into a sink or swallow hole in the Alachua prairie. By the choking of this outlet a lake was formed which, at its greatest extent, was eight miles long and in one place four miles wide and of sufficient depth to permit a number of FIG. 46. Diagrammatic section from Devil's Mill Hopper (northwest of Gainesville) to Alachua sink, Florida. The Devil's Mill Hopper is 115 feet deep but does not quite reach the level of underground water, D E. B is Alachua sink, whose bottom is filled with water to or a little above the level of underground water. C is a small sink above the water table, which does not contain water. If the opening at the bottom of the Devil's Mill Hopper becomes clogged, the sink will fill up to the surface and become one of the deep, small, circular lakes frequently found in the region. (Modified after Sellards.) freight steamers to ply upon it. After existing for about twenty years the underground passage from the swallow hole was opened again, and the lake gradually disappeared. Caverns. Caverns occur in limestone regions and are usually connected with swallow holes by more or less distinct passages. They have been formed, with few exceptions, by the solvent power of the water which poured through the swallow holes and joints, or seeped through the rocks from the surface, and, to some extent, through abrasion by the sediment carried by the subterranean streams. Since water circulates most rapidly along joints and bedding planes, it is in such positions that most rapid solution takes place, and it is here that caverns occur. In certain spots, owing to the presence of numerous open joints or to the solubility of the rock, large domes are formed. Solution is usually most effective in. forested regions, since the humus affords a large and constant supply of carbon dioxide, without which water is but slightly solvent. Caves may, however, be formed by carbonated waters ascending from below; an example of which is the interesting and extensive Wind Cave in the Black Hills of South Dakota, which was formed by hot water coming up from a great depth and gradually enlarging the joints and fissures in its ascent. In regions of thick limestone, caves at different levels, called THE WORK OF GROUND WATER 71 " galleries," occur (Fig. 47) ; as, for example, in Mammoth Cave, Kentucky. These galleries are the result (i) of the presence of layers of relatively insoluble rock upon which the underground streams flow until they dissolve and erode out a wide passage. If this layer is worn through after a time, or a joint is enlarged, permitting the water to reach a lower soluble layer, it may descend until a second relatively FIG. 47. Diagram showing the formation of the galleries of limestone caves by the lowering of the valley (indicated by dotted lines) to which the underground water dissolving them flowed. insoluble layer is encountered. If this process is repeated, several galleries will result. The lowest level at which caves may be formed is that of the lowest surface stream into which the underground water is discharged. (2) Migration from one level to another may also result from the intermittent lowering of the valleys (Fig. 47) of the surface streams into which the underground waters of the caverns flow. The galleries of caves may divide and reunite, forming a network of channels at the different levels. It has been estimated that in Kentucky alone there are 100,000 miles of underground passages. Natural Bridges may be formed by the partial caving in of the roofs of caverns, or by the enlarging of two swallow holes opening to the same underground stream. Natural bridges are also formed in other ways (pp. 91, 112). ^-- Cave Deposits. After a cave has been abandoned by the stream which formed it, the water entering is confined chiefly to small seepage. At this stage much of the water is removed by evaporation so that solution gives place to deposition. The deposits in caves are usually in the form of stalactites and stalagmites. The former begin as a thin film of lime around the outside of a drop of water which evaporates on the roof of a cavern. Upon this additional lime is left by other drops until a stalactite, resembling an icicle, is suspended from the roof of the cavern. The accumulations of lime which form where the PHYSICAL GEOLOGY FIG. 48. Stalactites and stalagmites in Marengo Cave, Indiana. (D. Appleton and Company.) water evaporates on the floor of the cavern are known as stalagmites (Fig. 48). By the union of the stalactites and stalagmites pillars are formed. Caverns are also formed in other ways (pp. 209, 298), but the great majority are formed from solution. Karst. Karst is used as a descriptive term for any lime- stone region which has been etched and eroded by water into FIG. 49. Block diagram of a karst (limestone) region, , f , F; illustrating the effect of solution. Sink holes AAA drain a rOU S h surtace (* ! f the surface and discharge their water through under- 49)- The name IS ground channels to an open valley. Surface streams are derived from Karst lacking, and the main valley has steep sides. The spring , 'A f at C may be very large. A fault is also shown at C. (Modified after De Martonne.) the Adriatic, where THE WORK OF GROUND WATER 73 such a surface is developed upon a nearly pure limestone. It is a desolate region in which vegetation is scanty, except in swallow holes (dolines), where the small amount of insoluble matter yielded by the rock accumulates and furnishes a soil for plants. The drain- age is, for the most part, subterranean; and the surface is etched out into a network of narrow channels between which blade-like masses of rock rise. It is pitted with swallow holes and, where important streams cross the karst land, they flow in deep gorges, rather than in ordinary valleys. Landslides. Landslides may result from a number of conditions, one of which is often associated with underground water. Soil and subsoil tend to move down a hillside when they become charged with water. If this movement is insensible it is called " creep " ; if FIG. 50. Conditions favoring landslides. The sensible " slumping " strata AC and BD are clay or shale wnich > wnen wet > ! . are slippery, so that sliding is likely to occur. or sliding. Railroad tracks may be gradually moved down hill and trees be tilted by the slow movement of hillside creep. The conditions favorable for a landslide are a steep slope upon which soil rests, or steeply dipping rock which has been undercut at the base, artificially or by streams, so that the upper layers are unsupported (Fig. 50). When, under either of these conditions, the soil or rock becomes saturated with water, its weight is increased, and, moreover, the water, acting as a lubricant, lessens the friction which pre- viously prevented FIG. 51. Diagram showing a valley which has been the soil or rock from deepened by glacial erosion, leaving steep slopes unsup- sliding. Such was the cause of the Mt. ported on each side. Fractures may develop at AB, and a portion of the side may slide into the valley. Greylock, Massachu- setts, landslide, in which a great mass of soil and glacial debris slid down the steep mountain side after a period of excessive rainfall; and of the landslide in Quebec, where the rock hillside slipped 74 PHYSICAL GEOLOGY FIG. 52. Landslide, Turtle Mountain, British Columbia. (Photo. Hopkins.) along a plane of steeply dipping slate which had been lubri- cated by underground water. In mountainous re- gions, where the val- leys are deep and the slopes steep, condi- tions are extremely favorable for land- slides (Figs. 51, 52). One of the most de- structive of such slides occurred on the Ross- berg in Switzerland in 1806. Here the rocks high up on the mountain slid suddenly into the valley, burying the village of Goldau and causing the death of several hundred people. Masses of rock, some of which were as large as houses, were spread over the valley for two or three miles. Evidences of many prehistoric landslides are to be seen in Switzer- land, as well as in other mountainous regions. At Siders a land- slide is spread out for several miles across the Rhone valley, and some of the hills formed from the material of the slide are almost 200 feet high. So marked is this land- slide topography that it forms the boundary between the French and German-speak- ing people in the Rhone valley. Conditions favor- able for landslides were created artifici- ally in the excavation of the Culebra Cut in the Panama Canal, where the rock will continue to slide peri- FIG. 53. A lake in eastern France formed by a landslide. The character of the material of the dam is shown in the foreground. THE WORK OF GROUND WATER 75 odically until a gentle slope is formed. During a single year (1911) nearly 36 per cent, of the total material excavated had been brought in by landslides. Landslides may dam streams, forming lakes (Fig. 53) or rapids. Lake Oechenen, in the Kandersteg valley of Switzerland, and the Cas- cades of the Columbia River were formed by landslides broken from the high mountains a few centuries ago. The rounded hills and basins FIG. 54. Landslide topography which has much the appearance of a moraine. Kandersteg valley, Switzerland. sometimes produced by landslides are very similar in appearance to those formed by glaciers (Fig. 54). Moreover, the heterogeneous clays and angular bowlders of which they are composed resemble glacial debris. The rocks of landslides, however, instead of being scratched, as is true of glacial bowlders, often show impact marks , formed by the striking of one rock against another in their violent descent down the mountain side. CONCRETIONS Although concretions are usually of little geologic importance, they occur so frequently in the rocks and sediments of the earth and excite so much interest that they deserve some attention (Fig. 55). 7 6 PHYSICAL GEOLOGY Concretions are masses varying greatly in shape and in size from less than a pinhead to more than 10 feet in diameter, and are formed by the gradual segregation of mineral matter. The shape, as has been said, varies greatly. Some concretions are spherical, some are flat, and others curved. The odd shapes which resemble animals (Fig. 55) are usually produced by the growth of two or more concretions until they join. The center of attraction may be a fossil or a bit of mineral, FIG. 55. Clay-stone concretions of various shapes. They are composed largely of lime carbonate and occur in clay. but in the majority of specimens no nucleus can be detected. In some formations (for example, the Arikaree, Miocene, in Nebraska) they may, by their abundance, so strengthen the loose sands and clays containing them as to form a resistant bed which stands as cliffs wherever cut by streams. Concretions usually occur in definite beds in a formation, and it is sometimes possible to trace such beds for several miles. They occur in rocks of every age, from the most ancient to those now forming on the bottoms of lakes and seas. THE WORK OF GROUND WATER 77 Composition of Concretions. Concretions are seldom of the same composition as the containing rock; those occurring in limestone are apt to be of silica; in clays and shales, of lime or iron carbonate; in sand and sandstones, of iron oxide or lime carbonate. Lime concretions, or clay stones, are probably more abundant than any others. When concretions of limestone and iron carbonate (clay ironstones) are much cracked in the interior and the cracks filled with calcite or quartz, they are called septaria (Fig. 56). In sandstone iron concretions of two kinds may occur: " spherical," in which a spherical shell surrounds a core of sand, and " pipestem," which, as the name implies, are cylindrical. The former are probably formed as the result of the chemical change of some iron mineral in the rock, such as pyrite, which renders the latter soluble. After being thus changed, " it spreads outward as a drop of ink does on blotting paper. Evapora- tion takes place around the outer margin of the solution, iron oxide is precipitated, and the first ring or shell is formed." (J. Geikie.) Pipestem concretions are formed where soluble iron compounds are oxidized about the tubes produced by the roots of plants. It is probable that certain masses of gravel in southern California which now stand up as hills have been cemented together by a kind of concretionary i FIG. 56. A polished section of a sep- tarium. The white veins are calcite, the darker portions chiefly lime carbonate. action. 1 The flint nodules that are so abundant in the chalk of southeastern England some- times had their beginnings in sponges which secreted a siliceous skeleton, and in other fossils. Upon this small quantity of silica as a center, other silica taken from the sea water was added to form the nodular flints. Since by a microscopic examination the structure of the chalk in which the nodules lie can be traced, it is evident that the flint nodules were formed in the chalk mud of the ocean floor, rather than on top of these sediments. Time of Formation. Lime concretions or clay stones are formed by the gradual accumulation of lime carbonate, and during their growth they inclose portions of the sediments in which they lie. They are often formed before the rock containing them is hardened (indurated), as is shown by the facts that (i) they are often cut by joints and (2) when they contain fossils, these remains are seldom flattened by the pressure of the overlying rocks as are those in the surrounding shale. Although many of the concretions which occur in sedimentary rocks were formed while they were in an un- consolidated state and before they were deeply buried, there is no doubt that some were formed after the sediments had been consolidated into rock. Oolitic Limestone (Greek, oon, egg, and lithos, a stone), so-called be- cause of its resemblance to fish roe, may be almost completely corn- Arnold, R., Jour. Geol., 1907, Vol. 15, pp. 560-570. PHYSICAL GEOLOGY posed of minute concretions (Fig. 57). Limestone of this origin (p. 249) is often widespread and many feet in thickness. It is, however, held by some investigators that the most of the oolitic limestone is the product of microscopically small algae (plants) capable of secreting lime. Geodes. Geodes differ from concretions in that they are formed in cavities of the rock and from without inward (Fig. 58). When lava contains steam cavities, silica may be deposited on the walls of the cavities and, by slow addition, may in time fill them. In this way agates are formed, the colored layers of which are due to coloring matter carried in and deposited with the silica. Other geodes are formed by the force of crystallization in the following way : if silica begins to crystallize in the cracks of a crushed fossil embedded in a rock, a shell, for example, the fragments of the shell may be forced farther and farther apart by the force of crys- FIG. 57. A hand specimen of oolitic limestone. (U. S. National Museum.) FIG. 58, A geode broken in two. Cheyenne River, South Dakota. THE WORK OF GROUND WATER 79 tallization until a hollow sphere, many times larger than the original fossil, may result, lined with crystals. In some geodes of this sort the fragments of the fossil may be entirely dissolved away. REFERENCES FOR UNDERGROUND WATER SPRINGS BOWMAN, I., Forest Physiography, pp. 41-61. DE MARTONNE, E., Geographic Physique, pp. 342-347. FULLER, M. L., Occurrence of Underground Waters: Water-Supply Paper, U.S. Geol. Surv. No. 114, 1905, pp. 18-40. GREGORY, H. E., Underground Water Resources of Connecticut: Water-Supply Paper, U. S. Geol. Surv. No. 232, 1909, pp. 60-76. SLIGHTER, C. S., The Motions of Underground Water: Water-Supply Paper, U. S. Geol. Surv. No. 67, 1902. WEED, W. H., Formation of Travertine and Siliceous Sinter by Fe gelation of Hot Springs: Ninth Ann. Rept., U. S. Geol. Surv., 1889, pp. 613-676. WOODWARD, H. B., The Geology of Water Supply, pp. 79-95. ARTESIAN WELLS CHAMBERLIN, T. C., Artesian Wells: Geology of Wisconsin, Vol. I, 1883, pp. 689- 701. CHAMBERLIN, T. C., The Requisite and Qualifying Conditions of Artesian Wells: Fifth Ann. Rept., U. S. Geol. Surv., 1885, pp. 125-173. DARTON, N. H., Geology and Underground Water Resources of the Central Great Plains : Professional Paper No. 32, U. S. Geol. Surv., 1905, pp. 190-372. KARST DE MARTONNE, E., Geographie Physique, pp. 462-472. GEIKIE, J., Earth Sculpture, pp. 266-277. KATZER, F., Karst und Karsthydrographie, 1909. KNEBEL, W. V., Hohlenkunde mil Beriicksichtigung der Karstphdnomene, 1906. CAVES BLATCHLEY, W. S., Indiana Caves and their Fauna: Twenty-first Ann. Rept., Ind. Geol. Surv., 1897, pp. 121-212. HOVEY, H. C., Celebrated American Caverns. M ARTEL, E. A., Les Abimes. MATSON, G. C., Water Resources of the Blue Grass Region: Water-Supply Paper, U. S. Geol. Surv. No. 233, 1909. SHALER, N. S., Aspects of the Earth, pp. 98-142. CONCRETIONS AND GEODES ARNOLD, R., Dome Structure in Conglomerate: Jour. Geol., Vol. 15, 1907, pp. 560- 570. GEIKIE, J., Structural and Field Geology, pp. 120-124. CLELAND GEOL. 6 8o PHYSICAL GEOLOGY GRABAU, A. W., Principles of Stratigraphy, pp. 467-475; 718-721. GRATACAP, L. P., Opinions upon Clay Stones and Concretion : Am. Naturalist, Vol. 18, 1884, pp. 882-892. MERRILL, G. P., Rocks, Rock- feathering, and Soils, pp. 35-37. NICHOLS, H. W., New Forms of Concretions: Field Columbian Museum, Geol. Series, Vol. 3, No. 3, 1906, pp. 25-54. SHELDON, J. M. C., Concretions from the Clay Stones of the Connecticut Valley. GENERAL HOBBS, W. H., Earth Features and their Meaning, pp. 180-194. RIES AND WATSON, Engineering Geology, pp. 295-357. TOPOGRAPHIC MAP SHEETS, U. S. GEOLOGICAL SURVEY, ILLUSTRATING THE WORK OF GROUND WATER Arredondo, Florida. Greenville, Tennessee North Carolina. Bristol, Virginia^Teni^essee. Williston, Florida. Weingarten, MissourT*4llinois. Standingstone, Tennessee. Princeton, Kentucky. Lockport, Kentucky. Kingston, Tennessee. Waterloo, Illinois. CHAPTER IV THE WORK OF STREAMS IT is difficult to over-emphasize the importance of streams, since they carry off the excess of rainfall above evaporation, with the excep- tion of the ground water which enters into chemical composition with rocks or is discharged in underground courses directly to the seas (p. 56). The quantity of water carried in streams is therefore enormous. It has been estimated that the rivers of the world annually discharge 6500 cubic miles of water; a volume which, if spread over Massachusetts, would cover it three quarters of a mile deep. The water of flooded streams is derived largely from -rainfall, while the chief source is spring water, when they are low. FACTORS IN STREAM EROSION Material Carried by Streams. In walking up a small valley one can readily discover the sources of the gravel and sand in the stream bed, and of the mud which renders the water turbid. The small particles which have been broken from the rocks of the banks by the various agents of the weather, and the larger fragments which have been loosened by frost are continually being carried down into the bottom of the valley by gravity (hillside creep, p. 73) and washed down by rains ; deposits of sand and clay through which the valley is cut in places furnish an easy supply during floods ; the solid rock of the valley sides, when undercut by the stream, falls into the water; and some sediment is obtained from the bed over which the stream flows. How the Sediment is Moved. Streams accomplish their work of removing this load of sediment (i) by pushing along the larger of angular rocks, (2) by rolling the rounded and smaller pebbles, and (3) by carrying in suspension the finer sand and clay, as well as such thin, flat particles as mica flakes. This ability of running water to carry fine particles in suspension is due to the fact that the smaller the volume of an object, the larger in proportion is its surface. This 81 82 PHYSICAL GEOLOGY being the case, a slight upward current, formed by the deflection of the water from the irregularities of the stream bed or side, will lift small particles and carry them onward until they >- again fall to the bottom, or are caught up by an- ^r _j *~ ^^ ** other current (Fig. 59). ^ / "X^ ) "^*V^T~Ul. ) ^ e ec ^* es anc ^ cross cur ~ rents of a river are espe- FIG. CO. Diagram showing how upward cur- n rr i i rents are produced by irregularities on the bed Cially effective in this work of a stream. during high water. In this way, sand and clay, after many short journeys, are ultimately carried to the ocean. The quantity of sediment carried by a stream depends upon its volume and velocity and on the amount and nature of the accessible material. Factors Determining the Velocity of Streams. The velocity of a stream depends upon (i) the slope of its valley, (2) its volume of water, (3.) the amount of its load, and (4) the shape of its channel. It is greatest in the middle of the stream, and some distance below the surface. If the volume of a stream is increased eight times, its velocity is doubled, since the velocity varies as the cube root of the volume ; if the amount of the sediment is decreased, the velocity is increased ; and, other things being equal, a stream following a straight channel flows faster than one in a winding course, because it loses less energy in friction with its sides. A stream which is ordinarily clear is often muddy when swollen, both because of the greater run-off which enters it, and because of the large amount of sediment which it is enabled to tear from its bed and banks on account of its greater velocity. If the velocity of a stream is increased several times, its power becomes almost incredible. It has been shown that a current moving six inches a second will carry fine sand; one moving 12 inches a second will carry gravel; four feet a second, stones of about two pounds weight; eight feet a second, stones of 128 pounds; 30 feet a second, blocks of 320 tons ; if a stream can ordinarily move a pebble of one ounce, it can move a stone of four pounds when doubled by a flood. This fact is expressed in the law that the transporting power of a stream varies as the sixth power of its velocity. Keeping the above law in mind and remembering that a heavy object loses about one third of its weight in water, it is easy to understand the cause of the destructiveness of such floods as that which overwhelmed Johnstown, THE WORK OF STREAMS 83 Pennsylvania, in 1889, and swept away large rocks, twenty-ton loco- motives, and massive iron bridges as easily as, under ordinary cir- cumstances, the river could move sand. A fall of one foot in a mile is quite sufficient to carry a river steadily onward ; one foot in a thousand feet will make a fairly rapid river; one in two hundred, a torrent. Water Wear. The pebbles and sand carried by the streams are worn away by their impact against the bed rock and by striking against each other. The result of such wear is the production of rounded stones. In mountain streams the angular fragments from the talus are rounded before they have been carried a mile. Solution. In addition to the sediment carried by the force of the current, the waters of every river contain a large amount of mineral matter in solution. This is largely obtained from springs, but also from the run-off and by the solution of the stream bed. The amount in any stream varies with the season, being greater in proportion to the volume of water in dry than in wet seasons, since in the former the water is largely underground water. The small river Thames, England, carries to the sea about 348,230 tons of dissolved minerals a year, and the Mississippi River carries 113,000,00x3 tons. " The Rhine carries enough carbonate of lime to the sea each year for the annual formation of 3,320,000,000 oyster shells of the usual size." (A. Geikie.) It is estimated that in every 5000 years rivers carry their own weight of minerals in solution to the sea. The weight of the dissolved matter carried to tidewater by the streams of the United States (270,000,000 tons) is more than half that of the sediment (513,000,000 tons). 1 " The tons per square mile per year removed from different basins show interesting comparisons. In respect to dissolved matter the southern Pacific basin heads the list with 177 tons, the northern Atlantic basin being next with 130 tons. The rate for the Hudson Bay basin, 28 tons, is lowest; that for the Colorado and western Gulf of Mexico basins is somewhat higher. The denudation estimates for the southern Atlantic basin correspond very closely to those for the entire United States." (Dole and Stabler.) Vertical Erosion (Corrasion). 2 By erosion (Latin, erodere, to gnaw away) streams are able to cut down their valleys. This may be 1 Water-Supply Paper, U. S. Geol. Surv. No. 234. 2 The terms corrasion, abrasion, corrosion, erosion, and denudation have sometimes been used rather loosely in geological and geographical literature. In this work corrasion (Latin, corra- 8 4 PHYSICAL GEOLOGY accomplished by (i) the mere impact of the water, especially if the rock is easily disintegrated (Fig. 60). The effect of clear water upon striking loose sediment with great force is well shown in hydraulic mining. This principle was also employed in the leveling of a portion of Seattle, where a high hill was cut down by means of a power- ful stream of water. (2) In thinly bedded rocks, such as shales FIG. 60. A bank undercut by clear water. (U. S. Geol. Surv.) (p. 250), the stream bed may be deepened by " lifting " ; that is, the shale, broken by joints, is separated by the water along the bedding planes (p. 234) ; and the fragments are thus floated off. The effect of this process alone in regions underlain by shales may be of the greatest importance. "Lifting" is especially effective when the stream beds have been exposed to the weather at low water. At such times, temperature changes or frost may loosen much material in the bed, which is picked up and removed during high water. Water without sediment has little effect in eroding thick dere, to rub) and abrasion are used as synonyms, meaning the detachment of rock particles as a result of wear ; corrosion (Latin, corrodere, to gnaw) is used for the work done by solution ; erosion is used to include both corrasion and corrosion, as when we say a river erodes its valley , or a sea erodes its shores. The term denudation is reserved for the lowering of a land surface by any agency. THE WORK OF STREAMS bedded rocks, as is apparent on the brink of Niagara Falls where the thousands of tons of water which pour over them hourly are unable to remove the r ^ .. soft algae which cover the rocks, as the water is filtered by Lake Erie. (3) When swift streams are supplied with tools (Fig. 61) in the form of sand and pebbles, their erosive power becomes greatly in- creased. Weathering and Vertical Erosion. - *. j FIG. 61. Bowlders in a stream bed. Here the bowlders valleys are usually form a pavement which hinders the erosion of the valley. wider at the top than at the bottom. This is due to the fact that while the valley is being deepened by erosion it is also being widened in several ways. The rock is loosened by the various agents of the weather and carried to the stream by rainwash and wind. Normally, valleys are most rapidly widened in temperate regions, since there the soil freezes and thaws frequently so that " creep " (p. 73) plays an important role. Valleys cut in sand or clay are often widened to a considerable degree as a result of the pressure of the overlying sediment, which forces the unconsolidated sand or clay at the base to " flow out," causing a slumping of the upper portion. Animals walking on the slopes, falling trees, the cutting of the stream against its sides are among the agents which help to loosen the material of the valley sides and thus tend to widen the valley. If erosion is very rapid as compared with the work of the agents of the weather, steep-sided gorges barely wide enough to accommodate the stream will result : such are the gorge of the Aar at Meirengen, Switzerland, the picturesque gorges of Watkins Glen and Ausable Chasm, New York, and the canyon of the Virgin River in Arizona. Usually, however, young valleys are V-shaped, the wearing back of the sides more than keeping pace with the deepening of the valley. 86 PHYSICAL GEOLOGY Base Level of Erosion. 1 If a stream is swift, it continues to deepen its valley as it flows from the higher lands to the sea, until at or near the mouth, the bed will be at, or even slightly below, sea level. (The bed of the Mississippi River is locally as much as 100 feet below sea level.) The entire length of the valley, however, will not be deepened to the level of the sea, since as its slope (gradient) is diminished, the ability of the stream to erode its bed also decreases, and before sea level is reached the stream will have ceased to deepen its valley in its upper course. When this condition is attained, the stream is said to be at base level; that is, it has reached the lowest level to which a stream can wear a land surface. As the stream approaches base level, its current becomes less and less rapid, so that the deepening of the last few feet of the valley may take longer than all the rest. If the land is raised and the gradients of the streams are increased, they will again cut until a new base level is reached. If on the other hand the land is lowered, base level will be reached more quickly. During their histories streams usually reach a number of tem- porary base levels. If, for example, a stream flows into a lake, it cannot cut lower than that level; and if the lake remains in existence for a long time, the stream will excavate a broad valley where it enters the lake. Again, if a stream flows over a stratum of hard rock in its lower course while its upper course is in less resistant rock, the depth to which it cuts in the hard rock will be the temporary base level, and a broad valley will be developed above the resistant rock, while the latter will constitute the steep-sided narrows so characteristic of the scenery of eastern Pennsylvania. Effect of Load. Whether a stream carrying sediment will erode or deposit depends upon its velocity and upon the amount of mate- rial. If its velocity is great, the sand and gravel will be used as tools with which to cut down the stream bed, or widen it. If, however, the velocity is sufficiently decreased, as frequently occurs when a side stream with a steep gradient flows into a master stream with a 1 The term base level has been used in several senses, the difficulty arising because of the fact that as commonly used the surface described is a slope and not a level plain. It has been suggested that "base level" be limited to the level base with respect to which normal sub- aerial erosion proceeds ; to employ the term grade for the balanced condition of a mature or old river ; and to name the geographical surface that is developed near or very near the close of a cycle a "peneplain" or "plain of gradation." (Davis, Wm. M., Geographical Essays, P- 387.) As used in this volume a base level is the lowest possible slope to which a region can be cut by running water. Thus a stream in a canyon may cut its channel to base level ages before it develops a plain at that level. A peneplain is any extensive tract of land reduced to essential planeness (base level) by the erosion of running water. THE WORK OF STREAMS gentle grade, it may drop its load. Decrease of volume due to evap- oration and to the absorption of the water by the soil, such as takes place when a river flows through a dry region, may reduce the stream's velocity to such an extent that it is unable to carry its load of sediment. The Platte River of Nebraska is a typical example of such a river. Its headwaters have a small amount of sedi- ment in proportion to the volume of water, and it is there- fore able to cut a deep canyon in its upper course ; but in passing over the dry and thirsty plains it loses so much water that it is not only unable to degrade its bed, but even de- posits much of its load during the dry season. A stream which flows over such sandy plains has shallow, crooked channels and is con- stantly shifting its course by cutting away the banks in some places and forming bars in others. When the load is so great that it is deposited in the channel, the latter may become too small for the water of the stream, in which case the water will break out and follow a new course. If this is repeated many times a network of small, shallow streams, called a braided stream (Fig. 62), may result. The Colorado and Platte rivers have about the same gradient, but the former receives less sediment in proportion to its volume and consequently is able to cut a great canyon, while the lower Platte flows in a broad and shallow valley. FIG. 62. Braided stream, Kandersteg valley, Switzerland. 88 PHYSICAL GEOLOGY When a stream has developed a slope which gives it just sufficient velocity to carry its load, leaving no energy for deepening its bed, it is said to be graded. If a stream has less sediment than it can carry, it will remove material from its bed. If it is unable to trans- port all the sediment brought to it, part of this will be left as a deposit, the channel will be raised, and the gradient will be increased until the stream becomes swift enough to carry away its load. When a stream is at a temporary base level (p. 86) above a fall or rapid, there are often smooth reaches where the stream is at grade. If the land through which a river flows has not been elevated or depressed for a long period of time, few falls will exist, and it will be at grade for long stretches. If, however, a long-continued uplift or several uplifts have occurred, even large rivers may be unable to erode their beds to grade. Even when the last uplift was so remote that the large rivers have been able to develop well-graded courses, the tributaries may, and usually do, have a steep slope. Factors Affecting the Rate of Erosion. The rate of erosion of a stream depends upon a large number of factors, (i) Loosely com- pacted rocks, or rocks with a soluble cement, are easily eroded. If, for example, the grains of a sandstone are held together with lime, the solution of the cement will cause the grains to fall apart and thus render the work of the stream easier. (2) Rapid erosion is further favored if the rock has numerous joints and is thin-bedded (p. 24). Usually sedimentary rocks are more readily eroded than massive, crystalline rocks (p. 330) such as granite. (3) The greater the velocity of a stream, the greater, other conditions remaining the same, will be the erosion. Since the velocity of a stream depends upon the volume of water as well as upon the slope of its bed, the cutting power will be greater during floods (p. 82). (4) Under any of the above con- ditions erosion will be favored if the stream has sufficient sand and gravel with which to cut its bed but not so much that a large part of its energy is expended in carrying it. When the amount of sedi- ment is increased without an increase in the volume of water (p. 86), or when the quantity of sediment remains constant but the volume of water decreases, erosion may cease and deposition take place. Rapid erosion by abrasion requires some sediment, but not too much, a steep slope, and a considerable volume of water. Scour and Fill. A stream at flood may be deepening (degrading) its channel where its velocity is great, at the same time that it is build- ing up (aggrading) its flood plain where the velocity is slight. After THE WORK OF STREAMS 89 the flood has subsided the channel thus deepened may be entirely filled with sediment. This process is called scour and fill. The Missouri River sometimes scours out its channel to depths of from 70 to 90 feet and later fills it again. It is evident that the deepening of the beds of such rivers is largely confined to high water. A fail- ure to understand scour and fill has led some observers to assign a great age to stone implements found deeply buried in river gravels (p. 680). Lateral Erosion. When a young river is deepening its valley it flows in a narrow channel between steep banks, but since its course is seldom straight it tends in places to cut more on one bank than on the other, with the result that as it cuts downward it also cuts sidewise, thus widen- ing its valley. By the time grade is reached the valley walls will have flared open, but FJG ^ _ Unsymmetrical valley formed as a will be steeper on the of the dip of the rock, outside of each curve. Unsymmetrical valleys are formed (i) in this way and also (2) by the greater hardness of the rock on one side of the stream than on the other (Fig. 63) (where the strike of the rock parallels the course of the stream). When two neighboring streams have ceased to degrade their beds, they will cut laterally and may in time wear away the divide which separates them, thus causing one to flow into the other. FEATURES DUE TO STREAM EROSION Falls and Rapids. Falls and rapids result from a number of causes, (i) Regions in which a harder layer of rock overlies a softer one fur- nish most favorable conditions for the formation of falls (Fig. 64). When a stream, in deepening its valley, encounters a harder bed of rock lying in the position shown in the diagram (Fig. 65), the less resistant beds are worn more rapidly than the harder ones, and a rapid will result first, which upon further erosion will become a fall. Falls become lower and lower in the course of time, until the resistant beds form mere ledges in the stream bed and the falls cease to exist (Fig. 65). Niagara Falls (Fig. 66) have gradually cut back until now they are seven miles from their original position. The recession of these 90 PHYSICAL GEOLOGY falls and their verticality are due to the fact that the strata which compose the higher land consist of massive limestone, about 80 feet thick at the falls, which are underlain by soft and easily weathered FIG. 64. Falls of the Genesee River, Rochester, New York. (Photo. C. R. Dryer.) and eroded shale. When the water plunges over the limestone, it wears away the soft rock beneath more rapidly than the hard capping stratum, leaving the latter projecting. Fragments are continually falling from this overhanging ledge and are used by the water as tools FIG. 65. Diagram illustrating the recession of a waterfall formed by a resistant bed that dips up the stream. (Modified after Salisbury.) to excavate the shale further. This erosion is also aided materially by blocks of ice in winter. The height of the falls is about 165 feet, and the gorge which has been excavated is from 200 to 400 yards wide and THE WORK OF STREAMS FIG. 66. An ideal section of Niagara Falls, showing how the soft shales are being worn away, leaving the limestone above unsupported. (Gilbert.) about 300 feet deep in places. The rate of cutting of the Canadian Falls has been about 4.5 feet a year since 1842, while that of the American Falls, because of the smaller volume of water, is as small as 0.2 foot a year. A natural bridge may be formed when the water above a fall per- colates through a joint or crack athwart the stream and thence along a bedding plane or approxi- mately horizontal crack, emerging under the fall as a spring. If the cracks are enlarged by solution and erosion, a tunnel large enough to accommodate the entire volume of the stream may be formed, and a natural bridge result (Fig. 67). (2) When the fall of a river in working up stream passes the mouths of tributaries falls develop in them also. The beautiful Min- nehaha Falls of Minnesota are an example of falls formed in this way. (3) In mountainous regions, where the main streams have deepened their valleys rapidly their tributaries are often unable, A B because of their smaller volume, to keep pace with them and therefore flow into them over falls or rapids. To this cause the " roaring brooks " of New England are for the most part due. (4) The Atlantic coast, from New York southward, is bordered by a low- lying plain (Coastal Plain, p. 224) composed of soft, unconsolidated sands and clays. To the westward, this belt joins a belt of older and harder rocks (Piedmont Plateau) along a line roughly parallel with the coast. II 1 \ c II 1 1 ) :":'.*'.. ';''- :''./.:...'.' '.":.'-.'-': :';-;':''".'."'. >';-' .' : ." '-'''''.'' '-'.-, ''':'': FIG. 67. Diagrams illustrating the formation of a natural bridge by the widening of a joint or other crack B athwart the stream, through the solution of the lime- stone by water which reappeared as a spring under the fall at C. In the process of time a tunnel sufficient to carry a large part of the volume of the stream was ex- cavated, and finally the entire volume of the stream. When this was accomplished a natural bridge (shown in the cross section) spanned the valley. 9 2 PHYSICAL GEOLOGY When streams on their way to the sea pass from the hard to the soft rocks they flow over rapids or falls, because the less resistant rocks are cut down more easily than the hard. The boundary between the Coastal and Piedmont plains is for this reason called the " Fall Line," and it is here that many cities are located, both because the falls fur- nish water power and because they deter- mine the head of navigation. (5) When a stream in cutting its bed en- counters a hard rock mass, the erosion of FIG. 68. A fall formed when resistant rock is en- {-fog valley is retarded countered by a stream. The rock CD is hard gneiss, u . , while that represented by lines is softer schist. The line at tftat P olnt > bu * AB is the course of the stream. may continue farther down the valley. A fall or rapids (Fig. 68) will naturally be formed at such a place, and the hard rock mass will constitute a temporary base level which will prevent the stream from deepening its bed above the fall. As a result of the lateral erosion of the stream and of the action of the weather, the valley above the fall may be greatly widened, forming arable land. A case somewhat similar to the above is that of the falls of the Yellowstone, which are the result of the presence of lava, made more resistant by thermal action (Fig. 69). When rocks are FIG. 69. The falls of the Yellowstone River. The rock is lava, and the falls at A and B are due to the superior hardness of the lava at these points. ^ less jointed or fractured, in one portion of a valley than in another, they are less affected by erosion and may produce a fall or rapids. (6) Falls also result where rocks have strongly vertical joints, as vertical joints in homogeneous rocks have the effect of vertically inclined beds. THE WORK OF STREAMS 93 (7) The numerous falls of Switzerland were formed much as in (3), but are largely due to the erosion of the main valleys by glaciers so that the tributary streams enter their mains over falls. These side valleys are called " hanging valleys." Exceptions Falls not the Result of Erosion. (i) A lava stream (Fig. 70) may dam a valley and thus produce a fall. Many examples of this sort might be cited. (2) Limestone (travertine) may be deposited in streams in such quantities as to dam them, form- ing falls (Fig. 71) and even ponding back the water to produce lakes. Topolic Falls in Dalmatia, east of the Adriatic, afford an illustration of the construction of a travertine dam. These falls are 70 feet high and are advancing down- stream. (3) When tributary streams with steep gradients carry a large quantity of coarse debris, they may deposit their loads in the main stream in such amounts as to form temporary rapids. Landslides also accomplish the same result. The Cascades of the Columbia River were formed thus. (4) When a stream is forced out of its valley by landslides (p. 73), glacial deposits (p. 155), or in any other way, falls may result. Potholes. When for any reason a strong, permanent eddy is produced in a stream, as at falls or rapids, pebbles and stones are given a rotary motion as they are carried through the eddy and wear down the stream bed in this place, tending to produce circular holes. These "potholes" (Fig. 72), " washtubs," "giant's caldrons," or " kettles," as they are called, occur in hard granites as well as in shales and limestone, and may be seen in the bed of almost any rapid FIG. 70. Falls formed as a result of the damming of a river channel by lava. (Modified after H. E. Gregory.) 94 PHYSICAL GEOLOGY FIG. 71. Travertine Falls near Davis, Oklahoma. The travertine which has been deposited to form these dams comes from springs containing large quantities of calcium carbonate in solution. The lime carbonate is deposited as travertine when the carbon dioxide escapes. Forty-four dams occur in this creek within a mile. (Oklahoma Geol. Surv.) stream. They vary in diameter from a few inches to ten feet or more and in depth to forty or more feet. The size of a pothole depends upon the velocity and volume of the current and the length of time during which the eddy remains at the same point. By the deepen- ing and coalescing of potholes the channels of streams may be mate- rially deepened, streams sometimes accomplishing their greatest work of erosion in this way. In the Alps there is scarcely a gorge through- THE WORK OF STREAMS 95 out the length of which one cannot see the polished surfaces and regular curves which are the traces of more or less com- plete potholes. Canyons. Can- yons are deep valleys with steep sides. They are formed where the down-cut- ting of a stream (corrasion) greatly exceeds the weather- ing back of the slopes. The conditions favoring the formation of such valleys are (i) a rock capable of maintaining a steep face, such as resistant rock on which the trickling water cannot act quickly, or a firm, permeable rock into which a large part of the water soaks, leaving little for erosion ; and (2) a rapidly cutting stream. (3) An arid climate is more favorable than a moist one, since the work of the weather will be at a minimum in the former. Canyons are nevertheless formed in regions of heavy rainfall (Fig. 64). When a stream approaches base level and ceases to corrade its bed rapidly, the walls of its canyon will be weathered back until in time they form a broad, open valley. FIG. 72. Potholes in gneiss, Shelburne Falls, Massachusetts. FIG. 73. A generalized block diagram of the Grand Canyon of the Colorado. The youthful stage of the region is shown in the fact that the streams have as yet accomplished but a small part of the work to be done. The Colorado valley is a young valley. The cliffs of the canyon are formed of resistant beds, while the slopes are of weaker beds. One of the grandest canyons in the world is the Grand Canyon of the Colorado in Arizona (Fig. 73), which was formed under condi- tions most favorable for steep-sided valleys. The river flows through CLELAND GEOL. ^ PHYSICAL GEOLOGY a high plateau, 6000 to 8000 feet above the sea, in which it has cut a trench a mile deep in certain places. The climate is arid ; the gradient of the valley is steep ; the amount of sediment is sufficient to furnish tools for cutting, but not so great as to overload the stream ; the rocks are sandstones, limestones, and shales, overlying granite. The Grand Canyon in Arizona is about 220 miles long and may be described as a valley within a valley, since, in certain localities, the upper portion, cut in the softer, sedimentary rocks, is eight to ten miles wide, while the lowest part cut in the hard granite is barely wide enough to hold the river. The total depth of the canyon is almost a mile. The canyons of the tribu- tary streams branch again and again as theyare followed back, and are miniatures of the Grand Canyon. The gorge of the Niagara River, Au- sable Chasm, and Watkins Glen, in New York, are examples of canyons developed in a moist region. In these cases the valleys are all postglacial and have been cut so rapidly that their sides have been but little widened by the weather. In Ausable Chasm (Fig. 74) the verticality of the walls has been maintained in places by vertical joints. Instances of Rapid Erosion. The Duna, a river of eastern Prussia, blocked by an ice jam in 1901, was forced to take a new course. In thirty-four hours it was able to cut a gorge one meter to three and a half meters deep and four meters to eight meters wide, representing an excavation of 2250 cubic meters of material. The bottom of the Sill tunnel in Austria was provided with a pavement of granite slabs more than a yard thick. Great quantities of debris were swept over this pavement at a high velocity, and so rapid was the abrasion that it was found necessary to renew the granite slabs after a single year. Effect of Deforestation on Rivers. When forests are cut down or the vegetation on the hills is killed, the latter being sometimes the FIG. 74. Ausable Chasm, Chazy, New York. This is a young valley. (U. S. Geol. Surv.) THE WORK OF STREAMS 97 case in the vicinity of smelters, erosion may be very rapid. This is well shown in Potato Creek (Figs. 75, 76) in the Ducktown copper FIG. 75. Potato Creek, Tennessee, a stream overburdened with waste and aggrading. (See also Fig. 76.) (U. S. Geol. Surv.) region of Tennessee, where the waste from the bare slopes is too great for the stream to remove and is piled up along its course as a flood plain (p. 1 19). In this creek the waste has accumulated for a number of years at the rate of a footer more a year (Fig. 76, A, B), and has FIG. 76. Generalized diagram showing the effect of deforestation on Potato Creek, Tennessee. A shows the former condition of the valley, and B the condition after the timber had been killed and the stream loaded with sediment. The telephone poles were buried to their cross arms. built up a flood plain in which telephone poles are buried almost to their cross arms, while highway bridges and roadbeds have been either buried or swept away by floods. 9 8 PHYSICAL GEOLOGY The effect on stream flow of for- ested and deforested (Fig. 77) areas is well illustrated near Bilt- more, North Caro- lina. The David- son River has its upper drainage basin in the Pisgah for- est; the Tuckasegee River in a defor- ested land that has been logged, burned over, pastured and farmed. The two areas drained are of geologically the same age and structure ; the headwaters of the streams are found within the same range of mountains ; the rainfall of the two areas is the same ; the steepness of slope of the two watersheds is about the same. Yet the Tuckasegee, though the larger river, shows greater fluctuation in discharge than the Davidson; and the Davidson is practically free from sediment, while the Tuckasegee bears gravel and sand which it often spreads over fertile lands. Growth of Valleys. It is possible to study the growth of a valley in almost any region. Water does not flow down a slope in sheets for long distances, but FIG. 77. Rapid erosion of deforested land and one method of preventing further erosion. (U. S. Geol. Surv.) FIG. 78. A young valley, western Nebraska. work of the stream has only begun. The THE WORK OF STREAMS 99 FIG. 79. Block diagram showing the manner in which the divide between two streams is narrowed. soon finds depressions where it accumulates into streams. Even though a slope were perfectly uniform, a slight heterogeneity of soil or rock would permit the water to remove more material in one place than in another , and thus begin the excavation first of a gully, and later, by prolonged erosion, of a ravine which still later would develop into a broad valley. A valley is length- ened at its upper end and is cut back by the water which flows in at its head (Fig. 78), the direction being determined by the greatest volume of water which enters it. This is called headward erosion. A valley is widened by rainwash, lateral erosion (Fig. 79), and in other ways (p. 89). Its length depends upon the distance to which its stream can cut inland. At the beginning a valley has running water only during and immediately after rains, but later, when it has cut below the water table (p. 56), a permanent stream flows through it (Fig. 32, p. 57). Tributary streams tend to turn in the direction of their main (Fig. 80), a feature which is often most pronounced late in their history. Valleys Formed in Ways Other than by Stream Erosion. Although the great majority of valleys are de- veloped by stream erosion, some were already formed for the streams which FIG. 80. Map showing the usual flow through them. The popular relation of tributary streams to the not i on tnat canyons, such, as that of main stream into which they flow. . . . A . the Colorado River in Arizona, were formed by great cataclysms which rent the earth and produced the deep fissures now occupied by streams, is without foundation. Streams, however, do occasionally flow into fissures formed by 100 PHYSICAL GEOLOGY the fracturing of the surface during earthquakes, but they are so few as to be unimportant. Some great valleys, nevertheless, were ready made for the rivers which flow through them. The Great Valley of California, through which the San Joaquin and Sacra- mento rivers flow, was formed, not by stream erosion, but by BLACK FOREST FIG. 81. Section across the Vosges and Black Forest, Germany, showing the graben in which the Rhine flows. (Penck.) the sinking of the land along a valley-like depression, or by the uplift of parallel mountain folds, and is called a structural valley. Into such a depression streams may flow from the high lands on the sides and unite (unless the region is arid) to form a river system. The Great Basin region of Utah is also a structural valley, but because of the aridity of the climate no streams flow through it. The River Jordan and the Dead Sea are in a valley R FIG. 82. Diagram A illustrates the development of parallel consequent streams on a sloping surface. Diagram R is the same region after the streams have become ad- justed to the structure of the underlying rocks. The streams entering the main at right angles are subsequent streams. The main stream flows through its water gap in the hard ridge. The gaps on either side were eroded by former streams but no longer have streams in them and are called wind gaps. formed by the sinking (faulting, p. 261) of a long and comparatively narrow block of the earth's crust. Such a valley is called a rift valley. Owen's valley in California and a portion of the Rhine valley in Germany (Fig. 81) are other examples of valleys due to faulting. Glaciers excavate valleys in the solid rock, which may afterwards become occupied by streams, but these are usually merely THE WORK OF STREAMS IOI ancient river valleys which have been widened and deepened by the ice (p. 129). The Direction of Val- leys. The direction of stream valleys depends upon a number of condi- tions, some of which can be illustrated by a hypo- thetical case. If a portion of the bottom of a shallow sea is raised above sea level, the land, under these conditions, will have no established stream valleys. When, then, the rain falls first upon such new land, it will gather in places where there are depres- sions and form large or small lakes. Elsewhere, rivulets will flow down the slope, joining here and there in their descent until a stream of considerable length develops. Streams of this sort, whose position and direction are deter- mined by the slope of the original land surface, are called (i) consequent (Fig. 82 A), since their direction is a consequence of the topography of the country, without regard to the char- acter of the rock through which they pass. The streams on the Atlantic Coastal Plain are chiefly FIG. 83. Diagram A shows a region in which a stream flows at grade. Diagram B shows the same region after it has been slowly upwarped athwart the course of the stream. The river is shown as having been able to deepen its valley as rapidly as the elevation occurred. A stream with such a history is an antecedent stream, since it was able to maintain the course it had prior (antecedent) to the deformation of the surface. 102 PHYSICAL GEOLOGY consequent streams. As streams deepen and lengthen their valleys, their tributaries may encounter new kinds of material and find that some are more easily eroded than others, with the result that they gradually develop valleys in the less resistant rocks. In such case, the position and size of the branch streams are determined by the nature of the underlying rock and not by the original slope of the surface; the valleys being cut in the weaker strata, while the harder strata stand up as ridges or even mountains. The Shenandoah valley of Virginia, the Lehigh val- ley of Pennsylvania, and the Hoosic and Hoosa- tonic valleys of Massa- chusetts and Connecticut are examples of valleys of this type. Valleys formed in this way are called (2) subsequent (Fig. 82 B), the process being known as structural ad- justment. It will readily be seen that if streams drain adjoining regions, the one whose course is most generally confined to the more easily eroded beds will grow more rapidly and so may cut headward until it captures branches or even the entire upper courses of streams less favorably situated. Such a process is called stream piracy (p. 107). If the land is warped up athwart the course of a consequent stream whose direc- tion is so well established that it is able to degrade its bed as rapidly as the elevation takes place, thus keeping its old course, the stream is called antecedent (Fig. 83 A and B}. (3) Another factor which sometimes determines the direction of a stream is faulting (p. 261) (Fig. 84). In regions of pronounced faulting, such as the Adirondacks, the courses of many streams may, FIG. 84. Direction of drainage determined chiefly by faulting and jointing, near Lake Temiskaming, Ontario. (After Hobbs.) THE WORK OF STREAMS 103 for considerable stretches, follow lines of dislocations. (4) Where the rock over which a stream flows is strongly jointed (Fig. 85), the joints are sometimes followed to some extent by the smaller tribu- taries. Larger streams, however, are less affected, usually showing little evidence of this influence. (5) When streams flow through structural valleys (p. 100), their direction is necessarily predeter- mined. (6) In a region underlain by horizontally bedded rock, the FIG. 85. Fall Creek, South Dakota. Showing the effect of jointing on the course of a stream. valleys extend in many directions without systematic arrangement and are described as dendritic (treelike). Such a river system is in striking contrast to one developed in a region of tilted strata in which the beds vary in their resistance to erosion. In such a region the tributaries have a trellised appearance (Fig. 92, p. 107). Basins and Divides. All the land surface which is drained by a river and its tributaries is called its hydro graphical or drainage basin, and the boundary between two river basins is termed the divide, since the water falling on it is divided, part flowing into one river system and part into the other. A part of the Great Conti- PHYSICAL GEOLOGY nental Divide, which separates the basin of the Mississippi River which empties into the Gulf of Mexico and that of the Snake River which finally discharges its waters into the Pacific Ocean, is in the Yellowstone National Park. A divide may be a sharp ridge, as, for FIG. 86. Block diagram illustrating the formation of outliers and the erosion of a plateau. The fronts of the High Plains in Nebraska and elsewhere are being cut back in this way. example, in the Kicking Horse River basin of British Columbia, where the divides between the tributaries have been worn down to knifelike ridges which in many places are not a foot in width; or a flat plain, so level that the location of the divide is uncertain. Such a divide is the height of land between the Great Lakes and Hud- son Bay, where the same swamp often drains both north and south. The position of the divide between the Orinoco and Amazon rivers in South America is, perhaps, even more uncertain. Divides are seldom stationary, since the streams on the opposite sides do not usually cut head- ward or laterally with equal rapidity. The divide between two tributaries of the same river may also be narrowed by the lateral erosion of the streams until it dis- appears (Fig. 79). By an increase in the number of tributaries, ridges are cut into hills. In this way the Seven Hills of Rome were sculptured, and many of the conspicuous buttes of the western United States (p. 328) were separated from the higher plains (Figs. 86, 87). FIG. 87. Eagle Rock, Nebraska. (U. S. Geol. Surv.) THE WORK OF STREAMS 105 Elevations Due to Unequal Hardness. The term hogback is given to narrow ridges which stand above the general level of a region, because of the greater resistance of a steeply dipping layer of rock and of the greater erosion of the softer rock (Fig. 88 A, B). They are especially conspicuous on the flanks of mountains. When regions in which the rocks are folded have been subjected to erosion the harder beds stand up as mountain ridges. In this way the Appa- w s w s w FIG. 88. Photograph and section of a hogback near Canon City, Colorado. The ridge is due to the superior strength of one main and two subordinate strata. SS are strong and WW weak rocks. (After Brigham.) lachian Mountains (p. 477) were formed. The difference between a hogback and such mountains is largely one of height, width, and extent. Where sheets of lava cover softer beds, as is not uncommon in the southwestern portion of the United States, flat-topped, isolated hills, called mesas or tables (Fig. 89), are formed by the headward cutting of tributary streams. Any harder bed of horizontal rock io6 PHYSICAL GEOLOGY FIG. 89. Black Mesa. (U. S. Geol. Surv.) overlying softer beds will produce such hills, the name " mesa " being used to designate the shape of the hill, not the kind of rock. In the western United States the word butte is used for any steep- sided hill and is also loosely used for any conspicuous elevation. Outliers. When a part of a formation is separated from the main body by erosion (or, occasionally, by faulting), it is called an outlier (Fig. 90). It is, therefore, simply a remnant of a more ex- tensive bed or series of beds. Outliers are usually short-lived, since they are objects of attack on all sides by erosion. Outliers FIG. 90. In the diagram outlier A was formerly united to By but was separated from it by erosion. often occur scattered along the front of prominent escarpments; as, for example, near the border of the High Plains in the Middle West (Fig. 86). Rock Terraces. In a region underlain by alternate hard and soft beds, such as in the Colorado Plateau, the resistant rocks may form rock terraces and the softer rocks slopes in the river valley or canyon. The "steps" or "benches" of the walls of the Grand Canyon of the Colorado (Fig. 73, p. 95) are among the most striking features of this remarkable valley. Rock terraces may also result THE WORK OF STREAMS 107 from the elevation of the land, since when the gradient of a river is increased it is able to cut a gorge in its old valley floor, leaving rock terraces on the two sides. Stream Piracy. Because of the more rapid headward cutting of one stream than another there is a continual though usually slow FIG. 91. One of three diagrams showing the development of topography in a region where the underlying strata are inclined (dip) and vary greatly in their resistance to erosion. The region is conceived to be reduced to a peneplain with low ridges of harder strata. (Modified after Davis.) absorption of the tributaries of one river system by another and also a struggle for existence among the tributaries of each river system. A stream which has cut headward so rapidly as to divert the headwaters of another stream to itself is said to behead the latter, and the act is spoken of as stream piracy (Figs. 91, 92, 93). FIG. 92. In this (second) diagram the peneplain has been elevated and the streams have cut deep valleys and picturesque water gaps. The direction of the tributary streams is determined by the strata, and a "trellised" drainage system results. The result of stream piracy is well shown in the difficulties experi- enced by a commission appointed by the Argentine and Chilean gov- ernments to determine a disputed boundary in the Andes between io8 PHYSICAL GEOLOGY the two republics. Since no map was available when the boundary was first fixed, this was stated as following the crest of the mountains, as it was believed that this was permanent and was identical with the divides between the Atlantic and Pacific rivers. Later a dispute arose as to the exact boundary, and the survey made to settle the ques- tion showed how inaccurate this belief was. It was found that the Chilean rivers, with their short, steep routes to the Pacific Ocean, had captured the upper courses of nearly all of the Argentine rivers, obliging them to make sudden turns and to flow through the deep gorges which lead to the Chilean coast. Another example of stream FIG. 93. In this (third) diagram some of the subsequent streams are seen to have cut back until they have captured part of the drainage of the parallel, consequent streams, leaving wind gaps. When this region is reduced to base level it may again have the appearance shown in Fig. 91. (Modified after Davis.) piracy is to be seen in the Kaaterskill Creek of New York, which, because of its shorter course to the Hudson, has captured the lakes at the headwaters of the Schoharie Creek which flows on a gentle gradient by a circuitous route to the Mohawk River. Many of the " wind gaps " (passes without streams flowing through them) of the Blue Ridge in Virginia were eroded by streams flowing to the sea, whose headwaters were captured by other streams which, although following longer courses, were able, because of their greater volume of water, to deepen their valleys more rapidly than those flowing through these gaps. The Cumberland Gap, through which passed many thousands of the early immigrants to Kentucky, has such a history. Conditions favorable for river capture occur in regions of tilted beds (Figs. 91-93) in which there is a marked difference in the strength of the rocks. The larger branches follow the outcrops of the weaker beds, THE WORK OF STREAMS 109 and their tributaries join them at right angles, because all except the master streams are subsequent rivers (p. 102). In such regions, the larger streams cut rapidly in the weaker rocks and often behead the streams that flow across the hard beds. After a stream has been captured its new grade will be steeper than before, and it is likely to cut a trench in its old valley, leaving the remnants of the latter as terraces. In regions of horizontal rocks stream capture is also common. If one of two streams heading toward the same point has a straighter and steeper course, or a greater volume of water, or a load of sediment sufficient for rapid cutting but not so great as to cause deposition, it may cut back more rapidly than the other and in time capture the headwaters of the latter. THE EROSION CYCLE The terms youth, maturity, and old age are used to express the characteristics of valleys, and are helpful since they are as descrip- tive of them as the same terms applied to human beings. " They have reference not so much to the length of their history in years as to the amount of work which streams have accomplished in compari- son with what they have before them." Youth. Young valleys are V-shaped, with steep sides, and are occupied by rapid streams unless the land is low. Since they have had but a short life, rapids and waterfalls are often numerous ; the divides are wide and ill-drained, as the frequent occurrence of marshes and lakes usu- ally indicates. The Grand Canyon of the ^, . , FIG. 94. Block diagram showing a region in the V^olorado, the steep youthful stage of its erosion cycle. Sea level is repre- gorge of the Niagara sented by the bottom of the diagram. River, and all narrow, steep-sided, or V-shaped valleys are in youth. A region is said to be youthful (Fig. 94) when sufficient time has not elapsed for streams thoroughly to dissect and drain it ; in other words, the streams have the larger part of their task before them. The Red River valley of North Dakota and Minnesota is such a region, since it has not long been no PHYSICAL GEOLOGY subjected to stream erosion. It was formerly the site of a lake (p. 656) whose bed was covered evenly with sediment. After the lake was drained the bed was exposed to erosion, and a drain- age system was de- veloped whose stream courses were deter- mined by the in- equalities of the FIG. 95. Block diagram showing a region in the bottom. Later in its mature stage of its erosion cycle. The bottom of the history new tribu- block is sea level. . ... tanes will erode side valleys, the main valleys will be widened, and a mature topography will result. Maturity. A mature valley is deep, but has flaring sides and gently rounded upper slopes. A region in full maturity (Figs. 95, 96) is in decided contrast to a youthful region. Instead of few tribu- taries and consequently wide divides, the land is thoroughly "dissected by valleys, the divides are narrow, the valley sides are less steep than in youth, and the streams are accomplishing their greatest work both in erosion and transportation. In such a region the rainfall runs al- most immediately into the streams; lakes have practi- cally disappeared, having been drained by the cutting down of their outlets or filled by stream sediment and organic matter. In this stage the relief is greatest, and arable land is at a minimum; roads are difficult and must follow either the valleys or FlQ< 96> _ Ma p showing the stream courses the narrow divides, and the in a mature region. THE WORK OF STREAMS III inhabitants are isolated. There are many such regions in the United States; for example, large portions of West Virginia, southeastern Ohio, eastern Kentucky, and Tennessee are in maturity. As a rule a master stream reaches maturity earlier than its tributaries, and in its lower course earlier than in its upper course. A region in maturity may be traversed by a stream which flows through a broad, old val- ley, and a youthful region may be traversed by a mature stream. Old Age. Continued erosion will gradually cut down the valley sides (Fig. 97) to gentle slopes, lower the divides, and thus tend to reduce the surface to an undulat- ing plain. The sluggish streams will meander (p. 121) in wide valleys. The region is then in old age (Figs. 98, 91). An abso- lute plain may, perhaps, never be reached, since elevations will FIG. 97. Diagram showing the profile of young, mature, and old valleys. be left here and there because of some favoring condition, such as (i) hardness of rock or (2) a favoring position with reference to the drainage of the plain. Such hills or mountains rising above the general level of the surface are called monadnocks, from a mountain of that type in New Hampshire. Portions of Kansas have passed through youth and maturity and are now in the stage of old age. The time required for the production of a base-leveled condi- tion or for " pene- planation " is called the cycle of erosion. It will take, perhaps, one hundred thousand times as long to pass from maturity to old age as from youth to maturity. It will be seen from the above that the age of a region is not recorded in years but in the work accomplished or to be accomplished. Effect of Elevation and Depression on Streams. If a region is elevated after it has been reduced to base level (peneplain), the streams will be quickened and will again be enabled to deepen their valleys. If the streams meandered (p. 121) on the peneplains, they may intrench themselves in their old courses until they flow through CLELAND GEOL. 8 FIG. 98. Block diagram showing a region in old age. Sea level is represented by the bottom of the block. 112 PHYSICAL GEOLOGY deep, meandering rock gorges. When a stream has thus intrenched its meanders, the evidence is strong that it has been rejuvenated. Many examples of intrenched meanders are to be seen in Europe and in the United States. In the latter, Pennsylvania, Kentucky, and Utah furnish excellent and striking examples. The great natural bridges of Utah, one of which has a height of 305 feet and a span of 1 FIG. 99. Map showing the course of the Ardeche River, France. The origin of the natural bridge by the perforation of the neck of the meander is evident. 273 feet, were formed by the perforation of the necks of intrenched meanders, as was also that of the Ardeche River, France (Fig. 99). If a region is uplifted before the erosion cycle is completed, the rivers will deepen their courses, leaving their former broad flood plains (p. 128) standing as terraces or " benches." A section of such a valley will show a valley within a valley. If a region is more ele- vated near the ocean than further inland, the upper courses of the streams will be " ponded," unless they are able to deepen their valleys as rapidly as the land is elevated. This differential movement of the earth's surface is called warping. Streams which hold their THE WORK OF STREAMS courses in spite of differential elevations, as has been seen (p. 102), are called antecedent streams. If a region underlain by tilted rocks which vary in composition, some resisting erosion more than others, is reduced to base level and then raised, the sub- sequent erosion is such as to give cer- tain proof of its earlier history. An interesting example, in which, however, the river encountered granite (Fig. 100 A and B) rather than tilted rock, is found in the history of the Gunnison River in Colorado. When the Rocky Mountains were being uplifted to their present posi- tion, the streams which now drain them began to cut their valleys. Among them the Gunnison River followed along the depression of the plateau and began to deepen its bed. Its course happened to lie over a great mass of granite, buried be- neath softer strata. The river, having a steep gradient, rapidly cut its way through encountered the granite FIG. 100. Two block diagrams showing the effect of erosion upon resistant and weak rocks. The streams in A have approximately the same slope and are deepening their valleys in strata of the same kind. B shows that the stream on the right encountered resistant granite which was both eroded and weathered more slowly than the weaker rock. As a consequence, the stream on the right has cut a deep and steep-sided gorge, while that on the left has cut a broad valley with gently sloping sides. the soft surface rocks and finally Since its valley was already deep when this occurred, it was unable to turn aside from the hard rock and continued to cut its way through it until the picturesque Black n 4 PHYSICAL GEOLOGY FIG. 101. Drowned river valleys. Chesapeake and Delaware bays and Albe- marle Sound were formed by a lowering of the land which permitted the sea to fill the valleys. Canyon, more than 2000 feet deep, was excavated. The Un- compahgre River, which joins the Gunnison after flowing in approximately the same direc- tion for some distance, was born at the same time. It has flowed, however, over soft material which could be readily eroded, and has been able to excavate a valley several miles in width which in one place is separated from the narrow Black Canyon by a narrow ridge of granite. If a region is depressed, the velocity of the streams will be lessened, and the condition of old age will be hastened. Drowned river valleys (p. 227), such as the Delaware, the St. Lawrence, and Chesapeake Bay (Fig. 101),. are the result of the sinking of the land in the lower courses of the rivers. PENEPLANATION When a base-leveled region (peneplain) has later been up- lifted and dissected by erosion, the evidence of the former base-leveled condition is to be seen in the horizontal sky line presented by the higher hills (Fig. 102). The effect of erosion on an elevated peneplain under different conditions of rock structure is shown by a study of (i) southern New England, (2) the Appalachian region, and (3) eastern Canada. (i) The Peneplain of Southern New England. Southern New England is underlain on each side of the broad Connecticut valley by hard, crystalline rocks ; the valley is in part composed of sand- stone and in part of lava. Before the present elevation took place, erosion had been active for so long that even the lavas, granites, THE WORK OF STREAMS FIG. 102. The peneplain of the Rhine district near St. Goar, in which the Rhine has cut a shallow valley. (Photo. D. W. Johnson.) gneisses, and schists had been cut down to a comparatively level plain, above which stood some hills a few hundred feet high, such as Mt. Monadnock, Mt. Greylock, and Mt. Wachusett. When this peneplain was raised and the streams again began to erode, the weak sandstones were quickly cut away, leaving the trap rocks standing as the Holyoke and other trap ranges of the Connecticut valley, and the old crystalline rocks bounding the " valley " as the high- FIG. 103. Peneplain with several monadnocks in the distance. Camp Douglas, Wisconsin. (Sankowsky.) Near lands. In the highlands, streams have cut deep and usually narrow valleys. The higher hills have about the same altitude, except that the surface of the ancient peneplain rises to the northwest ; on Long Island it is at sea level, but its height increases to an altitude of about 1500 to 2000 feet in Vermont and New Hampshire. Mt. Il6 PHYSICAL GEOLOGY Monadnock, Mt. Greylock, Mt. Wachusett, and some others, as has been stated, rise as monadnocks (Fig. 103) several hundred feet above the ancient plain. (2) The Appalachian Peneplain. The Appalachian Mountain region, from the Hudson River to Alabama, is underlain by rocks differing in their resistance to erosion, which have been bent into broad folds and, in places, broken by faults (p. 25) (Fig. 104). In ancient times (Cretaceous, p. 516) the folds were planed off by ero- sion, leaving the outcropping strata in long, more or less parallel lines, resistant beds alternating with weaker ones. The surface of this (Cretaceous) peneplain is now seen in the approximately level crests of the ridges, showing that the base-leveling of the region had been almost completed. Upon this plain the rivers took their courses to the sea : the Delaware, Susquehanna, and Potomac flowing to the Atlantic across the strata without regard to their structure; the New River of Virginia and the French Broad of North Carolina FIG. 104. A generalized section across the southern Appalachian Mountains. Peneplains are shown by the dotted lines. flowing to the west; while the southern part of the region was drained to the south by the large Appalachian River. An up- warping along a north-south axis occurred which diminished the ve- locity of some streams and increased that of others, thus favoring stream capture (p. 108). The Potomac, Susquehanna, and Dela- ware rivers, continuing to flow in approximately their old channels, cut the deep gorges or water gaps at Harpers Ferry, the Delaware Water Gap, and near Harrisburg. The tributaries of these rivers, such as the Shenandoah and Lehigh, cutting more rapidly in the weaker limestone beds, have excavated broad, subsequent valleys, more or less at right angles to their mains, leaving the resistant strata standing up as mountains (Fig. 105). The gradient of the south- westward, as well as that of the eastward-flowing streams was in- creased and resulted in the headward cutting of one of these until it captured the headwaters of the southward-flowing Appalachian River. A later warping in northern Alabama and Mississippi along an east-west line caused a tributary of the Ohio to cut headward and capture the stream which had formerly robbed the Appalachian THE WORK OF STREAMS 117 River. In this way the Tennessee originated, made up of parts of three rivers which formerly had different courses. After the first elevation, the region remained at approximately the same level for a long time, as is shown by the accordant altitudes of FIG. 105. The Water Gap near Harrisburg, Pennsylvania. The horizontal sky line shows the surface of the ancient peneplain. (Maryland Geol. Surv.) the plainlike valleys between the mountains, but long before the ridges could be reduced, another uplift (Fig. 104) occurred which caused the streams to deepen their beds to the present level. The last uplift must have been relatively recent, since the new valleys are as yet comparatively narrow. (3) The Laurentian Peneplain. The great hunting and fishing region of North America is that vast area almost surrounding Hud- son Bay, which stretches from Lake Superior and the St. Lawrence River on the south to the Arctic Ocean on the north, and from the shores of Labrador on the east to Lake Win- nipeg on the west (Fig. 106). This is known as the " Lau- rentian shield " (p. 389). When one stands on almost any eminence in this region, he finds that he is on a great FIG. 106. The horizontal sky line of the Laurentian peneplain with an incised valley. (Photo. T. C. Brown.) n8 PHYSICAL GEOLOGY plain dotted with lakes, in which, especially along the margins, the streams flow through deep valleys over falls and rapids. The ruggedness of the southeastern margin of the peneplain, where it borders the St. Lawrence, is due to the many valleys which have been cut in it and which have given rise to the rough region known as the Laurentian Mountains. The complicated and dis- torted rocks of the region vary greatly in composition, but have, nevertheless, been reduced to a common level, with the exception of the residual domes and ridges (monadnocks) of the peneplain. The interior peninsula of Labrador is so level that in an area of 200,000 square miles there is not a difference of general level of more than 300 or 400 feet. " The Canadian shield can be described as an ancient peneplain which has undergone differential elevation; has been denuded, and subsequently slightly incised around the uplifted margin." (Wilson.) The Adirondacks were in part reduced to base level at the same time, but in the eastern portion either the surface was not base- leveled, or subsequent movements have raised it and given it vary- ing altitudes. Rate of the Denudation of Continents. The land surface of the United States is being lowered at an average rate of about one inch in 760 years, or of one foot in a little more than 9000 years. The total amount carried to the sea each year from the United States is approximately 270,000,000 tons of dissolved matter and 513,000,000 tons of suspended matter. " If this erosive action had been con- centrated upon the Isthmus of Panama at the time of the American occupation, it would have excavated the prism for an eighty-five foot level canal in about seventy-three days." l How the Load of Streams is Measured. Estimates such as the above are obtained by measuring the amount of water discharged by rivers, together with the minerals in solution and the insoluble silt, and pebbles which are either carried in suspension or rolled along the bottom. The volume of water discharged by a river is found by multiplying the number of square feet in its cross section by the ve- locity a second to obtain the discharge a second. The quantity of silt is found by filtering samples of water from various portions of the section at different times of the year, and the quantity of soluble material is determined by evaporating samples of the water after filtering. The difficulty in obtaining accurate results is due to the 1 U. S. Water-Supply Paper No. 234, p. 83. THE WORK OF STREAMS 119 fact that the velocity and volume of rivers fluctuate often from day to day, and the quantity of silt varies with the velocity. Moreover, the material in solution in a cubic foot is greater at low than at high water, since the proportion of spring water is then greater. It is also difficult to measure the quantity of material roiled along the bottom. It is believed, however, that notwithstanding these diffi- culties, the estimate of the rate of denudation of the United States of one foot in about 900x5 years is accurate within 20 per cent. DEPOSITION Causes of Deposition. Streams bearing a full load will deposit their sediment when their velocities are diminished, (i) A stream flowing from a steep to a gentle gradient will deposit its coarser sedi- ment. (2) When a stream emerging from a straight, narrow channel flows into a wide, winding one, its current is diminished by friction with its bottom and sides, and deposition may take place. (3) When tributary streams with steep gradients flow into slow-moving main streams, they may deposit a part of their load. (4) Since the velocity of a stream increases with its volume, it is evident that, if the volume is diminished in any way, as by seepage or evaporation, its ability to carry sediment will be correspondingly decreased. Consequently, rivers in arid regions are often depositing streams (p. 81), even when they have steep gradients. (5) When slow-moving streams, carrying much fine sediment, meet any obstruction, such as a stranded log or a tree which has fallen from the bank, the slight check to the current produced in this way may cause the formation of a sand bar or island. Many of the islands of the lower Mississippi River began as " snags." (6) When a stream reaches a body of still water, either a large lake or the ocean, all of the sediment soon finds a resting place.' The goal of all sediment is the sea, but in its journey ocean- ward it makes many halts, forming the alluvium of the river valley. Flood Plains. Flood plains are formed by graded streams, as a result of both lateral erosion and of deposition during overflow. A broad, flat valley may be formed in this way. Since rivers nor- mally first reach base level where they enter the sea, their flood plains are usually widest there. The lower Mississippi flood plain (which is, perhaps, more correctly described as a delta) is five to eight miles wide and is bounded on the east by clay bluffs 100 to 300 feet high, and on the west side, as far as the Red River, by less prominent 120 PHYSICAL GEOtOGY banks. The downstream slope of a flood plain varies with the vol- ume of the water in the stream and its load. The slope of the flood plain of the lower Mississippi, for ex- ample, is only two to three inches a mile, while streams carry- ing coarse material may build up flood plains with slopes of 50 to 75 feet to the mile. Flood plains are highest near the river and slope gradually away from it (Fig. 107). This is due to the fact that, at flood, the coarser and more abundant material is deposited where the silt-laden main current is checked by contact with the slow-moving waters of the sides. FIG. 107. The flood plain of a river. The natural levees on each side are shown, as is also the structure of flood-plain deposits. FIG. 108. Meanders, Owens valley, California. (U. S. Geol. Surv.) THE WORK OF STREAMS 121 The fertility of the Nile valley in Egypt is due to the thin layer of silt which is spread over the flood plain each year. If the sedi- ment deposited on a flood plain is coarse, the plain will be infertile. Meanders. After a river has become more sluggish and is con- sequently unable to cut downward, it may undercut its banks on the outside of its curves and thus widen its valley floor. As the outside of a curve is cut away, the inside is filled with sediment to flood level, and a strip of land is thus formed. In this way, as well as by deposi- tion during overflow, a broad, flat valley is developed which, as has been said, is called a flood plain because covered by water during floods. On such a flood plain a river will take a still more winding or meandering course (Fig. 108). The origin of these meanders is easily conjectured. Imagine a perfectly straight stream flow- ing through a level alluvial plain. If, under such conditions, a tree is blown over into the stream, a rock falls from the bank, or a tributary stream forces the current against the opposite bank, or brings in gravel and builds a natural jetty, the current will be deflected a little, the bank will be undercut, and the chan- nel changed at this point (Fig. 109). The stream will then strike the opposite bank obliquely a little further down its course, wearing it away at this point, and thus, one after another, new meanders will be formed. A single obstruction may, therefore, affect the oscillations of the current for an indefinite distance down its course. The length of the Mississippi River (Fig. no), from the mouth of the Ohio to the Gulf of Mexico, is 1000 miles, so meandering is its course, although the direct dis- tance is only 600 miles. One of the plans for improving the Missis- sippi is to straighten the channel by cutting off the curves. Oxbow Lakes. Once initiated (Fig. 109), meanders tend to be- come more pronounced in form, changing from an open loop to one which is horseshoe-shaped. The neck of land separating one bend from the next may become more and more narrow (Fig. in) until, 1- FIG. 109. Diagram showing the initiation of meanders. (Modified after Salisbury.) 122 PHYSICAL GEOLOGY in time of flood, the river may straighten its course by cutting a channel across this narrow strip, leaving horseshoe or oxbow lakes, or bayous, which are soon separated from the new and shorter channel by deposits of silt. Brooks tend to develop a greater number of bends than larger rivers, since they are easily deflected by accidental disturbances, such as a fallen tree or a landslide, while a larger river tends to obliterate its smaller irregularities and to develop the larger ones. As a result, we find many close-set meanders in small brooks, while in large riverjs there are a small number of well-spaced meanders which grow to large size before FIG. i io. Meanders of the Mississippi they are cut off. Many ex- River. The successive positions of the amples might be cited of cities river in 1883, 1895, and later are shown. d m j d h b fc Ihe movement of the meanders down- & . stream and their tendency to increase are of meandering rivers, which have shown. (After Salisbury.) been left far inland by the cut- ting off of the meanders on which they were situated. Because of their changing channels rivers make very poor political boundaries. In the course of time the " oxbow " lakes formed by the " cut- offs " are destroyed, as they are apt to be filled with sedi- ment when the stream is at flood, and at other times sand is blown in by the wind, and vege- tation takes root FlG> Iir> _ An oxbow lake formed by the cutting there. through of the neck of a meander. THE WORK OF STREAMS 123 Natural Levees. A study of a topographic map of the lower Mississippi River shows that it flows between banks which rise ten or more feet above the surrounding swamps, and occasionally constitute the only dry land for long distances. Such embankments are called natural levees (Fig. 107). They are gradually built up in time of flood when the water is swift and contains much sediment. The current in the channel is sufficient to carry the sediment onward, but its FIG. 112. Map showing the changes in the course of the Hoang Ho on its delta (shaded). The river is useless for navigation because it is so changeable, and its waters are restrained only by an elaborate system of dikes and canals. (Richtofen.) velocity is checked when it comes in contact with the slow-moving flood water on the sides, sediment is deposited, and an embank- ment is thus erected above the swamp. Natural levees are often strengthened and heightened artificially to prevent floods, but it is readily seen that during a flood a river may break through its levees, spread over its swamps, and perhaps change its course. The Mississippi River broke through its levees in 1912, causing great destruction of life and property. The levees of the Hoang Ho in I2 4 PHYSICAL GEOLOGY China have been increased artificially so that, in places, the surface of the river is 30 feet above the surrounding plain, but in spite of man's efforts it has often changed its course and is called " China's Sorrow " because of its great destructiveness (Fig. 112). In 1904 the mouth was 250 miles north of its position 40 years before (p. 133). Natural levees are sometimes high enough to turn the courses of the tributary streams for long distances; thus the Yazoo travels for 200 miles parallel to the Mississippi before entering it, and the FIG. 113. Shallow basins formed as a result of the building of natural levees along the stream are filled at flood time with water from the stream and from the sides of the valley. (Minneapolis topographic sheet, U. S. Geol. Surv.) St. Francis for 100 miles. Lakes are sometimes formed when a stream in winding through a valley builds up its levees and thus incloses basins between them and the banks of the valley (Fig. 113). Alluvial Cones and Fans, (i) In Arid Regions. In desert re- gions streams are fed chiefly by tributaries whose sources are in the mountains where the rainfall is greater than on the arid plains. At rare intervals heavy downpours (cloud-bursts) may occur on the lower courses which, though often of only a few minutes' duration, may fill the valleys, producing torrents of great erosive power. But ordinarily such streams rapidly lose volume as they flow out on the thirsty land, as their lower courses are seldom fed by springs. Dur- ing certain seasons, when the rainfall in the mountains is heavy, some desert rivers are a hundred miles longer than at other times. Streams flowing from high lands into deserts quickly drop their sedi- ment at the mouths of their gorges, both because their gradients are diminished and because their velocity is decreased as water is lost THE WORK OF STREAMS 125 FIG. 114. An alluvial fan near Salt Lake City. (U. S. Geol. Surv.) by evaporation and by absorption into the porous soil. In this way a pile of waste is accumulated, half cone-shaped, with a base varying in diameter from a foot to forty or more miles. Accumula- tions such as this are called alluvial cones when steep, or fans (Fig. 114) when the slope is not great. In general they are composed of coarser materials at the apex and progressively finer ones toward the base, since a stream first drops the larger debris with which it is burdened when its velocity is checked. TV* ci-r^Qmc flrtwino- FlG - H5- The San Joaquin valley and Tulare Lake, Ihe streams Mowing CaUfornia ; The basin of Tulare Lake is due chiefly to Over alluvial cones or tne building up of alluvial fans across the San Joaquin fans seldom have valley by Kings River and Los Gatos Creek. 126 PHYSICAL GEOLOGY single channels throughout their courses, because, as they lose volume, they are unable to carry all of their load and therefore deposit it along the sides of their channels, so narrowing them that the water breaks through the banks and forms other channels. This process may be repeated again and again until at the base of the cone a stream has been divided into a number of distributaries. These distributaries tend to keep the fan or cone symmetrical. The angle of the slope of these accumulations varies (i) with the rapidity with which the velocity of the stream is UNCONSOLIDATEO SEDIMENTS Porous sand Quartzite Limestone ay gravel above ground- and gravel below water table ground-water table Impervious Porous sand and d Crystalline rock FIG. 116. Cross section of a typical valley in an arid region. Beneath the alluvial slope gravel predominates, but towards the central flats it gives way to alternate layers of sand and clay. Water is obtained when wells reach the porous sediments, as at c, d, and e. The dotted line shows the base of the alluvial slope. (U. S. Geol. Surv.) diminished; (2) with the kind and amount of the sediment; and (3) with the size of the stream. The slope of cones and fans of large streams usually is less than that of small torrents, which may be as steep as from 5 to 15 degrees. An alluvial fan sometimes causes the formation of a lake by build- ing a dam across a river. Where Kings River enters the San Joaquin River of California it has deposited a fan which has dammed the San Joaquin, forming the shallow Tulare Lake (Fig. 115). Piedmont or Alluvial Plains are formed by the coalescing of adjoining fans. The slope of such plains may be so uniform that the angle is not easily detected by the naked eye by one traveling THE WORK OF STREAMS 127 over the region (Fig. 116). Almost any topographic map of a desert basin, however, shows that the slope of Piedmont plains is usually considerable. In desert regions oases are often found on alluvial fans, since water can be obtained here from wells or from the mountain streams. The principal settlements of Utah are on the alluvial slopes at the foot of the Wasatch Mountains, and many of the cities of Persia and Turke- stan are situated on alluvial fans. (2) In Humid Regions. Alluvial cones and fans are also de- posited in moist regions where a main stream is unable to remove the rock and silt carried into it by its tributaries. In such cases, the FIG. 117. Lake Brienz and Lake Thun were formerly one lake, but have been separated by an alluvial fan upon which Interlaken is situated. cone or fan forces the main stream over to the opposite side of the valley, compelling it to undercut its bank. This may cause the for- mation of rapids, with a shallow lake above. The Liitschine River, in the Lauterbrunnen valley in Switzerland, has built an alluvial fan which has divided the lake into which the river flows into two parts, Lake Brienz and Lake Thun (Fig. 117). Fans in humid regions may be of considerable extent, and are well-developed in portions of the Rhone valley in Switzerland and in the larger valleys of the French Alps. Since they are well-drained and usually fertile, they are often the sites of villages. Alluvial Terraces. Terraces are not uncommon in river valleys and are composed either of rock, when they are called rock terraces (p. 128), or of stratified clay, sands, and gravels, when they are known CLELAND GEOL. 9 128 PHYSICAL GEOLOGY as alluvial terraces. The latter are fragments of sediments which once filled the valleys to their level, and may be accounted for by meandering and swinging streams, slowly degrading valleys which had previously been aggraded ; in other words, by streams slowly eroding their flood plains. Such a change from deposition to erosion may be the result of one or more of several causes, (i) If a region is elevated so as to increase the velocity of the streams, deposition is succeeded by erosion. (2) This is also true if the volume of water in a stream increases without a corresponding increase of sediment. Such a condition may result when a moist climate follows a dry one, or when a stream captures the headwaters of another stream. Alluvial terraces in many dry regions appear to indicate oscillations between dry conditions, when soil and rock waste were washed down from the mountain sides into the valleys, and moist conditions, when the deposits formed in the valley bottoms were dissected because the load of the streams had been diminished. This resulted from the fact that during wet years the soil was held in place by the flourish- ing vegetation ; while during the dry years, although the rainfall was less, the amount of waste removed was great because of the disap- pearance of the vegetation which formerly bound the weathered rock. (3) If the quantity of sediment is decreased, as occurs when a stream ceases to erode at its head, deposition gives place to erosion. (4) As a valley lengthens, so much of its load may be dropped in the upper and newer portions of its flood plain that it is enabled to degrade its older flood plain. FIG. 1 18. -Rock terraces due to uplift. <5) When a region suffers successive uplifts, so that a stream is unable to cut away its former flood plain before its grade is again increased, terraces will be formed which correspond on the two sides of the valley (Fig. 118). (6) If a degrading stream decreases in volume, it will not be able to occupy the full width of its valley and will cut a narrower valley in the older one. The last- mentioned cause, although perhaps the one which first suggests itself, appears to have been rarely effective in the formation of terraces. The close of glacial times (p. 663) seems to have been especially favorable for valley filling, because of the overloading of the streams THE WORK OF STREAMS I2 9 which derived their material more or less directly from the glaciers as well as from the rapid erosion of new gorges. The depression of the land in many places, as in the Connecticut valley, reduced the velocity of the streams, and occasionally ice jams of long duration also caused deposition. The deposits thus built in valleys have since been partly removed, thus causing the formation of terraces. Discontinuity of Terraces. The terraces on the two sides of a valley do not necessarily agree in height. This is due to the fact that, in swinging to and fro across its valley, a stream not only cuts laterally but also at the same time degrades its bed (Fig. 119 A, ), the flood plain often being higher on one side than on the other. In the Brat- tleboro, Vermont, region, for example, the stream appears to deepen its valley about 12 feet in each swing. (E. H. Fisher.) If a stream meanders entirely across its valley, it will destroy its flood plain, but if it fails to make a complete swing, a fragment will remain as a terrace. When in its meanderings a stream encounters a rock ledge (Fig. 119) in its valley floor, the lateral cutting may be retarded to such a degree that it will begin to swing to the opposite side of its valley before completing its usual lateral movement. In this way a portion of the flood plain will be preserved as a terrace. When other rock ledges are encountered in its further swings across the valley more terraces will be left, and the " meander belt " will be narrowed. The theory of defending rock ledges affords a better explanation than any other for many of the terraces of the New England valleys. Fig. 119. Block diagrams showing the origin of stream terraces defended by rock ledges. The terraces H, K, A, B, C, D, E, F, G, owe their preservation to the presence of rock ledges which prevented the stream from cutting them away as the valley was deepened. The relation of rock to alluvium on the right of the block diagram is also shown in figure B. (Modified after E. H. Fisher.) 130 PHYSICAL GEOLOGY Characteristics of River Deposits. A cross section through a river deposit does not show a homogeneous deposit of stratified sedi- ment, but rather lens-shaped masses of coarse sands and gravels at different levels, buried in stratified sands and clays (Fig. 107). When traced up or down the valley, these deposits are found to lie in long and comparatively narrow belts. They represent the former channels of the aggrading river, where the current was strong enough to remove all but the coarser material of its load. The finer deposits the mud and fine sand were laid down in the more sluggish water on either side of the channel and on the flood plain. Beds of muck, marking the sites of shallow lakes and swamps, are also common. Scale of Miles. ^^ 25 50 75i8r\ DELTAS Deltas are formed where streams enter either lakes or seas. If the body of water into which the river flows is large, all of the sedi- ment carried in by the stream is dropped, and the bottom is gradually built up at the river's mouth. Since sediment settles much more quickly in salt than in fresh water, it is dropped more quickly in the ocean. Because of the low gradient, a river often splits into sev- eral channels (Fig. 120) as it enters its delta, the branches being known as distributaries. The shape of a delta, as the name implies, is usually that of the Greek letter of that name, with one angle of the triangle pointing upstream. Growth of Deltas. The rate of growth of a delta depends upon (i) the amount of sediment carried by the river, (2) the depth of the sea or lake, (3) the strength of the waves or currents, and (4) the stability of the bottom of the sea or lake where the deposition is tak- ing place. Deltas are apt to be largest in seas in which the tide is weak, since under such conditions practically all of the sediment is dropped soon after it reaches still water. When the ocean bottom at the mouths of rivers is subsiding, the upbuilding of the bottom may be insufficient to compensate for the subsidence. The Mississippi FIG. 120. The delta of the Mississippi River. THE WORK OF STREAMS 13 1 delta has been built upward and outward in spite of subsidence; the sinking which produced the Chesapeake and Delaware bays, however, was so rapid that estuaries were formed (p. 114). The Mis- sissippi River is extending its delta at the remarkable rate of one mile in sixteen years; the Rhone has added a mile to its delta in Lake Geneva since Roman times. It leaves the lake as a clear stream, but gathers sediment from its tributaries in France, with which it builds another delta at its mouth at the rate of about a mile a century. In 220 B.C. the town of Pu-tai, China, stood one third of a mile from the sea, but in 1730 it was 47 miles inland, and to-day it is 48 miles from the shore. (King.) Many of the " points " in lakes are deltas which have been built out by streams. Structure of Deltas. A section through a delta shows approxi- mately horizontal beds of fine material at the bottom, which do not differ greatly from other deposits at a similar depth where no delta occurs. These are termed the bottom-set beds. Above these are the ^ IGp I2I- Ideal section of a delta built into , ,. j f quiet waters of constant level. The lower horizontal steeply inclined fore-set beds are caUed bottom . set> the inclined> fore _ set) and beds, composed of coarser the upper, top-set. (After Barrell.) sediments which have been swept outward by the currents and waves and may have a slope approaching the angle of repose (Figs. 121, 122). The top-set beds are nearly horizontal and are laid down upon the fore-set beds. These are usually the last deposits of the river in the upbuilding of the delta. The surface of a delta is comparatively level, but gradually rises upstream. In it large and small lakes may occur; the depressions in which they lie being those portions of the delta which, because of the accidental position of the distributaries, were not filled to the general level with sediment. Their life is necessarily short, since they are gradually being filled by accumulations of silt during floods, and by swamp vegetation. Deltas may be very extensive. That of the Ganges and Brahma- putra has an area of 50,000 to 60,000 square miles, with its head 200 miles from the sea. The length of the Mississippi delta is more than 200 miles, and its area is more than 120,000 square miles. The Orinoco delta has an area larger than that of New Jersey. The 132 PHYSICAL GEOLOGY head of the delta of the Hoang Ho is 350 miles from the coast. The Imperial valley in California is the result of delta building. The Gulf of California formerly extended 150 miles further northwest than now, and across it a delta was built by the Colorado River, so high as to shut off the upper part of the gulf and inclose a lake of salt water. This lake has almost entirely disappeared and its bed FIG. 122. Longitudinal section of a delta, showing the dipping, fore-set beds. (Photo. R. S. Tarr.) has become the Salton sink. Thanks to irrigation this basin is ex- tremely fertile. The depth of delta deposits is often great. A boring at New Or- leans encountered driftwood at 1042 feet, and depths of 500 feet are not uncommon in other deltas. It has been shown that in many cases the subsiding of deltas progresses at a pace about equal to the deposition. Deltas are usually noted for their fertility. The three most densely populated regions of the world, outside of cities, are the deltas of eastern China, India, and the Po River in Italy. This is true in spite of the fact that, because of their level surfaces, deltas are especially subject to floods. The great flood in the Mississippi River delta in 1912 destroyed many lives and millions of dollars' worth of property, and this was also the case with earlier floods. One of the notable examples of such easily flooded districts is the delta of THE WORK OF STREAMS 133 the Hoang Ho in China. This river is restrained by great dikes (p. 124), some of which are 30 feet above the level of the region; but notwithstanding these precautions many disastrous floods have occurred. For several hundreds of years previous to 1852 this river emptied into the Yellow Sea. In that year, when in unusual flood, it broke through its north levees and emptied into the Gulf of Chihli, some 300 miles farther north. This is only one of the many shiftings which this river has made during its history (Fig. 112). During a flood in 1887 many villages were destroyed, and the loss of life through drowning and famine exceeded 1,200,000 people, more than the entire population of Nebraska. DEPOSITION IN LAKES BY STREAMS AND BY OTHER AGENTS Mechanical Deposits. Streams deposit their loads when they flow into lakes, forming deltas (p. 130) at their mouths and covering the bottom of the lake with the finer silt, which is carried farther out since it remains in suspension longer. Lakes may in time be entirely filled by the growth of their deltas, first becoming swamps and then level meadows through which the streams may flow in meandering courses (Fig. 123 A, B). Meadows of this history are abundant in regions which have been glaciated, such as Michigan, New York, and Minnesota. Lakes are shallowed by the waves cutting back the cliffs along their shores and carry- ing out into them the material thus derived. It is thus seen that as soon as a lake comes into existence, agencies arise which tend to obliterate it; sediment begins to fill it, and the outgoing stream commences to deepen the outlet and thus in time to drain it. FIG. 123. Map A shows a lake being filled in with sediment carried by streams. Map B shows the same lake converted into a marsh, with the streams flowing in meandering courses. 134 PHYSICAL GEOLOGY Lakes equalize the flow of streams, preventing floods, and also act as filters. Chemical Deposits. In addition to such mechanical deposits as those described, chemical deposits are also found in lakes. Lime is sometimes deposited, and iron in the form of limonite (p. 686) is precipitated. In some of the lakes of Sweden and Canada iron of this origin is so abundant as to be of economic importance. Organic Deposits, (a) Diatoms. Dredgings in lakes show that the bottoms are sometimes covered with thick deposits of diatoms (microscopic plants which secrete siliceous tests, p. 581). Since these organisms multiply with great rapidity, they may form extensive deposits, called diatomaceous earth. (b) Marl. Calcareous deposits in the form of marl may accu- mulate to great depths in lakes. This is a white, or gray, clay-like deposit which is composed largely of calcium carbonate. It is formed either by the accumulation of shells, or through the agency of certain plants (algae) which extract carbon dioxide from the water and thus cause the deposition of the lime dissolved in the water. Marl is formed only where small quantities of clay are washed into the lake, since, if large quantities are carried in, the deposit would be termed mud. Deposits of marl may be a score or more feet in depth and are often overlain by peat. In regions where limestone is not accessible, marl is sometimes used in the manufacture of Portland cement. (c) Peat. A brown deposit, called peat, composed of the partially decayed remains of plants, sometimes accumulates in swamps, marshes, and shallow lakes. Peat forms most rapidly in cool, moist climates where, although the vegetation may not grow rapidly, the low temperature retards decay. Under favorable conditions it also accumulates in warm countries. In Florida, for example, there are considerable areas of peat. Extensive areas of peat occur in the United States, such as that of the Dismal Swamp of Virginia and North Carolina. In Massachusetts, it is estimated that there are 15,000,000 cubic feet of peat. One tenth of the surface of Ireland is underlain by peat, and large areas in Europe and elsewhere are pro- vided with it. Peat is dried and used for fuel in some regions where it occurs in great abundance, and where its extraction is easy. Playas. In desert regions, where no permanent lakes occur, streams sometimes reach depressions when their volumes are increased during the wet season or by cloud-bursts, and form temporary, shallow lakes which may cover large areas. The largest in Nevada THE WORK OF STREAMS 135 is in the Black Rock desert and is 450 to 500 square miles in area, although seldom more than a few inches deep. Such temporary desert lakes are called playas. Their beds, when dry, are covered with fine clay and sand, and sometimes with gypsum and salt. On the mud of ancient playa beds the footprints of extinct animals have been preserved (p. 379). Salt Lakes. A salt lake may be formed (i) by the cutting off of an arm of the sea by a delta, as in the case of the Salton Sea, California (p. 132), or by an elevation of the sea bottom, which isolates a body of water. Under such conditions, the water will, at first, have the same composition as sea water. If, however, the water flowing into the lake exceeds the evaporation of its surface, it will gradually be freshened. Such was the history of Lake Cham- plain. If, on the other hand, such a lake has been formed in a desert region where evaporation is excessive, the water will become more salty as time goes on. The Caspian Sea was formerly con- nected with the Black Sea, but is now isolated and is growing more salty. (2) Salt lakes are also formed by the concentration of fresh water. Basins in arid regions which do not receive enough water to cause them to overflow may, in time, become saturated with salts of various kinds. The streams bring in common salt (NaCl), gypsum (CaSO 4 -2fI 2 O), Epsom salt (MgSO 4 -7H 2 O), and calcium carbonate (CaCO 3 ), which they obtain from the rocks over which they flow. These salts may accumulate in the lake as evaporation proceeds, until the water becomes so concentrated that they are precipitated. Iron oxide and calcium carbonate will be deposited first; upon further concentration, gypsum, which is insoluble in strong brine, will be precipitated ; then common salt and Glauber salts (Na 2 SO 4 ), in the order of their solubility. This order is often interfered with under certain conditions. Cold weather, for example, will cause the precipitation of Glauber salts (Na 2 SO 4 ) before the common salt has all been precipitated. If the evaporation of the surface of the salt lake does not equal the amount of water received during a wet season, the deposition of gypsum and salt will cease, and the beds of salt may be covered by the sediment brought in by the streams. With the recurrence of the dry season the deposit of gypsum and salt will commence again. Many alternations of mud and salt are encoun- tered in wells sunk on the margins of salt lakes. In some salt lakes most of the salt has been deposited, and the liquid remaining, called 136 PHYSICAL GEOLOGY " bittern," contains chiefly Epsom and Glauber salts. The Dead Sea is such a lake. (3) The salt of some salt lakes has been attributed to an accumu- lation of wind-blown salt. Perhaps the best example of a salt lake in which this origin is evident is furnished by a lake in northern India (Sambhar Lake). This lake is situated in an inclosed basin more than 400 miles inland and appears to receive the greater part, if not all, of its salt from dust-laden winds which sweep over the plains between it and an arm of the sea during the dry months. Analysis of the air during the dry season shows that at least 3000 metric tons of salt are carried over the lake annually, an amount sufficient to account for the accumulations of salt in the lake. Alkaline Lakes. Alkaline and borax lakes differ from salt lakes in that they contain a predominance of sodium carbonate or borax. The source of this carbonate and borax, as in the case of common salt, is the rocks over which the streams which feed such lakes flow. Origin of Rock Salt. Deposits of salt underlie many hundreds of square miles of sedimentary rocks in New York and other states. The thickness of the salt beds varies greatly, the thickest reported in New York consisting of 325 feet of solid salt. The greatest salt deposit known is that at Stassfurt, Germany, which is 4794 feet deep. Since salt and gypsum occur together, it is believed that such deposits have been formed as a result of the evaporation of salt lakes. One objection to this theory is the great thickness of some beds and their purity. In the case of such deposits it is believed that an estuary or lagoon was separated from the sea by a bar over which water was carried during storms or perhaps at high tide. If the region in which this occurred was hot and arid, it is conceivable that salt might be deposited to the depth of the lagoon or estuary. If such a basin should slowly subside, a bed of salt of great thickness could result. Such remarkably thick deposits as those in Louisiana, where the bottom has not been reached at a depth of 2000 feet, requires a still further modification of the theory. Extinct Lakes. Upon their disappearance lakes leave behind them proofs of their former existence. If they were of comparatively short duration, as would be the case if they had been formed by ice jams (p. 186), their former presence might be attested by (i) the deltas deposited by the streams which flowed into them, as well as by (2) the stratified sand and clay which were spread over their beds. When their life was long, (3) wave-cut terraces, (4) sand bars and THE WORK OF STREAMS 137 FIG. 124. Deltas formed in Lake Bonneville by the Logan River, Utah. (U. S. Geol. Surv.) spits, deltas, and other thick deposits are left (Fig. 124). One of the most remarkable lakes of this sort was Lake Bonneville (Fig. 125), of which the Great Salt Lake is a withered remnant. This lake at FIG. 125. Contour map of the shore terraces of Lake Bonneville, Utah. The terraces were cut and built at different lake levels. (U. S. Geol. Surv.) 138 PHYSICAL GEOLOGY its greatest extent covered 19,750 square miles and was 1000 feet deep. At this time it had an outlet to the north which carried the excess waters to the Pacific. During this period, too, great terraces were cut and immense deltas were built. From Salt Lake City one can see these terraces on the lower slopes of the mountains and from them can learn the former levels of the lake. A change in climate finally reduced this extensive lake to the present relatively small Great Salt Lake, which has an area of 2000 square miles and an average depth of 15 feet. Since the water now contains 18 per cent, of salt, it is so dense that the bather is required to exert no effort to keep his head above water, as it is impossible to sink. REFERENCES FOR THE WORK OF STREAMS GENERAL CLELAND, H. F., North American Natural Bridges, with a Discussion of their Origins: Bull. Geol. Soc. America, Vol. 21, 1910, pp. 313-338. DE MARTONNE, E., Geographic Physique, pp. 413-442. GILBERT, G. K., Re-port on the Geology of the Henry Mountains: U. S. Geog. and Geol. Surv. of the Rocky Mountain Region, 1877, pp. 99-150. HAUG, E., Traite de Geologic, pp. 406-436. RUSSELL, I. C., Rivers of North America. SALISBURY, R. D., Physiography (Advanced), pp. 114-203. SALISBURY AND ATWOOD, The Interpretation of Topographic Maps: Professional Paper, U. S. Geol. Surv. No. 60, 1908. SHALER, N. S., Aspects of the Earth, pp. 143-196. TAYLOR, F. B., Niagara Falls Folio : U. S. Geol. Surv. No. 190, 1913. FLOOD PLAINS DAVIS, W. M., The Development of River Meanders : Geol. Mag., Vol. 10, 1903, pp. 145-148. JEFFERSON, M. S. W., Limiting Widths of Meander Belts : Nat. Geog. Mag., Vol. 13, 1902, pp. 373-384- CYCLE OF EROSION DAVIS, W. M., Geographical Cycle, Geographical Essays, 1909. DAVIS, W. M., Base Level, Grade, and Peneplain : Jour. Geol., Vol. 10, 1902, pp. 77-109. DAVIS, W. M., The Peneplain: Am. Geologist, Vol. 23, 1899, pp. 207-239. STREAM PIRACY BOWMAN, I., A Typical Case of Stream Capture in Michigan : Jour. Geol., Vol. 12, 1904, pp. 326-334. DARTON, N. H., Examples of Stream Robbing in the Catskill Mountains : Bull. Geol. Soc. America, Vol. 7, 1896, pp. 505-507. DAVIS, W. M., Stream Contest along the Blue Ridge: Bull. Geog. Soc. Philadel- phia, Vol. 3, 1905, pp. 213-244. THE WORK OF STREAMS TERRACES 139 DAVIS, W. M., The Terraces of the Westfield River, Massachusetts: Am. Jour. Sci., Vol. 14, 1902, pp. 77-94. DAVIS, W. M., River Terraces in New England: Bull. Harvard Collection, Museum of Comparative Zoology, Vol. 38, 1902, pp. 281-346. DODGE, R. E., The Geographical Development of Alluvial River Terraces : Pro- ceedings Boston Soc. Nat. Hist., Vol. 26, 1894, pp. 257-273. FISHER, E. F., Terraces of the West River, Brattleboro, Vermont: Proceedings Boston Soc. Nat. Hist., Vol. 33, 1906, pp. 9-42. INTRENCHED MEANDERS DAVIS, W. M., The Seine, the Meuse, and the Moselle : Nat. Geog. Mag., Vol. 7, 1896, pp. 189-202; 228-238. DEPOSITION BARRELL, J., The Geological Importance of Continental, Littoral, and Marine Sedi- mentation: Jour. Geol., Vol. 14, 1906, pp. 316-356; 430-457; 524-568. SALT LAK^S HARRIS, G. D., Rock Salt, Its Origin and Importance: Bull. La. Geol. Surv. No. 7, 1907. TOPOGRAPHIC MAP SHEETS, U. S. GEOLOGICAL SURVEY, ILLUSTRATING THE WORK OF RUNNING WATER Regions in Topographic Youth Regions in Topographic Maturity Casselton, North Dakota. Charleston, West Virginia. Fargo, North Dakota. Briceville, Tennessee. Bright Angel, Arizona. Lancaster, Wisconsin-Iowa-Illinois. Kaibab, Arizona. Becket, Massachusetts. Bisuka, Idaho. Arnoldsburg, West Virginia. Milan, Illinois. Monterey, Virginia- West Virginia. Niagara Falls, New York. Great Falls, Montana. Regions in Topographic Old Age Stream Piracy Caldwell, Kansas. Kaaterskill, New York. Butler, Missouri. Lake, Yellowstone National Park, Morrilton, Arkansas. Wyoming. Gloversville, New York. Lykens, Pennsylvania. Rejuvenated Streams and Entrenched Meanders Lockport, Kentucky. Huntingdon, Pennsylvania. Harrisburg, Pennsylvania. Ravenswood, West Virginia-Ohio. 140 PHYSICAL GEOLOGY TOPOGRAPHIC MAP SHEETS, ILLUSTRATING STREAM DEPOSITS Alluvial Fans Braided Streams Cucamonga, California. North Platte, Nebraska. Sierraville, California. Kearney, Nebraska. Desert Well, Arizona. David City, Nebraska. Parker, California-Arizona. Gothenburg, Nebraska. Disaster, Nevada. Natural Levees Donaldsonville, Louisiana. Baton Rouge, Louisiana. Hahnville, Louisiana. Flood Plains and Meanders Terraces St. Louis, Missouri. Cohoes, New York. Butler, Missouri. Lacon, Illinois. Lake Providence, Louisiana. Hartford, Connecticut. Jefferson City, Missouri. Mountain Home, Idaho. CHAPTER V THE WORK OF GLACIERS WHEN viewed from an eminence, a mountain glacier has the appear- ance of a river of ice flowing down a valley to a point where it ends abruptly and a stream emerges from beneath it and courses toward the sea. If the climate is cold, as in Greenland, glaciers may even reach the sea, where their shattered fronts are carried away as icebergs by the ocean currents. GENERAL CONSIDERATIONS Distribution and Size of Glaciers. Glaciers exist on high moun- tains, even in the tropics. In temperate regions they abound on high ranges, especially on those against which moisture-laden winds blow; as, for example, the Sierra Nevada, the Cascade ranges, the Alps, the Caucasus, the Andes, and the Himalayas. Mountain glaciers vary in size from those which barely extend beyond their cirques (hanging, clifF, or corrie glaciers) to the great Seward Glacier of Alaska, more than 50 miles long and 3 miles wide where narrowest. In the Alps there are 2000 glaciers, the largest of which, the Aletsch Glacier (p. 187), is more than 10 miles long, although the majority are less than a mile. These Alpine glaciers vary in width from a few hundred feet to about one mile. u The thickness of ice in the Alpine glaciers must often be as much as 800 to 1 200 feet," the depth usually being least at the lower end. Great glaciers are confined to polar regions (continental glaciers, p. 168) and to high mountains of the temperate zones. Position of the Snow Line. The level on the earth's surface above which some of the snow of one winter lasts through the fol- lowing summer, thus forming areas of " perpetual snow " or snow fields, is called the snow line. Its position depends upon the normal temperature of the region as well as upon other factors. In general, the snow line varies little from the line on which the average tempera- ture is 32 F. Near the equator it is 15,000 to 19,000 feet above the sea, while in polar regions it is almost, or quite, at sea level. In 141 142 PHYSICAL GEOLOGY intermediate regions the height increases toward the equator. In the Alps the snow line is 8500 feet above the sea ; in the western United States and in British Columbia the higher mountains are covered with perpetual snow; in Massachusetts it has been shown by kites that glaciers would exist at an altitude of 11,470 feet; and it is estimated that glaciers would develop in the Scottish Highlands if the average temperature were lowered three degrees. The position of the sun with reference to a mountain range in- fluences the height of the snow line. In the northern hemisphere, for example, other things being equal, the snow line will be lower on the north side of a mountain than on the south side, since the former receives heat from the sun fewer hours each day. Certain forms of topography also favor the retention of snow. For instance, snow gathers to greater depths in deep ravines than on a level surface, as it is blown in by the wind and protected from the sun's heat so that it may remain from one winter to the next. A moist climate also favors a low snow line on account of the greater snowfall, since more time is required to melt, or evaporate, a thick layer of snow than a thin one. On the Himalayas the snow line is 3000 to 40x30 feet lower on the south than on the shaded north side, because of the greater amount of snow precipitated there by the moist, south winds from the Indian Ocean. The few inches of snow which fall on the north slope may be melted in a few warm summer days, while the several feet of snow on the south side may not disappear, even when subjected to a longer period of warmth. In dry climates the snow may disappear entirely by direct evaporation. As far as temperature is concerned portions of Siberia are under glacial conditions, but the climate is so arid and the snowfall so scanty that the snow which falls is soon evaporated. Formation of Ice in Snow Fields. Snow differs from ice in being composed of fine crystals, loosely consolidated and separated from one another by air, whereas ice consists of crystals in contact. In a snow field there is every gradation from fluffy snow to granular snow or neve and finally to solid ice. The change from one state to the other is well shown in snowdrifts of the temperate zone, which become granular if they exist for a few months, the granules being about the size of small hailstones. If they exist still longer, the drifts are rep- resented by small mounds or ridges of solid ice. The transforma- tion from snow to neve and then to ice is very gradual and is accom- plished (i) by the pressure of the overlying snow which forces the air from between the snow crystals and thus tends to compact them; THE WORK OF GLACIERS 143 (2) by rain and the water from the upper layers of the melting snow, which soaks down into the snow, freezes, and expels the air; and (3) by the growth of the snow crystals. It is in this way that the coarsely granular snow seen in drifts in the early spring and in the neve of snow fields is produced. The growth of the crystals is accomplished partly at the expense of the smaller crystals which lose bulk by evaporation, while their larger neighbors take the mois- ture given off to increase their own size, and partly from the thaw water which bathes them. Neve passes insensibly into snow, on the one hand, and into ice on the other. A crystallographic study shows that ice is made up of crystals, the external form of which has been obliterated by pressure and as a result of their growth. Ice is, there- fore, a crystalline rock, like marble, and is classed as a rock. Snow does not accumulate indefinitely above the snow line; a part melts and runs off, a part is evaporated, and a part is carried away by glaciers. It has been estimated that if glaciers had ceased to drain the snow fields at the beginning of the Christian era, the Alps would now be buried under a mantle of snow about 5000 feet thick. MOUNTAIN GLACIERS Formation. When ice has accumulated to a considerable depth it tends to spread, much (so far as external appearance is concerned) as does a mass of stiff molasses candy ; and if it rests on an inclined surface, it tends to move down the slope. When the ice in an ice field begins to move it is called a glacier. If we study typical glaciers, such as those in the Alps, in Glacier National Park, in British Columbia, or in Alaska, we find that in general they are similar but show individual differences. We find, upon following a glacier to its head, that it begins in a broad amphi- theater (Fig. 126), called a cirque (French for amphitheater), above the snow-covered floor of which rocky walls rise precipitously, often to a height of several hundred feet. In this amphitheater snow gathers to great depths, often to hundreds of feet. The snow comes from the frequent storms which rage there and from the accumula- tions on the walls of the cirque, from which it is swept in by winds or carried by avalanches. Cirques are therefore the feeding grounds of mountain glaciers. In them one finds every gradation, from snow which is freshly fallen, through granular neve or half-formed ice, to compact ice. From the cirque the solid ice of the glacier moves CLELAND GEOL. IO 144 PHYSICAL GEOLOGY slowly down the mountain valley until it reaches a point where the melting equals the forward movement (p. 159), the size of the glacier depending (i) upon the area of the neve field drained by it, (2) upon the amount of the precipitation, and (3) upon the rate of melting. Some- times glaciers flow between forests and even cultivated fields, as, for example, in the valleys of Grindel- wald and Chamonix, f where glaciers lie FIG. 126. Cirques or reeding ground, and medial moraine ... of the Breithorn Glacier. (Photo. L. E. Westgate.) within a few hundred feet of the homes of the inhabitants. In New Zealand a glacier from the Mt. Cook range discharges its debris in the midst of subtropical vegetation. Cirques. One of the most striking and beautiful features of the Alps in Switzerland, of the Selkirks in Canada, of the Rocky Mountains of the United States, and of other high mountains of the temperate re- gions are the ragged crests (Fig. 127) sepa- rating the gigantic semicircular cirques, which hang high up on the mountain sides. These cirques domi- nate the high moun- FIG. 127. Cirques and small glacier, bt. Christophe, France. tains and correspond to the limit of per- petual snow of the Glacial Period (p. 141). Their walls are rough and precipitous, while their floors are comparatively smooth and THE WORK OF GLACIERS level, the former having been roughened by the attacks of the frost and other weathering agents and the latter having been scoured by glaciers into rounded surfaces. Most of the lakes of the high mountains, which give such scenery much of its charm, rest in cirques. They lie in basins, formed either by dams of glacial debris (moraines) left by glaciers, or in depressions cut into the solid rock of the cirque by the ice (rock basins). An understanding of the origin of cirques is, therefore, necessary for an appreciation of the scenery of high mountains. Origin of Cirques. If the average temperature of a mountain region is being lowered as a result of a change in climate, the drifts of snow which accumulate in ra- vines and spots sheltered from the full heat of the sun may last from one season to the next. On account of the weight of the snow and for other causes (p. 142), the lower layer will be compressed into ice and will slowly move down the slope. This movement will separate the moving mass of snow and ice from the snow which rests upon the upper slope, near the valley wall. The rock wall will thus be partially exposed, and a crevasse, called the Bergschrund (German for mountain gap or fissure), will be formed (Fig. 128). The Bergschrund, in fact, marks the line where the real downward motion of the ne"ve begins. Crevasses of this sort vary in width from two or three feet to more than 80 feet, and play an important part in the formation and enlargement of cirques. One such Bergschrund, 150 feet deep, which extended down to the rock bottom of the cirque, was explored, and its floor was found to be composed of rock masses, partly or completely dislodged from the wall of the cirque. During the days of summer the water from the melting snow drips down into the crevasse, wetting the rocks and filling the cracks. As soon as the sun sets the temperature of such regions is rapidly lowered and the water filling the cracks and joints freezes, forcing the blocks from the sides. Since the cracks at the base of the rock wall are more com- pletely filled with water than are those in the upper portion, the greatest disruptive effect is at the bottom of the crevasse, thus tending to produce and maintain vertical walls. As the cirque is enlarged by the wedge work of the ice on the rock in the Bergschrund, the crevasse also moves back. The circular form of the cirque results from the movement of the snow and ice away from the surrounding walls toward the center of the depression. FIG. 128. The Bergschrund of a glacier. Swiss Peak, British Columbia. (Photo. L. E. Westgate.) 146 PHYSICAL GEOLOGY The development of cirques is apparently not necessarily limited to the heads of former stream valleys, although this is generally the case, but they may have their origin in a somewhat different way. If the drifts on a mountain slope last year after year until late in the spring, it will be found that their edges are usually bordered by fine soil which is slowly being removed by water and deposited in deltas at the lower margins of the drifts. This fine soil is the result of the alternate freezings and thawings of the water in the cracks and pores of the rock, which is thus finally broken up into small fragments. In this way, by nivation, a niche, the beginning of a cirque, may be formed on a moun- tain slope. Development of Cirques. A high re- gion which has been partly cut into cirques " resembles nothing so much as a layer of dough from which biscuit have been cut." As the amphi- theaters or cirques on the two sides of a mountain ridge en- large, they finally en- croach on each other, first forming a nar- row, ragged, comb- like ridge (Fig. 129 A, 5), and somewhat later, as the separat- FIG. 129. A shows mountain valleys formed by stream erosion. B shows the same valleys after they have been occupied and strongly eroded by glaciers. The main valley has become U-shaped, and the side valleys have become hanging valleys with strongly developed cirques. An attempt has been made to show the prob- able approximate deepening, in feet, by glacial erosion. ing walls of the cirques are partially quarried away, producing tooth- like peaks. Fate of Cirques. If changes in climate cause a glacier gradually THE WORK OF GLACIERS to wither back into its cirque and finally to disappear, the character- istic features of the abandoned cirque are slowly obliterated ; land- slides and talus descending from the cliffs are heaped upon the bottom, filling the lakes and covering the bottom; the morainic (p. 159) or rock ridge at its entrance is breached by a gorge cut by the out- flowing stream ; side valleys are developed ; and the resulting topog- raphy presents few features to indicate that it was developed from a cirque. Ablation. The surfaces of glaciers are constantly being lowered by direct evaporation and by melting ; those of the Alpine glaciers are lowered from 18 to 25 feet during the summer months, that of the Mer de Glace having been lowered twenty-four and a half feet in 1842. Since the advance of a glacier depends upon the thickness of its mass, it follows that when ablation is excessive the front will retreat. A retreating glacier is, consequently, thinner and, unless its valley walls are vertical, narrower than when it was advancing. SURFACE OF MOUNTAIN GLACIERS The surface of a glacier is usually rough (Fig. 130) as a result of a number of causes. (i) Irregularities Due to Tension. Because of the brittleness of the ice mass, glaciers are broken by cracks called crevasses. Some of these are the result of the more rapid motion of the center than the retarded sides, which produces strains un- der which the ice fractures. The cre- vasses formed in this way are diagonal and extend up the valley (Fig. 136 C, p. 151). When a glacier emerges from a nar- row portion of its valley longitudinal cracks are developed, and the tension on a Curve produces trans- p IG I30 _ Surface of a glacier showing seracs and verse crevasses which crevasses. 148 PHYSICAL GEOLOGY rise obliquely from the bottom, since the latter portion of the ice is retarded by friction with the bed. Crevasses when first formed are usually separated from one another by rela- tively level surfaces, but since their upper portions are soon wid- ened by melting, the intervening ice often becomes blade-like in its sharpness, so that the surface of the gla- cier presents a maze of sharp ridges. Such a ridge of ice is called a serac (Fig. 131). In FIG. 131. The Aletsch Glacier, Switzerland. crossing a glacier, such as the Mer de Glace, these sharp, steep ridges the chief difficulties encountered are of ice. The most conspicuous roughness of a glacier's surface develops where there is a sudden change in the slope of the bed (Figs. 132, 133). In a river this would produce a waterfall, and in a glacier it produces an ice/all. Such icefalls make travel on a glacier extremely difficult and dangerous. The ice passes over the fall slice by slice, the fall (as in a river) remaining station- ary. Below the fall the blocks heal together, but the resulting surface is extremely rough, although it gradually^ becomes FIG. 132. Longitudinal sections of a glacier Smoother. showing icefalls formed where the slope of the bed ki j of a glacier increases suddenly. (After Heim.) is not to be under- stood that a glacier is much fractured in all parts. The absence of cracks on portions of the Aar Glacier is shown by the fact that a pond 20 feet deep and covering 10 acres existed for 24 years and was carried a distance of 600 feet. (2) Irregularities Due to Streams and Ice Tables. The surfaces THE WORK OF GLACIERS 149 FIG. 133. Denver Glacier, Alaska, showing an ice- ^> feeding grounds, and lateral and medial moraines. (Photo. F. B. Sayre.) of glaciers become irregular in other ways besides fracturing. Water from the melting ice forms rivulets, which erode and melt channels in the ice. When such a . stream reaches a cre- vasse it plunges down, forming a circular shaft called a moulin (French for mill). As the ice moves on, this opening is closed and a new one formed in its place. Thus a series of inactive moulins, in various stages of preserva- tion, are left extend- ing down the glacier from the active one. The active moulin, however, may be said to remain stationary or confined to narrow limits, and may, in time, excavate potholes (p. 93) many feet in depth in the rock beneath the glacier (Fig. 134). Since the ice of a glacier varies in com- pactness it melts unevenly, and this also tends to produce a rough surface. The surface of a glacier is also roughened by the irregular melting of the ice, due to the accumulation of debris. If a fragment of rock which has fallen on the ice is too thick to be heated through by the sun it will protect the ice beneath from melting. Because of this it may in time stand on an ice pillar several feet in height, forming an ice table (Fig. 135). After a time the pillar may become so high that the sun will be able F to melt it. The protecting cap of rock will FIG. 134. A giant pot- f b r hole formed in the bed of a then be undermined and will slide ott, on glacier by the water, sand, the south side in the northern hemisphere, and gravel carried through a an(J w j n then be rea( j tQ cauge the forma . crevasse. Near Chnstiama. (After A. Geikie.) tion of another column. 150 PHYSICAL GEOLOGY FIG. 135. Ice pillars protected by slabs of rock. Parker Creek Glacier, California. (After Russell.) The portions of a glacier over which dust or thin layers of earth are spread will be melted more rapidly than those not so covered, since the dark dust absorbs heat more rapidly than does ice. In this way dust wells and other irregular hollows several inches in depth are formed, the depth depending upon the diameter of the hollow and the angle at which the sun's rays strike it. The great drifts of snow which had to be removed each spring during the construction of a railroad in Norway were scat- tered over with fine dust in order that they might be more quickly melted by the sun. If, however, dust is more than an inch thick it prevents the under- lying ice from melting and forms dirt cones. Often the greatest irregulari- ties on mountain glaciers are the long lines of rock debris (surface moraines) which may be of considerable thickness and which usually rest upon high ice ridges formed by the protecting cover of the former. The water from the moulins and that which reaches the bottom of the glacier in other ways, as for example that melted from the lower surface of the glacier by friction, that which comes from the springs in the valley through which the glacier is moving, and that which seeps through the cracks of the ice, all emerges from a tunnel in the end of the glacier as a single stream, often of considerable size. These streams flow even throughout the extreme winters of glaciated regions. MOVEMENT OF GLACIERS The Swiss early had reason 'to believe that glaciers move, as was shown when two glaciers advanced over fields and meadows, up- setting barns and filling the quarries from which the citizens of Bern obtained their marble. A recent example of this sort occurred in 1909-1910, when the advancing Child Glacier in Alaska threatened to destroy a $1,400,000 steel bridge. Rate of Movement. It was not, however, until 1827 that any serious attempts were made to determine the rate at which glaciers move. In that year Hugi built a hut upon the Aar Glacier in Switzer- THE WORK OF GLACIERS land and noted its position from year to year. In fifteen years it had moved 1428 meters, or about 100 meters a year. Forty- four years later the remains of the hut were found 2408 meters lower down the valley. Since these first measurements careful surveys have been made from time to time, and it has been found that the motion of Alpine glaciers seldom exceeds one third to two thirds meter (one to two feet) a day. In 1861 the heads of three guides with some hands and fragments of clothing appeared at the foot of the Bossons Glacier on whose neve they had been buried beneath an avalanche forty-one years before. So perfect was the preservation that they were easily recognized by a guide who had known them in life. The rate of movement had been eight inches a day. In large glaciers, however, the rate is much more rapid. It is estimated that the Child Glacier in Alaska moves about 30 feet a day during the summer, and a large glacier which drains the snow fields of north Greenland is said to have moved more than 60 feet in a single day. These latter figures are exceptional and apply only to very large and thick glaciers. Large glaciers, however, do not always move faster than small ones, since other conditions may counterbalance the greater thickness. Differential Movement of Glaciers. By placing stakes in a straight line across the surface of a glacier and a vertical row on a I < 1 / ; / O 1 I ooe ABC FIG. 136. Diagrams showing the movement of glaciers. A, a line of stakes placed in a straight row across a glacier becomes more and more curved each day. By a line of stakes placed in a vertical row on the exposed side of a glacier becomes more and more inclined. C shows the formation of marginal fissures produced by the pulling of the more rapid central portions upon the slower marginal portions. side exposure, it was found that the middle of a glacier moves faster than the sides (Fig. 136 A} and the top faster than the bottom (136 A, B). In one glacier, while the top moved 6 inches, the middle moved only 4.5 inches, and the bottom 2.5 inches. The reason for the slower motion of the sides and bottom is evidently to be found in the friction with 152 PHYSICAL GEOLOGY the walls and bed of the valley through which the glacier flows. It follows from the above that the rate of movement will be reduced if the bed of the glacier is rough, and that a smooth bed will favor rapid motion. It has been found that the line of swiftest motion is not always in the middle of glaciers, but that as in the case of rivers, although to a lesser degree, it is deflected from side to side, being nearer the outside of a curve. Factors Influencing the Rate of Movement. The rate of move- ment of a glacier increases with (i) the slope of the bed upon which it rests, but depends even more upon (2) the slope of the upper surface of the ice and upon (3) its thickness. A general inclination of the upper surface of a glacier is necessary for glacial movement, although for short stretches the surface of the ice may even have a backward slope. The beds of valley glaciers slope in the general direction of the movement of the ice, but there are many local exceptions, as is shown by the deep basins in valleys formerly occupied by glaciers. The great ice sheets of North America moved to the south over a land surface which for many miles sloped towards the north, i.^., in the direction opposite to that of the movement of the ice. In all such cases the upper surface of the ice must have sloped in the direction of the glacial movement. The velocity of a glacier is greater in summer than in winter, and at midday than at night ; that is, when the glacier is melting more rapidly and is most thoroughly saturated with water. The Mer de Glace, France, moves in summer at an average rate of 27 inches a day in the middle and 13 to 19 inches a day near the sides; in winter the rate is about half as much. Other factors influencing the rate of motion of a glacier, besides the slope of the ground, the slope of its upper surface, and the quantity of water with which the ice is sat- urated, are the amount of load in its basal portions, which tends to retard the rate, and the straightness of its course and the smoothness of its bed, which tend to increase it. Lower Limit of Glaciers. Glaciers move down their valleys until they reach a point where the melting (ablation) equals the forward movement (Fig. 137). When the melting exceeds the forward move- ment, the glacier is said to retreat; when it is less the glacier advances. It is evident that the lower limit of a glacier will not be fixed (except when it reaches the sea) unless the conditions of temperature and snowfall remain constant. Since both temperature and snowfall THE WORK OF GLACIERS 153 usually vary from year to year and, more widely, in cycles, 1 the ends of glaciers are seldom stationary for long periods. If the depth of the snow in the cirque increases during a single year or a number of years, the glacier will advance ; while if the snowfall is slight or the average temperature high so that little snow can accumu- late, the glacier will retreat. For example, because of the hot summer of 1911 practi- cally all of the glaciers of the Alps were in retreat in 1912, one (the Brenva Glacier on Mt. Blanc) receding 50 me- ters. The Muir Glacier in Alaska has retreated seven miles in the past twenty years, and the glaciers of the Chamonix valley in the, Alps, one quarter to one half of a mile since 1812. In 1858 there was a harbor in Bell Sound, Spitsbergen, at the head of which was a strip of lowland and beyond this a low, but broad glacier. In 1860-1861 the glacier advanced over the lowland, filled up the harbor, and extended far into the sea. It is now one of the largest glaciers in Spitsbergen. A large glacier responds to excessive or deficient snowfall more slowly than a small one, and several years may elapse before it shows the effect of such changes. An unusual cause of rapid glacial advance is recorded from Alaska, where the ice fronts of a number of glaciers have moved forward as a result of earthquake shocks. During an earthquake in 1899 the mountains from which the snow supply of these glaciers is derived were so vigorously shaken that great avalanches of snow and rock were thrown down on the neves. This increased supply caused all of the glaciers in the region affected to advance. They did not all, 1 There appears to be a climatic cycle of 35 years during which a series of cold or rainy years is followed by years which are warmer or drier. FIG. 137. The Rhone Glacier, showing crevasses and front. 154 PHYSICAL GEOLOGY however, show a simultaneous forward movement. The reason for this is to be found in the size of the glaciers. It was discovered that the smaller glaciers advanced more quickly and had, indeed, in 1909 already passed completely through the period of advance, while the long glaciers were at that date (1909) either still advancing or merely beginning to advance. 1 TRANSPORTATION OF MOUNTAIN GLACIERS Surface Moraines. We have seen (p. 29) that in regions where frost is an active agent of the weather, talus, composed of angular rock 5, fragments of various sizes, rests at the bases of cliffs. If the bottom of a valley with high and steep sides is occupied by a glacier, these fragments will fall upon its surface ; and as the ice moves on all parts of the glacier's side, will pass under the cliffs which supply the debris. In the process of time, a regular ridge of angular rocks and soil will rest upon the edge of the ice. Such a deposit is called a lateral moraine (Figs. *3 8 > J 39> I 4)- The ice of glaciers does not commingle at their confluence, but the masses move on side by side, the two adjacent lateral moraines uniting to form a medial moraine (Figs. 138, 139, 140). In this way, by the union of several branches, a glacier may be covered with several medial moraines. A medial moraine may also be formed when a glacier passes over an elevation in its bed from which it scrapes off rock fragments. After the glacier has passed this point the debris 1 Physiography and Glacial Geology of the Yakutat Bay Region, Professional Paper No. 64, U. S. Geol. Surv., 1909. FIG. 138. Map showing the formation of the Mer de Glace by the union of several gla- ciers. The development of lateral and medial moraines is illustrated also. THE WORK OF GLACIERS 155 f. may be exposed by the melting of the surface of the ice and continue to the end of the glacier as a medial moraine. It will readily be seen that the material of the various surface moraines of a glacier may differ widely in composition, since they were derived from the rocks of many parts of the valley. The Baltoro Glacier of Hindu Kush has fifteen moraines of different colors. (Bonnersheim.) Since the surface moraines are usually sufficiently thick to protect from the sun's rays the ice upon which they rest, they are generally situated on ridges of ice, sometimes 50 to 80 feet in height. After a time the ridges become so high that the morainic material slips off, thus widen- ing the morainic belt. After several repeti- tions of this process the medial and lateral moraines may cover completely the lower end of a glacier. The size of some of the rock frag- ments carried on the surfaces of glaciers is very great. One such bowlder con- fc^ ^lVV^l>~ / ^"\* i f^f^l) v~/^V' tained 244,000 cubic Kfn /IT .. ^ FIG. 139. Diagram and cross section of a mountain feet (Forbes), which glader Lateral morajnes are seen to produce medial IS equivalent to a moraines. The movement of the superglacial material to squared Stone 122 feet f rm englacial and finally subglacial material is shown. In CO f *H Icefalls occur near the confluence of the glaciers on the K' 3 ' right, and 3 6 feet high. The weight of material which a glacier can carry on its surface is limited only by what it may receive, and the very weight of the surface load will hasten the movement of the glacier. Upon the disappearance of a glacier, these great rock masses are often left in unstable posi- tions and are then known as balanced bowlders, or rocking stones. 156 PHYSICAL GEOLOGY All bowlders transported and deposited by glaciers are given the general name erratics. Figure 141 shows a balanced bowlder. _ Subglacial Material. The bottom por- tions of glaciers con- tain stones andground- up rock. This ma- terial is derived, either directly from the rock bed or from the super- glacial material which reaches the bottom through the crevasses. The subglacial mate- rial is usually much FIG. 140. Aar Glacier showing lateral and medial moraines, and cirques. (Photo. L. E. Westgate.) worn. En glacial Material. Between the sur- face and the bottom of a glacier some debris may be carried. This is derived in part from the superglacial material which has not reached the bottom , , through the crevasses, in part from that which gathered on the surface of the snow or neve and was subse- quently covered, and in part from that which was scraped off an elevation in the bed. The englacial material may become superglacial by abla- tion, and subglacial by gradually settling or by the melting of the lower ice. All will be deposited when the FIG. 141. Balanced bowlder, Hoosac Mountain, i r Massachusetts. The bowlder is so nicely balanced that tne glacier is although of great weight it can be made to vibrate with reached. little effort. THE WORK OF GLACIERS 157 EROSION BY MOUNTAIN GLACIERS Plucking and Abrasion. Glaciers accomplish their work of erosion in two ways, (i) The ice secures a hold on the material of its bed either by freezing about projecting points of rock or by being pressed into the joints and other cracks by its great weight. As the glacier moves on rock fragments are pulled out and carried along. This process is called plucking, and by it a glacier may remove a great quantity of material in much-jointed rock. The process, however, is of little effect on rocks which have few joints ; and conse- quently one some- times finds that a glacier has been able to deepen its valley more easily in hard granite and gneiss than in the softer limestone, because the former were much fractured, per- mitting the plucking out of blocks, while the latter being less broken was little affected. (2) Glaciers also deepen and widen their valleys by abrasion. The tools which accomplish the work of abrasion are the rocks which have been torn from the bed by plucking and those which have reached the base of the ice from the surface through crevasses. Hold- ing these rock fragments in a firm grasp and pressing with great force, estimated to be 48,600 pounds to the square yard in portions of the Aar Glacier, a glacier acts as a gigantic file, cutting down pro- jecting points, and deepening and smoothing its bed. The hard pebbles scratch and the bowlders groove the bedrock ; while the clay and rock, ground fine by the process, polish the surface, producing the smoothed and striated appearance so characteristic of glaciated rocks. It will readily be seen that a thick glacier, because of its great weight, will be able to erode its bed more rapidly than a thin one, FIG. 142. Roche moutonne'e, Central Park, City of New York. (U. S. Geol. Surv.) 158 PHYSICAL GEOLOGY and that a very thin glacier or one with clear ice at its bottom may, indeed, not only be unable to wear down its rock bed, but may even override loose sand or glacial drift. In moving over a projection in its valley a glacier smooths off the side upon which it impinges (the stoss side) and plucks angular frag- ^ ments from the lee side, often leaving it rough and jagged. It also smooths rough sur- faces into forms (Figs. 142, 143) which, because of their rounded shapes, have been given the name roches mouton- n'ees (French for rock sheep). If one looks down a glaciated valley, the smooth stoss slopes of the roches moutonnees are very striking. If, however, one glances up the valley, the rough lee sides of the roches moutonnees are often so conspicuous as to seem to con- tradict the statement that glaciers deepen their valleys by abrasion. Effect on the Material Carried. Not only are the rock beds of glaciers grooved by bowlders, scratched by pebbles, and polished by FIG. 143. Diagram showing a projecting rock smoothed and rounded on the side upon which the glacier impinged (stoss), and rough- ened on the lee side by plucking, a roche moutonnee. FIG. 144. Glaciated pebbles. (After Blackwelder and Barrows.) clay, but these tools are, in turn, scratched and polished (Fig. 144). If one axis of a glaciated pebble is much longer than the other, it is usu- ally found that the striations are parallel to the longer axis. This is due to the fact that the pebble was held in this position in the ice, as it offered less resistance in this way than in any other. If the axes of a pebble are approximately equal, it may have scratches running in many directions, as this shape would enable it to turn more easily in the ice and therefore to be carried onward in various positions. THE WORK OF GLACIERS 159 Since much of the rock of the valley floor over which a glacier moves as well as the pebbles which it holds in its grasp are ground to powder, it is not surprising to find the water of glacial streams so turbid with sediment as to be spoken of as glacier milk. The light color of such streams differs from the yellow color of ordinary streams because the former carry freshly ground, unoxidized rock, and the latter the prod- ucts of weathering. The Aar Glacier, for example, a comparatively small glacier of the Alps, is estimated to discharge 280 tons of rock flour a day during a summer month. Factors Influencing the Rate of Erosion. As has been said, glaciers accomplish little erosion in passing over smooth surfaces. The amount of morainic material carried by Greenland glaciers, for example, is surprisingly small. This is due to the fact that they have moved over their beds so long as to render them comparatively smooth. If, however, a glacier moves over a bed whose surface is sufficiently rough to permit the ice to tear away fragments by pluck- ing (p. 157), it is likely to deepen its bed rapidly, since under these conditions a large surface is exposed to the wear of the debris held in the base of the ice. Erosion is also favored by the incoherence of the material over which the ice moves, by the weight of the ice, and by a rapid rate of movement; but is unfavorably affected by an over- loading of the basal portion of the glacier with debris and by the resistance of the rock. DEPOSITS OF MOUNTAIN GLACIERS Terminal Moraines. At the lower end of a glacier, where the melting equals, or nearly equals, the advance, all of the debris (the superglacial, englacial, and subglacial) is deposited as a terminal moraine. Terminal moraines (Fig. 145) are usually crescent or horseshoe-shaped, concave towards the glacier, and often form con- spicuous hills in valleys once occupied by glaciers. The heights of terminal moraines vary greatly, since the quantity of material de- posited in them depends upon a number of factors, (i) The length of time during which the front of a glacier remained stationary is important. If a glacier advances 600 feet a year and for a number of years melts back at the same rate, it is evident that each year all of the debris carried on, in, and under it will be left at the same place, with the exception of that which is carried away by the stream which flows from it. If, on the other hand, the ice melts back 600 feet a year, CLELAND GEOL. II 160 PHYSICAL GEOLOGY while it advances 500 feet, it is evident that comparatively little debris will be left at any one spot, and no conspicuous hills or ridges will be formed. (2) The velocity of the glacier, (3) the quantity of material transported by it (p. 154), and (4) the amount carried away by the stream which flowed from its end are also determining factors in the size of terminal moraines. They sometimes reach a height of several hundred feet, but heights of 100 or 200 feet are more common. If the front of a waning glacier halts for considerable periods at different points, a series of terminal moraines (also called recessional moraines) will be left. The material of terminal moraines usually consists of a heterogene- ous mixture of large and small pebbles and bowlders of different kinds, FIG. 145. Moraine near Dansville, New York. (Photo. H. L. Fairchild.) embedded in clay and sand. Occasional patches of stratified sand and gravel from the water of the melting ice also occur. All the glacial debris is called drift, the unstratified is called till or bowlder clay, and the stratified (sorted and laid down in water) is called stratified drift. On glaciers which move between precipitous walls supplying great quantities of talus, the lateral moraines will be large ; and upon the disappearance of the ice, especially if the retreat be slow, a high ridge of unstratified drift will be left on each side of the valley. Some of these are a thousand feet or more in height. The terminal moraine of such a glacier may be comparatively insignificant. Ter- minal moraines are breached by streams and are sometimes entirely removed by them. Sometimes, however, the moraine constitutes an effective dam for many years, behind which a picturesque lake lies. THE WORK OF GLACIERS 161 Thousands of mountain lakes owe their existence to such morainic dams. Submarginal Moraines. Another kind of moraine is formed under the sides of a glacier by the movement of the ice from the center to the sides. This should not be confused with the lateral moraines of the surface (p. 154). In valley glaciers which receive little superglacial debris these submarginal moraines maybe thicker than the surface moraines. The presence of polished and striated pebbles and bowlders in such moraines is abundant proof that the drift composing them had been carried between the ice and its bed. Ground Moraine. A glacier may be so full of debris in its basal portion that it is unable to carry all of it. Under such conditions some of the load is deposited and is overridden. Such deposition takes place (i) where the ice is thinning near the end, as this makes its movement less rapid, so that it is unable to carry all of the load which it has acquired in its progress through a rough valley. Such deposi- tion also occurs (2) after a glacier has passed over a projection in its bed, as the bottom of the ice is then heavily loaded with the debris which it has plucked or abraded from the obstacle. In a valley formerly occupied by a glacier there is usually a layer of compact till composed of clay and much-worn pebbles. This deposit is known as the ground moraine and was derived either from the bottom of the advancing ice, as described above, or from the base of the ice upon its disappearance. It is usually thickest near the terminal moraine and thinnest near the head of the glacier, while over portions of the valley it may be entirely lacking. Since conditions in valley glaciers favor erosion rather than deposition, their ground moraines are seldom important, being in contrast in this respect to continental glaciers (p. 171), whose ground moraines are of con- siderable thickness, although seldom attaining the depth of terminal moraines. The Work of Glacial Streams. The streams which flow from beneath glaciers or from their sides are supplied with pebbles from the moraine and an abundance of rock particles derived from the rock ground to fine flour between the ice and its bed. With such tools they are able to deepen their channels as long as they have sufficient velocity. The streams from certain glaciers emerge from their fronts in deep gorges which they have cut in the rock. The Lammer Glacier of Switzerland and the Mer de Glace are examples. It is doubtful, however, if the deepening which is such a marked feature of valleys 162 PHYSICAL GEOLOGY long occupied by glaciers, has been accomplished to any great extent by subglacial stream erosion. If the streams which issue from glaciers are ponded by terminal moraines, lakes are formed in which they deposit their loads. If they have a free course, however, they will carry their loads of rock flour and peb- bles farther down the valleys. The coarse gravel will soon be dropped, but the finer material may be car- ried some distance. When, however, a stream thus loaded reaches a more gentle FIG. 146. The union of the Rhone and Arve rivers near Geneva, Switzerland. The water of the Rhone, hav- ing been filtered by Lake Geneva, is clear and blue, while that of the Arve is grey with the rock flour carried into it by glacial streams. To the right is seen the cement works for recovering the Arve sediments. (Hobbs, Earth Features.) grade, it may lose so much velocity that it becomes overloaded and compelled to take a braided course (p. 86). The stratified deposits laid down in valleys by glacial streams are called valley trains (p. 178). As has been stated (p. 1 59), streams flow- ing from glaciers are milky with rock flour, while those which gather their water from the land sur- face may be yellow with the clay of the weathered rock which they bear along, but streams filtered by FIG. 147. Block diagram showing a valley blocked by a moraine; the stream having been diverted from its old course has cut a steep-sided, postglacial gorge. lakes are clear. At the confluence of the Rhone and the Arve (Fig. 146) a striking contrast is seen between the clear water flowing from Lake Geneva and the turbid water of the Arve which has its source in the glacier of that name. For a THE WORK OF GLACIERS 163 short distance after their union the waters of the two streams flow side by side, but gradually they merge. Since streams are no longer overloaded after the retreat of their glaciers, they begin to erode the alluvial deposits of their former flood plains and in this way form the terraces which so often border stream valleys in glaciated regions. As the streams deepen their beds in their partially filled valleys, they occasionally fail to find their former channels, and after excavating broad valleys in the recently deposited gravels may cut narrow gorges into solid rock. This is shown in the diagram (Fig. 147), in which a stream flows from its alluvium-filled valley (in the background) into a deep, postglacial gorge in the foreground. LANDSCAPE MODIFIED BY GLACIAL ACTION Characteristics of Glaciated Valleys. A striking feature of moun- tain valleys which have been subjected to the long-continued erosion of thick glaciers is the flatness of the floors and the steepness of the valley sides, as contrasted with the V-shaped valleys cut by streams. A cross section of a valley which has been shaped by glaciers is typically a gigantic U, sometimes more than 3000 feet deep and three miles wide. The tributary streams of such valleys usually enter them over falls. The high, tributary valleys are called hanging val- leys (Fig. 148), and their occurrence is proof that the main valley has been deepened by glacial action. This peculiar relation between the main valley and its tributaries can best be understood by following the history of a valley from the time it was first occupied by a glacier until it again became free from ice. When a main valley is occupied by a thick glacier, it will in time be deepened and broadened, especially near the bottom, and the valley sides will at the same time be oversteepened. This excavation is termed overdeepening. Since the main valley is well filled with ice, it is evident that the glaciers of the tributary valleys will not be able to lower their beds far below the surface of the main glacier. Consequently, when the glaciers disappear from the valleys, the side valleys will no longer enter the main valley at grade, but by falls. In other words they have become hanging valleys. In this way those steep-sided, picturesque valleys were formed for which Switzerland and British Columbia are famous. The many falls of the valleys of the Yosemite, California (Fig. 148), and Lauterbrunnen, 164 PHYSICAL GEOLOGY THE WORK OF GLACIERS 165 Switzerland, and many other of the high falls of the world are of this origin. Hanging valleys of a different origin have been dis- cussed elsewhere. The courses of valleys are straighter after glaciation than before. This is due to the fact that the glaciers, because of their rigidity, cut off the " spurs " on the inside of the curves. The line on the side of a U-shaped valley above which glacial erosion was not effective is marked by a change in slope, forming a sort of " shoulder." Such " shoulders " are of some economic importance in Switzerland, since they usually afford good pasturage a.nd are favorite spots for hamlets, as they are not subject to the severe cold of the deep valleys. The " shoulders " are usually about 1000 feet above the bottom of the valleys in the Alps, but are sometimes as much as 3000 feet above. Mature Glaciated Valleys. It is evident that valleys which were formerly occupied by glaciers will not be U-shaped unless the glaciers were at work for a long time, and every gradation can be seen between them and V-shaped valleys, in which the inside of the curves (spurs) have been little cut away and the beds are still broken by falls. In valleys long subjected to glacial action the spurs are cut away, the bottoms widened, and the sides smoothed. In the upper and middle portions where the weight of the ice and its movement were greatest, basins may have been formed in which lakes now rest. Lake Chelan, Washington, is probably such a lake, as are also the beautiful lakes of the Scottish Highlands. Even maturely glaciated valleys may not have graded beds, since, under certain conditions (p. 157), a glacier erodes one portion of its bed more deeply than another. Destruction of Features of Glaciated Valleys. The characteristic features of glacial valleys are destroyed in process of time by the work of erosion and weathering, very much as are those of cirques. Talus slopes accumulate at the bases of the cliffs ; landslides sometimes cover considerable areas of the bottoms with debris ; the streams from the mountains build out alluvial fans and cones; and in the course of time the " shoulders " are worn away by weathering and the action of the rills which tumble over them. The streams from the hanging valleys cut down their beds so that they enter the main valleys through deep canyons. It is possible by these criteria roughly to determine the length of time since the disappearance of the glacier from the valley. 1 66 PHYSICAL GEOLOGY Fiords. 1 The coast of Norway is noted for the long, narrow bays, called fiords (Fig. 149), which may be navigated for many miles. FIG. 149. Fiord, Grenville Channel, British Columbia. (U. S. Geol. Surv.) Into these fiords streams enter from hanging valleys over falls. Soundings show that while the end towards the sea is very deep, it is not so deep as at some distance inland. The maximum depth of the Sogne fiord in Norway (Fig. 150) is 4000 feet, and that of three ...._ others is 2550, 2298, and 1800 feet. The greatest depth occurs where the fiord is bounded by moun- tain masses of great extent and elevation. There seems little doubt that fiords are valleys which were greatly deep- ened by glacial erosion. Their increased depth from the outlets inward is due either to the greater erosion of the glaciers some distance inland, where they were presumably thicker and their erosive power consequently greater; or to the piling up of morainic matter where they entered the ocean ; or probably both cooperated to produce the result. Whether or not the glaciers actually cut the valleys below sea level has not been proved. 1 It has been maintained that fiords owe their characteristics to earth movements and not to glacial action, and, in fact, that fiords occur in non-glaciated regions. According to this theory areas were fractured along certain belts as they were being raised to form plateaus. These belts of more or less shattered and fissured rocks are supposed to have subsided, with the formation of steep-sided troughs. In support of this theory it is pointed out that fiords are arranged along a kind of angular network believed to be caused by intersecting lines of frac- tures. (Gregory, J. W., The Nature and Origin of Fiords, 1913.) FIG. 150. Map of Sogne fiord, Norway. THE WORK OF GLACIERS 167 During the process of the glacial deepening of the Scandinavian fiords the land was higher than now. This was followed by a period of great submergence and later by a reelevation of a few hundred feet. Fiords are common in Greenland, Alaska, British Columbia, southern Chile, and Patagonia. PIEDMONT GLACIERS When mountain glaciers reach the plain at the foot of the mountains from which they flow, they spread out, as they are no longer confined by valley walls, and coalesce to form piedmont (foot of mountain) * FIG. 151. Model of the Malaspina Glacier. The dark margin is the moraine- covered area upon which a forest of spruce, cottonwood, and alder grows. (Model by Lawrence Martin. Copyright, University of Wisconsin.) 1 68 PHYSICAL GEOLOGY glaciers. A typical example of such a glacier is the Malaspina in Alaska (Fig. 151). It is formed by the union of several glaciers which move down the valleys of the St. Elias range upon a nearly flat plain. The area of the united glacier is nearly 1500 square miles, about the size of Rhode Island. The lateral margin where the ice is probably 1000 feet thick is covered with a belt of morainic matter a few feet thick and several miles wide, on which grows a luxuriant vegetation. Extensive areas of bushes are found and, near the outer edge, trees some of which reach a diameter of three feet. On the surface of the nearly stagnant glacier are numerous ponds in which stratified deposits are laid down. The central portion of the glacier is com- paratively free from debris and is much broken by crevasses into which streams from the melting ice flow. Piedmont glaciers are rare at the present time, but were much more numerous during glacial times, when they existed at the foot of the Alps, the foot of the mountains of western North America, the southern Andes, and elsewhere. CONTINENTAL ICE SHEETS Up to this point mountain glaciers have been discussed because, on account of their small size and accessibility, they are more easily studied and their phenomena are better known than are those of the great continental glaciers such as now cover Greenland and the Ant- arctic Continent. At one time ice sheets covered the northern portions of North America and Europe. These were of great extent, those of North America covering an area estimated at 4,000,000 square miles; of long duration, and probably of great thickness. The stratified and unstratified drift so conspicuous in many of these once glaciated regions was formerly believed to have been transported to its present position and the underlying rock scratched and polished, by a great flood (Mosaic flood) which swept down from the north, carrying with it pebbles and bowlders which striated and grooved the rocks over which they were borne. The term drift is a relic of this ancient theory. One cannot obtain a clear conception of the con- ditions which existed in North America and Europe during the Glacial Period (p. 645) without a study of the existing continental glaciers of the polar regions. Greenland. Greenland is a continent 1400 miles long and 900 miles wide. Of this area, fully three quarters are covered at all times with ice, the only inhabited portion of the continent being a narrow THE WORK OF GLACIERS 169 70 60 50 40 30 20 strip of land along the coast, generally from 5 to 25 miles wide (Fig. 152). The great snow desert of the interior is devoid of life, with the exception of lowly forms, such as the microscopic red plants (Sphcerilla nivalis) which sometimes exist in such abundance as to give a red color to the snow and produce the " red snow " of both mountain and conti- nental glaciers. All Greenland explorers give nearly identical de- scriptions of the interior. Near the coast the ice sheet rises with comparative ab- ruptness, being steeper on the east than on the west coast, while the central por- tion is nearly flat (Fig. 153). The gradient of the surface, as a whole, gradually de- creases as the interior is approached, and "the mass thus presents the form of a shield with a surface cor- rugated by gentle, almost imperceptible undulations, lying more or less north and south." (Nansen.) The highest recorded point in the interior is about 9000 feet above the sea, although it is possible that unexplored portions may reach an altitude of 10,000 feet. A hundred miles from FIG. 152. Map of Greenland, showing a con- tinental ice sheet. FIG. 153. A section across Greenland, showing the profile of the ice and the prob- able configuration of the land. the coast no depressions or elevations in the ice mark the presence of valleys or mountain ranges beneath ; but within 50 or 70 miles of 170 PHYSICAL GEOLOGY the coast broad, shallow, basinlike depressions appear which, when followed seaward, grade into great glaciers which flow from the central mass into the sea through valleys. The scenery of these broad de- pressions resembles, on a grand scale, the gathering grounds of Alpine glaciers. The ice of the depressions which give rise to these separate tongues of ice is probably a mile in thickness, 1 but on the mountain ridges it is much thinner. The smooth almost flat central portion gives place near the margins to a surface of a decidedly different character. Here, where the motion of the ice is more rapid, being greater in one portion than in another, the surface is much broken by crevasses which make travel well-nigh impossible. Near the coast, where the mountains protrude through the ice as islands (nunataks), the whiteness of the surface is broken by patches and lines of rock waste derived from these projections. Nunataks are often surrounded by deep ditches, due to the absorption of the sun's heat by the dark rock walls and its radiation from them, with the consequent melting of the adjacent ice. As has been said, the great interior plateau is drained by glaciers which descend through valleys. Many of these reach the sea, where their fronts are broken off and carried away as icebergs. Some of these glaciers are among the largest known. One of the most remark- able is the Humboldt Glacier, which has a breadth of more than 50 miles where it enters the ocean. Its front rises precipitously from the level of the water to a height of 300 feet, and the total thickness above and below the water level at this point is probably 2700 feet. Some of these glaciers fail to reach the sea but spread out on flat plains. In such glaciers it is seen that the ice is stratified and that the white upper layers are in marked contrast to those near the base, which are often so filled with debris that it is difficult to tell where the ice ends and the ground moraine begins. This loading of the basal portions of the ice and the almost total freedom of the surface from debris should be borne in mind when the work of the ancient ice sheets is considered. The rate of movement of the Greenland glaciers near their ends is sometimes more than 50 feet a day, but in the interior of the ice sheet the rate may be as slow as an inch a day or practically zero; in other words the motion is from the center of the ice sheet outward, the movement being caused by the weight of the ice. Moreover, the 1 Geikie A., Textbook of Geology. THE WORK OF GLACIERS 171 movement is locally concentrated and therefore increased where the ice finds a relatively narrow outlet between inclosing ridges. The contour of this buried continent can be conjectured from the fact that the greatly indented coast with its numerous islands re- sembles that of Norway, so that it is probable that a rough, moun- tainous surface like that of Norway underlies the ice. The Antarctic Continent. The Antarctic Continent is larger than the whole of Europe and differs from Greenland in its greater height, in the greater severity of its climate, and in the absence of a strip of ice-free land bordering the ice. The interior, as in Greenland, is dome or shield-shaped and was found by Amundsen to be 10,500 feet above the sea at the pole. Above the general level of the shield, mountains rise to heights of 15,000 feet. The excess ice is drained by valley glaciers, as well as by the Great Ice Barrier, a floating ice shelf which in Victoria Land forms an ice cliff many miles long and varying in height from 280 feet to places so low that it can be used as a wharf. ANCIENT GLACIATION The proofs that glaciers at one time covered a large area in North America are so conclusive and the belief is now so universal that it seems remarkable that the theory was not advanced until 1846 (by Agassiz), and that nearly thirty years elapsed before its general accept- ance by geologists. DEPOSITION (i) Bowlders. One of the most striking proofs that a region has been covered with glaciers is to be found in the occurrence of bowlders in the soil and on the surface, which differ in composition from the un- derlying rock and consequently were not derived from it. When traced to their parent ledges, some of these bowlders are found to have been carried several hundreds of miles over hill and dale. In regions of rough topography, glacial bowlders are often found at a much higher level than the ledges from which they came. For example, bowlders of quartzite have been found on Mt. Greylock, Massachusetts, which must have been carried into a valley and then to the top of the mountain, a vertical distance of almost 3000 feet. Bowlders of jasper conglomerate, composed of bright red pebbles of jasper embedded in white quartz, have been found from the northern 172 PHYSICAL GEOLOGY to the southern confines of Ohio, and when traced northward are seen to have been derived from a deposit on the north shore of Georgian Bay in Canada. Native copper from the copper deposits of Lake Superior has been found in the drift many miles to the south. Pieces of copper transported in this way were often picked from the drift and made into ornaments by the ancient mound builders. In New England trains of bowlders have often been traced to outcrops which have some distinctive character (Fig. 154). In any one deposit the number of kinds of rock is usually not very great, although in some'cases one may find as many as twenty different varieties in a single bed of till. The bowlders are often angular and scratched, and in both of these features differ from stones which have been rolled about by streams. It should not be inferred from the above that no rounded and un- scratched bowlders occur in glacial debris, for sometimes the angular and scratched bowlders are in the minority. (2) Unstratified Drift. As has been said in connection with mountain glaciers, all of the debris transported by the ice is included under the general term drift; till or bowlder clay being the unstratified t j ie drift, and that carried and later deposited by streams being called stratified drift (p. 178). Till (Fig. 155) is composed of a heterogeneous mixture of bowlders of many sizes embedded in clay. The mix- ture of coarse and fine material and the lack of stratification proves without question that such deposits were not made by streams. The stratified drift (Fig. 156) was deposited by melting waters under various conditions. Drift is not confined to the valleys of glaciated regions but is spread over hills as well, being in a measure independent of topography. The deposition of drift may render a region either rougher (Fig 157) or smoother (Fig. 1 58). If, for example, in passing over a region some- FIG. 154. Map of "bowlder train" from Iron Hill, Rhode Island. The direction of the movement of the ice is generalized. (After Hobbs.) THE WORK OF GLACIERS 173 what cut up into valleys, the ice sheets filled them with drift, compel- ling the streams upon the retreat of the ice to make new valleys, the FIG. 155. Till. Note the heterogeneous mixture of large and small bowlders and fine clay. (Pennsylvania Geol. Surv.) result may be a smoother surface. Many portions of the northern United States have been modified in this way. On the other hand, the irregular dumping and piling up of drift in moraines may roughen the landscape. The northern half of the peninsula of Michigan, which in certain places now rises from 1000 to noo feet above the surface of the Great Lakes, seems to have no rock standing more than 250 to 300 feet above the lakes, there being from 700 to 800 feet of drift on its higher parts. The average thickness of the drift in southern Illinois is somewhat less than 30 feet; in north- western Ohio, north- ern Indiana, and FIG. 156. Stratified lake clay resting on till. PHYSICAL GEOLOGY northeastern Illinois the average is almost 200 feet, and in the south- ern peninsula of Michigan about 300 feet. Borings near Cleveland, Ohio, show that the drift extends 470 feet below lake level. FIG. 157. Diagram showing a region made rougher by glaciation. The drift deposited under the ice was often compacted by the great weight of the ice mass into a dense bowlder clay which is excavated with difficulty. Mention should be made of an area in Wisconsin and adjacent por- tions of Iowa, Illinois, and Minnesota, which was not covered by the ice FIG. 158. A region made smoother by glacial debris. The uncertainty in coal mining (black bands) in such a region is shown. sheets, the driftless area (Fig. 159 A). In this area the rock is deeply weathered, and rock pillars are not uncommon ; the drainage is per- fect, the streams being without swamps, lakes, or waterfalls. It differs markedly in these respects from the adjoining regions (159 B). This freedom from ice was due to a combination of causes : its position with reference to the centers from which the ice moved, the higher ground to the north, and the presence of the deep Michigan and Superior basins which diverted the flow of the ice. Moraines. The drift was laid down either as terminal, ground, or lateral moraines, or as stratified deposits. (A) Terminal moraines (see p. 159) were formed where the ice front remained stationary or nearly so, for a long time, so that its forward movement was almost or quite equal to the melting at its margin, sometimes being slightly in excess and sometimes slightly less. Under such conditions it will readily be seen that the glacial debris would be left in extremely irregular deposits, unless the drift had been uniformly distributed throughout the ice, which apparently was seldom the case. The term " recessional moraine " is often used to indicate those moraines which were formed during the various halts of the ice as it retreated to the north, " outer terminal moraine " THE WORK OF GLACIERS I7S being reserved for the moraine formed at the time of the greatest ex- tension of the ice. In this discussion the term " terminal " will be used to include both. " The surface of these moraines (Fig. 145, p. 160) is a jumble of elevations and depressions, which vary from low, gentle swells and shallow sags to sharp hills a hundred feet or so in height, and deep, steep-sided hollows. Such tumultuous hills and hum- mocks, set with depressions of all shapes, which are usually without outlet and are often occupied by marshes, ponds, and lakes, surely cannot be the work of running water. The hills are heaps of drift, lodged beneath the ice edge or piled along its front. The basins were left among the tangle of morainic knolls and ridges as the margin of the ice moved back and forth. Some bowl-shaped basins were made by the melting of a mass of ice left behind by the retreating glacier and buried in its debris." (Norton.) Moraines of the Last Great Ice Sheet in North America. These moraines usually occur in belts three to ten or fifteen miles in width, their position being marked in Minnesota, Wisconsin, and other states by thousands of lakes. In regions of little topographic relief moraines may be the most conspicuous features of the landscape. Some of the moraines have been traced several hundreds of miles and if the correlations are correct have been identified over a distance of a thousand miles or more (Fig. 160). FIG. 159. A shows the drainage of a portion of the "driftless area" indicating that it had a normal development. It is in decided contrast to the adjoining glaciated area B. CLELAND GEOL. 12 176 PHYSICAL GEOLOGY At its greatest extent (Fig. 565, p. 646) the ice may have stretched on the east as a great wall from Massachusetts to northern Labrador, discharging icebergs throughout its length. The terminal moraine Forms the backbone of Long Island and stretches across northern Moraines of the Wisconsin Ice Sheet and the limit of maximum glaclation FIG. 1 60. Map showing moraines and direction of ice movement (indicated by arrows) of the last continental ice sheet. The lobate character of the moraines is pro- nounced. The "driftless area" was not covered by ice. (After U. S. Geol. Surv.) New Jersey. From there its direction is northwest to the state of New York and from there to the southwest as far as northern Ken- tucky and almost to the southern tip of Illinois. From Kansas it extends a little west of north into North Dakota, from which point it has a general east and west direction except in the mountainous regions. THE WORK OF GLACIERS 177 (B) Ground Moraine. The till which covers regions between moraines and which constitutes by far the largest area of the gla- ciated surface is called the ground moraine. Its topog- raphy varies widely, being usually rolling and inter- spersed with swamps, but sometimes nearly flat over large areas. The ground mo- raine is of variable thickness, being thinner in Canada (Fig. 161) than in the United States, since the ice in its movement to the south car- ried away much of the ma- terial derived from the under- lying rock, leaving little to be deposited on the melting of the ice. Drumlins, where they occur, are conspicuous fea- tures of the ground moraine. These are smooth, elliptical hills composed of till and have their longest axes parallel to the movement of the ice (Fig. 162). They are not by any means found everywhere in the ground moraine, but FIG. 161. Map showing the effect of glacia- tion in different parts of North America. The area of maximum deposition lies chiefly south of the Great Lakes, where the ice was relatively thin and its erosive power was generally feeble. (Modified after Dryer.) FIG. 162. Drumlins. Wayne County, New York. (Photo. H. L. Fairchild.) are abundant in portions of central and western New York, Wis- consin, eastern Massachusetts, and elsewhere. The islands in Boston Harbor are drumlins. There is still much doubt as to the precise I 7 8 PHYSICAL GEOLOGY mode of their formation, but there is abundant evidence that they were developed beneath the margin of the ice and were built up by the addition of successive layers of till. Stratified Drift. An estimate 1 has been made, based upon a study of the Malaspina Glacier of Alaska (p. 168), that at certain stages of the withdrawal of the great ice sheets the Mississippi River had a volume sixty times greater than at present. When one con- siders the number of streams flowing on, under, and in front of the ice, whose combined volumes made the greater rivers of the time, we can understand the abundance of stratified deposits, such as kames, eskers, deltas (laid down in temporary lakes), outwash plains, and like deposits, so common in the glaciated portions of North America. Outwash Plains. The streams which flowed from the fronts of the continental ice sheets were usually heavily charged with silt, FIG. 163. Outwash plain. New York. (Photo. H. L. Fairchild.) sand, and gravel, which they obtained from the ground and terminal moraines (p. 175), so that they were able to aggrade their beds, often leaving thick deposits. If they flowed through well-defined valleys, the loads were deposited in the valley bottoms and fond valley trains. If a number of streams issued from the ice plain, or in a shallow valley, they gradually raised its deposited their surplus loads. In this way the streams quickly built their beds above the level of the surrounding regions and were in consequence forced to shift their positions to lower places, forming braided streams (p. 86). When valley trains grew to such an extent that they overlapped, an outwash plain resulted (Fig. 163), much as 1 O. D. von Engeln. THE WORK OF GLACIERS 179 FIG. 164. Kettle holes. (U. S. Geol. Surv.) alluvial slopes are formed from the growth of adjacent alluvial fans (p. 124). Outwash plains differ from valley trains in being shorter and wider. Outwash plains are closely associated with terminal moraines. The longer the front of a glacier remained stationary, the more favorable were the conditions for the accumulation of gravel, since the streams then had an abundant supply of debris which they continued to de- posit. The material of outwash plains and valley trains is sand and gravel, the coarser material being nearest the moraine and the finer further away. Outwash plains may be very extensive, and since they are composed of sand and gravel are usually infertile, sometimes to such a degree as to form miniature deserts. The sandy, desert- like plain south of the terminal moraine on Long Island is an outwash plain. Outwash plains and morainic plains may be " pitted " ; that is, the general level may be broken by more or less rounded depressions. These pits are called kettle holes (Fig. 164) and were usually devel- oped from the melting of blocks of ice which had been buried in the drift as the ice retreated. In the outwash plain of the Hidden Glacier in Alaska kettle holes are seen to be forming; and "their development is due to the melting out of ice from beneath the plain, although in no case was the ice actually seen." (Tarr.) Terraces. The valleys of the rivers of Ohio, Indiana, and Illi- nois which carried off the water of the melting ice are now bordered by terraces of stratified drift, and the conspicuous terraces of the Connecticut and Merrimac rivers (p. 127) and their tributaries are remnants of deposits of stratified drift laid down either by over- i8o PHYSICAL GEOLOGY loaded glacial streams or by streams which were overloaded as a result of the erosion of till soon after it was uncovered by the ice. Deltas. When glacial streams entered bodies of water they rapidly built out deltas (Fig. 122). Ancient deltas of this origin often afford the best evidence of the former existence of a glacial lake. In western Massachusetts, for example, a glacial lake of large extent was formed by the damming of the Hoosic River by the ice sheet. Although it did not exist a sufficient length of time to permit the waves to cut back the shores to form cliffs, yet the heavily loaded streams which flowed into it built conspicuous deltas. Eskers. Eskers (Fig. 165) are narrow, usually winding ridges of gravel and sand, ten or more feet wide at the summit and from a few feet to fifty or more feet high, and resembling abandoned railroad grades. They usually follow valleys but sometimes extend across the country with little regard to the topography, even when the hills stand 200 or more feet above the valleys. Single esker ridges in Ireland, Scandinavia, Finland, Maine, and elsewhere have been traced many miles. Usually, however, they are less than a mile in length. Eskers are believed to have been formed beneath glaciers by sub- glacial rivers which flowed in tunnels beneath the ice, and are most FIG. 165. The long, narrow, winding ridge is an esker. It is composed of strati- fied sand and gravel. (Photo. F. B. Taylor.) THE WORK OF GLACIERS 181 abundant where subglacial streams emptied into bodies of water. Under such conditions the outlet of the stream from the glacier would be more readily closed by the delta which formed rapidly where the sediment was dropped as the stream emerged from the ice into the lake, and the tunnel under the ice would thus be gradually filled with sand and gravel. Most eskers were probably formed, for the most part, in connection with the melting of stagnant ice, since it is evi- dent that had the ice been even in slight movement the winding ridges would have been destroyed. Kames. In glaciated regions groups of sand and gravel hills and ridges, as well as isolated, conical hills with high and steep sides, are not uncommon. These hills of strati- fied drift are called kames (Fig. 166). They are often con- fused with eskers and indeed the two are, in individual cases, so closely associated and shade into each other so perfectly that it is difficult to state whether the deposit is a kame or an esker. Kames are composed of stratified sand and gravel, while drumlins ^ ame . s are c mpose or sana /. na rave y na quently are always characterized by rounded slopes, (p. 177) are composed of till. They are often excavated for building and road material, and are favored sites for cemeteries. The same origin cannot be assigned to all the deposits that are classed as kames. '"'Some were formed at the margin of the ice, where the streams issuing from beneath under pressure heaped up their loads against the ice front. Upon the melting of the ice these deposits assumed a more or less irregular surface, depending upon the character of the ice front. Kames of this origin are especially common near terminal moraines, and some of the conspicuous knolls and hills of moraines are often, individually, kames. Isolated kames may have been formed from the deposits of small lakes resting in depressions on the surface of FIG. 166. Kame. North Adams, Massachusetts. Kames are composed of sand and gravel, and conse- 1 82 PHYSICAL GEOLOGY the glacier. As the ice melted these would be inverted to form mounds. Sand and gravel carried into moulins (p. 149) whose sub- glacial passage had been closed would produce such hills. When stagnant ice occupies deep valleys, drainage along the sides may give rise to large deposits of sand and gravel, which may be left in some- what the form of a terrace with a kame topography when the ice has disappeared. Outwash plains and valley trains sometimes begin in kame areas. Relation between Stratified and Unstratified Drift. It should not be understood that stratified and unstratified drift always have topographic forms which distinguish them, or that they can always be clearly separated. The mingling of the unstratified and stratified FIG. 167. Diagram showing the relation between stratified and unstratified drift. In this figure the rough moraine leads to a sandy outwash plain on the left. On the right is stratified drift which was laid down in a temporary pond between the terminal moraine and the retreating ice front. deposits (Fig. 167) is readily comprehended when it is remembered that the edge of the ice sheet probably seldom remained in the same position long, but oscillated back and forth during its advance and retreat. In this way till has been covered with sand and gravel which in turn has been overridden by the ice and covered with till. Moreover, when temporary lakes existed between the ice front and its moraine, stratified deposits were laid down in the midst of the unstratified. EROSION BY CONTINENTAL GLACIERS The amount of erosion formerly ascribed to continental glaciers was probably excessive. There is no doubt that the ice sheets modified the topography over which they passed, in some cases profoundly, but in general the more pronounced features of the landscape were little changed by erosion, although large areas were altered to a greater or less degree by the irregular deposition of drift. Effect on the Underlying Rock. Previous to the appearance of the continental ice sheets the surface of the rock of North America was deeply weathered (p. 651), much as it is now in the southern states. Consequently, when the ice covered and moved over this " rotten " rock (p. 27) and soil, it found an abundance of material THE WORK OF GLACIERS 183 which it carried along for a time and later dropped, either as hetero- geneous, unstratified till, or as stratified sands, clays, and gravels. The rock under- lying the drift is often smoothed and striated (p. 157) (Fig. 1 68), differ- ing from that of the non-glaciated regions in this par- ticular, as well as in the fact that the surface rock is usu- ally fresh and does not pass gradually into soil, as the rot- ten rock has been F IG - 168. A rock surface polished and striated by removed by the glacial action. (Photo. L. .E. Westgate.) glaciers. The scratches and grooves (Fig. 169) on the surfaces of glaciated rocks usually have a common direction (with some variation) and show, as do the glacial bowlders or erratics (p. 156), the direction of the movement of the ice. Harder por- tions of the rock being less easily smoothed by the ice, project slightly above the gen- eral surface and also show by the greater abrasion on one side (stoss) the direction from which the ice The effect of different on came, erosion kinds of rock is not always in proportion to their softness, al- though the softer the rock the more easily it will be worn away. More material may actually be removed from a hard but much-jointed granite by the FIG. 169. Rock ground and polished by glaciers. The excavation on the right is artificial. i8 4 PHYSICAL GEOLOGY " plucking " of the blocks of the rock than is removed from a soft limestone in which joints are poorly developed (p. 157). Under certain conditions glaciers may have little effect on the underlying forma- tions, as is shown by till and even sand and clay deposits (Fig. 170) laid down by an earlier ice sheet, which were but slightly affected when over- ridden by a later ice sheet. In Switzerland glaciers have overridden Alpine landslides without carrying away many blocks. It is possible, however, that such unconsolidated deposits as those just described were frozen when the ice moved over them. Modification in the Shape of the Hills. The shape of the hills in glaciated regions sometimes shows the direction from which the ice came, the side upon which the ice impinged, called the stoss side, having TIG. 170. Till overlying lake a more gentle slope than the other, clays, showing that a lake first ex- tne lee side (Fig. 171). Hummocks isted and that the ice sheet advanced /* i 111 i 11 over the clays without being able f rock eroded b ^ g^Ciers and known to remove them. Williamstown, as roches moutonnees (p. 158) are Massachusetts. well-developed in many places, but may be especially well studied in portions of Canada. Effect of Glaciation on Drainage. In general it can be said that glaciation tended to disturb the preexisting drainage, with the FIG. 171. Diagram showing the effect of glacial erosion. The stoss side suffers more than the lee side, and the slopes are more gentle. The direction of ice movement is shown by the arrow. result that land which in preglacial times was as thoroughly drained, for example, as portions of West Virginia and Kentucky to-day, became swampy, with abundant lakes and ponds. THE WORK OF GLACIERS 185 /. Lakes and Ponds. A glance at any good map of the United States shows that south of the limit of glaciation lakes are almost absent except (i) those formed by rivers in their meanderings (p. 120); (2) those in limestone regions (e.g., Florida) (p. 69); and FIG. 172. Map showing the great abundance of lakes in a portion of a glaciated region. Ashby Quadrangle, Minnesota. (3) those formed by wave and current action along the coast (p. 220). This condition is in decided contrast to that in the glaciated portions (Fig. 172). The depressions in which lakes and ponds occur were FIG. 173. The Fulton chain of lakes in the Adirondacks, New York. First, Second, Third, Fourth, Fifth, Sixth, and Seventh Lakes are evidently the result of the partial filling of a preglacial river channel with glacial drift. 1 86 PHYSICAL GEOLOGY FIG. 174. Lake Marjelen, formed by the damming of a valley by the Aletsch Glacier. formed in several ways. The rock may have been scooped out by glaciers, form- ing rock basins (p. 145). Many such exist in mountainous regions affected by glaciation. River valleys may have been dammed by drift so as to form large lakes, such as Lake Geneva and Lake Constance in Switzerland. The many lakes which add so much to the attractiveness of the Adiron- dacks are the result of the repeated damming of old river courses (Fig. 173). The uneven surface of the drift is often dotted with lakes and ponds which rest in the inequalities formed by the irregularly de- posited material. Basins may be pro- duced by a combination of the above methods. The finger lakes of central and western New York (Cayuga Lake and Seneca Lake), Lake Chelan, Wash- ington, and Lake Como, Italy, are the result of the deepening of old river valleys which lay in the direction of the movement of the ice and of the damming of their outlets. The bodies of water thus formed are long and remarkably deep. Many temporary lakes were formed between the ice front and moraines and also when glaciers moved up a slope, thus preventing the waters from taking their natural course. Glacial Lake Agassiz is such an example (p. 656). Valley' glaciers sometimes dam their tributary valleys, thus forming lakes in them. Lake FIG. 175. Map showing the probable preglacial course of the Genesee River (shaded) and the pres- ent drainage. THE WORK OF GLACIERS 187 Marjelen, Switzerland (Fig. 174), owes its existence to the dam formed by the Aletsch Glacier. 2. Rivers. To form a conception of the effect of glaciation on a well-drained region, imagine a mature country, such as portions of West Virginia (Fig. 95, p. no), invaded by glaciers advancing from the north. It is evident that the north-south valleys would be the ones most likely to be deepened, since they are in the direction of the movement of the ice, and that the east-west valleys would prob- ably be entirely or partially filled with drift, leaving basins which would be occupied by lakes upon the disappearance of the ice. In many cases, the streams would keep their old, wide, preglacial courses for a portion of their length, but in other parts would occupy deep, narrow, rock gorges which they had eroded after they were forced out of their old channels and had cut down through the drift (Fig. 175). Waterfalls and rapids often occur at the points where streams have been diverted from their u , ! , j - r preglacial valley and to cut a new one. (Modified after old channels by drift el f neman } (p. 163). Many of the manufacturing centers of New England and other northern states owe their establishment to the existence of such natural water power. In portions of New York, Ohio (Fig. 176), Michigan, Indiana, Minnesota, and other northern states, and in Canada, the preglacial drainage has been greatly modified by glacial action. In certain areas the streams have new courses; the old valleys are so filled with drift that no evidence of them is to be obtained except by borings. FIG. 176. Map showing the present course of the Ohio River and the course which it had previous to glacial times (shown by arrows). Because of the deposi- tion of drift the Ohio was forced to abandon its wide, 1 88 PHYSICAL GEOLOGY ICEBERGS Formation of Icebergs. On account of the muddiness of the water in front of a glacier which enters the sea (Fig. 177) and also because of the danger from the frag- ments of ice which continually break off from the glacier without warning, it has been impossible to determine definitely how great icebergs are formed. Ice breaking from that portion of the front of a glacier which is above the water produces small bergs, but large ones do not usually have this origin. The two figures (Figs. 178, FIG. 177. Nunatak Glacier, Alaska, entering the water of a fiord and discharging icebergs. (U. S. Geol. Surv.) 179) show two theories of the origin of icebergs. The first theory (Fig. 178) holds that near sea level a glacier is cut back by wave action and melting, leaving a pro- jecting ice foot some distance beneath the surface of the water, which gradually thins toward the end. Great icebergs which suddenly appear from the water some distance from a glacier are believed by the adherents of this theory to have come from the ice foot, from which they had been broken by the buoyancy of the water. The second theory (Fig. 179) holds that the upper part of a glacier which enters the sea will project beyond the lower part, both because of the more rapid movement of the top than the bottom and because of the melting of the ice by the water. In proof of this it is stated that large masses, extending to the very top of the ice front, shear off and sink vertically into the water, disappear for a few seconds, and then reappear, almost to their original height, before they turn over. If the glacier projected under the water to within 300 feet of the surface, it would cause the mass to turn over at once. According to this theory most of the small bergs consist of masses broken from the ice precipices ; larger ones are formed when a piece shears off and sinks into the water; and ice detached under the water may also form bergs. A third theory (Fig. 180) holds that the front of the glacier is broken off by the buoyancy of the water. THE WORK OF GLACIERS 189 FIG. 178. Size and Work of Icebergs. Bergs from Greenland seldom stand 200 feet out of water, and a height of 100 feet is more usual; but icebergs have been reported from the Antarctic which are of great size, being several miles long and 500 feet or more high. Icebergs vary greatly in shape (Figs. 181, 182), those of the Antarctic regions being frequently of a tabular form, while those from Greenland are usually picturesquely irregular. If icebergs were regular in shape and without debris their thickness could be easily determined, since in the case of solid ice the part which appears above the water, is only one ninth of the mass. The .principal geological effects of icebergs are two: they abrade the bottoms of the shallow seas where they strand, and they transport their load of debris until it is dropped as *** the ice melts. Most of the load is lost before it has been carried 100 miles, but some of the debris of Green- land icebergs is deposited on the New- foundland Banks. It is stated that in the Baltic Sea bowlders which have FIG. 180. FIGS. 178-180. Diagrams illus- trating the theories of the forma- tion of icebergs. Fig. 178. Ice- bergs formed by the breaking off been dropped from icebergs are often an d floating of the foot of a glacier found upon vessels which have been as the upper portion is eroded and sunk a few vean; melted back by the waves. Fig. 179. nk a tew years. - Icebergs formed by gravity, since it is held that the upper part of a GLACIAL MOVEMENT glacier will project beyond the lower r . ,.. r . . part, both because of the more rapid There is great difference of opinion movement of the top and because of concerning the mechanics of glacial the melting of the ice. Fig. 180. movement, and the problem may be Icebergs broken from the glacier as i j i i j it enters the sea, by the buoyancy considered as one yet to be solved. of the water (/) Viscosity Theory. One of the early theories held that the motion of glaciers is due to the semiplastic or viscous nature of ice (Forbes), which permits it to move upon a slope very much as do such substances as thick tar or sealing wax, the force which urges it forward being its own weight. Experiments have been performed which appear to show that, in small masses, ice 190 PHYSICAL GEOLOGY will not yield to pressure without breaking, even when the pressure is very slowly applied. If this is true under all conditions ice is not a viscous substance as has been supposed, and the theory fails. (2) Expansion and Contraction. According to this theory a glacier downhill moves as a son 'd body, simply through alternations of temperature. When a mass suffers a rise in temperature it expands, the mo- tion taking place in the direction of least resistance, that is, down the bed. When the temperature falls, contraction will ensue; but since gravity opposes a backward movement a gradual creeping down the bed occurs. The creep of sheet lead on a roof illustrates this action. Since ice is a poor conductor of heat, it is FIG. 181. An iceberg. The vessel gives an idea of the size. 1 FIG. 182. Iceberg, Labrador. The dark bands of debris were probably horizontal in the glacier. (Photo. F. B. Sayre.) evident that such rapid movement as is often observed could not result from this cause. (j) Re gelation. A theory (Tyndall) based upon the fact that broken ice heals under pressure, even at melting temperatures, holds that the movement of glaciers is accomplished by the repeated frac- THE WORK OF GLACIERS 191 turing and later freezing together (regelation) of the surfaces of the fractures when they again come into contact. Under the influence of pressure a glacier is continually yielding to fractures of all sizes, but after changing the position of its parts as a result of the down- ward movement of the broken fragments, it is again united by regela- tion. The effect of this constant rupture and regelation is thought to cause the glacier to behave like a plastic or viscous body. That it is not plastic is indicated by the failure of the Mer de Glace, moving at a rate of only two feet a day, to withstand a change of slope in its bed of even two degrees without fracturing. (4) Melting and Pressure. The lowering by pressure of the melting point of ice forms the basis of another theory. (Thompson.) At the points of greatest pressure in a glacier melting occurs, and the stress is relieved. The water thus formed moves to a point where there is less pressure and immediately freezes. The forward motion of the whole is, therefore, if the theory is correct, effected by a con- tinual process of alternate melting and freezing. (5) Growth of Granules. Since the crystals or granules of glacial ice vary from one seventh of an inch or less to an inch or even four inches in diameter, it has been held that the growth of the granules of the ice is a leading factor in its movement. (6) Other theories of an importance perhaps equal to those pre- sented have been suggested, but none seems to explain all of the ele- ments of the problem. Some of the difficulties have doubtless arisen from the desire to ascribe all of the phenomena of glacial motion to a single cause. The movement of glaciers will undoubtedly be found to be far from simple and to depend upon a number of factors, no one of which is alone competent to produce the characteristic move- ment of large bodies of ice. REFERENCES FOR GLACIERS '. : v - .>. .. . '. :'J :. ' -.'';. V^ *"' ,'"" ^ :r -fr-" EXISTING MOUNTAIN GLACIERS Folios of the U. S. Geol. Surv., in Montana, California, Washington, Oregon, and Wyoming. GILBERT, G. K., Glaciers and Glaciation: Harriman Alaska Expedition, Vol. 3. MARTIN, L., The National Geographic Society Researches in Alaska : Nat. Geog. Mag., Vol. 22, 1911, pp. 537-561. RUSSELL, I. C., Glaciers of North America. RUSSELL, I. C., Malaspina Glacier: Jour. Geol., Vol. I, 1893, pp. 219-245. SALISBURY AND ATWOOD, Topographic Maps: Professional Paper, U. S. Geol. Surv. No. 60. CLELAND GEOL. 13 192 PHYSICAL GEOLOGY TARR, R. S., The Yakutat Bay Region, Alaska : Professional Paper, U. S. Geol. Surv. No. 64, 1909. TYNDALL, J., The Glaciers of the Alps, 1860. PRESENT CONTINENTAL GLACIERS HOBBS, W. H., Characteristics of Existing Glaciers, 1911. NANS EN, F., The First Crossing of Greenland, 1890. PEARY, R. E., Northward over the Great Ice, 1898. SHACKLETON, E. H., The Heart of the Antarctic, 1910. GLACIAL EROSION CHAMBERLIN, T. C., The Rock Scourings of the Great Ice Invasions : Seventh Ann. Rept., U. S. Geol. Surv., 1885, pp. 174-248. CHAMBERLIN AND SALISBURY, Geology, Vol. i, 1906, pp. 281-307. DAVIS, W. M., Glacial Erosion in France, Switzerland, and Norway : Proceedings Boston Soc. Nat. Hist., Vol. 29, 1900, pp. 273-322. DAVIS, W. M., The Sculpture of Mountains by Glaciers : Geographical Essays, 1909. DAVIS, W. M., Die Erkldrende Beschreibung der Landformen, 1912. GEIKIE, J., Earth Sculpture, pp. 212-249. RESULTS OF GLACIATION CHAMBERLIN AND SALISBURY, Geology, Vol. 3, 1906, pp. 327-446. GEIKIE, J., The Great Ice Age. SALISBURY, R. D., Glacial Geology in New Jersey : N. J. Geol. Surv., Vol. 5, 1902. TARR, R. S., The Physical Geography of New York State, 1902, pp. 103-154. WRIGHT, G. F., The Ice Age in North America. WRIGHT, W. B., The Quaternary Ice Age, 1914. DRUMLINS, ESKERS, AND KAMES ALDEN, W. C., Drumlins of Southeastern Wisconsin : Bull. U. S. Geol. Surv. No. 273. 1905- FAIRCHILD, H. L., Drumlins of Central Western New York : Bull. N. Y. State Mu- seum No. in, 1907. GREGORY, J. W., The Relation of Eskers and Kames : Geog. Jour., Vol. 40, 1912, p. 169. GLACIAL MOVEMENT CHAMBERLIN AND SALISBURY, Geology, Vol. i, pp 308-323. OGILVIE, ALAN G., Some Recent Observations and Theories on the Structure and Movement of Glaciers of the Alpine Type : Geog. Jour., Vol. 40, pp. 280-294; (es- pecially for bibliography). PRESTON, T., The Theory of Heat, pp. 279-300; (especially for the theories of regelation, and expansion and contraction). REID, H. F., Mechanics of Glaciers: Jour. Geol., Vol. 4, 1896, pp. 912-928. THE WORK OF GLACIERS 193 TOPOGRAPHIC MAP SHEETS, U. S. GEOLOGICAL SURVEY, ILLUSTRATING GLACIERS AND GLACIAL EROSION Mountain Glaciers Cirques and Glacial Valleys Shasta, California. Chief Mountain, Montana. Chief Mountain, Montana. Philipsburg, Montana. Glacier Peak, Washington. Cloud Peak, Wyoming. Cloud Peak, Wyoming. Hayden Peak, Utah. Mt. Stuart, Washington. Kintla Lakes, Montana. TOPOGRAPHIC MAP SHEETS ILLUSTRATING GLACIAL DEPOSITION Drumlins Moraines Outwash Plains Sun Prairie, Wisconsin. St. Paul, Minnesota. Brooklyn, New York. Boston, Massachusetts. Harlem and Brooklyn, New York. New York City and Weedsport, New York. New York City and Vicinity. Vicinity. Waterloo, Wisconsin. Minnetonka, Minnesota. Elmira, New York. Syracuse, New York. Lake Geneva, Wisconsin. Whitewater, Wisconsin. Northville, South Dakota. CHAPTER VI THE OCEAN AND ITS WORK THE oceans and seas cover about 72 per cent, of the surface of the earth. The average depth of the oceans is about two and one half miles, that of the Pacific being somewhat greater than that of the Atlantic ; the average height of the continents, however, is only a little more than 2000 feet. If all the dry land above sea level were washed into the sea, it would fill only one fortieth of that depression. Soundings to a depth of 32,088 feet have been made in the Pacific Ocean near Mindanao, P. L, and to a depth of nearly 28,000 feet near Japan (Tuscarora Deep). Within 70 miles of Porto Rico the ocean bottom descends to 27,366 feet, and within 10 miles of the Bermuda Islands depths of 17,460 feet are encountered. These great depths are not of wide extent, but are almost as limited as are the great heights of the continents. Moreover, the greatest depth of the oceans is practically the same as the greatest mountain height, each being about six miles. There are, however, few such excessive differences in elevation in short distances on the land as there are differences in depth in the ocean, although Mt. Everest (29,002 feet) is within 60 miles of the nearly sea level Ganges plain, and the vol- cano Fuji in Japan rises 12,400 feet directly from sea level. GENERAL CHARACTER OF THE OCEAN Topography of the Ocean Floor. In order to gain a true concep- tion of the topography of the ocean bottom, it must be borne in mind (1) that stream erosion, which is continually at work on the land and which tends to roughen its surface, is absent on the ocean bottom ; (2) that minor depressions which may exist temporarily tend to be rapidly filled by the sediments brought to the ocean from the land and by the material carried in solution, some of which is precipitated directly and some absorbed by animals to form their shells and skele- tons, only to be left upon the ocean floor after their death. '-Bordering practically all of the shore lines of the oceans is a belt of water which has a depth of less than 600 feet and is from several miles 194 THE OCEAN 'AND ITS WORK 195 to 200 miles wide. This gently sloping sea floor is known as the continental shelf or submarine delta (Fig. 183). The continental shelf is economically of great importance, since the waters lying above it are the great fishing grounds of the world. From its outer edge the sea floor slopes abruptly, so that within a few miles there are depths as great as 6000 feet, while beyond the slope is gentle but gradually increases, until within a distance of from 50 to 100 miles it attains a depth of 12,000 feet. At this depth or lower, the greater part of the ocean bottom is a great monotonous plain, so nearly flat that if the APPALACHIAN FIG. 183. Profile showing the continental surface from the Appalachian Mountains to the deep sea. water were removed, the greater part of it would appear to the eye to be almost perfectly smooth. Irregularities of the Ocean Floor. The irregularities which exist on the ocean bottom are (i) depressions on the continental shelf which are extensions of river valleys (p. 227) ; (2) the steep slope at the outer edge of the continental shelf; (3) volcanic cones, built up from the depths of the sea ; (4) precipices, due to faulting (some in the Mediterranean being 3000 to 5000 feet high) ; (5) well- defined, wavelike ridges, corresponding to mountains on the land ; and (6) broad plateau areas which rise several thousand feet above the deeper portions. Such a plateau extends beneath the Atlantic Ocean from Iceland to a point in the South Atlantic almost opposite the southern extremity of Africa. It reaches the surface in Iceland, the Azores, St. Paul, Ascension, and Tristan de Cunha islands, but, for the most part, lies 6000 feet or more below the surface. Composition of Ocean Water. The water of the oceans contains about three and one half per cent, of mineral matter in solution, more than three fourths of which is common salt (NaCl). Of the total mineral matter in solution, the salts of sodium, magnesium, and cal- cium constitute 97 per cent. Almost every known element has been found by analysis to be dissolved in sea water, and they are all more or 196 PHYSICAL GEOLOGY less radioactive. The composition of the salts which occur in ocean water is as follows 1 : Common salt, NaCl 77-76 Potassium sulphate, K 2 SO 4 . . . 2.46 Magnesium chloride, MgCl . . . 10.88 Magnesium bromide, MgBr 2 . . .22 Magnesium sulphate, MgSO 4 . . 4.74 Calcium carbonate, CaCO 3 . . . .34 Calcium sulphate, CaSO 4 ... 3.60 100.00 In a discussion of the composition of sea water not only the dis- solved mineral matter should be considered, but the dissolved gases as well, since oxygen is essential for the existence of marine organisms and for oxidizing dead matter of organic origin. In addition to oxy- gen, nitrogen and carbon dioxide are present. In fact, the ocean probably contains from eighteen to twenty-seven times as much carbon dioxide as the atmosphere and is the great reservoir of this gas. It is not, however, equally distributed, but is more abundant in polar seas than in equatorial, since cold water has a greater capac- ity for it than warm. Temperature of the Ocean. The temperature of the surface of the ocean varies with the latitude, from a mean annual temperature of 80 F. at the equator to one of 40 F. at the poles. Since the rays of the sun do not penetrate the water to great depths, it is probable that the seasonal changes are not felt below 50 feet. The tempera- ture at the bottom of the ocean is surprisingly cold, being about 29 F. at the poles and 35 F. at the equator. This layer of cold water is very thick; for if we consider water above 40 F. as warm, the layer of warm water is nowhere more than 4800 feet thick, and is usually considerably less. The low temperature of the deep water is due to the movement of the waters from the polar regions, which slowly creep toward the equator along the ocean bottom, so that we find in the tropics, at great depths, the low temperatures which are en- countered only on the surface in the Arctic and Antarctic regions. Exceptions to the rule that the temperature decreases from the sur- face downward are found in such seas as the Mediterranean, the Gulf of Mexico, and the Red Sea. In these seas the temperature of the bottom is approximately the same as that at the bottom of the strait separating them from the ocean, and the surface temperature is almost constant, being practically the average temperature of the surface in winter. In the Mediterranean, for example, the tem- perature at a depth of 6000 feet is 55 F., while in the Atlantic at 1 Data of Geochemistry, Bull. U. S. Geol. Surv. No. 491, p. 23. THE OCEAN AND ITS WORK 197 3000 4000 FIG. 184. Diagram showing the peculiarity of temperature of the Mediterranean. that depth it is 35 F. This ATLANTIC MEDITERRANEAN difference in temperature is due to the failure of the cold waters which slowly move on the ocean bottom from the poles toward the equator, to reach the confined basin of the Mediterranean (Fig. 184). Distribution of Marine Life. -There is little doubt that the marine life of the past ex- isted under conditions similar to those of the present, with the exception, perhaps, that in the early history of the world the great depths were less inhabited than now. In the seas of to-day the greatest abundance of animal life is found in the shallow waters of the continental shelf, where food, supplied both from the sediments brought in by the streams and by the plants that grow there, is plentiful. However, some animals are able to withstand the enormous pressure of the water at great depths, al- though the abundance and variety is small compared with that which flourishes in the shallow waters. When the oceanic telegraphic cables are raised for repairs, marine animals are often found attached to them. Warm waters are more favorable to organisms than cold, although even in the waters bordering the Antarctic Continent the fauna is often varied and plentiful. At and near the surface of the ocean microscopic and other small organisms appear in great numbers, and on the bottom numerous forms of life are frequently found, but in the thousands of feet of water which lie between the bottom and a few hundred feet of the surface of the deep seas there is an almost total absence of life. There is no portion of the land surface on which life is so nearly absent. This is in contrast with the shallow waters, where life is probably much more abundant than on any portion of the dry land. Certain species are restricted to muddy bottoms ; some to sandy; some thrive best in clear, quiet waters out of reach of land sediments ; while others are most abundant where the water is in motion. Plant life is limited to the depth to which light penetrates and is, consequently, scarce in bottoms lying at depths greater than 100 to 200 feet. Since the presence of ammonium carbonate in water 198 PHYSICAL GEOLOGY aids marine organisms in the formation of their calcareous shells and skeletons, and since this compound is most abundant in warm waters, it is probable that when the shells of fossils are thick, the water in which they lived was warm. Thus the existence of thick-shelled, Paleozoic fossils in the rocks of the Arctic region indicates that when they were alive, the waters in that region were much warmer than now. It is evident from the above that in order to understand the life and physical conditions of the remote past a knowledge of the habits and conditions of life of animals now living is necessary. Age of the Ocean. Attempts have been made to estimate the age of the ocean from the quantity of salt dissolved in it. Such estimates are based on the as- sumption that all of the salt of the ocean has been derived from the weathering and erosion of rocks and has been carried to the seas by streams. The simplest form of the problem assumes that the age of the ocean may be determined by dividing the total amount of salt in it by the amount of this mineral carried to the sea each year by streams. The amount of salt in the ocean can be determined with considerable accuracy, since the composition of sea water varies little in different parts of the world, and the approximate total volume of the ocean is known. There are, however, a number of doubtful elements in the problem, (i) The amount of salt discharged by rivers may have varied from time to time. The rate of discharge has undoubtedly been hastened through human agency. The importance of this factor is seen in the fact that 14,500,000 metric tons of common salt are mined or extracted from salt wells yearly. If this is annually returned to the ocean, it is evident that the present rate of discharge is higher than in the past. (2) The salt blown upon the land from the ocean is considerable and must be deducted from the total carried in. (3) The salt received by the decomposition of the rocks by marine erosion (p. 202), and from volcanic ejectamenta must be subtracted. (4) Much salt once in the ocean is now stored within the sedimentary rocks. When the known factors are considered, it is " inferred that the age of the ocean, since the earth assumed its present form, is somewhat less than 100,000,000 years." 1 The amount of calcium carbonate in the oceans cannot be used as a basis for an estimate of their age, since some of it is precipitated upon reaching the salt water, and much of it is used by animals and plants for their skeletons and shells. MOVEMENT OF THE WATER Wave Motion. Since marine erosion is accomplished chiefly by wave action, it is important to know something of the theory of wave motion, of the height and force of waves, and of the depth to which they are effective. Storm waves are set in motion as a result of the friction between the wind and water. The water appears to move forward, just as do the waves in a field of grain which is agitated by the wind. If a pebble is thrown into a pond on a calm day, waves Clarke, F. W., Bull. U. S. Geol. Surv. No. 490, p. 142. THE OCEAN AND ITS WORK 199 are set in motion and any floating object is seen to rise and fall as the crests and troughs of the waves pass under it, but it is not borne along. As each wave glides under the object, it is moved forward a short dis- tance, but as soon as the crest has passed beneath it, it comes back to its former position. In storm waves, however, the friction of the wind drives some of the surface water along and thus produces sur- face currents (p. 217). The height of a wave is the vertical distance between the trough and the crest, and the wave length is the distance from crest to crest. The wave length in heavy storms varies but little from 600 feet, although waves more than twice that length have been observed in the southern ocean. Each particle of water in a wave moves in a vertical orbit (Fig. 185), i.e., if a wave is ten feet in f y h a' b' c' FIG. 185. Diagram illustrating the orbital movement of water in waves. The particles of water move forward in the crests and backward in the troughs, each particle moving in a closed orbit. height the diameter of the orbit is ten feet. In open seas storm waves may be 20 to 30 feet high, and waves of 50 feet have been reported ; it is, perhaps, doubtful if waves exceeding 50 feet in height are ever developed in the open ocean. Waves 10 to 15 feet high are propa- gated at a rate of about 60 miles an hour. Wave motion is propagated indefinitely downward, but rapidly de- creases from the surface to the bottom (Fig. 185), so that at compara- tively shallow depths even sand is not disturbed ; the force of wave motion is one fifth at 65 feet (20 m.), one fiftieth at about 190 feet (50 m.), and perhaps not effective below 230 feet (70 m.). The depth to which agitation extends is in the ratio which the length bears to the height. Thus, a wave 30 feet long and 10 feet high would move the water 6 inches vertically at a depth of 10 feet, whereas a wave of the same height and three times the length would agitate the water 1 8 inches below the bottom of the wave. (Wheeler.) In violent storms it is possible that there is some motion at 3000 feet, but, in 2OO PHYSICAL GEOLOGY general, the mechanical action of the waves is not perceptible at depths greater than 600 feet. This last estimate is based upon the occurrence of ripple marks to be found upon the sand of the ocean bottom. Storm waves sometimes travel great distances, even thousands of miles, preserving their length and velocity, but diminishing in height until they become gentle swells. The Breaking of Waves. As a wave nears a shelving shore its length is decreased and its height increased. The breaking of the wave is the result of friction with the bottom, which retards the lower part, while the crest, continuing with its previous speed, finds itself without support and " breaks." This tumbling crest is called a breaker or roller. Since waves of the same height break in about the same depth of water, a line of breakers is formed. If the ocean bottom descends gently, the water of the breakers rushes upon the shore, and gravity then draws it back down the beach and along the bottom beneath the incoming wave as the un- dertow. On pebbly (shingle) beaches FIG. 186. Diagram showing the j r i_ 11 i i_ directions of the various currents pro- the grinding ot the pebbles as they duced by a wave moving in the direc- are moved forward by the waves and tion AB 9 a shore current BE, and carr i e d back by the undertow can undertow BC, and a reflected wave . , /) be heard, even when the waves are small. When waves strike a coast obliquely, a shore current (p. 217) is produced (Fig. 186), and on coasts where the prevailing direction of the storms is fairly constant, the importance of currents of this origin in transporting sand and gravel is very great. Force of Storm Waves. The force of waves varies with their height, but it is difficult to reduce the force of impact with which a breaking wave strikes a clifF to an exact mathematical calculation. Their strength is, however, influenced by the force of the wind which generates them, by the depth of the water over which they have moved, and by the distance which they have traveled. Experiments at Cherbourg, France, showed that the force of storm waves on that THE OCEAN AND ITS WORK 2OI coast varies from about 600 to 800 pounds a square foot. The force of the impact of waves 10 feet high on certain harbor walls and piers was determined to be 1.36 tons a square foot. The average wave pressure on the coast of Scotland for the five summer months is 61 1 pounds a square foot and for the six winter months is 2086 pounds. At Dunbar in the North Sea the pressure is sometimes three and a half tons a square foot. Height of Storm Waves. On an islet off the coast of Oregon (Tillamook Rock) which is exposed to the sweep of the ocean, the waves of a storm in 1912 dashed against the lighthouse with such force and to such a height as to break the heavy glass of the lantern 132 feet above the sea. During another storm on the same islet a mass of concrete, weighing half a ton, was thrown to a point 88 feet above sea level. During the construction of the breakwater at Plymouth, England, blocks of stone weighing from 7 to 9 tons were removed from the seaward side of the breakwater at low-water level, carried over the top, a distance of 138 feet, and piled upon the inside. During a heavy gale three and three quarters million tons of shingle are estimated to have been taken from Chesil Bank, England, and carried seaward by the waves. The height to which the water of storm waves is thrown is often very great. At Alderney breakwater, in England, the spray from the breaking waves was thrown upward to a height of 200 feet. At Hastings, England, water was thrown as high as the top of a large hotel, and pebbles were lifted from the beach and carried across the wide promenade into the bedroom windows of the houses fronting the sea. It is stated that windows in the Dunnet lighthouse, Scotland, were broken at a height of 300 feet above high-water mark, by stones swept up the cliff by sheets of sea water during heavy gales. Tides. Tides must be considered in a discussion of the work of the ocean, since they are an important, though usually inconspicu- ous agent. Tides are produced by a combination of the attraction of the sun and moon, and of the rotation of the earth ; and every part of the ocean experiences two high and two low tides each day. Al- though the tide in mid-ocean is only about three feet high, its height becomes greatly increased when it approaches shallow shores or enters funnel-shaped bays or estuaries. In the Bay of Fundy, Nova Scotia, for example, the difference in height between low and high tide is sometimes greater than 50 feet. Because of its effect on the level of the water, the tide permits a wide vertical range for the work of waves on shores. Tidal Currents. Tidal races or currents, such as that at Hell Gate, in the City of New York, are not infrequent in narrow straits, and are often effective in erosion. The race at Hell Gate is due to the fact that the tide rises higher in Long Island Sound than in the bay of New York harbor, and to the further circumstance that the time of the high tide is different on the two sides of the strait. The inlets of barrier islands 202 PHYSICAL GEOLOGY (p. 221) and the channels (thoroughfares) back of them are kept open largely by tidal scour, and the deep waterways in some bays are sometimes maintained in the same way. The work accomplished by tidal currents consists more in the transportation of material prepared by the waves than in the actual wear of the coast. The importance of tides to man is considerable. Many of the important harbors of the world could not be entered without tides. This is shown by the fact that ships must wait until the water is deepened by high tide before entering. The washing out of harbors by the tides twice a day is of great sanitary importance. The produc- tion of power from tides has not as yet been financially successful, but the possibility of the use of tidal power in the future in the gen- eration of electrical energy is worthy of mention. Tidal Bores. When a tide enters the mouth of a river which is obstructed by the form of the entrance and by the shallows, its progress may be so retarded that its waters will, for a while, be prevented from passing up the valley. When its height finally becomes great enough, it rushes up in one or more great waves, which are called bores. In the Tsientang River, China, and in the Amazon River, Brazil, waves 20 or more feet in height are said to have been developed at times in this way. Smaller bores occur in other rivers. These waves are characteristic of but few rivers and are not of daily occurrence in any, but in such rivers as those cited they sometimes tear out the banks, destroy forests along the shores, and wash away islands. Earthquake Waves. Because of their great length, waves generated by earth- quakes (p. 292) rise to great heights when they reach shelving shores. Such a wave 10 to 30, or perhaps more, feet in height struck the coast of Japan in 1896, killing 26,975 people, destroying $3,000,000 worth of property, and changing the shore line in many places. Because of their infrequency, earthquake waves are of little impor- tance in marine erosion as compared with storm waves. Ocean Currents. The great currents of the ocean, such as the Gulf Stream, per- form a very slight work of erosion or transportation, but are of vast importance in modifying past and present climates of those regions near which they pass. This is due to the influence of the winds, since they convey the warmth of the poleward cur- rents and the cold of the equator-moving currents to the adjacent lands. MARINE EROSION Factors in Marine Erosion. The impression one receives on seeing a wave strike a rocky shore is that the blow and the weight of the water are the only forces which are important in marine erosion. This, however, is an error, (i) When a wave is dashed against a cliff, every crack and cranny is more or less filled with water, and the hydrostatic pressure exerted tends to force the walls of the fissure apart. This force sometimes amounts to three tons on the square foot ; a force which, often repeated, must accomplish an important THE OCEAN AND ITS WORK 203 work. (2) Moreover, the air in the fissures, even above the reach of the waves, is suddenly compressed and forced into the minute cracks as the waves dash against the cliffs. Upon the withdrawal of a wave the pressure is suddenly released, and the air and water rush out with a suction which, when frequently applied, may loosen and dislodge large blocks of rock. An often-quoted example is that of the Eddystone lighthouse, England, in which a securely fastened door was driven outward as a result of the partial vacuum produced by the withdrawal of a wave during a storm in 1840. Blocks of stone in well-built sea walls are sometimes started from their places, partly at least in this way. (3) The rocks which are broken or quarried from sea cliffs by the impact of the waves and in other ways become tools with which the waves are able to accomplish their greatest work of erosion. As these are lifted by the waves and hurled against the cliffs, they act as hammers which beat to fragments the rocks against which they strike. A high cliff is affected in the same way as a lower one, but is usually cut back more slowly, because as the waves undercut it, the talus (p. 29) falling from above may accumulate in quantities greater than the waves can quickly remove. Under such conditions the energy of the waves may be largely expended in grind- ing to pieces and removing the talus. Sea cliffs, however, weather back more rapidly than cliffs inland, as they are wet with spray and are usually undermined by springs and are comparatively free from talus. When a cliff descends precipitously into deep water the waves merely wash up and down and, having no tools with which to cut, wear it back very slowly. (4) The spray thrown up by the waves also has an erosive effect upon certain rocks, since it washes away the weaker ones and dissolves others which it can affect chemically. In this latter way silicates are broken down and limestones are dissolved. Shore Ice. In high latitudes shore ice protects the shore during the winter months, and even when loosened by the summer thaw it prevents the waves from breaking against the coasts with their full force. Shore ice, nevertheless, is important in the erosion of coasts in regions where it forms. During the winter a broad shelf of ice develops, whose thickness is usually much greater than that which would result from the direct freezing of the sea, which even in polar regions seldom exceeds 8 or 10 feet. The thickness of 30 to 60 feet to which this shore ice or ice foot forms is the result partly of the direct freezing of the ocean water, partly of the accumulation of snow on the ice, which is converted into ice by the water from the waves, and 204 PHYSICAL GEOLOGY partly of the action of storms in heaping up the ice. Shore ice may hold a load of pebbles, both on its upper surface and near the bottom : the former falling on the ice from the cliffs, as a result of the loosen- ing of the rocks by frost ; the latter being obtained from the beach which is frozen to the bottom of the ice. Therefore so far as the position of the debris is concerned, shore ice resembles glaciers (p. 156). During storms this ice is broken into great rafts or floes, and large masses are driven upon the shores by the force of the wind and waves, while in calmer weather they are moved backward and forward by the tides. The stones embedded in the bottom of the ice grind and crush the rocks over which they are pushed, scratching and polish- ing rocky shores very much as glaciers polish and scratch the rocks over which they move (p. 183). As in the case of glaciers, the rock tools which accomplish this work are themselves ground to powder (P- I S9)- It is probable that many of the striations on the rocks of the coast of Labrador, and even on coasts as far south as Newfound- land, were produced by floe ice and not by glaciers. The striae made by the former, however, seldom have a uniform direction. Ice in Lakes. Ice has much the same effect in protecting and erod- ing the shores of lakes as in the seas, but the protection which it af- fords is probably relatively greater, because the waves are usually less effective. The absence of strong shore lines in some glacial lakes (such as those which formerly existed in New York and Massachusetts) FIG. 187. Diagrams showing the effect or ice shove , . , , , in producing "walled lakes." (After Hobbs, Earth which may have been Features.) in existence for a long time, may have been largely due to a protecting fringe of ice which prevented the waves from cutting back the shores. If a lake freezes over completely and is repeatedly subjected to considerable changes in temperature, it may, by the expansion of the water in refreezing, produce a strong " push " on the shores. The expansion of the ice which accomplishes this result is produced as follows. Water freezes at 32 F., and in so doing expands one ninth in volume, but when the temperature of the ice becomes lower than 32 it contracts. This causes the ice to pull away from the shores THE OCEAN AND ITS WORK 205 or to crack. The water which rises in the cracks soon freezes, and when the temperature is again raised, the ice will expand so that the surface will be too large for the lake in an amount equal to the width of the cracks, and will either override the shores or push them up by horizontal pressure. If the shores are marshy they may be ridged or arched up into gentle folds. Such a push may make a ridge or wall about a lake if the shores are of sand and gravel. " Walled lakes " are not uncommon in Canada and in the northern United States (Fig. 187). The ridging may be increased to some extent by the ice driven up by the waves in the spring. RESULTS OF MARINE EROSION The erosive work of the ocean is constant ; in storms the waves strike with great violence, at other times more gently, but always some work is being accomplished. The conspicuous work of the waves is on the cliffs which border the sea. The rapidity with which cliffs are worn back and the sharpness of their profile depend upon a number of factors : (i) the hardness or softness of the rock, (2) the presence of cracks and joints, (3) the position of the beds, (4) the depth of the water, and (5) the height of the waves, the work of the waves being confined to a belt extending a little above high tide and slightly below low tide. If the water at the base of a cliff is deep, the incoming waves do not break. Moreover, since no rock fragments are available for battering the shore, such a wall may endure many centuries with little change. On those portions of the Outer Hebrides where no gravel exists, barnacles are said to be as abundant on the wave-swept cliffs after a storm as before. Since seaweeds often flourish upon the shores where the waves are very active, they are important in protect- ing the rocks upon which they grow. Effect of Erosion on Different Materials. It is evident that soft chalk or glacial drift will be worn back much more rapidly than hard granite. At Cape de la Heve, France, where the chalk cliffs are 300 feet high, the shore is being cut back at a rate of about one yard a year, and the lighthouse stationed there has been twice set back. The annual loss of these cliffs, for a distance of 142 miles, is estimated to be about five and one half million cubic yards. (Wheeler.) So effective is the marine erosion of some chalk cliffs that some of the streams flowing over them are unable to deepen their beds with sufficient rapidity to 206 PHYSICAL GEOLOGY FIG. 188. Chalk cliffs on the coast of France. The waves have cut back the cliffs so rapidly that the streams enter the sea from hanging valleys. keep pace with the wearing back of the cliffs and consequently fall over them from hanging valleys (Fig. 188). The wear on granite cliffs, on the other hand, is often so slight that the battering of the waves for a century is scarcely perceptible. Along the coast of Marble- head, Massachusetts, granite, well within reach of the waves, still bears glacial striae, showing that thou- sands of years of wave wear have not been effective on this hard rock. Since low-lying, sandy shores are apt to lie in places where sand is accumulating, they usually suffer less than rocky and pre- cipitous shores. On such a coast, how- ever, a slight change in the currents such as that due to un- usual or prolonged storms may cause the shores to be cut away rapidly, as has been true of Coney Island, New York, and along the New Jersey coast, where the former sites of houses and hotels are now covered by the sea. FIG. 189. Undercutting of massive granite by wave action. THE OCEAN AND ITS WORK 207 Influence of Joints and Other Planes on Erosion. The profile of a cliff is largely determined by the nature and trend of the divi- sional planes of the rock of which it is composed (Fig. 189), especially of the stratification planes and joints. If stratified rock is not strongly jointed and dips toward the sea (Fig. 190 A], the cliff formed will incline in the same direction. In such a case the wave moves up the slope with little resistance, since an overhanging cliff is absent. When the strata dip gently FIG. 190. A y cliff formed in seaward-dipping strata without strong joints. B, cliff formed in strongly jointed seaward-dipping strata. SI SI A B FIG. 191. A y cliff formed in landward-dipping strata without strong joints. B, cliff formed in strongly jointed, landward-dipping strata. SI X" JL II 1 I PL 11 1 1 SI. 1 1 II __-r-r /I ' ' II I ' /yf 1 1 1 ' 1 1 1 1 /f 1 ' ' I | i 1 I'M II 1 ' A B FIG. 192. A, profile of a cliff formed in horizontal strata without strong joints. By profile of a cliff formed in strata with strong joints. FIGS. 190-192. Diagrams showing the profiles of cliffs formed by wave erosion. towards the sea and a porous stratum rests on an impervious one, landslides may occur, when the porous stratum is undermined. If the strata are inclined towards the land (Fig. 191 A) overhanging cliffs will be formed, since as one layer is worn back another equally overhanging, is exposed. It is on such cliffs that the waves are most effective. If the strata are horizontal the base of the cliff is excavated, but as the upper part is in the form of a stair (Fig. 192 A), the waves have little effect. It should be borne in mind in this connection that if the joints of the rock are better developed than the stratifi- cation planes, the profile of the cliff will depend largely upon their direction, so that CLELAND GEOL. 14 208 PHYSICAL GEOLOGY an overhanging cliff (Figs. 190 B, 193) will be the result of joints inclining inland; a sloping cliff, of joints that incline toward the sea (Fig. 191 E) ; and a vertical cliff, of vertical joints (Fig. 192 B). In overhanging cliffs the dismemberment sometimes begins at the top of the cliff, where the agents are not the waves, but the rain, frost, etc. In such cases the work of the sea consists largely in keeping the base of the cliff free from talus. The height of a sea cliff depends, to some extent, upon the rapidity of marine erosion, since if weathering is more rapid than the work of the ocean, talus will accumulate at its base and protect the shore. In this connection the im- portance of springs and seep- age from underground water FIG. 193. The effect of marine erosion on 111 i i i j strongly jointed beds. Nantucket, Massachu- should not be overlooked, setts. (Photo. S. Powers.) for they often assist in under- mining cliffs. Loose material, such as sand or glacial deposits, will not form cliffs unless the erosion is very rapid. Coves and Headlands. The irregularities which result from marine erosion may in general be classed as headlands and crescent- shaped beaches called coves, and are brought about (i) by the un- equal resistance of the rock, the softer being cut away more rapidly than the harder. Such a condition results when vertical or steeply dipping strata, composed of hard and soft beds, lie at right, Or at considerable ^ IG> I94> Block diagram showing coves formed in . , weak strata, a harder stratum and a lava dike projecting angles to the coast as headlands. (Fig. 194), and a similar shore line is produced when a rock is much more jointed or fractured in one portion than in another. (2) Where the force of the waves is greater on certain parts of shores than on others, THE OCEAN AND ITS WORK 209 coves and headlands may also result. Coves are not cut back indefinitely, since after a time the headlands protect them from the full force of the waves and equilibrium is established. When this condition is attained, the headlands and coves are worn back at an equal rate. It is evident from the above that wave action is able to develop small irregularities of coast line, but not great ones. Sea Caves and Blowholes. Caves (Fig. 195) are often developed on rocky shores where the rock is strong enough to form a roof. If the rock is weak, chasms develop. Such chasms or gullies sometimes extend across narrow headlands, converting the outermost parts into islands. Caves occur at the bases of cliffs and are formed in one of several ways, or by a com- bination of them : (i) by the beating of the waves, espe- cially if the water near the shore is neither too deep nor too shallow and if there is a supply of debris which can be used in the work of ex- cavation ; (2) by quarrying along joints. (3) Since the level of underground water near the coast is sea level, solution caves are not un- common at bases of cliffs in FIG. 195. Sea cave, Watermouth, England. The sea worked along some fault or plane of weak- limestone strata. Such caves ness m tne slate. The enlargement of the cave are often enlarged by the was assisted by the cleavage planes^ (E A. N. ' y Arber, The Coast Scenery of North Devon.) waves. (4) If a weak bed of horizontal rock is at sea level and is subjected to the attack of the waves, it affords especially favorable conditions for excavation by waves. In the development of caves hydrostatic pressure and the compression and expansion of the air are important forces. Fingal's Cave has been thus quarried out of the lava of the south shore of the island of StafFa. It extends inland 200 feet, the floor being below 210 PHYSICAL GEOLOGY FIG. 196. Perce Rock, Gaspe, Canada. (Photo. S. Powers.) sea level and the roof more than 50 feet above. Sea caves are ex- cellent indicators of ancient sea levels (p. 214). Sea caves occasionally extend inland and open to the surface of the ground, sometimes behind headlands one hundred or more feet in height, and at considerable distances from the shore. During quiet weather these openings appear on the surface as deep wells, but during storms the water is sent through them with great force, sometimes throwing spray high into the air, and they are consequently known as blowholes, spouting horns, etc. Blow- holes are sometimes formed simply by the landward extension of sea caves whose bottoms, as well as roofs, usually have a strong upward in- clination inland. They are also formed when, in the land- ward cutting of a cave, a vertical joint is encountered which, when enlarged by hydrostatic pressure and the compression and expansion of air, is drilled to the surface. Arches are not uncommon features on some coasts. They are formed (i) by the uniting of two caves on opposite sides of a head- FIG. 197. A sea arch. When the roof falls the point will become an island. (De Martonne.) THE OCEAN AND ITS WORK 211 FIG. 198. Stacks, Skye, Scotland. land, as is illustrated by Perce Rock in Quebec (Figs. 196, 197)5 <> r (2) by the partial collapse of the roof of a cave. Stacks. Waves sometimes quarry along strong joints, leaving isolated portions of cliffs in the form of chimneys or stacks (Fig. 198). Stacks are also formed by the falling in of the top of a sea arch (Fig. 199). High stacks and chim- neys are most common in hori- zontal or gently inclined beds, where the strike (p. 253) coincides with the general trend of the coast. The Old Man of Hoy on the coast of the Orkney Islands is a well- known example. This is an angular column of red and yellow sand- stone, more than 600 feet high. Many examples of such structures are to be seen on the rocky shores of New England and Nova Scotia, in the Bermuda Islands, and on the shores of Lake Superior. If the rock is resistant, the stacks withstand the battering of the sea for many years, and as the sea cliffs re- treat, may be left be- hind as rocky islets. Marine Terraces. As waves cut back a shore, they develop a submarine terrace (Fig. 200) which ex- . tends from the base tic. 199. The Burgermeister Gate: A in 1864, and r , .. , , B in 1899 after the arch had fallen leaving a stack, of the cliffs and slopes (Drawing after Andersson.) gently seaward unfl 212 PHYSICAL GEOLOGY FIG. 200. Plain of marine denudation, Yorkshire, England. (Photo. J. W. Gregory.) it ends abruptly in deep water. The width of such a terrace depends upon the distance that the waves have cut into the land the wave- cut terrace and the distance to which the terrace has been built out by the material worn from the cliff and carried out to the edge of the rock terrace by the undertow the wave-built terrace (Fig. 201). The depth of the water over the outer edge of the " cut and built " terrace depends upon the size of the waves which prevailingly beat against the shore. In small lakes it is slight, while in larger lakes it may be twenty or more feet in depth. The floor of the North Sea between Great FIG. 201. Section showing the wave-cut terrace CB, Britain and Europe and the wave-built terrace EC, the whole constituting i r i At-]o n tj the wave cut and built terrace. a few miles west of Ireland is believed by some geologists to be a plain of marine denudation. In eastern Patagonia, southern Australia, and other places the sea beats against cliffs from 200 to 1000 feet high, a fact which implies that marine erosion has cut them back tens or perhaps scores of miles. THE OCEAN AND ITS WORK 213 Striking Examples of Marine Erosion. The almost complete destruc- tion by the sea of the village of Dunwich, Eng- land, within historic times, affords an excellent ex- ample of rapid marine erosion under favorable conditions. The village was built upon sand and gravel which formed at the shore a cliff 50 feet high. In the time of Henry II the village is described as " of good note and abounding with much riches," but in Queen FIG. 203. Map of Sharp's Island, Chesa- peake Bay. In 1848 it contained 438 acres and supported a summer re- sort and a number of people throughout the year ; in 1900 (lines) the area was 91 acres, and in 1910 (solid black) 54 acres. If the rate of erosion continues, the island will disappear be- fore 1930. (U. S. Geol. Surv.) FIG. 202. Map showing how rapidly the island of Helgoland has been destroyed by the sea. The length of the shore line at different times is given. This island is used by Germany as a naval base, and its shores are artificially protected. Elizabeth's time it was reduced to one fourth of its former size. Records show that " at one time a monastery, at another several churches, then the old port, then four hundred houses at once, and gradually the jail, the town hall, the high roads, and even the ancient cemeteries, the coffins of which were for sometime exposed in the cliff, were all swept away by the devouring sea." The erosion of the cliff has now ceased, as it is protected by a bank of shingle. The port of Ravenspur, England, where Henry Boling- broke landed in 1399 to depose Richard II, has entirely disappeared, and no one knows exactly where it stood. Portions of the English coast, where the cliffs are from 200 to 250 feet high, have receded at an annual rate of 14 feet. In 1831 a volcanic island called Graham's Island composed of volcanic ash, appeared above the Mediter- ranean Sea near Sicily. After reaching a height of 200 feet above the sea and a diameter of a mile, the volcano became extinct, and so rapidly and thoroughly have the waves worn it down, that not even a shoal remains to indicate its former position. On Cape Charles, Virginia, it has been necessary to build three successive lighthouses on account of the encroach- ment of the sea. The first was built in 1827, 700 feet from the shore line of that time; this was abandoned in 1863, and the whole site has now been washed into the sea. 2I 4 PHYSICAL GEOLOGY The second was built in 1864, also about 700 feet from the shore, but this now stands on the edge of the water and has been abandoned for a new tower still further inland. A remarkable case of marine erosion is exemplified in an island in the North Sea, Helgoland, whose circumference has been reduced from 120 miles in the ninth century to 45 in the fourteenth, 8 miles in the seventeenth, and to an islet only 3 miles in circumfer- ence at present. The remnant has probably survived because of its greater height (170 feet) and because of the somewhat more resistant character of the rock (Figs. 202, 203). Sea-captured Streams. When streams on ap- proaching the seashore turn and run parallel to it for some distance be- fore entering it, they are sometimes cut in two as a result of the more rapid erosion of the coast at some one point (Fig. 204 A, B). Streams which have been recently af- fected in this way enter the sea over falls. Raised Beaches. - Shores that have been raised (Fig. 205) are sometimes marked by sea cliffs, beaches (Fig. 206), sand spits, and bars, un- less the elevation took place so long ago that stream erosion and the FIG. 204 A, B. Block diagrams showing how a stream may be captured by marine erosion. weather have destroyed these. On the coast of Scotland the beaches rise one above another to a height of 100 feet, and the old sea caves are sometimes used as stables. The raised beaches of Norway and Scotland are occupied by villages, and without them the shores would often be deserted. On the coast of THE OCEAN AND ITS WORK 215 FIG. 205. A raised beach showing the coarse bowlders of the old beach and the ancient sea cliff. (Photo. F. B. Sayre.) California well-marked marine terraces are found 1500 feet above sea level. On the coasts of South America and elsewhere the recent elevation of the land is also proved by their presence. Ancient Plains of Marine Denudation. In Labrador and Cali- fornia ancient marine terraces are well marked. The ancient plain of marine denudation on the east coast of India is an unusually fine example of such a plain which has re- sulted from the long- continued action of the waves. The two striking features (Fig. 207) of the plain are (i) the evenness of the surface and (2) the steep-sided hills, the former islands, which rise above it. As the ancient shore is approached, the outliers (p. 106) (ancient islands) are more numerous ; those which are near the old shore are tied to it by old sand and pebble bars. Moreover, sea caves (Fig. 208) are not uncommon in the ancient sea cliffs. Seaward of the marine plain is the coastal plain in which cuestas (p. 225) have been developed. FIG. 206. A plain of marine erosion with ancient islands, east coast of India. (Photo. S. W. Gushing.) 2l6 PHYSICAL GEOLOGY FIG. 207. An ancient plain of marine denudation, with the former islands stand- ing above the plain as hills, is shown in the diagram. The accordant level of the hills indicates an ancient peneplain. (Modified after S. W. Gushing.) The New England Marine Plain. As has been seen (p. 215), the presence of cliffs at the shore line shows that locally marine erosion may become more effective than subaerial (the work of weather, wind, and streams). The sea, however, can work only against the shore, while the effect of stream erosion and of the weather is to reduce the whole surface of the land (p. 114). The work of the sea, though powerful, is limited to the shore line, and plains produced by marine erosion are of small extent compared with the extensive plains carved by the subaerial agencies. It has been suggested, however, that at certain times, planation by the sea may become more effective than usual over much broader areas. After a region has been worn down to such an extent that the soft beds of rock are reduced to base level, leaving the harder as hills, subaerial erosion works very slowly, and the amount of sediment carried to the sea by the streams is so small that the littoral currents expend little energy in moving it. Under such conditions the surface of the land is lowered very slowly, while marine erosion is relatively much more effective. If such a stage is combined with some submergence the sea has an added advantage, and its action is concen- trated on the residual hills and T- o A L r TU uplands remaining from subaerial tic. 208. An ancient shore line. 1 he rim v . 1111- of the cave is 65 feet high. East coast of erosion ' The land bordering the India. (Photo. S. W. Gushing.) Atlantic coast of North America is thought by some to have been under conditions such as these during a long period of time (Cretaceous and Tertiary). 1 The uplands of New England and New Jersey and the resistant ridges of the Appa- 1 Barrell, J., Bull. Geol. Soc. Am., Vol. 24, No. 4, 1913, pp. 688-696. THE OCEAN AND ITS WORK 217 lachian Mountains presented an irregular front to the sea, upon which marine erosion was concentrated. As a result, a plain of marine denudation many miles wide was cut. Upon subsequent oscillatory elevation, with many halts, lower plains were cut. Consequently, in traveling from western Massachusetts to Long Island Sound, instead of a much-dissected, gently sloping peneplain, one finds first the high, rugged moun- tains which were not attacked by the sea; then a deeply dissected, slightly sloping FIG. 209. New England plains of marine denudation, according to Barrell. The dotted lines A, B, C, D are the successive levels of the sea. high plain, now almost completely destroyed ; and, successively, lower plains, better preserved, until the sea is reached. The highest plain if restored would reach an elevation of from 2300 to 2400 feet in western Massachusetts ; and a total of seven originally well-developed plains may be recognized, the lowest at a height of 700 feet. Below this are four plains of fainter development. If this theory is correct, the so-called New England peneplain (p. 114) is really a combination of several surfaces of marine denudation (Fig. 209). TRANSPORTATION Littoral or Shore Currents. The sediment carried into the ocean by streams, as well as that eroded from the shores by waves, is usually soon carried away by currents produced by waves, wind, and tides. When a wave strikes a shore at right angles to its trend, the water thrown upon the shore returns as the undertow (p. 200) and may carry the sediment to great depths. The debris at the foot of the cliffs is, however, not immediately transported to the deep water, but is moved back and forth by the waves and the undertow, and is thus ground finer and finer with time. Since the velocity of the undertow rapidly decreases with the depth of the water, only the finer sand can be carried a considerable distance. Consequently, one usually finds coarse pebbles (shingle) near shore, and progressively finer sediment farther out. When a wave strikes a shore obliquely, a portion of the water returns immediately as undertow (Fig. 1 86, p. 200), and a portion moves along the shore and forms a littoral or shore current. The zone of breaking waves is the road of shore drift, and it often happens that it is the waves produced by storms rather than those of the 2i8 PHYSICAL GEOLOGY prevailing winds which determine the direction of the greatest shore drift. Large particles are not carried far by the shore currents, but the finer sand may be transported many hundreds of miles. Tidal Currents. Tidal currents are often of great importance in the removal of sediment (p. 221). When the tide flows through narrow passages, as between islands, or in V-shaped bays, swift cur- rents are developed which erode and carry away the mud, sand, and gravel which come within their reach. Some tidal currents run so strongly that divers are unable to stand against them. The out- going tide has greater power than the inflowing, since the latter mov- ing in as a great wave fills the bays above their normal level and backs up the water of the rivers, often for long distances. On ac- count of this accumulation of water an outflowing current begins along the bottom before the tide is wholly in, and when the tide changes this adds to the strong current which has already begun. Such strong, outflowing currents tend to keep the channels deep and open, and carry the mud and sand into deeper water. The transporting and erosive powers of the outgoing and incoming tides are, how- ever, sometimes almost equally strong, as was shown by an examination of a steamer which was sunk off the mouth of the Gironde River in France. The vessel rested on her keel in 36 feet of water. At the end of the ebb tide the sands were so scoured as to leave the hull supported only in the middle, but at the end of the flood tide the vessel was again completely covered, the sand beds extending 100 yards fore and aft of the vessel and 50 yards from each side. (Partiot.) Tides not only scour out channels but may also cause the deposi- tion of the sediment which the rivers are carrying to the sea. It often happens that sand flats are formed at the entrances of bays. If a point projects on the side of the river mouth first reached by the incoming tide, the tidal flow may carry the sediment far beyond the mouth of the river ; but if no such point exists, the entrance may be- come more or less choked. FEATURES RESULTING FROM TRANSPORTATION Beaches. When the sea has cut a rock terrace so wide that a strip of sand and gravel is left between the cliff and the sea, a beach is formed. Along coasts exposed to strong waves the breadth of the wave-cut terrace must be much wider before sand is left on it to form a beach than in quiet water, since in the former the sand and gravel may be swept away as fast as formed even when the terrace is several hundred feet wide. Wide beaches are usually first formed within THE OCEAN AND ITS WORK 219 slight recesses of the coast, where the littoral currents deposit their load. Such beaches are crescent-shaped. Near the base of a sea cliff bowlders or coarse gravel will be found, but as one goes from the cliff FIG. 210. A bayhead beach. Conception Bay, Newfoundland. (U. S. Geol. Surv.) the material of the beach is seen to become pebbly and finally to consist of fine sand. The horizontal distance over which a pebble travels before it is ground to sand is very short. Bayhead Beaches. The detritus worn by the waves from the cliffs and from the bottom where the water is shallow, and that brought to the sea by streams is in part carried into deep water, where it is im- mediately deposited and, in part, is swept along the beach by shore or littoral cur- p IG> 2 ii. Lagoon inclosed by a storm ridge. rents. As the waste (Photo. De Martonne.) 220 PHYSICAL GEOLOGY is carried along it does not conform closely to the shore unless the indentations are comparatively slight. When it is swept into a shallow, sheltered bay or cove, it may form a bay head beach (Fig. 210). When such a beach is attacked by storm waves a ridge is sometimes thrown up on the seaward edge, forming a dam behind which a shallow lake or marsh is formed (Fig. 21 1). An interesting fact in connection with these pebble beaches is that sometimes during a single gale an entire ridge may be moved as much as 30 feet. Bars and Spits. When littoral drift reaches an abrupt bend in a shore, as, for example, at the entrance of a bay which extends some distance inland, it does not follow the bend, but usually continues in the direction in which it has been moving. It therefore passes from shal- low water to deep, where it drops its load. Since the portion of the littoral cur- rent that carries the sand is usually narrow, the material dropped into the deep water is gradually built up in the form of an embankment, like a railroad fill, which may, in time, extend entirely across FIG. 212. Map showing an incomplete bar the bay. Currents do not almost shutting the larger lake from the sea, build bars above the level of and a complete bar across the smaller lake. the w but wayes do Delta filling is well shown. (Atwood.) so by washing the sand and gravel from the slopes of the bar to the top. As soon as a portion of the sand is exposed above the water, it may be blown into dunes by the wind. Since dune topography is rough, such sandy stretches often have an uneven surface. Often a bar is never completed (Fig. 212), since the rivers flowing into the bays have sufficient volume and current to keep a channel open. The scouring action of tidal currents may also be able to re- move sediment to deep water as rapidly as it is brought in by the shore currents. Incomplete bars, when built above the surface of the water by waves, are called spits, and when curved by the force of the tidal current at right angles to the drift are called hooks Contoxn- interval 10 feet THE OCEAN AND ITS WORK 221 (Fig. 213). Sometimes the end of a hook is curved so far around as to form a loop. Provincetown harbor, Massachusetts, is an example. Bars are often of great disadvantage to navigation, since they so shallow the water that vessels are compelled to wait until high tide before they can pass over them. In other cases constant dredg- ing, maintained at great expense, is nec- essary to keep a channel open. Spits and hooks sometimes serve as breakwaters and are of consider- able value to ship- ping in time of storm. The effect of the formation of bay- FIG. 213. Hook Bay near the north entrance to head beaches and of Chignik Bay, Alaska. The hook was formed by shore currents. (Atwood, U. S. Geol. Surv.) bars by the shore currents is to shorten the coast line and give it a smoother outline. Sand Reefs or Barrier Beaches. When the water offshore is shallow, the waves drag bottom and build up a ridge of sand or gravel some distance from the shore, which is as high as the storm waves can lift the material. After the surface is reached the height is further increased through the piling up of sand dunes by the wind. Sand reefs or barrier beaches (Fig. 214 A, B) are therefore formed on shelv- ing shores, along a line to which material is brought seaward by the undertow and landward by the drag of the waves. Such sand reefs are separated from the mainland by narrow lagoons, or if they have been in existence for a long time by marshes (p. 223). Sand reefs are approximately parallel to the low shores which they border. They are seldom continuous for many miles (Fig. 215), but are broken by " inlets " which are kept open by tidal scour or by water which is brought into the lagoons by the streams, or by a combination of the two. Inlets occur at intervals of from two to twenty miles on the Atlantic coast of the United States. After a sand reef is formed, it sometimes happens that a second reef is built up in front of it, leaving 'PHYSICAL GEOLOGY FIG. 214. Formation of barrier islands or sand reefs. These are built near the line of breakers, off shallow, sandy shores. The lagoon in 214 B is shown to be nearly filled with sediment and organic matter. a lagoon between it and the first reef. One of the most remarkable barrier beaches is off the coast of Texas and extends without a break for a hundred miles. On the Atlantic coast of North America, from New Jersey south, the barrier beaches are so well-developed that it has been proposed to make a protected waterway by deepening the lagoons back of them. If this is accomplished, vessels will be able to sail from New York to Florida, practically free from storm waves, being protected almost the en- tire distance by sand reefs. The barrier beaches off the coast of New Jersey are especially favored as pleasure re- sorts because of their mild temperature in win- ter and cooling breezes in summer. In some places the barriers are growing and in others they are being washed away. Whether they grow or waste depends upon FIG. 215. Barrier beaches on the coast of Texas. Matagorda Bay has been formed by a barrier beach, and Galveston is situated on one. THE OCEAN AND ITS WORK 223 whether or not the supply of sand is too great for the waves to remove. The lagoons back of sand reefs are gradually filled by the sediment carried in by streams from the mainland, by the sand blown in by the winds, and by the accumulation of marsh vegetation. Back of the sand reef on which Atlantic City, New Jersey, is situated, peat has accumulated to a depth of one or more feet over a wide extent. In time a marsh-filled lagoon will become dry land, and the sand reef will be joined to the mainland. Sand reefs are sometimes hardened by the deposition of lime car- bonate between the sand grains until they form rock reefs. A no- table case of this kind occurs off the coast of Brazil. Tied Islands. Islands are sometimes tied to the mainland by sand and gravel brought by littoral currents. This is accomplished in one of two ways, (i) If littoral currents exist which move parallel to the shore in opposite directions, some- times simultaneously and sometimes successively so that they carry material to the same point, which is gener- ally a strait separating an island from the mainland, a tongue of land consisting of sand and gravel may unite the island to the mainland. (2) Islands are also tied to the mainland (Fig. 216) by the extension of sand spits from either the mainland or the island or from both. Many examples of islands tied to the mainland in one of these ways might be cited. Gibraltar, an island tied to Spain by a narrow sand beach called the " neutral ground," and Nahant, off the coast of Massachusetts, are familiar examples. Examples of the Constructive Work of the Sea. The work of the sea, as we have seen, is constructive as well as destructive. It is stated that on one portion of the coast of England (the estuary of CLELAND GEOL. 15 FIG. 216. Tied island, southern Italy. 224 PHYSICAL GEOLOGY the Humber) about 290 square miles have been added to the coast, while on another (Fens of Lincolnshire), the area of the land has been increased more than 1000 square miles. It is stated that for every square mile washed away from portions of this coast, three square miles have been added on others. Moreover the sea-built land is, on the whole, richer than that which was destroyed. A tele- graph pole erected at a point on the English coast in 1873 was 3 feet inland in 1902. At Atlantic City, New Jersey, portions of the sand reefs are being built out while others are retreating. Hotels have had to be moved forward so as to be kept near the sea. The history of the town of Rye, England, is instructive as showing that the land may be attacked by the sea at one time and later be increased at the same point and by the same agent. This town was once de- stroyed by the sea, but the site is now two miles inland. SHORES The shores of the oceans may, in general, be classed topographi- cally as smooth or rough, or according to origin as those resulting from elevation or from submergence. To understand the configura- tion of a shore one must keep in mind (i) that the effect of deposition on the ocean bottoms is to smooth out all inequalities and to produce a monotonous plain which slopes gently from the beach to the edge of the continental shelf, and (2) that the effect of erosion on high land is first to roughen it. Smooth Shores. When a sea bottom on which sediment has long been accumulating is raised to form land, a smooth, approximately flat plain will be exposed. The low, level plain of Yucatan, which slopes beneath the water so gently that vessels cannot approach in safety nearer than three miles from the coast, so that all freight must be taken to land in shallow boats, is a good example. The land bordering the Atlantic and Gulf coasts of the United States south of New York is a somewhat broken, level plain, through which streams flow sluggishly to the sea. The underlying strata dip gently towards the sea and are composed of unconsolidated sands and clays containing marine shells. This plain varies in width from a fraction of a mile to 500 miles, extending from the Fall Line on the west to the shore on the east. The Fall Line marks the boundary between the new, unconsolidated sands and clays of the Coastal Plain and the harder, ancient rocks of THE OCEAN AND ITS WORK 225 the Piedmont Plateau (p. 91). The name indicates that the streams flow over falls or rapids where they pass from the hard rocks of the old land to the easily eroded sediments of the Coastal Plain. The greatest coastal plain in the world forms the north and west parts of Siberia and has a maximum width of more than 1000 miles. The plain is low and poorly drained. If England, eastern Europe, and the intervening sea floors were raised 300 feet, England would be united to the mainland, the Baltic would be changed to a chain of lakes, and the North Sea would be reduced to a gulf. If this should happen, the ancient shores could be readily determined by the elevated sea cliffs, sea beaches, wave-cut terraces, and sand spits ; while the raised sea bottoms would consti- tute coastal plains. The new shores would be smooth with few in- dentations. Cuestas. The material of land newly raised from the sea has a dip seaward, due both to the original inclination of the sediments and also to that which was brought about during the process of uplift. If the beds of recently raised coastal plains differ somewhat in resistance, the streams will in time give a zonal character to the topog- raphy, the harder beds standing higher than the softer ones. There will thus result FIG. 217. Diagrams A and B illustrate the development of cuestas. As the weak stratum of the coastal plain was cut away more rapidly than the firm, the latter formed rather steep slopes facing inward, and long, gentle slopes towards the coast. 226 PHYSICAL GEOLOGY alternating bands of lowland and highland, the lowland being bordered on the sea- ward side by infacing cliffs formed by the harder beds. The low ridges thus developed have a steep descent on one side and a gentle slope on the other and are called cuestas. Examples of coastal plains with this banded arrangement are not uncommon. In Alabama the Appalachian Mountains are bordered by the " Black Prairie," a belt of lowland formed in easily eroded limestone. Next to this is a ridge (cuesta) which ascends rather abruptly 200 feet above the lowland, composed of more resistant lime- stone (Fig. 217 A y B). The geological structure of the Ghats in India (Fig. 207. p. 216) shows the formation and characteristics of such topography. Very ancient coastal plains with resulting cuestas constitute a large part of New York, Ohio, and other states. Rough Shores. By marine erosion a shore may be slightly rough- ened, but it is not possible for waves unaided to make irregular shores like those of the coast of Maine, Nova Scotia, Wash- ington, northern Europe, British Columbia, and the coast of the Adriatic. Such shores are formed by the sinking of the land or the raising of the sea level. When a region is partially sub- merged the higher hills be- come islands or peninsulas, and the valleys become estu- aries or bays. Consequently, rugged coasts bordered by high, rocky islands (Fig. 218), are evidences of subsidence. An interesting example is to be found in northeastern North America, where the coast line between New Brunswick and Portland, Maine, is 2000 miles long, although a straight line between the points is only 200 miles in length. Another characteristic of a sunken coast is the existence of sub- marine valleys. On the coasts of Europe and North America sound- ings have shown that the valleys of rivers extend far out on ancient coastal plains (Fig. 219), now the sea bottom. The Hudson River FIG. 218. Portions of the coast of Maine, showing the effect of subsidence. The valleys have become bays, and the hills peninsulas and islands. THE OCEAN AND ITS WORK 227 (Fig. 220), for example, formerly extended across the continental shelf into the deep sea, as is shown by its deep, submarine channel. The St. Lawrence, Potomac, and other rivers also have submarine channels. The bays of the Coastal Plain are the result of a slight subsidence after the newly made land had been cut up to some extent by streams, and are consequently merely drowned valleys. Bays are sometimes formed by the ele- vation of the sea bottom on SCALE OF MILES./'! 1*0 20 30 40 50 FIG. 219. Chesapeake and Delaware bays are drowned river valleys, the ancient submerged channels of which can be traced out to sea. (After Dryer.) FIG. 220. Map showing the sub- merged channel of the Hudson River. This channel can be traced about 125 miles beyond the present mouth of the river. (After Dryer.) one or two sides of an area in which there was no such movement. The Gulf of California had such an origin. Bays are made also by the settling of great blocks (fault blocks, p. 267), as is true on the coast of the Red Sea. Examples of Irregular Coasts. The character of irregular coasts depends upon several factors. Coasts of Folded Regions. If the region is folded, with the axes of the folds parallel to the coast, the bays and islands will have a like direction. A typical example of such a coast is to be found on the northeast shore of the Adriatic Sea (Fig. 221), with its elongated islands, its constricted straits, and narrow bays; all of which are parallel to the coast. When the folds are perpendicular to the shore, a rugged coast with projecting points and deep indentations results (Finisterre, Spain). 228 PHYSICAL GEOLOGY Fiord Coasts. The coasts of high, glaciated regions are characterized by narrow, branching bays of great depth (p. 226, and Fig. 150), with precipitous, almost vertical sides, called fiords (p. 166). It has been shown (p. 167) that fiords are valleys greatly deepened by glacial erosion, which have prob- ably been drowned by a sinking of the land. Fiords occur only in high latitudes. Deeply Indented Coasts in Non-glaciated Regions. In northwestern Spain, Brittany, Ireland, and elsewhere are fiord-like coasts which, however, have not suffered from glacial erosion. These funnel-shaped bays were produced by the drowning of deep valleys formed by stream erosion. Such in- dentations differ from fiords, not only in their origin, but also in their V-shaped cross section and in the fact that they gradu- ally deepen seaward, while the deepest portions of fiords are some distance inland. Coasts of Slightly Sub- merged Coastal Plains. The coastal plains of the United States have been described (p. 224) as re- cently raised portions of the ocean bottom. After having been cut up to some extent by erosion a slight submergence occurred, which drowned the valleys, thus forming Chesapeake Bay, Delaware Bay, etc. Proofs of Elevation and Depression. Although the coast of Maine is quite typically that of a region of submergence, there is evidence that considerable elevation followed the period of greatest sinking. This evidence is to be seen in the marine clays which are found above the present sea level, as well as in the abandoned shore lines which are now far above the tide. FIG. 221. Portion of the east coast of the Adriatic. The folds of the rock largely determine the direction of the straits, islands, and peninsulas. THE OCEAN AND ITS WORK 229 The Island of Capri, off the coast of Italy, offers an unusual ex- ample of submergence within historic times (Fig. 222). In ancient times a sea cave, now known as the Blue Grotto, was used by the Romans as a resort from the oppressive heat of certain seasons. In order to obtain light an opening was cut in the roof. Since that time the land has sunk so that even the artificial opening is now partly submerged. The blue color of the grotto is due to the refraction of the sun's ,-, _ . f , *IG. 222. Section of the Blue Grotto, Island of rays in the water, by Capri, showing proof of subsidence. (Modified after means of which the Von Knebel.) red rays are lost. In some of the caves of the Bermuda Islands (Fig. 223), stalactites hang from the roof and extend into the sea water which partially fills the caves. Stalactites obviously could not have been formed in water and therefore prove the former greater elevation of the island. The temple of Jupi- ter Serapis at Poz- zuoli, near Naples, proves that the coast has suffered, first an elevation, then a de- pression, and finally a reelevation almost to its former level. The evidence is to be found in three col- umns of the temple, whose surfaces have been roughened for a height of from 12 to 21 feet above the base by boring mollusks (Lithodomus) which live only in sea water. The temple was, of course, built on land. It was then submerged by the sinking of the coast, so that the columns were immersed in the sea to a height FIG. 223. Diagrammatic section through a cave in the Bermuda Islands, showing one proof that subsidence has taken place. Stalagmites and stalactites are formed only in the air. 230 PHYSICAL GEOLOGY of 21 feet above the base. At this time the Lithodomi bored into the stone and made their homes there. The lower 12 feet of the columns were buried in sedi- ment and therefore escaped damage. Later the land was again raised, and the columns are now some distance from the shore (Fig. 224). Such evidence, although interest- ing as showing recent changes in sea level, is of minor im- portance as compared with the occurrence of, strata con- taining marine shells at heights of 14,000 feet or more above the sea. The Stability of the Atlantic Coast of North America. The statement often made that the coasts of Nova Scotia, New England, and New Jer- FIG. 224. Three columns of the temple or , , , , , T . o AT i TL j iT j sey have recently undergone a gradual Jupiter Serapis near Naples. Ihe dark and , .. . . . . rough band above the figure is the portion ^^dence and that this movement which was perforated by boring mollusks. The 1S stlU m progress, rests upon the lower portion of the columns was protected by following evidence. 1 Stumps of mud and the upper portion projected above trees are found in salt marshes ; salt the sea. water is found overlying fresh-water peat ; marshes have increased in size ; dikes erected to keep out the tide are themselves covered at high tide; a bench mark at Boston is now three-fourths of a foot nearer the mean level of the sea than when it was placed there three quarters of a century ago. When each case is care- fully studied it is found either that the apparent sinking is due to local causes, or that no definite conclusions can be drawn. The evidence from marshes is especially uncertain, because when drained they settle ; when sand dunes encroach upon them, they are compacted and their surface is consequently lowered ; when a bar behind which fresh-water marshes and forests exist is cut through by waves (Fig. 225), the marsh will be invaded by sea water, the trees will be killed, and salt-water peat may in time cover the fresh-water peat. Changes in the direction or velocity of ocean currents may also bring about local differences in sea level. The apparent lowering of the bench mark near Boston is doubtless due to the narrowing of the bay as a result of the artificial filling in of the marshes. Such a constriction of the channel would 1 For a more complete statement see : D. W. Johnson, Science, Vol. 32, 1910, pp. 721-723, and Fixite de la Cote Atlantique de VAmerique du Nord, Annales de Geographic, Vol. 21, 1912, pp. 195-212. THE OCEAN AND ITS WORK 231 i rlG. 257. block diagram showing the appearance the Dead bea ot Palestine and origin of a graben and horst> is a graben in which a fault block has sunk 2600 feet below the level of the plateau, depressing it in places below the level of the sea. Reverse or Thrust Faults. In reverse faults the hanging wall should be considered as having moved up, and we thus find that instead of a stratum being separated as a result of the faulting, the two ends overlap so that the older beds are pushed over the A B C FIG. 258. A fold passing into a thrust fault. (Heim.) 264 PHYSICAL GEOLOGY younger ones (Fig. 258 C). The result of thrust faulting in the Selkirks of Canada and in many other regions has been to move strata over and upon others that were formerly hundreds of feet higher. Occasion- ally, recumbent folds can be traced into thrust faults (Fig. 258 A, B). This is not surprising, since it is evident that both are due to lateral pressure ; the movement being so great in the latter that the strain could not be relieved without breaking. Thrust or reverse faults involve a shortening of the earth's crust, as has been said, and differ in this respect from normal faults which are the result of a stretching of the crust. The fault plane usually approaches the horizontal more nearly in thrust faults than in normal faults; that is, the hade of the former is greater than that of the latter. Examples of Thrust Faults. Many examples of thrust faults might be cited. A great thrust fault (Bannock overthrust) l extends approximately 270 miles from northern Utah into Idaho, in which the older strata have slid over the younger (horizontal displacement) FIG. 259. Thrust faults and folds in the southern Appalachians. a distance of at least 12 miles. In the same region other faults with a throw of 15,000 to 20,000 feet have been described. In Massa- chusetts, 2 a thrust fault has been described in which the strata have slid along a fault surface for 15 miles. The southern Appalachian Mountains (Fig. 259) are broken by many faults that run parallel to the system. In Virginia and Georgia 15 or more parallel thrust faults occur, running from northeast to southwest, along which the older strata have been pushed over the younger. One of these faults has been traced for 375 miles, and its greatest horizontal displacement is at least n miles. Vertical and Horizontal Faults. * If faulting has taken place along a vertical (90) joint or other fracture, the arrangement of the strata may give the appearance at the surface of either a normal or a reverse fault, depending upon whether the right or left side moved down (Fig. 260). In many cases, along the same side of a given fault line the movement may have been upward in one place, downward in another, and without evident movement at another. In some cases, a fault occurs along bedding planes and is called a 1 Richards and Mansfield, Jour. Geol., Vol. 20, 1912, pp. 681-709. 2 J. Barrell. THE STRUCTURE OF THE EARTH 265 bedding fault. It should not be understood from the foregoing that the movement in faults is always merely up or down. Often the movement has both a vertical and a horizontal component, and occa- FIG. 260. Diagrams showing A, horizontal strata ; B, the strata displaced by a vertical fault; and C, the fault scarp obliterated by erosion. sionally the vertical movement is inconsiderable and the horizontal important. This was true of the fault that caused the San Francisco earthquake (p. 275), and it has been observed in mines, where the A B C FIG. 261. Diagram illustrating a dip fault and the effect of erosion upon the outcrop. faulted surfaces can be studied, that evidences of horizontal move- ment are more often met with than those of vertical. Influence of Faults on Topography. If faulting did not take place long ago, it is evident that a cliff or fault scarp will be present, the B FIG. 262. Diagrams showing the effect of an oblique fault upon dipping beds, and the outcropping of the stratum on the surface after the fault scarp had been planed by erosion. 266 PHYSICAL GEOLOGY height and prominence of which will depend upon the amount of the faulting and its recency (Figs. 261, 262, 263). Fault scarps are also formed when weak and resistant rocks are brought into contact FIG. 263. The effect of faulting on the outcrops of anticlinal and synclinal beds before and after the erosion of the fault scarp. by faulting. The weak beds are worn away more rapidly than the hard, so that after prolonged erosion the latter may form cliffs, even though they are on the downthrow side (Fig. 264). The original fault scarp is usually soon eroded to such an extent that the con- A.__ B figuration of the land gives little or no indication of the existence of a fault. When, however, a resistant bed such as a sill of lava (p. 326) or a quartzite stratum is present ?'' ~ and is exposed by erosion, the I ' L fault will again be the indirect FIG. 264. Diagram showing a cliff BAG cause of a cliff (Fig. ^264). In formed by faulting. As erosion wore away the Scotland the relatively soft weaker rock a cliff resulted from the presence rQcks of the centra l lowlands of the strong bed CD. Upon further erosion . . . . the bed GF disappeared and the cliff F was have been brought up against formed. the relatively hard rocks of THE STRUCTURE OF THE EARTH 267 the highlands, producing a strong line of demarcation (Fig. 265). Such a topographic form is called a fault-line scarp or cliff. Sometimes the topography of a region is determined largely by faulting, resulting in steplike hills or mountains. This is well shown in a portion of Oregon (Fig. 266), in the Great Basin region of FIG. 265. Fault in the Scottish Highlands. The strata on the right are weaker than those on the left. Utah (p. 354), the Colorado Plateau, and the Connecticut valley of Massachusetts and Connecticut. In central Sweden the criss- cross valleys and lakes of angular and zigzag outline are, either directly or indirectly, due to the faulting of rhomboidal-shaped blocks. The greatest example of faulting now apparent in the topography of the continents is the Great Rift valley of east Africa, which consists FIG. 266. Diagram of a mountain formed by faulting. The slope of the surface of the original block is shown in outline. (Modified after Davis.) of a series of grabens (p. 263) in the bottoms of which lakes (Albert, Tanganyika, etc.) are aligned. The magnitude of the displacement is indicated by the depth of Lake Tanganyika, which is nearly 4190 feet, the bottom of the lake lying 1600 feet below sea level. This great rift extends from Abyssinia southward for some 1500 miles. In the Adirondacks of New York the " latticed drainage " is to a 268 PHYSICAL GEOLOGY great extent produced by faulting which caused the streams to flow in fault valleys. Minor Features of a Fault Fracture. When a fault fracture is visible, it is often found that it is represented by a zone of angular rock fragments, often cemented together to form a fault or crush breccia (Fig. 357), sometimes several yards wide. Some important gold and silver deposits occur in the filling of such breccia (p. 371). When the breccia is very resistant, as is the case when the filling is quartz, it may stand in relief after the surrounding strata are denuded, resembling a dike. The side of a fault surface is often polished and striated by the move- ment of the walls, so that it is possible to tell the direction of the movement from the striations. Such a surface is called a slickenside (Fig. 267). It resembles a glaciated surface but is usually more glazed. Detection of Faults. Topography, as has been seen, is not al- ways a safe guide for the detection of faults, since their presence is not always indicated by cliffs. The most satisfactory evi- dence of a fault is to be ob- tained when a geological map of a region is made, and it is found that all of the neigh- boring formations end upon a more or less straight line. A sudden change in the soil which is made apparent in the degree of fertility of neighboring fields, and the pres- ence of rapids in streams which cross a fault are also indicative of the presence of a dislocation. Springs often occur along fault lines, and when a number of them are aligned they indicate, but do not prove FIG. 267. A slickenside surface formed by lateral movement along a fault plane. The rock of one side of the fault was removed. (U. S. Geol. Surv.) FIG. 268. Diagram showing "drag dip" near a fault. (Modified after W. N. Rice.) THE STRUCTURE OF THE EARTH 269 the presence of a fault. Dikes of lava also sometimes occur in fault fractures. In regions of horizontal rocks faults can often be traced along the surface some distance from the fault line (p. 277) by the upturned edges of the strata (drag dip), produced by the friction along the fault surface of the edges of the strata during the faulting (Fig. 268). Origin of Faults. In seeking an explanation for normal faults the fractures along which the movements occurred must first be found. These are often joints which, as has been seen, have probably been formed by tortional strains (p. 259). Normal faults indicate a stretched condition of the crust, which permitted the unequal settling of blocks bounded by joints or other fractures. In the formation of grabens and horsts (Fig. 257, p. 263) there appears to have been first a compression which arched the strata, followed by a relief of the pressure, which ^^^^^ permitted the set- tling of the blocks of the arch. Reverse faults are evidently FIG. 269. Diagram showing a fault shading into a fold. the result of lateral (After De Martonne -) pressure and are best developed in highly folded and distorted rocks. Thrust faults often pass horizontally into folds, and verti- cally they sometimes shade gradually into more or less gentle folds (Fig. 269). The origin of the force which produced folding will be taken up more fully under the discussion of the formation of mountains. Rapidity of Fault Movements. At irregular intervals dislocations of from a fraction of an inch to 20 or more feet have taken place in a few minutes along fault planes. In Owen's valley, California, in 1872, a slipping occurred along a line 40 miles long which resulted in a throw of from 5 to 20 feet. Along a fault 50 miles in length a dis- placement of as much as 30 feet occurred in Japan in 1891. Such faulting always produces earthquakes (p. 281). In some mines faulting is observed to be taking place continually, but at a slow rate. The total displacement resulting from such movements, if long-con- tinued, will necessarily be great, but no surface features may mark its presence, since erosion may cut away the upthrow side as fast as it is formed. 270 PHYSICAL GEOLOGY CONFORMITY AND UNCONFORMITY When strata are deposited one upon another in unbroken succes- sion and without disturbance, they are said to be conformable (Fig. I, p. 23). When, however, one set of beds is deposited on another which has been above the sea and there eroded, the two beds are said to be unconformable, and an unconformity, represented by the erosion sur- face which separates the strata, is said to exist. Unconformities may exist between stratified and igneous or metamorphic rocks (Fig. 322, FIG. 270. A diagram showing an unconformity or erosion interval, AB. p. 326). Usually, an unconformity is marked by some change in the relative dip of the beds, one set resting upon the upturned edges of an older series (Fig. 270). Such an unconformity is spoken of as an angular unconformity, because the strata below the uncon- formity meet those above it at an angle. An unconformity, however, sometimes occurs between strata which have not suffered any rela- tive change in dip (Fig. 271) ; i.e., both the younger and the older FIG. 271. Diagram showing unconformities. The stratum C is separated from the strata B and D by unconformities. The unconformity between B and C is not readily noticeable, because the strata of both are horizontal. An unconformity exists also between the glacial drift A and the limestone B. (Near Milwaukee, Wisconsin.) strata may be horizontal. In such cases the unconformity can usually be recognized by old erosion channels or by basal conglom- erates (p. 240), but occasionally such a break is difficult to detect. Importance of Unconformities. Unconformities are of much importance in the study of the geology of a region, because of the geological history which they reveal. From the unconformity THE STRUCTURE OF THE EARTH 271 in Fig. 272 we learn (i) that there was a long period of quiet, during which the lower series of beds were deposited continuously on the ocean bottom. This was followed (2) by a period of folding or tilting and of elevation, so that these beds were raised above the level of the sea. (3) The strata were then subjected to erosion for such a long period that the upturned edges were worn to a peneplain. FIG. 272. An unconformity in which the lower beds are inclined, while the upper are horizontal. Wyoming. (U. S. Geol. Surv.) How long this erosion interval lasted cannot be told from any one geological section. (4) The land was later depressed and became sea bottom, so that (5) sediment was laid down on the old land sur- face. (6) Reelevation again converted the sea bottom into land, and streams are now at work carrying the rock back to the sci. Overlap is a term used in describing an unconformity in which the younger strata cover a larger area than the older ones and conse- FIG. 273. Diagram showing the overlapping of the strata AEDC due to the encroach- ment of the sea upon the ancient land surface AC. (Modified after Grabau.) CLELAND GEOL. 1 8 272 PHYSICAL GEOLOGY quently overlap the older strata (Fig. 273). They were deposited when the water in which they were laid down had a greater extent than it had when the older strata were deposited. CONSTITUTION OF THE EARTH'S INTERIOR The ancient Greeks and Romans in speculating about the under- world came to the conclusion that its heat and other igneous phe- nomena were due to the work of imprisoned giants. Our present theories are less fanciful, but because of the inaccessibility of the deeper portions of the earth's interior and our failure to reproduce the conditions there, our knowledge of its constitution is far from satisfactory. Zone of Variable Temperature. The temperature at the surface of the earth is variable because of the changes in daily and seasonal temperature, but at a depth which in Java and India is about 12 feet and in New York about 50 feet the temperature is constant through- out the year. The Interior Heat of the Earth. Below the level where seasonal change in temperature occurs, the temperature increases with the depth. This fact has been determined by well borings, by tunnels, and by mines, but since the deeper borings are only a little more than a mile in length, the statements in regard to the rate of increase at great depths must necessarily be theoretical. The rate of increase is usually estimated at i F. for 60 to 75 feet, but it varies so widely at different places that to strike an average is difficult. In the St. Gothard tunnel (Italy to Switzerland) it is i F. for 82 feet ; at Calumet, Michigan, the average for 4939 feet is i F. for 103 feet ; in the British Isles it varies from i F. for every 34 feet to i F. for every 130 feet. A boring in West Virginia to a depth of 5386 feet showed an increase of one degree for 80 or 90 feet for the upper half, and of one degree for 60 feet in the lower half. Near Leipzig, Germany, a boring 5560 feet deep showed an average of one degree for 56 feet. In South Dakota the artesian wells show a depth of from 17.5 to 45 feet for each degree. The rapid increase shown in South Dakota, however, is probably due to folding in the water-bearing stratum. The presence of hot igneous bodies would also increase the tempera- ture gradient. Since the temperature gradient follows the surface configuration, the level of the Simplon tunnel which connects Switzerland and THE STRUCTURE OF THE EARTH 273 Italy was made higher than the grade of the railroad required in order that a too great temperature might be avoided, but even with this precaution the heat in some portions was so intense as almost to stop the work of excavation. The heat of the interior of the earth is known also from lavas which reach the surface through volcanoes and fissures (p. 299). Copper wire, which melts at about 2200 F.,has been fused when thrust into molten lava. In one case, where a lava stream from Vesuvius overflowed a village, brass was decomposed into its component metals. It is estimated that the initial temperature of lava, when it issues from Vesuvius, is probably more than 2000. THEORIES OF THE PHYSICAL STATE OF THE EARTH'S INTERIOR If the temperature of the earth increased uniformly from the surface downward at a rate of one degree for every 60 feet of descent, a temperature of 3000 degrees would be reached at a depth of about 34 miles. Such a temperature would be sufficient to melt all but the most infusible rocks under the conditions existing at the surface of the earthy but since the rocks at this depth are subjected to the enormous pressure of the overlying rocks, the conditions are very different. (i). Internal Fluidity Theory. Based on the assumption that the heat increases regularly from the surface downward, it has been held that the earth is a molten globe covered with a thin crust 25 to 30 miles thick. Although the belief in a molten in- terior has a wide popular acceptance, there are a number of serious objections to it. (a) Effect of Increasing Density. The rate of increase in temperature is probably not uniform, but diminishes with the depth. This is evident from the fact that the average specific gravity of the rocks of the earth's surface is about 2.8, while ^hat of the earth as a whole is 5.5, so that the specific gravity of the central portion must be at least equal to iron, which is 7.7, and is probably higher. This increase in density is due to some extent at least to the great pressure of the overlying rocks, but may also be due to the concentration of heavy metals within the center of the earth. It has been suggested that metallic iron is to be found in greater quantity there than on the surface. But whatever the cause, the effect of increasing density would be to increase the conductivity of the rock, so that instead of a uniform increase of one degree for each 50 or 60 feet of descent, at a greater depth 70 feet might be required for a change of one degree, then 80 feet, then 100 feet, and so on. The increasing con- ductivity of the rock alone would, consequently, carry the temperature necessary to fuse rocks much deeper than 30 miles.. (b) Effect of Pressure. Another serious objection to the theory of a molten in- terior is to be found in the effect of pressure on the melting point of rocks. Since nearly all substances expand upon melting, they remain solid when subjected to pressure which prevents expansion, and in order to melt them it is necessary to raise the tem- 274 PHYSICAL GEOLOGY perature so high as to overcome the effect of the pressure. It is evident for this reason that it would be necessary to go deeper to find the fluid interior than if the pressure of the overlying rocks was slight. At a depth of 50 miles a temperature of 3500 F., though sufficient to melt almost any rock at the surface, might not be high enough to overcome the enormous weight of the overlying rocks, and since a still greater pressure is encountered at the increased depth necessary to obtain a higher temperature, it is evident that the melting point, for this cause alone, might never be reached. t (c] Rigidity of the Earth. Two further objections to this theory have caused its abandonment by scientists. The first of these is that the earth is not pulled out of shape by the attraction of the moon and sun, as would be the case if it were substantially a molten globe. On the contrary it is shown to be more rigid than glass or steel. The second objection consists in the fact (Milne) that the velocity and character of earthquake waves (p. 283) suffer an abrupt change at a depth of about 30 miles, being transmitted at a more rapid rate below this level than on the crust, showing that the nucleus is more rigid than the overlying rocks. (2) Solid Interior. The second theory, as has already been indicated, is that the earth is substantially a solid because of the increasing conductivity and pressure. (3) Gaseous Center. Upon the assumption that below a depth of 190 miles the temperature of the earth is at the critical temperature of all substances (the tem- perature above which a substance can exist only in a gaseous state), it is held that the solid crust passes into a liquid zone, which in turn passes gradually to a gaseous magma. (Arrhenius.) The gaseous magma is potentially but not actually a fluid with a tem- perature above the fusion point of all substances. The rigidity of the earth, according to this theory, may nevertheless be greater rather than less than that of steel. (4) Radioactivity and a Solid Center. A fourth theory is in direct contradiction to the preceding and holds that the temperature of the interior is derived from the heat given off by the radioactive minerals of the earth's crust. According to this theory a radioactive crust, 30 to 45 miles thick, supplies all of the heat for the interior of the earth, and below a depth of about 45 miles the earth has a temperature of only about 1560 C. (Strutt.) So many elements of doubt enter into the above theory that it should merely be considered as suggestive, although all theories must take into account the enormous amount of heat generated in this way. (5) Subcrust Theory. Another theory which has been generally abandoned holds that between the solid crust and the solid center is a fused or semifused layer. Summary. Any hypothesis of the constitution of the earth's interior which is in accord with the known facts must hold (i) that the earth is a globe which increases in density from the surface toward the center; (2) that the temperature of the interior is in- tensely hot, perhaps 20,000 C. at the center, or even higher; (3) that the rigidity of the earth as a whole is greater than that of steel. REFERENCES FOR STRUCTURE OF THE EARTH CHAMBERLIN AND SALISBURY, Geology, Vol. I, 2d ed., pp. 486-589. GEIKIE, J., Structural and Field Geology, 3d ed., 1912. LEITH, C. K., Structural Geology, 1913. REID, H. F., Bull. Geol. Soc. America, Vol. 24, 1913, pp. 163, 186. CHAPTER VIII EARTHQUAKES EARTHQUAKES or tremblings of the earth's surface, when severe, are the most terrifying phenomena of nature, with the possible ex- ception of violent volcanic eruptions. The tremblings of the earth vary greatly in their intensity, from those which cause great destruc- tion of life and property to those which can be detected only by deli- cate instruments. Although severe earthquakes occur only at irregular intervals, specially constructed instruments called seismo- graphs (Greek, seismos, earthquake, and graphein, to write) show that the earth is never free from minor vibrations. Such minor trem- blings are produced by water waves, by changes in atmospheric pres- sure, by readjustments due to the lightening of the earth's surface by erosion and its weighting by sedimentation, by the strains pro- duced by the attraction of the moon and sun, and in other ways. Destructive earthquakes are, however, of only occasional occur- rence and arise from disturbances within the earth's crust (p. 281). The San Francisco Earthquake. The earthquake which in 1906 shook California and wrought such havoc in San Francisco was the most disastrous to property of any in North America within historic times, although the loss of life was slight. Much of the destruction, however, was due to the fires which were started as a result of the shocks, and to the breaking of the water mains of the city, which made it impossible to extinguish the flames. The shock came without warning, as is usually true of great earth- quakes, and lasted less than one minute. It was followed by others of less intensity during the day and for several weeks afterwards. Where the shock was severe trees were injured, some being broken off, some overturned, and some split from the ground upward ; build- ings were shifted horizontally and often badly broken; animals were thrown from their feet and persons from their beds. It was found that the greatest intensity of the shock was along a fault line (p. 267), and that in general the violence diminished with distance from the 275 2 7 6 PHYSICAL GEOLOGY fault on either side. " The rate of diminution, with the exceptions to be mentioned presently, may be expressed by saying that at five miles from the fault only a few men and animals were shaken from their feet, only a few wooden houses were moved from their founda- tions, about half the brick chimneys remained sound and in condition for use, sound trees were not broken, and no cracks were opened which did not immediately close. At a distance of twenty miles only an occasional chimney was overturned, the walls of some brick build- ings were cracked, and wooden buildings escaped without injury; FIG. 274. The slipping of alluvial soil toward the Salinas River, as a result of the San Francisco earthquake. the ground was not cracked, landslides were rare, and not all sleepers were wakened. At seventy-five miles the shock was observed by nearly all persons awake at the time, but there were no destructive effects ; and at two hundred miles it was perceived by only a few persons." l The exceptions to the gradual diminution of intensity occurred on the artificially filled or " made " land in the city, and where there were tracts of deep alluvial soil, especially where ground of this character was saturated with water (Fig. 274). Such ground behaved during the earthquake very much like " jelly in a bowl," the sudden shock causing it to be thrown into waves. 1 G. K. Gilbert. EARTHQUAKES 277 This earthquake was the result of shocks produced by renewed slipping along an old fault line which had been known to geologists for some time. The " earthquake topography " of such a fault line FIG. 275. Map of the fault line formed during the San Francisco earthquake in 1906. It was the vibrations set up by the faulting along this line that produced the earthquake. (U. S. Geol. Surv.) is described on page 265. The fault line along which the slipping occurred has been traced about three hundred miles on the land (Fig. 275), and the principal movement was found to be horizontal 2 7 8 PHYSICAL GEOLOGY instead of vertical (Figs. 276, 277) and to measure from 8 to 20 feet. Some vertical movement also occurred, but it was inconsiderable. Distribution of Earthquakes. A study of the distribu- tion of earthquakes gives a clue to their cause. They occur (i) in volcanic re- gions, where the earth's crust is sub- jected to high tem- perature and to strains produced -by explosions ; (2) along belts of young and growing mountains, where strains are being relieved from time to time; (3) along coasts, where the sea bottom descends steeply from the shores, especially where they are bordered by high mountains. The conditions last mentioned are well illustrated in Japan, where the earthquake records since the beginning of the seventeenth century show that FIG. 276. Fence parted by an earthquake fault, 1906. The fault fracture is inconspicuous, although the horizontal displacement is eight and a half feet. Near San Francisco. (U. S. Geol. Surv.) FIG. 277. Diagrams illustrating the nature of the fault which produced the San Francisco earthquake. (After Gilbert.) a severe shock has occurred on an average of once in two and a half years. By far the greater number of these have been accompanied by a movement along the scarp of the great Tuscarora Deep, which lies a short distance off the coast. (4) Earthquakes EARTHQUAKES 279 also occur where there has been an overloading with sediment. The great earthquake of the Mississippi Valley in 1811 may have been caused by a readjustment of the crust as a result of the great weight of sediment laid down there in recent geologic time. Boundaries were thrown into such confusion as a result of these shocks that it was necessary for the government to make a resurvey of 1,000,000 acres. FIG. 278. Earthquake regions of the Eastern Hemisphere, shown in black. (After Montessus de Ballore.) The earthquake " danger spots " of the United States are situated on the Pacific coast, in the Great Basin (Utah), and in the lower Mississippi Valley. New England has been remarkably free from severe shocks, although many slight earthquakes have been recorded. In the last-named region the danger is greatest where lines of fracture intersect. Over the world as a whole two zones are recognized in which earthquakes are most frequent : the Mediterranean belt, which 280 PHYSICAL GEOLOGY extends from Spain through the Himalayas to eastern China and from which 53 per cent, of the recorded earthquakes originated ; and the Pacific belt, which borders the Pacific Ocean and from which 41 per cent, of the recorded earthquakes came. Only 6 per cent, of the recorded earthquakes have occurred outside of these two belts, showing how rare severe earthquakes are over the greater portion FIG. 279. Earthquake regions of the Western Hemisphere, shown in black. (After Montessus de Ballore.) of the globe. The earthquake zones are shown in Figs. 278 and 279. It should not be concluded from the above that all parts of the earth- quake belts are equally affected. For example, along the line of the proposed Nicaraguan interoceanic canal, earthquake shocks are frequent and severe, while along that of the Panama Canal they have been few and slight, as is shown by fragile arches which have re- mained standing for many years. EARTHQUAKES 281 Summary of the Causes of Earthquakes. The earth may be caused to tremble in many ways, (i) Severe earthquakes have been produced by volcanic eruptions, but the disturbances thus caused are confined to comparatively small areas, as they are the result of steam explosions and of the fracturing of the rock as lava rises in the earth's crust. (2) The falling of the roof of a cave may produce a jar which will cause some damage. The earthquake shocks at Visp, Switzerland, which fissured buildings and caused landslides, were due to the col- lapse of cavern and tunnel roofs, and the earthquakes which are of frequent occurrence in the Karst region (p. 72) on the east coast of the Adriatic are of this origin. The jar produced by the fall of an overhanging rock which formed the brink of a fall has been sufficient to break windows several hundred yards distant. The above causes (i) and (2) are unimportant, and their effect is small. (3) The great earthquakes of the world are a result of the fracturing of the rock of the earth's crust, or of the vibrations produced during faulting, (a) They may be due to the jolting of earth blocks whose movement begins and ends suddenly; and also when thick delta deposits suddenly slump an earthquake may be produced, (b) They are also due to the vibrations produced during faulting by the friction of one block as it rubs against another. This method may be il- lustrated by rubbing the closed fist on a table, or by rubbing two blocks of wood together, (c) They may be produced by a simple breaking of the rock. It has been suggested that some at least of such fracturing " may have relation to sudden deformation by rock flowage." (Leith.) l It is evident from the above that great earthquakes are most likely to occur in growing regions ; for example in young mountains, where the strains have not yet been relieved. Displacements. The amount of the movement of the earth along faults in the production of earthquakes varies greatly. After the California earthquake (1906) it was found, as already stated, that the movement was horizontal and varied from 8 to 20 feet. The movements of the crust in the Sumatra (East Indies) earthquake of 1892 were also horizontal, the total slip of the fault amounting to from n to 13 feet. No trace of the fault was visible at the surface, the proof of the movement being fur- nished by geodetic measurements. Vertical movements are perhaps more common than horizontal, although they are usually accompanied by some horizontal movement. The Japanese earthquake of 1891, for example, was produced by a fault which has been 1 Leith, Structural Geology, 1913, p. 69. 282 PHYSICAL GEOLOGY traced 40 miles, on one side of which the ground sank from 2 to 20 feet, while on the other side (the east) the wall of the fissure was moved 13 feet northward at the same time. The portion of the Alaskan coast affected by the earthquake of 1899 was found to have been displaced vertically in amounts varying from zero to 47 feet, the average being between 5 and 12 feet. The great earth- quake of Owen's valley, California, in 1872, was produced by a fault which has been traced 40 miles, whose vertical displace- ment at the surface (throw) was from 5 to 20 feet. If a fault caused the Charleston earthquake (Fig. 280), no evidence of such movement appears at the surface. 1 I ' I ' I V:V- ;; ; X;V.-: ;;.-.;> : ; -V;V-^ ;;;.;! -;.V; FIG. 280. Diagram showing the absence of surface evidence of a fault, because of the presence of a thick, unconsolidated bed of sand above the solid rock. It is sometimes found that the dis- placement along a fault not only varies in amount at different places, but also that along the same fault the downthrow side in one portion is on the right, for example, and at another on the left. Such a fault is called a hinge fault. Many earthquakes originate beneath the sea, some of which have been very destructive. Off the coast of Greece the telegraphic cable broke at the moment of an earthquake in 1873, and upon sub- sequent examination it was found that the break was seven miles from land, and that the water which formerly had been 1400 feet deep at this spot was 2000 feet in depth after the shock. Submerged preci- pices 3000 to 5000 feet high occur in this region and are doubtless fault scarps whose formation caused many earthquakes. Many records are extant of vessels which were made to vibrate by submarine earthquakes, to such a degree that the crew thought that they had struck a reef; loose objects rattled about, and in some cases men were thrown to the deck by the violence of the shock. Depth of the Plane or Point of Origin. Wherever it has been possible to determine the direction of the emergence of the waves of great earthquakes, it has been found that they converge at a depth of less than 12 miles and usually less than 5 miles; that is within the zone of fracture. The point or place of origin is called the focus. EARTHQUAKES 283 Earthquake Waves. When the earth is shaken by an earthquake two sets of vibrations are started, one which follows the surface of the earth and another which passes through it, the former traveling more slowly than the latter, which passes through the 8000 miles of the diameter of the earth in from 20 to 22 minutes. Earthquake instruments (seismographs, p. 287) situated on the side of the earth opposite an earthquake shock show three series of vibrations : (i) pre- liminary tremblings, 20 to 22 minutes after the shock, followed by (2) the strong vibrations of the principal shock, and finally by (3) a Vibrations not far distant Vibrations at epicentrum FIG. 281. Diagram showing the path of earthquake waves and the vibrations which they produce. (After Sieberg.) series of feeble vibrations (Fig. 281). Some of the waves are (i) compressive or longitudinal and have the same nature as the vibra- tions which travel through a liquid, and some are (2) transverse and vibrate at right angles to the direction of the transmission of the shock. Such waves as the latter can be propagated only in a solid. The velocity of the earthquake waves which pass through the earth is uniformly about 375 miles a minute, on the assumption that this movement is along a straight line. This indicates a rigidity of the earth's interior of one and a half times that of steel. The velocity of the surface waves varies with the rock through which they pass, and other conditions ; that of the earthquake at Naples in 1857 being nine or 10 miles a minute, while in Germany in 1874 the rate was 28 miles 284 PHYSICAL GEOLOGY a minute. The velocity depends to a large degree upon the density and elasticity of the rock, being much slower in sand and loose sand- stone than in slate, schist, or granite. This has been shown experi- mentally by noting the velocity of shocks produced by explosions of gunpowder, and it has been found that the velocity is 825 feet a second in sand, and 1088 feet a second in slate and schist. In all such experiments, however, account must be taken of the presence of fis- sures and whether or not the fissures are filled with water. Amplitude of Vibration. By the amplitude of vibration is meant the distance each rock particle is moved from its position of rest during an earthquake (Fig. 282). It is a common notion that the amplitude is very great, but measurements show that they are minute, FIG. 282. Wire model showing the motion of an earth particle during an earthquake. an amplitude of 20 millimeters (three fourths of an inch) being suffi- cient to destroy a city ; one of 10 millimeters (three eighths of an inch) constituting a severe earthquake ; and one of 5 or 6 millimeters being adequate to shatter a chimney. Amplitudes much greater than the above have been recorded. It should be remembered in this con- nection that it is the suddenness of the shock that makes it effective. This can best be illustrated by a simple experiment. If a stone or metal slab upon which a marble rests is struck a sharp blow, the marble will be thrown into the air, but it is evident that the actual movement of the particles composing the slab, and through which the vibrations were transmitted to the marble, was a very small frac- tion of an inch, the projection of the marble being due to the great suddenness of a small movement. This phenomenon is well illus- EARTHQUAKES 285 FIG. 283. Pier driven into the ground by an earthquake shock. trated in some earthquakes. In one case (Calabria, Italy), the stone- work of a well was thrown out of the ground, and in its new position resembled a small tower. In an Icelandic earthquake in 1896 persons lying on the ground near a cliff were thrown over the edge. More commonly stones are thrown into the air and overturned (Assam, India). Sometimes heavy objects such as gravestones are embedded more deeply in the ground (Fig. 283). The reason for this can be shown by another simple experiment. If a ball of soft clay upon which a pebble rests is subjected to a sudden upward movement, the pebble will be embedded, to some extent, in the clay. The amplitude of the vibration of a rock particle should be distinguished from the earth waves which are produced in loose alluvium. For example, during the earthquake which shook the Mississippi Valley in 1811, and which was probably the most violent that has taken place in North America since its settlement by Europeans (although not the most destructive because of the sparseness of the population and the character of the buildings), the ground is described as having been thrown into great waves, so that the branches of the trees inter- locked as the waves passed under them. Ip this case, the ampli- tude of the vibrations of the rock upon which the thick alluvial soil rested probably did not exceed a few centimeters. Vorticose and Twisting Movements. After earthquakes, pictures are often found with their faces toward the wall, furniture has been turned partly or completely around, statues have been twisted on their pedestals and chimneys have been partly turned about. No one cause can be assigned to such movements. In many cases the turning was due to a simple motion backward and forward ; in others the rotation prob- ably resulted from " a combination of shocks from separate faults." (Hobbs.) The latter is given as the cause of the turning of a bronze angel in Belluno, Italy, through an angle of 20, and the rotation of the statue of Queen Victoria in Kingston, Jamaica. FIG. 284. Diagram showing a railroad track bent during the Charleston earthquake. 286 PHYSICAL GEOLOGY The bending of railroad tracks (Fig. 284) and the zigzag position of rows of trees which were straight before the earthquake were pro- duced by the lateral shifting of earth blocks. Duration. The duration of a severe earthquake is very short. As has been stated, the shock which destroyed San Francisco lasted about one minute, and the movement along the 300 to 400 miles of fault rift probably did not consume two minutes. The great Assam (India) earthquake lasted only two and a half minutes, and the destruction was accomplished during the first 15 seconds. The de- structive shocks of the Charleston earthquake lasted a little more than half a minute. Between December 16, 1811, and March 16, 1812, at least 1874 shocks were felt in the Mississippi Valley, of which eight were severe. No individual shock, however, was of long duration. Frequency. Delicate instruments (seismographs) show that the earth is continually trembling in all parts, and it is probable that quakes severe enough to be felt are shaking the earth in some regions at all times. Certain portions of the world as, for example, parts of Japan and southern Italy are subject to frequent shocks. In the former, a severe earthquake occurs on an average of every two and a half years, and minor shocks four times a day. A careful record of the aftershocks of the earthquake at Messina, Sicily, in 1908, shows that 87 shocks were felt during the first four days and 862 during the following year, four of which were severe. All definite predictions as to the time and place of earthquakes are of little value. This is illustrated in the case of San Francisco. The earthquake rift or fault line was known before the earthquake. It was* believed that renewed faulting might occur at any time, but whether within one year or many years could not be foretold. Since earthquakes are the result of a relief of strain, it is evident that a region is likely to be immune from severe shocks for some years after it has been shaken, since the strains which produced the shock have been partly or entirely relieved, and a shock will not occur until strains have again accumulated. Areas Affected by Certain Earthquakes. The areas affected by earthquakes vary greatly in size. A region four times the size of Europe is said to have been affected by the Lisbon (Portugal) earth- quake of 1755 ; that shaken by the great Assam (India) earthquake of 1897 was 1,750,000 square miles, of which 150,000 were laid in ruins. An earthquake in 1891 shook three fifths of the entire area of Japan. EARTHQUAKES 287 The Charleston (South Carolina) earthquake affected an area 1000 miles in diameter. Instruments for Determining and Measuring Earthquakes. Earthquake instruments or seismographs have been established in many parts of the world, and from them the location and intensity of earthquakes are known. For example, seismographic records will be made in Germany, the United States, and elsewhere of an earth- quake in Java or the West Indies. Seismographs vary widely in construction, but since they all endeavor to show the direction of the vibrations, the essential feature consists of three pendulums arranged so as to vibrate in mutually perpendicular directions, the record being made on a sheet of paper which moves at a uniform rate (Fig. 28 1). 1 EFFECTS OF EARTHQUAKES Faults and Fissures. We have seen that earthquakes are usually the result of faulting. Sometimes the fault rift extends to the surface as an open fissure, but more often the fissure is closed. When deep alluvial soil is shaken, many cracks are often formed, as a result of the compacting of the loose material and of its slumping. Such fissures are especially likely to form in stream valleys parallel to their course (Fig. 285), since the alluvium is unsupported on the stream side and moves in that direction, if at all. As a result of such slumping cracks are formed and valleys are narrowed. Fissures formed during the Mississippi Valley earth- quake of 1811-1812 are still visible. One such fissure diverted the course of the Mississippi River so that an oxbow (p. 121) was cut off. FIG. 285. Diagrams showing the effect of earthquake shocks upon loose material. The bridge girder has remained in place, but the piers have moved inward at the bottom. 1 For a more complete description of seismographs CLELAND GEOL. 1 9 : : Hobbs, Earthquakes, pp. 257-275. 288 PHYSICAL GEOLOGY In Arizona the waters of several streams now flow into a fissure formed during an earthquake (Fig. 286). In some earthquakes fissures have opened and then closed again, entrapping people and animals, and engulfing houses. Changes in Level. It is usual to find that the level of the land has changed during earthquakes. As a result of the earthquake of 1811- 1812 in the Missis- sippi Valley, Reelfoot Lal f e ' 2 * mil . e f long and 5 miles wide, was formed, the trees still being visible on its bottom. In the same earthquake Lake tofc 1 " *i * - ^ Eulalie was drained. % ; ill In the Indian earth- quake of 1819, a lake about the same size as Reelfoot Lake was formed. There is also often a lateral shifting of the ground during an earth- quake, as in that at San Francisco, which moves the plains or mountains in one direction on one side and those on the opposite side in the opposite direction (p. 278). The changes of level on opposite sides of faults which produce falls and lakes have already been dis- cussed (p. 262) (Fig. 287) ; elevation is as frequent an accompaniment of earthquakes as depression. During the earthquake which shook lower India in 1819, an area 50 miles long and 10 miles broad was elevated 10 feet and is called the Mount of God. Over an area of 600,000 square miles the coast line of Chile and Patagonia is said to FIG. 286. Fissure produced at the time of an earthquake. Arizona. (U. S. Geol. Surv.) EARTHQUAKES 289 have been elevated dur- ing an earthquake in 1835. Landslides. One of the most obvious effects of a severe shaking of the earth is the produc- tion of landslides and the slumping of thick soil which rested on a slope. The hills about Kingston, Jamaica, for example, are scarred by landslides formed during the earthquake of 1907. As a result of an earth- quake in India in 1897, the hills were stripped of their forests by land- slides. This permitted erosion to proceed so rapidly as to overload the streams, with the result that the rivers, instead of flowing from deep pools over rapids, flowed in broad, shallow channels over a sandy floor. An earthquake in Greece in 1870 caused great landslides which dammed up some of the valleys and formed lakes, some of which are still in existence. Earthquake Topog- Scale of /1//e FIG. 287. Map of the Chedrang fault, India, showing the effect of faulting on drainage. The figures show the amount of vertical elevation in feet. The river in places flows along the downthrow side of the fault, and is ponded back in others. The tributary streams also are dammed, forming pools. Waterfalls are formed where the river crosses the fault. In one place the fault runs along the old and now dry bed of the river, while the stream itself flows in a depression on the downthrow side. The large pools are not formed by the fault scarp, but by the reversal of the original slope of the river bed by the unequal elevation of the land, there being no eleva- tion at the pools, but an elevation of more than 30 feet above each pool, and a lesser elevation below. (After Oldham.) raphy. A description of the fault rift along which occurred the movement which produced the San Francisco earthquake will serve, in a general way, for all such earthquake faults or earthquake topography. This line is well 290 PHYSICAL GEOLOGY marked for a distance of 43 miles and follows a system of long, narrow valleys, except where it traverses wide valleys for short dis- tances. In some places it passes over mountain ridges, sometimes a pass, but in some cases over the shoulder of a mountain. Along the fault line low, precipitous cliffs or scarps occur. Small basins or ponds, many having no outlet, are of fairly frequent occurrence and usually ^ e at t ^ e k ase ^ ^ e scar P s - Trough- like depressions bounded on both sides by scarps also occur, and are due to the subsidence of the ground or to an uplift on one or both sides. In the Japanese earthquake of 1891 the fault FIG. 288. Diagram showing a line showed itself in some places as a ridge ' as if made by a gigantic mole just beneath the surface (Fig. 288). Sounds. Accompanying or slightly preceding earthquakes, sounds, described as a hollow rumbling or grinding and sometimes as a roar, have often been noticed. These are produced by the breaking and grinding of the rock as it is thrown into vibrations, and by the falling and breaking of objects on the earth. Loss of Life. The destruction of life is more impressive than any other effect of an earthquake. In 1812 Caracas, Venezuela, was so severely shaken that 10,000 people were killed, while the loss of life in Lisbon in 1755 amounted to 30,000. In 1905 an earthquake in India (Kangra) destroyed 20,000 people, and it is estimated that in 526 A.D. between 100,000 and 200,000 were killed by the shocks which dev- astated the shores of the Mediterranean. In 1908 the Messina earthquake, described as the world's most cruel earthquake, destroyed 77,283 people; and more than 30,000 were killed in the Italian earthquake of 1915. Fish in great numbers are sometimes killed by earthquake shocks which affect the sea, lakes, or rivers. Effect on Underground Water. After severe earthquakes it is not unusual to find that some springs have become dry, that some have had their volumes increased or decreased, and that some have burst forth where none formerly existed. Along a fault rift which extends for 1 20 miles in Afghanistan and Beluchistan over mountain and valley, springs are found in abundance, the volumes of which are said to be augmented after an earthquake disturbance. So marked is this rift that it has long been utilized as a thoroughfare. The composi- tion and temperature of the water of springs is also sometimes changed EARTHQUAKES 291 as a result of an earthquake shock. It is evident that the cause of this disarrangement of the underground water is the opening and closing of fissures leading to water- bearing strata or joints, and to fault- ing which may open a water-bearing stratum to a fissure. It is not unusual to find sand or mud cones and " crater- lets " after earth- quake shocks (Fig. 289). These are formed by jets of water which were forced through fis- sures during the disturbance. The water forming these jets originated in a water-bearing stra- tum or in water-bearing strata, or in fissures and caverns. Gases. Gases, usually containing large amounts of sulphureted hydrogen (H 2 S), are also sometimes discharged, with or without water. These gases were imprisoned in the soil and escaped either as a result of the fissuring of the ground or by being forced out by the shaking together of the loose material. The sulphureted hydrogen was doubtless formed, for the most part, by the decomposition of animal and vegetable matter in the soil, just as is that which rises from the mud on the bottom of ponds when it is stirred with a stick. The escape of the sulphureted hydrogen was especially noticeable in the Mississippi Valley and Charleston earthquakes. Construction of Buildings in Earthquake Regions. A study of the effects of earthquakes on buildings has led to certain recommenda- tions concerning the location and construction of houses in earth- quake regions, (i) Artificially filled ground and deep alluvial soils should be avoided, since these are likely to be badly fissured and are, moreover, thrown into large waves by a shock. (2) A firm and FIG. 289. Craterlet formed during the Charleston earthquake. (U. S. Geol. Surv.) 292 PHYSICAL GEOLOGY stable foundation is of paramount importance, and particularly on soft and " made " ground. (3) Low structures, especially when well braced, with the beams and rafters attached firmly to the walls, are most desirable, because if a building does not vibrate as a whole, the parts act as battering-rams to throw over or break the walls. (4) Since the fires which almost invariably accompany earthquakes are often more destructive than the earthquakes themselves, it is im- portant that there should be ample fire protection. It is estimated that had the buildings at Messina been properly constructed at the time of the earthquake in 1908, 998 deaths out of every thousand would have been prevented. Effect of Earthquakes on the Sea. One of the most disastrous effects of earthquakes on low coasts is produced by the great sea waves (tsunamis) which sometimes follow the shocks. After the first severe trembling which shook Lisbon in 1755, the sea retreated from the shore, laying bare the bottom of the harbor, and then returned in a wave 60 feet high which completed the devastation of the city. This wave was destructive along the coasts of Portugal and Spain and was felt on the coasts of countries far distant. A great sea wave cost the lives of 27,000 people in Japan in 1896. The velocity of great sea waves and the distance to which they are propagated is well-known. In the Japanese earthquake of 1896 the wave which reached Honolulu, 3500 miles away, was 8 feet high at that place, and its mean velocity between these points was 68 1 feet a second. It was also recorded at San Francisco, to which point its mean velocity was 664 feet a second. The great sea wave from an earthquake in Peru, South America, in 1868, reached Honolulu, 5500 miles away, in 12 hours, and Japan, over 10,000 miles away, the next day. Because of their great wave length (sometimes 200 miles), great sea waves may not be sensible to vessels in mid-ocean and are never destructive until they reach a shallowing shore. Great sea waves are apparently not all due to the same cause. Some are probably produced by a sudden depression of a portion of the ocean bottom by faulting and a consequent drawing in of the ocean water. This causes the withdrawal of the water from the land, and the wave set in motion by the meeting of the water then spreads in all directions, devastating low-lying coasts. A sudden shock on the sea bottom is probably also competent to give rise to a great sea wave. Explosions of submarine volcanoes set waves in motion which may work great havoc on low coasts. Evidence that a Region has been Free from Severe Earthquakes. It is not always possible to tell whether or not a region has been EARTHQUAKES 293 subjected to earthquakes ; but some features such as pinnacled rocks with insecure bases and steep hillsides covered with soil give evi- dence that a region has been free from violent shocks for many cen- turies. For example, New England has probably been free from dev- astating earthquakes since glacial times, as the almost precipitous, soil-covered slopes of many hills and mountains show. This is also borne out by the occurrence of perched bowlders in very insecure positions. REFERENCES FOR EARTHQUAKES DAVISON, CHAS., A Study of Recent Earthquakes, 1905. DUTTON, C. E., Earthquakes in the Light of the New Seismology, 1904. FULLER, M. L., Our Greatest Earthquake: Pop. Sci. Monthly, Vol. 69, 1906, pp. 76-86. GEIKIE, A., Textbook of Geology, Vol. I, 4th ed., pp. 358-377. GILBERT, G. K., The Investigation of the San Francisco Earthquake: Pop. Sci. Monthly, Vol. 69, pp. 97-115. HOBBS, W. H., Earthquakes, an Introduction to Seismic Geology, 1907. LAWSON, A. C., et al, The California Earthquake of April 18, 1906: Report of the State Earthquake Investigation Commission, Carnegie Institution of Washington. TARR, R. S., and MARTIN, L., Recent Changes of Level in the Yakutat Bay Region, Alaska: Bull. Geol. Soc. America, Vol. 17, 1906, pp. 29-64. WRIGHT, C. W., The World's Most Cruel Earthquake: Nat. Geog. Mag., Vol. 20, 1909, PP- 373-396. CARNEGIE INSTITUTION PUBLICATION 87, Vol. i, Plate 15 and Atlas. CHAPTER IX VOLCANOES AND IGNEOUS INTRUSIONS A VOLCANO may be regarded as an opening in the earth's surface through which various gases and solid or molten rocks are ejected. The materials brought to the surface accumulate around the opening, forming a conical hill or mountain. The rapidity with which volcanic cones are built up is in contrast to the slowness with which other elevations are formed (p. 362), and they are able, consequently, to defy the agents of erosion during the period of rapid growth. Vol- canoes are, moreover, capable of destroying in a very short time re- liefs which erosion would be able to wear down only after centuries of work ; as, for example, when one destroys a large part of its cone in a few hours. However, although conspicuous, the work of volcanism, because of its limited extent is unimportant as compared, on the one hand with the movements which are raising the earth's surface, and on the other with erosion, which is lowering it and is universal in its effects. How Volcanoes Begin. The first step in the development of a volcano is the opening of a passage to the surface. This opening may be caused by the " blowing out " of a portion of the earth's crust with the resulting formation of a funnel. More usually, how- ever, it is a fissure through which gases, ash, and lava are ejected. The opening through which the material is ejected is usually en- larged by explosive action or by melting, and the lava and other ejectamenta tend to form a ring as they fall back to earth, until a hill is built up with a depression or crater in the summit. It is ap- parent, therefore, that the cone is not an essential part of a volcano, but is secondary to the vent, being merely a result of its action, not a cause. New Volcanoes. It is not often that man has seen the birth of a volcano, but a few well-known instances may be mentioned. On the shore of the Bay of Naples, amidst gardens and cottages, a vol- cano called Monte Nuovo had its birth in 1538, and in the course of 294 VOLCANOES AND IGNEOUS INTRUSIONS 295 two days built a cone to a height of about 500 feet. The eruption lasted only a week and has not been renewed since. An examination of the material of the cone showed that most of it was of volcanic rock, but that pieces of Roman pottery, fragments of the surface rock, and marine shells were also present. On a plain in Mexico between 2000 and 3000 feet above the sea, covered with fields of sugar and indigo, a fissure opened in 1759, from which rocks were thrown to great heights and about which several cones were built up, the smallest to a height of 300 feet and the largest, Jorullo, to that of 1300 feet above the plain. The eruption, which began in June, 1759, ceased in February of the following year. Volcanic cones have also been built up from the ocean bottom within recent times. In 1811 one such (Sabrina) was formed ofF the Azores, rising to a height of 300 feet above the sea. As it was composed of ash, it was soon washed away by the waves. Many of the great volcanoes of the world, such as Vesuvius, Etna, and Mauna Loa, began as submarine volcanoes many thousands of years ago and built up cones from abyssal depths. Classification of Volcanoes. Volcanoes are usually classified as active, dormant, and extinct. This classification is unsatisfactory, since a volcano which has long been considered to be extinct may become suddenly active, and volcanoes classed as dormant may never again be in eruption. For example, Vesuvius must have been regarded as extinct at the beginning of the Christian era, since it had been inactive so long that its crater was covered with vegetation, yet in a few days in the year 79 one half of its crater was blown off by a series of powerful explosions, and it has been intermittently active ever since. Volcanoes which have not been in eruption during his- toric times are said to be extinct; those which have been active in modern times, but are now inactive, are said to be dormant. All volcanoes may become active after a period of quiet, or may become extinct after a single paroxysm ; such, for example, as that of Monte Nuovo. MATERIALS ERUPTED The materials brought to the surface by volcanoes may be classi- fied as gases, solid matter, and lava flows. Gases. The difficulty in collecting gases from the crater of a vol- cano during eruptions renders our knowledge of them rather incom- plete. In fact, whatever information we have has been largely obtained from fumaroles or openings on the flanks of the volcano, and from the crater after eruptions have ceased. 296 PHYSICAL GEOLOGY The principal gases given off during volcanic eruptions are sul- phureted hydrogen (H 2 S), sulphur dioxide (SO 2 ), carbon dioxide (CO2), carbon monoxide (CO), hydrochloric acid (HC1), hydrogen (H), oxygen (O), nitrogen (N), argon (A), and water. It is stated that the gases emitted from Vesuvius in 1906 contained so much ammonia and hydrochloric acid that the glowing lavas were shrouded in a veil of ammonium chloride (NH 4 C1) vapor, and that the " pine tree " cloud of yellowish " smoke " which hangs over that volcano during eruptions consists chiefly of ammonia compounds. The glare of the red-hot lava in the crater is reflected from this cloud and gives the appearance of a burning mountain. The composition of the vapors depends upon the state of activity of the volcano ; chlorine is more abundant in the energetic phases, while sulphurous and carbonic gases characterize the dying out of activity. According to recent investigations, steam seems to be in smaller quantities than formerly thought. This contention is supported by the facts (i) that the amount of steam in craters decreases as the center of the crater is approached ; (2) that the white cloud which hangs over volcanoes during eruptions is a mixture of solids and gases, and not steam as it appears ; (3) that volcanic ash is invariably white and consequently has the appearance of steam when in sus- pension in the air; (4) that the volcanic cloud never produces rain- bows or aureoles. The great quantity of steam rising from some parasitic cones and from some lavas, however, is enormous. For example, it has been estimated that from one of the many parasitic cones of Etna suffi- cient steam was ejected during one period of one hundred days to form, if condensed, 462,000,000 gallons of water. The steam of fumaroles is, however, apparently largely of surface origin, as is shown by the increase in quantity after rains. Fragmental Materials. All of the substances thrown into the air by volcanic explosions, which fall to the ground in a solid state, are included under the term fragmental materials, and are classified as (i) dust, (2) ash, (3) cinders, (4) bombs, and (5) blocks of rock. These solid ejections are either portions of the rock which has been broken into pieces by the force of the explosions, or lava which was hurled into the air in a liquid condition but which solidified before reaching the ground. The size of the fragments varies from rocks weighing many tons to the finest dust, which may remain in the air many months. The term VOLCANOES AND IGNEOUS INTRUSIONS 297 ash applied to this fine material is misleading, since dust and ash are not the result of combustion, as the name seems to imply, but of the shattering of the rock or lava by explosions, the pulverization of lava by sudden cooling after it is hurled into the air, and the col- lisions between stones as they are hurled from the crater or as they fall back to the ground. No part of the work of volcanoes has a greater geological importance than the production of dust. Some of it is so fine that no watchcase is so closely fitted as to prevent its entrance. Near the vents it is sometimes scores of feet thick, and in regions several hundred miles away it is sometimes deposited to a depth of several inches. For example, in 1783 the dust from an Icelandic volcano was carried to Scotland, a distance of 600 miles, in sufficient quantity to destroy the crops. The larger particles are termed cinders and often -constitute the conspicuous deposits of the volcanic cone, the fine dust having been carried away by the wind. When a mass of molten lava is thrown into the air, it takes a more or less globular form and is called a bomb. Two kinds of bombs are common : one spindle or almond-shaped (Fig. 290), with an exterior only slightly cracked ; bomb, Aukland. the other with a surface cracked and broken, ft^ like that of the crust of a loaf of bread. The rotating liquid lava are cause of the difference is to be found in the cooled as they pass air< (After FIG. 290. Volcanic The degree of liquidity of the lava. The spindle- shaped bombs were formed from very liquid lava, and their shape was produced by their gyratory motion in the air, while the " bread-crust " bomb was formed from viscous lava which was little affected by the rotation, and which cracked in cooling, forming a glassy surface and a porous interior. Bombs vary in size from a few inches to several feet in diameter. When the ejected lava is blown full of holes by the expansion of the gas which it contains, it becomes so cellular that it is practically rock froth, the air cavities being sometimes eight or nine times greater than the inclosing glass, so that it is light enough to float upon the water. When it is in this condition, it is called pumice. After the eruptions of certain volcanoes situated on shores, great quanti- 298 PHYSICAL GEOLOGY ties of floating pumice have covered the neighboring waters so thickly as to be a menace to navigation. During the eruption of a volcano in Japan so much pumiceous material was thrown out that it was possible to walk a distance of twenty-three miles upon the debris floating on the sea. The size of the blocks of rock thrown out during eruptions varies greatly with differ- ent volcanoes. A 2OO-ton block is said to have been hurled a distance of nine miles from the volcano Cotopaxi in South America, and it is reported that a rock fragment 100 or more feet in diameter was ejected from the Japanese volcano Asama. Some volcanoes, however, throw out no rock fragments. The quantity of fragmental material ejected by volcanoes can best be shown by a few examples. It has been estimated that 4.3 cubic miles of material were ejected from Krakatao (p. 304) in 1883, and 28.6 cubic miles from Timboro in 1815. During the eruption of Sumbawa in the same year an area of nearly 1,000,000 square miles was covered by an amount of fragmental material estimated to be sufficient to make 185 mountains of the size of Vesuvius. Lava. All molten rocks which issue from the earth and also the solid rock which results when they cool are included in the term lava. Lava streams issue - either from the crater of a volcano by over- flowing or breaking through its rim ; from fissures or open- FIG. 291. Small craters, c. c, c, along a fissure, through . n , which lava has been extruded. in g s on lts flanks ? or through fissures in the earth's surface (Fig. 291), where there are no volcanic cones. When they issue from a volcano they flow down the steepest slope ; when they reach a gentle slope they spread out ; when some obstacle, such as a stone wall, is encountered, their progress is at first stopped, then they either overflow or overthrow it, or pass around its ends. Lava Streams. The surface of a lava stream, which at first glows like red-hot metal, cools quickly and blackens, but since the heat of the interior is kept in by the porous crust thus formed the deeper parts of the stream remain in a molten condition for a long time, oc- casionally foe several years. One can often walk across a lava flow a few days after it ceases to move, and while the deeper portions are still molten, without suffering any inconvenience. After the crust has hardened, the still molten lava of the interior may continue to VOLCANOES AND IGNEOUS INTRUSIONS 299 flow until it drains out, leaving a tunnel which may be several miles long (p. 309). A lava tunnel on Mt. Shasta, California, which is 60 to 80 feet high and 20 to 70 feet broad, has been explored nearly a mile without its end being reached. Effect of Composition on Fluidity. Lava varies greatly in com- position and fluidity. Some lava streams have flowed 20 to 30 miles or more, while others have solidified as soon as they issued from their craters ; some have flowed several miles, while others, with an equally high temperature and even greater volume, have moved a much shorter distance on an equal slope. This difference in the fluidity of lavas is due largely to their chemical composition and to their tem- perature. The basic, dark-colored lavas (p. 329) fuse at a lower temperature and are consequently more likely to flow long distances. The acid, usually light-colored lavas (p. 329) melt at a higher tem- perature and consequently become solid while still hot. They are therefore likely to solidify quickly. It is evident, however, that if a basic lava has a temperature which is but slightly above the melting point, it will be as stiff (viscous) as an acid lava at a high temperature. Temperature. The temperature of lava when it issues from the vent of a volcano is probably often greater than 2000 F. This is shown by the fact that copper wire, whose melting point is 2200, was fused in a Vesuvian lava stream which had already lost some of its heat (p. 273). The temperature of the lavas in Kilauea in July, 1911, was 1260 C. (2300 F.) ; that of Stromboli in March, 1901, was 1150 to 1176 C. (2102 to 2149 F.). Surface of Lava Flows. The sur- faces of lava flows vary greatly, some being so rough as to make walking danger- ous and difficult, while others are com- paratively smooth. A fluid lava will con- FlG> 292> _ p a hoehoe type of lava surface in the crater solidate with smooth of Kilauea, Hawaii. (U. S. Geol. Surv.) 300 PHYSICAL GEOLOGY FIG. 293. The rough aa surface of a lava flow on the volcano Colima. Mexico. and ropy surfaces, while a viscous one will become very scoriaceous. This latter condition is partially due to the gas in the lava, which instead of escaping freely to the air forms bubbles in the sur- face of the lava, just as air blown into soapy water forms a frothy surface ; the crust may also be broken to some ex- tent by the continued movement of the more liquid mass below, causing an ex- tremely rough sur- face when the mass hardens. The Ha- waiian word pahoehoe is used to designate the smooth type of lava, with the gently rounded, ropy surface which is characteristic of fluid lavas (Fig. 292) ; while another Hawaiian term aa is used for the rough, cindery surface (Fig. 293). Velocity of Lava Flows. The rate of flow of lava depends upon its fluidity and upon the slope over which it moves. In Iceland lava streams have flowed over surfaces which appear flat to the naked eye, while elsewhere they have consolidated on slopes which were almost vertical. A lava stream on Mauna Loa flowed fifteen miles in two hours, and the main stream from Vesuvius in 1906 descended the first steep slopes with a velocity of about two miles an hour. Such rates as the above are, however, rather unusual. The rate of flow is gradually reduced as the stream cools and as the slope diminishes. Lava often continues moving for a long time after the eruption ceases. A lava stream which began to move on Vesuvius in 1895 was found to be still in motion four years afterward. Nature of Lavas. The slag formed in an iron furnace is really an artificial lava, and from it much concerning the nature and behavior of lavas can be learned. When lava is spoken of as a molten rock it should be understood that, since rocks are composed of minerals varying in fusibility and solubility, it is really a liquid rock in which some mineral matter is dissolved in other mineral matter; i.e., it is a VOLCANOES AND IGNEOUS INTRUSIONS 301 mutual solution of mineral matter in mineral matter. Gases as well as mineral matter enter into the solution. This can best be illustrated by a well-known experiment. If crystals of snow, salt, and sugar are mixed together and compacted at a low temperature, an artificial rock will be formed in which the constituents can be recognized. If the temperature of this solid is now raised to about 32 F. the mass will become a liquid, even though the melting points of salt and sugar are very much higher. In this case, a rise in temperature sufficient to melt but one of the constituents is necessary, since this one is then capable of dissolving the others. If the temperature of such a solution is again lowered, the salt and sugar will not crystallize out until they are forced to take the solid form by the crystalliza- tion (freezing) of the water. It is evident that both in the process of solution and in that of crystallization the important factor is solubility, and that a temperature merely sufficient to melt one of the constituents is necessary. That lava should be con- sidered as a solution of vari- ous minerals is evident when cooled lavas are examined. If lavas were simply molten rocks in which the minerals had melted according to their fusibility, we should find that upon cooling the least fusible mineral would crystallize out first, then the others in the order of their fusibility. Such, however, is not always the case ; often the least fusible mineral is the last to take the solid form. This is due to the fact that the liquid mass is a solution in which the various minerals assume the liquid state, and upon cooling, the solid state, depending upon their solubility more than upon their fusibility, the least soluble rather than the most infusible crystallizing first. When a lava cools very slowly, as is usually the case when it is intruded beneath the surface, the molecules of which it is composed FIG. 294. Scoriaceous lava. National Museum.) (U. S. 302 PHYSICAL GEOLOGY tend to collect into crystals. When the process is long continued the point of saturation of the other minerals is reached, crystals are formed, and a rock composed entirely of crystals results. Granite and coarse-grained traps (p. 330) are such rocks. If the cooling is more rapid, rocks composed of fine crystals such as rhyolites and basalts (p. 331) may be formed. When, however, a lava flow cools so rapidly that no crystals or only a few can form, volcanic glass, or obsidian, is produced. Often a lava flow passes from a glassy to a crystalline state from the surface downward. When such cooling lava is under little pressure, the gases in the surface portions are able to expand, and often produce a surface which is called scoriaceous (Fig. 294) if cindery, or pumiceous if the pores are very numerous and small. TYPES OF VOLCANOES Because of their destructiveness volcanoes probably inspire greater interest than any other natural phenomenon, and it will consequently be well to discuss briefly the various types of volcanoes. It should, however, be remembered that the aggregate work of volcanoes is inconsiderable as compared with that of streams, the ocean, and other less conspicuous forces. The chemical composition of lavas, as will be seen, has a con- siderable influence upon the character of eruptions, but the principal factor is the physical state of the lava ; i.e., whether it is fluid or viscous and stiff. If the molten rock is so liquid that the gases can escape rapidly, they do not accumulate into great bubbles which throw the lava high into the air when they break. If on the other hand the lava is stiff, the gases gather into great bubbles which upon bursting throw out the lava as dust, cinders, and bombs. I. The Explosive or Vesuvian Type (i) Vesuvius. Vesuvius has been more carefully studied than any other volcano in the world and is yearly ascended by so many travelers that its value as an illustration is unsurpassed. Previous to 79 A.D. Vesuvius seemed to be extinct, but before the close of that year a great eruption occurred which destroyed the cities of Herculaneum and Pompeii, and laid waste a great extent of coun- try. During this eruption no lava was poured out, but a large part of the crater was blown off and the outline of the mountain greatly changed. Ash and dust were thrown to great heights and were carried VOLCANOES AND IGNEOUS INTRUSIONS 303 63 A.D A.D.78-I63I 1822 long distances by the wind. Pompeii, at the foot of the mountain, was buried beneath 25 or 30 feet of ash, and Herculaneum beneath 60 feet of mud and ash, the latter being covered by a layer of lava during a later eruption. So completely were these cities hidden that their sites were unknown for more than 1600 years. After this first historical eruption (79 A.D.), which is well described by the younger Pliny in a letter to Tacitus, 1 the volcano was occasionally eruptive until 1139. Then, for a period of almost 500 years, with the exception of one feeble eruption, the volcano seemed again to have become extinct, and the crater was choked with rubbish and covered with trees. In 1631 another violent eruption occurred (Fig. 295) ; fissures opened in the side of the mountain, through some of which steam and ash were thrown, and four streams of lava poured from the crater, three of which reached the sea. During this eruption the cone was reduced about 525 feet in height. During the eruption in 1906, the main lava stream flowed at a rate of a little less than two miles an hour where the slope was steep, but more slowly when passing over a lower t -r T71 ii' i rIG. 2QC. Section through grade. When wooden objects, such as Vesuvius, showing the changes trees, were encountered by this lava in the shape of the volcano stream, they were charred but not burned ; b tween 6 3 A - D - and l868 - some were broken off by the weight of the lava and carried on the surface of the stream. When a large object was reached the lava piled up behind it until it was moved aside, overflowed, or the stream moved around it. A summary of the sequence of events in a recent eruption is as follows : In 1904 Vesuvius was almost quiet, but soon explosions occurred of sufficient force to throw fragments short distances above the crater's rim. In 1905 a narrow stream of lava flowed from a fissure in the cone throughout the year. On April 4 of the following year a great cauliflower-shaped cloud of dust and gas rose from the crater, and lava streamed in small quantities from successively lower openings in the side of the cone. On April 7 an explosion occurred which sent a column of dust-laden gas four miles vertically into the air, and new and larger fissures opened through which lava flowed. The dust so weighted the roofs of the houses as to cause them to collapse with loss of life. One lava stream destroyed the town of Boscotrecase. 1 Translation in Shaler's Aspects of the Earth, pp. 50-56, or in Lyell's Principles of Geology, p. 603. CLELAND GEOL. 2O 304 PHYSICAL GEOLOGY N.W -/v: FIG. 296. A, Krakatao as seen from the north after the eruption of 1883. B, an outline of the crater of Krakatao. The dotted lines show the contour of the island before the eruption, the continuous lines as it is now. (2) Krakatao. The most stupendous volcanic explosion of mod- ern times occurred in the East Indies when the island of Krakatao, lying between Java and Sumatra, suddenly became eruptive in 1883. FIG. 297. Map of a portion of Alaska showing the thickness in inches of the deposit of ash from an eruption of the volcano Katmai in 1912. VOLCANOES AND IGNEOUS INTRUSIONS 305 This island was indeed not known to be a volcano, until in August of the above- mentioned year it became violently explosive (Fig. 296) and in two days blew away about one half of its surface, so that now the sea is 1000 feet deep where the central part of the mountain formerly stood. The amount of ash ejected was so great that the neighboring seas and land were in total darkness during the eruption. The ash was thrown to a height of 17 miles, and remained in the air many months, causing brilliant sunrises and sunsets throughout the world (p. 54). Ships 1600 miles away were covered with dust three days after the eruption; stretches of water with an aver- age depth of 117 feet were so filled with the debris as to be no longer navigable. The FIG. 298. A roof collapsed by the weight of ash from Katmai, Alaska, one hun- dred miles distant. The drift in front of the porch is volcanic ash. (National Geographic Magazine.) noise of the explosions was heard 2000 miles away, and the shock produced waves 50 to 80 feet high, which swept the adjacent shores, deluging 1295 villages and drowning about 35,000 people. The height and strength of the waves is well shown in the fact that a large vessel was carried one and one half miles inland and left stranded on land 30 feet above sea level, and that blocks of rock weighing 30 to 50 tons were carried inland two or three miles. (3) Katmai. The eruption of Katmai, a volcano in the Alaskan peninsula, in June, 1912, was one of considerable violence, but one which did little damage because of its situation in an almost uninhabited region. As will be seen from the map (Fig. 297), the fall of ash was 50 inches deep 30 miles from the volcano, and 6 inches deep 160 miles to the east of the mountain. So great was the amount of dust in the air that 306 PHYSICAL GEOLOGY 100 miles away total darkness prevailed for 60 hours (Fig. 298). The sound of the explosions was carried along the coast for 750 miles. (4) Mt. Pelee. The eruption of Mt. Pelee (1902) on the island of Martinique in the West Indies was remarkable because of two unusual features, (i) A great blast of highly heated air mingled with incandescent dust swept down one side of the moun- tain and overwhelmed the town of St. Pierre, killing all but two of its 30,000 inhabitants, one of these being a prisoner in an underground cell to which the air had access only through a small opening. The cause of this mortality was due almost entirely to the fine, hot dust which penetrated into all of the houses and, when breathed, resulted in almost instant death. The reason for the descent of the blast on one side only of the Cap StMartii U Perle- FIG. 299. Map of Mt. Pelee and environs, showing the portion of the island of Martinique devastated by the volcanic eruption of 1902. The breach in the crater wall is also indicated. (Hill, National Geographic Magazine.} mountain is readily seen when the form of the crater is studied (Fig. 299). With the exception of one place (opposite St. Pierre), where the Riviere Blanche had cut a deep gorge, the rim of the crater was several hundred feet high. When the explosion oc- curred its force, instead of being expended entirely upward, was partly directed through the gash in the side of the crater, and a great cloud of intensely hot gas, dust, and bombs moved down upon St. Pierre. (2) The second unusual feature was noticed after the principal eruption was over. It consisted in a spine of solid rock rising from the crater (Fig. 300), which began to grow in October, 1902, and reached an elevation of 1000 feet at the end of seven months. Much discussion has arisen as to the origin of this spine, but it is generally believed that it was formed by very stiff lava which solidified into a steep-sided column as rapidly as it was forced to the surface. Other VOLCANOES AND IGNEOUS INTRUSIONS 307 volcanoes are known in which the lava forced out of craters near the close of eruptions assumed the form of steep- sided cones. (5) Bandai-san. An eruption in which, so far as known, no lava was discharged took place in 1888 in Japan. For 1000 years Bandai-san, a volcanic cone 2000 feet high, had been dormant, when suddenly a terrific explosion blew away the greater part of the mountain (Fig. 301). Since this one explosion, the volcano has shown no signs of activity. The catastrophe was due to the heating of water which had perco- lated from the surface, and was in fact a steam explosion. A priest living on the mountain reported that the gases surrounding him were respirable. T ,, r , r i FIG. 300. The spine of Mt. Pelee, In all of the eruptions of the Martiniq j ue> French West Indies> I902/ explosive type, dust, cinders, (After E. O. Hovey.) and usually bombs are thrown out; earthquakes are prevalent previous to and accompanying the eruptions; sometimes lava is poured out from the crater or from ...... fissures in the mountain side. The greatest eruptions often occur after long periods of inactivity. II. The Quiet or Hawaiian Type FIG. 301. -Volcano Bandai-san. The . The Hawaiian type of volcano portion enclosed by the dotted line was IS in marked contrast to the CX- blown off during an eruption lasting less p l os i ve O r Vesuvian type, since in than two hours. The height of the cliff , r is about 1500 feet. the former eruptions are not ac- companied by severe explosions, but consist largely in the gentle welling-out of lava from the crater or from mouths in the sides of the mountain. In general, the features most characteristic of volcanoes of this type are: (i) their gentle slopes which do not average more than 7 degrees, (2) the large size of their craters or calderas? (3) the quietness of the erup- tions, (4) the fusibility of the (basic) lava which they discharge, and (5) the absence of severe earthquakes during and preceding eruptions. 1 See footnote on p. 309 for restricted use of term caldera. 308 PHYSICAL GEOLOGY The summit of Mauna Loa on the island of Hawaii is 13,675 feet high, while the volcano Kilauea on its flanks 20 miles distant is only about 4000 feet above the sea. Though forming one mountain the two volcanoes are entirely independent, having been joined by the gradual growth of the two cones. The surface near the summit of Mauna Loa is nearly flat for several square miles, and the crater can- not be seen until one is close upon it, the mean slope within a circle of five miles around the crater being about three degrees. If one conceives of the ocean as removed, this volcano (Mauna Loa) would tower above the floor of the sea as a broad-topped mountain, to a height of more than 30,000 feet, with a base many miles in diameter. Every island of the Hawaiian group is of the same nature and is usually built up by lava from several cones. With the exception of Iceland, the island of Hawaii is the largest pile of lava in the world. Crater of Kilauea. The caldera of Kilauea will be taken as a type of volcanoes of this class. On the top of the mountain is a great pit, three miles long and two miles FIG. 302. Map of the Kilauea caldera, Hawaii, in 1886. wide, surrounded by vertical, terraced walls (Fig. 302). The floor of the caldera is composed of a plain of black lava in which lies a lake of liquid lava of a bright orange color. The surface of the lake, except near the center, is covered by a scum of frothy lava. During eruptions, great volumes of this fiery liquid are thrown many feet into the air. From time to time the surface of the molten lake cools sufficiently to permit it to harden. The lava crust thus formed then cracks, and through the cracks jets and fountains of lava are ejected. The level of the lava floor does not remain station- ary, but gradually rises previous to an eruption, sometimes as much as 100 feet a year, VOLCANOES AND IGNEOUS INTRUSIONS 309 until the lava may reach to within 300 feet of the rim of the crater, but never (in modern times) overflowing it. After the eruption the floor may be 1000 feet below the edge of the crater. Eruptions. When the lava rises in the crater, it is evident that the pressure on its walls is greatly increased, since a column of liquid lava 50 feet high exerts a pressure of about 625 pounds to the square inch. The result of this increased pressure is either actually to fracture the mountain and thus to afford an avenue of escape for the lava, or to aid it to break and fuse its way through the porous lava of which the side of the mountain is built. During the eruption of Kilauea in 1840 lava first made its appear- ance five miles from the main crater; later it sank in this new crater and reappeared at other smaller openings farther down the mountain side ; finally, it was poured out on the surface still lower down and flowed in a molten stream to the sea. During the eruption of Mauna Loa in 1853, a fountain of lava 200 to 700 feet in height and 1000 feet broad burst out at the base of the cone as a result of hydrostatic pressure. Lava Streams. The flow of lava from Kilauea on one occasion " swept away forests in its course, at times parting and inclosing islets of earth and shrubbery, and at other times undermining and bearing along masses of rock and vegetation on its surface. It plunged into the sea with loud detonations. The burning lava, on meeting the waters, was shivered like melted glass into millions of particles, which were thrown up in clouds that darkened the sky and fell like a storm of hail over the surrounding country. The light was visible for over a hundred miles at sea, and at the distance of forty miles fine print could be read at midnight. " (J. D. Dana.) Such explosive action, however, does not always take place when lava reaches water, probably because of the cooler and more stony character of the lava. This was true of a lava stream from Vesuvius in 1794, which entered the sea so quietly that it was possible to watch its progress from a boat close to its front. The tunnels and caves on the Hawaiian volcanoes, caused by the draining out of the lava from below the hardened crust, are hung with lava stalactites 20 to 30 inches long, and stalagmites formed by lava dripping from above project from the floor. Such tunnels are sometimes buried beneath later flows and may later be utilized as outlets for lava, such as occurred during the Kilauea eruption just described, when the lava burst out near the foot of the mountain. An interesting form of lava found on Kilauea, called Pelee's hair, is composed of hair-like threads of lava glass, and in masses resembles tow. It is formed when the wind catches particles of molten lava, either from the lava froth or from the jets thrown up from the crater, and draws them out into glassy threads. Origin of Calderas. The craters of the Hawaiian volcanoes have been enlarged by the sinking in of their sides and, as has been said, are called calderas. Calderas l are also formed as a result of violent explosions which blow off the top of a cone, as was true of Vesuvius during the first historic eruption (p. 302). Calderas are craters of unusual size, varying from one to five or more miles in diameter. One of the most remarkable calderas in the world is that of Crater Lake, Oregon (Fig. 303), which is five to six miles in diameter and 2000 feet deep, the walls standing 900 to 2200 feet above the water. A small cone, called Wizard Island, rises a few hundred feet above the lake. The 1 Daly restricts the term caldera to great craters formed by explosions, such as that of Kra- katao. The word sink is suggested for the Hawaiian and Crater Lake (Oregon) craters, formed by the sinking in of the top of the mountain. PHYSICAL GEOLOGY presence of glacial striae on the crater rim proves that the sum- mit of the mountain was at one time much higher than now, and that a glacier moved down it over what is now the rim. It is believed that the crater was formed by the sinking in of the top of the mountain, the absence of volcanic ejectamenta about the mountain being proof that the mountain was lowered in this way rather than by an explosion. The craters of the moon are two to twenty times larger in diameter than those of the earth and may have been formed in the same manner as those of the Hawaiian type. Steep Lava Cones: Volcanoes of the Chim- borazo Type. It should not be concluded from a study of the Hawaiian volcanoes that all lava cones have a gentle slope. When the lava is viscous, steep-sided cones are formed. The great Ecuador volcano, Chimborazo (20,498 feet), is composed of lava and is, moreover, craterless. As the lava welled up from the vent, it left upon cooling no depression in the summit of the mountain. 6 2 U S When lava is very fluid and in great quantity, it may flow long distances and form compara- tively level plains many square miles in extent. The most remarkable lava plateau of this kind in North America (Fig. 304) covers an area of 200,000 to 250,000 square miles in Washington, Oregon, Idaho, and California (p. 583). It is a vast plain of black basic lava over which one may ride for many hours on a level surface. The lava which overspread this region poured out from fissures instead of from volcanoes, with little or no explo- sive action, and since it was very fluid, flowed for long distances, filling the valleys and covering the smaller hills. Some portions of the region were buried hundreds of feet deep, the greatest depth being estimated at 3700 feet. This great plain was not built up by a single great outpouring of lava, but by a number of flows, some of which followed each other in rapid succession. On the VOLCANOES AND IGNEOUS INTRUSIONS other hand, the surfaces of some of the flows were exposed to the action of the weather many years before the next outpouring occurred, as is shown by the thick layers of soil between the lava flows. Previous to the extrusion of the lava the region was a deeply dissected one, but the lava filled the valleys, buried the lower hills, and surrounded some of the mountains, leaving them as islands in a molten sea. The border of the lava plateau is very irregular, since ridges and spurs extend into it FlG . 304 . _ Lava fields in Washington, from the higher land, and it Oregon, Idaho, and California. in turn protrudes long fingers between the mountain masses. The edge of the sheet can best be compared to the shore line of a submerged coast (p. 226). Recent Icelandic Lava Sheets. Much of the nature of such lava plains as those described can be learned from a study of recent eruptions in Iceland, a region which exhibits marks of igneous activity in greater variety and magnitude than any other spot in the world. In 1783 lava welled out for several months from the great Laki fissure. This fissure is 20 miles long, and on it were formed more than one hundred low craters, from which sheets of lava were spread out on either side (Fig. 291, p. 298). From the place of eruption the lava stream flowed 47 miles on one side and 28 miles on the other, covering an area of 220 square miles to an average depth of 100 feet. The longest flow on record in Iceland is 90 miles, the slope of which is so gentle as to be almost imperceptible, the angle being only a little more than one half of a degree. In some cases lava has welled up from fissures in Iceland without the formation of cones; the longest flow of this class is 19 miles. In other parts the lava has built up great domes similar to those in Hawaii ; one of these is 4600 feet high, with an elliptical crater about three quarters of a mile across at its widest point. In 1913 a fissure three miles long was formed in Iceland from craters on which lava poured forth and covered the plains. In some cases, the lava shot up in a jet like a geyser ; in others, it flowed out like a fiery waterfall. CHARACTERISTICS OF VOLCANIC CONES Profiles of Volcanoes. The slope of a volcanic cone, as has been seen, depends upon the character of the material of which it is made. If it is composed entirely of cinders and ash, the slope will be at the 312 PHYSICAL GEOLOGY angle of repose, which may be as great as 30 or 40 for coarse ash or cinders (Fig. 305 A). The slope is more gentle, however, at the base of the cone, since the dust is carried farther from the summit than the coarser material and is washed farther still by rain and rills. If the cone is of lava, its slope will depend upon the fluidity of FIG. 305. Angle of slope of volcanoes. A cone com- ^g J ava Volcanic posed of ash will be steep, as much as 30, while that of , * r , lava may not be more than 9, as in the Hawaiian Islands. cones built ot basic lava usually have broad, flattened domes (Fig. 305 B), since such lava cools at a low temperature and consequently may remain liquid for a considerable time and flow long distances before solidifying. The Hawaiian and Icelandic volcanoes are examples. If stiff, viscous lava is dis- charged, the slope may be very steep (Fig. 307). Cones made of a combination of lava and ash are more common than any others and are usually steep-sided. The vol- cano Fuji, so often pictured by the Japanese artists, is of this type, as are many of the highest volcanoes of the world. A volcanic cone is seldom sym- metrical, since if it has been long in existence it has suffered many changes (Fig. 306). The irregular outline of Vesuvius, as has been seen, is the result of the blowing off of the greater part of an earlier crater, so that the present cone is partially surrounded by Mt. Somma, a remnant of the ancient crater. The slopes of Etna are roughened by scores of parasitic cones. The volcano Colima, in Mexico (Fig. 308), would be beautifully symmetrical were it not for a cone formed on the flanks of the mountain in 1869. This secondary cone is crater- less, showing that near the close of the eruption the lava was so FIG. 306. Outlines of volcanic cones : A y a cone formed of ash; B, a cone from which the top was blown by a great explo- sion; C, a caldera formed by faulting. VOLCANOES AND IGNEOUS INTRUSIONS 313 PHYSICAL GEOLOGY FIG. 308. Volcano Colima and a secondary cone on the left. stiff that it solidified as soon as it reached the surface. The profile of a cone depends also, to a greater or less degree, upon the force and direction of the wind during eruptions, upon the position of the crater, and upon the amount of erosion which it has suffered. Shape of Craters. The shape of the crater of a volcano depends both upon (1) the violence of the explosions, the diameter of the crater of an explosive volcano being, in general, proportional to the violence of the eruption; and upon (2) the character of the materials. A crater has steep, rugged inner walls when lava and coarse cinders are ejected (Fig. 309 A), but a much less steep slope when dust and fine ash are thrown out and fall back into it (Fig. 309 B). Erosion of Volcanic Cones. Up to this point we have discussed the phenomena of an eruption, the shape of cones and craters, and other features connected with recent volcanoes, but aside from these observations, we have learned little of the internal structure of volcanic cones. The structure is, however, revealed to us by an examination of ancient volcanoes which have been deeply eroded by atmospheric agencies or by the sea (Fig. 310). Great FlG - 3Q9- A, a cone formed of coarse i i i fragments; B, a cone formed of ash. explosions also, as we have seen (After Haug ) in the case of Krakatao, expose the internal structure to some extent, and it is also brought to light when the top of the volcano sinks in, as in the case of Crater Lake, Oregon (p. 309). VOLCANOES AND IGNEOUS INTRUSIONS 315 As long as a volcano remains active, the ravages of rain and torrents are repaired by the material ejected, but when it becomes extinct the work of denudation contin- ues uninterruptedly. The rate of erosion varies greatly, de- pending upon the nature and structure of the materials and upon the climate. Cones composed of coarse cinders are likely to endure a time than longer FIG. 310. Rocks, St. Paul Island; a volcanic cone dissected by the waves until the crater has been reached, forming a harbor, those of dust. They are more porous and therefore absorb the rain falling on them to so large a degree that little water is left for erosion. Even before a volcano becomes extinct, deep V-shaped valleys are cut into its sides. We find also that the dust and ash are in layers, and that sometimes black beds (Fig. 311) composed of disin- tegrated ash and humus, varying from a few inches to several feet in thickness, are interbedded with the ash. These black beds are ancient soils and prove that in the past the volcano ex- perienced many years of inactivity, which were followed by eruptions. After prolonged erosion it often hap- FIG. 311. A ravine (baranca) in the side of the vol- cano Toluca, Mexico. The light-colored deposit is vol- canic ash ; the dark bands are ancient soils which prove long periods of quiet after periods of activity. pens that long, wall-like bodies of hardened lava, called dikes (p. 324), are exposed. These dikes were formed during eruptions, when the force of the explosions or the pressure of the column of lava in 316 PHYSICAL GEOLOGY the vent was so great as actually to rend the cone. Into these cracks the lava was forced and cooled. It will readily be seen that a cone buttressed by dikes will be greatly strengthened, and that such a cone will be better able to withstand erosion than one composed entirely of fragmental materials. Necks and Plugs. After the upper portion of a cone has dis- appeared, the neck\ig. 312), as the compact lava or debris filling FIG. 312. Diagram illustrating the destruction of volcanoes. (After A. Geikie.) the vent is called, is exposed. The neck is composed either of lava or of the rocks or other fragmental materials which fell back into the crater and were consolidated to form a volcanic breccia. They vary in diameter from a few yards to two miles. Volcanic necks or plugs, when exposed by erosion, are often conspicuous features of the landscape. Many ex- amples are to be found in North America. From Mon- treal one can see several hills of this origin. In New Mexico, Arizona, California, and other western states of the United States volcanic necks are to be seen. They are not uncommon in por- tions of Europe, where they are frequently the sites of castles or churches (Figs. 313, 314). When erosion has succeeded in entirely tearing down a volcanic cone, it is often found that the neck pierced the surrounding rock without the aid of a fissure or fault, and that it is independent of the folds of the rocks. The great diamond mines of South Africa are FIG. 313. Volcanic neck upon which a chapel has been built. Le Puy, France. VOLCANOES AND IGNEOUS INTRUSIONS 317 FlG ' 3 ' 4 ' lcanic " eck u Mexico. (Photo. D. W. Johnson.) regi n ' New located in the necks of volcanoes, the brecciated rock of which is called "blue ground " and con- tains the gems. These latter necks have a diameter of 300 to 1000 feet. Age of Volcanoes in the United States. In regions of ex- tinct volcanoes every stage in the process of demolition may be studied (Fig. 315 A, B), from the perfect cone, whose slopes have as yet barely been touched by erosion, to that in which the only evidence that a volcano formerly ex- isted is to be found in a spot of igneous rock, a few feet or a few hundred feet in diameter, sur- rounded by sedimen- tary or other rock. The various stages in the erosion of vol- canic cones ,are well shown in the western United States, where every gradation may FIG. 315. Diagram A shows an active or recently extinct volcano with widespread lava flows at its base. Diagram B is the same region after prolonged erosion. The ash of which the cone was composed has been eroded away, leaving the volcanic neck pro- truding. The lava flows have been cut by erosion into flat-topped hills or mesas. In the section on the front of A the former successive positions of the streams are shown, their courses having been diverted as they were filled with lava from the volcano. (Modified after Davis.) PHYSICAL GEOLOGY be seen from young cones, such as Lassen Peak, California, which was active in 1914-1915, to those which have been worn down to their roots. Mt. Shasta, California, 14,350 feet high (Fig. 316), is a good FIG. 316. Mt. Shasta, California, a partly denuded volcanic cone. example of a volcano which has suffered much erosion, but Mt. Hood, Oregon, is still more worn, the sides being deeply trenched by ravines and only a part of the wall of the crater being left. DISTRIBUTION AND NUMBER OF VOLCANOES Number of Volcanoes. It is impossible to determine accurately the number of active volcanoes, since some that appear to be extinct may be merely dormant, and others that have recently been active and from which steam is still rising, may have been in eruption for the last time. It is, moreover, sometimes difficult to distinguish between independent and subsidiary vents. It is safe to say that there are approximately 325 active volcanoes, of which one third are on the continents. Distribution. A glance at a map of the world in which the vol- canoes are conspicuously indicated (Fig. 317) shows some striking features of their distribution. It is seen that they are not scattered VOLCANOES AND IGNEOUS INTRUSIONS 319 haphazard over the world, but are for the most part concentrated along lines or belts near the edges of the continents, and dot limited areas of the oceans. The volcanic belts are not continuous, how- /ft ' i r e 3 -Q .i -3 u s I K. CLELAND GEOL. 21 320 PHYSICAL GEOLOGY ever, but are interrupted in many places by areas in which no volcanoes occur. Although volcanoes are usually situated along the borders of con- tinents, this is not always the case ; some volcanoes in Ecuador, for example, are 150 miles inland, and in East Africa the volcano Kirunga is 600 miles from the coast. The most important of the volcanic belts almost encircles the Pacific Ocean, extending from the southern tip of South America northward along the Andes on the western coast of that continent, through Mexico, and along the western coast of North America to Alaska. From Alaska it curves westward and southward through the Japanese and Philippine archipelagoes to New Zealand and to the Antarctic volcanoes. The borders of the Atlantic, in contrast to those of the Pacific, are almost free from volcanoes. Two important belts, however, occur in this ocean ; one stretches from Iceland south to St. Helena and includes the Azores and other volcanic islands ; the other includes the West Indies and the shores of the Mediterranean Sea. Cause of Distribution. A study of regions of volcanic activity brings out the fact that they have recently undergone severe move- ments, or are actually being deformed at the present time. In other words, volcanoes are situated where mountain-making forces (p. 358) are active, and where, consequently, the earth is much fissured and fractured (p. 360). The fact that belts of active volcanoes are usu- ally found where mountain ranges are near or parallel to great deeps in the neighboring oceans has given rise to the belief that the eleva- tion of the strata of which mountain ranges or islands are composed is compensated by a sinking of the ocean bottom, and that as a result of these movements lava and ash are ejected to form volcanoes. It is to be noted in this connection that volcanic activity tends to die out in the older rocks and to appear in those of later date. It is evident from the above that the problem of the distribution of volcanoes is an important one, since on its solution must depend in a large measure the much more general one of the cause of volcanism. Ancient Volcanoes. The volcanoes of the past had as a rule a different distribution from those of the present. For example, Great Britain and central France were the scenes of intense volcanic activity ; the Connecticut valley, northern New Jersey, and many of the western states (Wyoming, Colorado, New Mexico, Idaho, and others) have ex- perienced great lava flows, or many and great volcanic eruptions. At a much earlier period in the earth's history (Pre-Cambrian) volcan- VOLCANOES AND IGNEOUS INTRUSIONS 321 ism appears to have been widespread in eastern and central Canada, and large areas in Wisconsin and Minnesota are underlain chiefly with volcanic rock. Throughout geologic history periods of unusual volcanism have been followed by others of comparative quiet. The last important period of volcanism preceded the advent of the Great Ice Age (p. 643), and it is possible that we are now living in the de- clining phases of the activity of that time. IMPORTANCE OF VOLCANISM TO MAN (1) Beneficial Effects. Volcanic regions are interesting not only because of the striking character of their phenomena and scenery, but also because of their economic value. Abundant springs are usually found in the neighborhood of volcanoes. The ashes from recent erup- tions often form a fertile and easily worked soil. When the surfaces of flows composed of dark-colored lavas (basic, p. 329) are decomposed, they furnish a soil which contains all of the elements needful for plant life, many of which are lacking in granite and other soils ; in Central America, certain regions are benefited far more than they are injured by the showers of volcanic ash, because of the increased fertility resulting from the minerals which these contain. Vesuvius is surrounded by a ring of villages in spite of the danger of eruptions, and the flanks of Etna support an extremely dense population. Of benefit to man, also, are the many lakes that rest in the craters of inactive volcanoes. The lakes of the Alban Hills near Rome, as well as Lake Bracciano and other lakes of Italy, are crater lakes. Crater Lake in Oregon (p. 309) is also a famous example. Lakes have also been formed by the damming of river valleys by lava streams. At the foot of Mt. Shasta, California, are rich tracts of alluvium, the sites of lakes formed in this way and later filled by de- posits which now constitute rich agricultural land. Many important ore deposits (p. 371) have resulted from the intrusion of molten rock. (2) Harmful Effects. Although volcanoes are sometimes indi- rectly beneficial to man, this does not compensate for the destruc- tion of life and property which result from an eruption. In addition to the destruction wrought by the fall of ash and the outpouring of lava, great disaster has been caused in other ways. Many times in the past, great floods have been brought about by the discharge of the water from lakes which rested in craters, and by the melting of 322 PHYSICAL GEOLOGY the snow and ice on and near the summits of the volcanoes. During an eruption of Cotopaxi in 1877 enormous torrents of water and mud produced by the melting of the snow and ice on the cone, together with great blocks of ice from the glaciers, rushed down the mountain, burying fields and villages beneath mud, lava, and ice for a distance of 10 miles. In contrast to the above is the existence of a great sheet of ice on Mt. Etna, which for nearly one hundred years has been protected from evaporation and thaw by a sheet of lava which over- flowed it without the heat being sufficient to melt it. As torrents of water rush down the side of a volcano, they not only erode it deeply but are also soon converted into streams of mud, when dust and ash are abundant on the cone. Herculaneum (p. 303) was buried in this manner, and in Java in 1881 torrents of mud and water from Galoon-goon flooded the rivers to such an extent that every village and plantation in this populous region was entirely de- stroyed for a distance of 24 miles. During a comparatively recent eruption of Vesuvius so much hy- drochloric acid was dissolved in the rain water which fell through the clouds of volcanic gases that the vegetation for miles around was injured by it. A subsidence of the land sometimes follows an eruption, as has been noted in the case of the Temple of Jupiter near Naples (p. 229). The sinking is probably brought about either by the withdrawal of molten rock from beneath the affected areas or by the weighting of the ad- jacent land by the ejected material, or by a combination of both. Volcanoes and Climate. 1 It has been shown that volcanic dust in the high atmos- phere decreases the intensity of solar radiation in the lower atmosphere. Therefore the average temperature of the earth is decreased when dust is present. From these observations some investigators have concluded that volcanic dust must have been a factor, possibly an important one, in the production of many climatic changes of the past. It has not, however, been shown that the periods of glaciation coincided with prolonged volcanic outbursts. SUBORDINATE VOLCANIC PHENOMENA There are a number of phenomena which are the direct result of heat and are usually connected with present or comparatively recent volcanism. 1 Bull. Mt. Weather Observatory, Vol. 6, Pt. I, 1913 ; Smithsonian Misc. Coll., Vol. 59, No. 29, 1913- VOLCANOES AND IGNEOUS INTRUSIONS 323 Mud Volcanoes. Cones built of mud with small craters in their summits are called mud volcanoes. They vary in height from a foot or two to more than a hundred feet ; some are continuously active and some are intermittent; some are quiet and a few are violently eruptive. In order that mud volcanoes may be formed it is necessary that (i) steam be present and that (2) it rise through a surface layer of clay which will make mud when wet. As the steam rises through the mud, it carries some up with it and so builds a cone. As such cones are composed of soft material, they have a short life, since they are readily destroyed by rains. The heat and steam necessary for the formation of mud volcanoes come from lavas which are present at a comparatively short depth, or may be produced by chemical action, such as occurs when sulphur is FIG. 318. Mud volcanoes, Lower California. (Photo. D. T. MacDougal.) oxidized. Mud volcanoes are found in the Colorado desert, in Lower California (Fig. 318), and in other parts of the world. The " paint pots " of the Yellowstone National Park, so-called because of their shape and varied colors, are miniature mud volcanoes. The eruptions, produced by the bursting of bubbles of steam, occur fre- quently, and can be safely and easily studied. Solfataras. Lava streams sometimes retain their heat hundreds of years after they have been poured out in sufficient amount to convert the water which percolates to them into steam. This is also true of the lava in the craters of volcanoes. Although meteoric waters probably furnish the greater amount of water which is returned as steam, yet the quantity of steam exhaled directly from lavas appears to be considerable in some cases. The term solfatara is used for a volcanic vent or area in which only gases and steam are discharged, the name being derived from the volcano Solfatara near Naples, which has been giving off only steam and gases since its last eruption (in 1198). ,-,. Geysers (p. 67) are found only in regions in which acidic lava is still hot, and hot, carbonated springs (p. 66) occur in similar situations, although their heat does not always have this origin. PHYSICAL GEOLOGY INTRUSIVE OR PLUTONIC ROCKS Igneous rocks have either been extruded on the surface in the form of volcanic products and lava flows, or they have failed to reach the surface and have consolidated beneath it. The latter are called plutonic (after Pluto, the Greek god of the lower world) or intrusive rocks. The quantity of lava which failed to reach the surface is probably many times greater than that which was poured upon it. The deep-seated intrusive rocks are never vesicular (full of gas blebs), since their contained gases were prevented from expanding by the over- lying pressure. They are coarsely crystalline because they cooled so slowly, owing to the fact that they were deeply buried, so that the crystals had time to grow. Such rocks are exposed at the surface only by the erosion of the rock strata which formerly covered them. Doubtless many such masses, now exposed at the surface, were at one time the deep-seated reservoirs from which the lava of volcanoes came. There are some rocks which link the extrusive and the plutonic rocks and may be classed simply as inter- mediate. The mechanics of igneous intrusions is discussed on page 334. It will be shown that intrusions probably work their way toward the surface largely by sloping (p. 337). When they have reached within a few thousand feet of the earth's surface, they take advantage of any planes of weakness, such as joints and faults, and continue their journey through fissures. /. Injected Masses Dikes. Dikes are masses of igneous rock which have FIG. 3 19- - A vertical, branching dike. hardened in more or less verti- (Photo. F. B. Sayre.) calcracks or fissures (Fig. 3 19). VOLCANOES AND IGNEOUS INTRUSIONS 325 They vary in width from a fraction of an inch to several hundred feet. Their length may be considerable ; one in the north of England runs from the coast inland for about 100 miles, and a length of 5 to 20 miles is not un- common. In Scotland a series of lava dikes run par- allel to each other for a dis- tance of from 20 to 30 miles, while on the coast of New England and in many other parts of North America they are very common. When the surrounding rocks decay more easily than the dike rocks, the latter project above the surface of the ground like walls (Fig. 320) and are sometimes used in Scotland as inclosures. Near Spanish Peaks, Colorado, a FIG. 320. Dikes, the Devil's slide, Weber's Canyon, Utah. (U. S. Geol. Surv.) dike stands as a great wall 100 feet high. Occasionally the dike rock weathers more readily than that which it cuts (Fig. 321), in which case the position of the dike may be indicated by a trench-like hollow. When the dike and the surrounding rocks are about equally resistant, no topographic features result. The texture of the rocks of dikes de- pends upon a num- ber of conditions : (i) if the fissure through which the ;>.':: : : . : :.y '/.'''/ '-'':':: "..'-.:->.y.>v V. = ~ 2 _ - = ' 'l ' 1 L T L_J H 1 1 = I , ' | 1 . ' | U -T- EC; d] ~rd FIG. 321. Diagram showing the effect of weathering upon two dikes (shown by horizontal lines), one of which is more resistant than the surrounding rock and the other less resistant. lava was forced was narrow, the rock of the dike is either glassy or finely crystalline; (2) if, however, the fissure was wide, the dike rock may be coarsely crystalline, with narrow margins of less crystalline or glassy rock. 326 PHYSICAL GEOLOGY Dikes are found cutting rocks of all ages, and they extend across the country without reference to topography. Sills. Lavas which have been forced between sedimentary strata and have formed sheets which have a small thickness as compared FIG. 322. Diagram illustrating the relation of the Palisades of the Hudson (vertical lines) to the strata in which this sill was intruded. with their extent are called sills (Fig. 322). Sills sometimes extend long distances along the same bedding plane, but often cut across from one stratum to another. They vary in thickness from a few feet to several hundred feet and sometimes have an extent of many square miles. When they have been exposed by erosion, they can be distinguished from extruded lavas by the absence of vesicular FIG. 323. The Palisades of the Hudson. The sheer face of the upper portion is due to the vertical jointing of the trap, and to the more rapid erosion of the weaker, underlying rock. (Photo. D. W. Johnson.) lava on the upper surface. The Palisades of the Hudson, which ex- tend for 30 miles along the west bank of the river as a bold cliff several hundred feet high, form a part of a sheet of intrusive lava (sill) which is underlain by sandstone and was formerly overlain by other sedi- mentary strata (Fig. 323). VOLCANOES AND IGNEOUS INTRUSIONS 327 Laccoliths (Greek, lakkos, a cistern, and lithos, stone). This term has been given to mushroom-shaped intrusions of lava which have been forced along bedding planes and have domed up the overlying strata. They are formed when molten rock rising through a pipe or fissure is unable to break through the overly- ing rock and spreads between the strata, lifting them and thus producing domelike elevations (Figs. 324, FIG. 324. Diagram illustrating the form and relations of dikes, A and D ; sills, C and E ; and a laccolith, B. 325). The difference between a sill and a laccolith is conse- quently a difference in the degree of the doming of the overlying strata. Laccoliths may be a mile or more thick and a number of miles in diameter. Mountains of considerable height have been formed in this way. The Henry Mountains of southern Utah, the Elk Mountains of Colorado, and many other elevations in the Rocky Mountains are laccoliths (Fig. 325). Laccoliths are composed of lava which was probably FIG. 325. Diagram of a laccolith, showing the rela- tion of the igneous intrusion to the overlying and under- lying strata. stiff and viscous and could consequently more easily lift the strata than force its way between them. The lavas of sills, on the other hand, were probably quite fluid and therefore could spread long distances. //. Subjacent Masses Stocks. The name stock is applied to large bodies of igneous rock lying in the midst of other formations. Stocks are usually circular or elliptical in outline and vary from a few hundred yards to many square miles in extent, usually increasing in size downward (Fig. 326 A, B). Since they are composed of more resistant rock than 328 PHYSICAL GEOLOGY that in which they were intruded, they often form knob-like elevations and are consequently often called bosses. Stocks resemble volcanic necks (p. 316), but are usually larger; the term neck, more- over, is employed only when there is evidence that it represents the chimney of a volcano. Batholiths (Greek, bathos, depth, and lithos, stone) are great irregular masses of igneous rock which stopped in their rise many feet from the surface of the earth, but have since been exposed by erosion. They are often many hundreds of square miles in area and may be FIG. 326. Section A and map B of a stock or boss. The granite intrusion being more re- sistant than the enclosing rock forms a hill. considered as merely very large and irregular stocks. In the aggregate these bodies cover many thousands of square miles, and although less striking are much more important than volcanoes. Some Effects of Intrusions. The rock with which a molten magma comes in contact is more or less changed ; the larger and hotter the intrusions the greater being the effect. This phenomenon will be discussed under metamorphism (p. 341). It is believed that some of the explosions which have taken place on or near volcanoes were due to the presence of molten rock at a short distance below the surface. Since igneous rocks are usually harder than those into which they are .intruded, they are often left in relief as buttes (p. 106) and bosses, as the land is reduced by erosion. Since igneous rocks are composed of minerals which differ in com- position and often in color and therefore expand and contract differ- ently when heated and cooled, we find in desert and tropical regions that granites and other igneous rocks exfoliate (p. 32) under the influence of diurnal temperature changes (p. 31), producing sphe- roidal bowlders which are often poised on rounded surfaces. These rocks resemble glacial bowlders, and the smooth surfaces on which they rest, roches moutonnees (Fig. 142, p. 157). VOLCANOES AND IGNEOUS INTRUSIONS IGNEOUS ROCKS 329 Igneous rocks, as we have seen, have consolidated from a state of fusion. The character of the rocks thus formed depends principally (i) upon the chemical composition of the molten mass and (2) upon the rapidity with which the magma cooled. Other conditions, such as fluidity and pressure, are likewise important. Subdivisions Depending upon Chemical Composition. Igneous rocks which contain a large percentage of silica (65 per cent, or more) are termed acid rocks, silica being an acid-forming oxide. Acid rocks are usually light-colored when crystalline, and are lighter in weight than basic rocks which contain much less silica (55 per cent, or less) and a correspondingly larger amount of the bases, such as potash, soda, lime, and magnesium. Basic rocks are usually dark-colored and fuse at a lower temperature (p. 299) than acid rocks. They are the common extrusive rocks and sometimes cover tens of thousands of square miles of the earth's surface, and when weathered often produce soil rich in plant food. Subdivisions Depending upon Texture. The term texture as applied to igneous rocks refers to their smaller features. When a rock is described as being granitoid, or having a granular texture, the reference is to one in which the crystals are distinct and are all of about the same size. A rock with afelsitic texture is one composed of a mass of very fine microscopic crystals. A rock is described as glassy when it is made up largely or in part of glass in which no definite crystals are to be seen. The rate of cooling appears to be the really important factor in deter- mining the texture of igneous rocks, although other conditions have considerable effect. The molten magma from which granitoid rocks were crystallized was so deeply buried that the rate of cooling was slow, thus giving an opportunity for the molecules of the same chem- ical composition to gather to form large crystals and, consequently, granitoid or granular rocks. Felsitic rocks 1 are the result of somewhat more rapid cooling, and are found on the margins of great masses of granitoid rocks which did not reach the surface, or in offshoots from them in the form of dikes. Glassy rocks are those which cooled so rapidly that the minerals had little opportunity to form. Rocks with a glassy texture, consequently, occur chiefly in surface flows and on the margins of dikes. 1 Felsite is also often the product of the devitrification of glassy rocks. 330 PHYSICAL GEOLOGY CLASSIFICATION OF IGNEOUS ROCKS I. Coarse-grained Igneous Rocks The rocks included in this group are those whose mineral grains are approximately of equal size and are large enough to be distinctly seen. Granite. Granites are composed of quartz, feldspar, and usu- ally of smaller amounts of either mica or hornblende. The grains of feldspar are usually easily distinguishable because of their shiny (cleavage) surfaces and their opaque white, gray, or red color. The quartz grains vary in tint from colorlessness to smoky gray, and can usually be recognized by their glassy luster and irregular frac- ture. Mica may be either muscovite or biotite, and may be told by its brilliant cleavage surface. The thin leaves, unless too small, can be easily separated with the point of a penknife. Hornblende occurs in green to black opaque grains or needles. Other minerals, such as pyrite or garnet and other less common minerals, may also be present. Numerous names are given to granites, some of them (commer- cial) depending upon their color and their desirability for building or monumental purposes, such as red, gray, yellow; while others are locality names. The color of the stone depends largely upon that of the feldspar, and upon the relative abundance of dark minerals. A red granite owes its color to its red feldspar; a gray color may be due either to the color of the feldspar alone, or to the combination of black hornblende or mica, and white feldspar. Syenite is a rock which may be described as a granite without quartz. It very closely resembles a granite, and is usually sold under the latter name. Diorite is a granular igneous rock composed of hornblende and feldspar of any kind, in which the amount of hornblende usually exceeds that of the feldspar, although the two may be in equal amounts. Its color is dark^ gray^r_greenish. Gabbro is made up of pyroxene (augite), with usually smaller amounts of feldspar of any kind. Gabbro and diorite are often dis- tinguished with difficulty, because of the similarity of hornblende and augite to the naked eye. Peridotite is a dark green to black rock, composed chiefly of such minerals as hornblende, olivine, and pyroxene (augite). VOLCANOES AND IGNEOUS INTRUSIONS 331 //. Compact or Fine-grained Igneous Rocks In this group are included rocks in which the grains are so fine that the individual crystals cannot be distinguished by the naked eye. They are intermediate between the granitoid rocks, composed of clearly distinguishable crystals, and the glasses. No definite line can be drawn between the two groups ; in some dikes, for example, a glass shades imperceptibly into a microcrystalline rock, and then into a coarsely crystalline or granitoid rock. This group is divided into two classes on the basis of color : (i) the light felsites and (2) the dark basalts. (1) Felsites vary greatly in color, but are not dark gray, dark green, or black. To the naked eye the rock has a flinty aspect, but with a lens it is often seen that it consists of mineral grains, too small for determination. When large crystals (phenocrysts) occur embedded in the fine-grained " ground mass," the rock is called a felsite porphyry. Porphyries contain feldspar phenocrysts (Greek, phain- esthai, to appear, and krystallos, crystal). If quartz is also present, they are known as quartz porphyries, and if hornblende is con- spicuous, they are called hornblende porphyries. Felsites occur in dikes and sheets. (2) Basalts form a very large and important group of igneous rocks. They are all heavy black, gray, brown, or greenish rocks of fine tex- ture, and have a wide distribution, covering many thousands of square miles of the earth's surface. The name trap is also used to include basalts and any dark-colored, heavy igneous rocks whose mineral constituents have not been determined. When the air cavities of vesicular basalts or of other igneous rocks are filled with minerals, the rocks are called amygdaloidal. This is one mode of occurrence of copper in some of the mines of northern Michigan (p. 396). ///. Glassy Rocks Rocks which are composed partly or wholly of glass are included in this group. They were formed as stated (p. 329), when molten rock solidified rapidly. They are therefore lavas which were either poured out on the surface, or in crevices where they were subjected to rapid cooling. One sometimes finds the sides of dikes glassy, while the interior is crystalline. The texture of glassy rocks is sometimes vesicular (Fig. 294) and sometimes pumiceous. 332 PHYSICAL GEOLOGY FIG. 327. A hand specimen of obsid- ian, showing the characteristic conchoi- dal fracture. (U. S. National Museum.) Obsidian or Volcanic Glass (Fig. 327) is pure, natural glass, entirely or nearly devoid of crystal grains. It is usually jet black in color, but is sometimes gray, green, red, or yellow. Because of the sharp edges which form when it is broken, it was highly prized by the Mexicans and other primitive peoples for the manufacture of sharp implements, such as knives and arrowheads. Pitchstone is a variety of ob- sidian in which the luster is resin- ous or pitch-like. The chemical and other differences between this rock and obsidian are slight. Pitchstones are variable in color. When conspicuous crystals are scattered through the rock, it is called pitchstone porphyry. FRAGMENTAL VOLCANIC ROCKS Rocks formed from the material thrown out by volcanoes are in- cluded under this head and are made by the consolidation of dust, ashes (material the size of a shot), lapilli (the size of a nut), and bombs (pieces the size of an apple, or larger). Tuff. When the rock is composed entirely of the finer kinds of volcanic detritus, it is called volcanic tuff. Rocks of this type are light in weight and usually loose in texture, although some are almost as compact as felsites. Tuffs contain fossils if the dust and ashes of which they are composed fell on a land surface covered with vege- tation, or in water in which marine organisms were living. Some of the rock through which the Panama Canal was cut is a tuff containing 1 marine shells. Tuffs are widely used in Mexico for building stones. Volcanic Breccia. This is a rock compose^ of angular fragments of volcanic rock, bombs, etc., which are cemented together with ash and dust. 1 For a more detailed study of igneous rocks the student is referred to L. V. Pirsson, Rocks and Rock-Minerals ; and J. F. Kemp, Handbook of Rocks. VOLCANOES AND IGNEOUS INTRUSIONS 333 FIG. 328. Basaltic columns in a lava flow near the city of Mexico. Columnar Struc- ture of Lava. A striking feature of many ancient lava flows whose lower portions have been exposed to observa- tion by erosion is their columnar struc- ture, the lava being broken up into an- gular columns which are often six-sided. If the lava sheet is horizontal, the col- umns are vertical (Fig. 328) ; if it has been intruded into a fissure (dikes), the columns are horizontal (Fig. 329). One may ob- serve similar joints in dried mud and starch, but in these substances the sides are much less regular. The explanation of columnar jointing is to be found in the contraction of the lava, resulting from cooling and loss of gas, and the conse- quent cracking of the rock. Since the least expenditure of energy is required to relieve the strain when three cracks radiate from* equidistant points at 329. A lava dike (depressed) showing the basal- tic jointing at right angles to the walls. Maine. Angles of I2O , the (Photo. F. Bascom.) formation of six-sided 334 PHYSICAL GEOLOGY columns usually results, and the direction of the columns is at right angles to the cooling surface. The reason for the horizontal position B FIG. 330. Diagrams showing the origin of basaltic jointing. In shrinking, the least number of cracks that will relieve the tension in all directions, A, is three. Similar radiating cracks from other centers complete the six-sided prism, B. When cracks fail to develop about some one point, a five-sided prism, C, results. (Modified after Chamberlin and Salisbury.) of the columns of vertical dikes and the vertical position of those of lava flows is thus explained (Fig. 330 A-C). AGE or IGNEOUS ROCKS The exact age of ancient volcanoes or of igneous intrusions can seldom be ascertained, but the relative age is often known. The relative age is determined as follows : a volcanic neck is clearly younger than the rocks which it penetrates ; a laccolith or sill is of later age than the beds in which it was intruded, and a lava flow is more recent than the formations over which it spreads. The eruption or intru- sion in each case could not have taken place before these rocks were laid down. If on the other hand pebbles of igneous rock are found in sedimentary rocks, we know that the rocks from which they were derived were at the surface before the sediments were deposited, or while the deposition was taking place. For example, if Devonian strata are cut by a volcanic neck, we know that the neck is younger than the Devonian, and if pebbles from this same neck are found in Middle Carboniferous sediments, it is evident that the lava was prob- ably intruded in early Carboniferous times. Sometimes the presence of fossils in volcanic tuff shows definitely at what time the eruption occurred. THEORIES OF VOLCANISM So many theories of volcanism have been offered that it is impossible in an introductory volume to do more than briefly indicate a few of VOLCANOES AND IGNEOUS INTRUSIONS 335 them. The theories may be classed under three heads : (I) those which assume a molten interior; (II) those based upon the assump- tion that the earth is solid from the surface to the center; (III) those holding that a few miles below the surface a zone of rock exists which is either molten, or at any rate in a non-crystalline condition. /. Theory Based upon the Assumption that the Interior is Molten The theory of a molten interior is now held by few geologists because of the many objections to it (p. 273). In the earlier days of geology when this theory had general acceptance, the difficulty of accounting for the independence of volcanic eruptions brought forth much discussion, and a number of modifications to the theory were sug- gested. If all lavas came from one great reservoir, it is evident that according to the law of hydrostatics eruptions would be simultaneous, or in two adjacent vents, from the lowest one. II. Theories Based upon the Assumption that the Earth is Solid (a) Heat by Friction. This theory is based on the fact that heat is developed by friction when rocks grind and crush each other. It is held that when great earth blocks (segments) move past each other, the pressure and friction along the lines of movement develop heat on a large scale. If fluxes (rocks which upon uniting with others produce a substance that will melt readily) are present to lower the melting point of the rock silicates, the heat may be sufficient to produce molten rocks and volcanoes. Since all rocks contain more or less water, steam under immense pressure will be developed upon the fusion of the rock. Explosions of the steam developed in this way are believed to be competent to drill channels to the surface, and to eject the molten rock through the chimneys thus formed. The intermittent action and extinc- tion of volcanoes, according to this theory, are dependent upon the movement of the earth's segments. It should be noted in this connection that no observations have been made of faults the walls of which are fused as a result of slipping. 1 (b) Formation of Lava Reservoirs by Relief of Pressure. This theory rests on the assumption that at moderate depths the heat of the earth is so great that the solid state of rocks is maintained only by the pressure of overlying rocks. If this assumption is correct, it is only necessary to show that the pressure of highly heated rocks can be relieved. This is thought by the advocates of the theory to be accom- plished when deeply buried, sedimentary strata are folded. If a stratum strong enough to sustain the weight of the overlying rocks is arched, and the underlying rocks are thus relieved of some of the pressure, the latter may melt. A volcano or fissure eruption may then occur if a crack to the surface is present through which the lava can force its way. The supply of lava would depend upon the amount of molten rock under the arch, and the extinction of the volcano would result from its exhaustion. Two strong objections to the theory are : (i) the difficulty of accounting for a tem- perature in sedimentary rocks high enough to fuse them, and (2) the difficulty of ex- plaining the presence of sedimentary rocks of basaltic composition (p. 331) of suffi- cient thickness under an arch to be a source of the lava of massive plateaus, 1 Schwartz, E. H. L., Causal Geology, p. 241. CLELAND GEOL. 22 336 PHYSICAL GEOLOGY (c) Liquid-thread Theory. 1 This theory assumes that the earth grew by the slow accession of meteorites (planetesimals), varying greatly in size and composition (p. 386), and that the interior, though solid, has become very hot as a result of the com- pression of the interior masses by the accumulation of the outer envelopes. In a globe composed of masses varying greatly in composition and fusibility, it is evident that some particles will be molten while others are still solid. As a result of the strains to which the earth is subjected, the liquid portions gradually move toward the surface, uniting in their upward course into larger and larger threads (Fig. 331). Because of their heat these threads finally reach the zone of fracture by fusing and fluxing. When such threads of lava attain the zone of frac- ture, they take advantage of any fissures or fractures which exist, and are poured out on the surface as fissure or volcanic eruptions. The intermittency of volcanic action is due, according to this theory, to temporary deficiencies in the supply, and the force of expulsion is produced especially by tidal and other stresses and by the slow pressure brought to bear on the threads of liquid rock, until the upper level is reached, where the expansion of the occluded gases begins to operate. This hypothesis, as has been pointed out, is based upon the assumption that the earth has never been in a molten con- dition, and that its interior is composed of a heterogeneous mass varying greatly in composition. If the observations upon FIG. 331. Diagram illustrating Cham- berlin's theory of volcanism. S is the surface of the earth ; aa', the zone of frac- ture ; a/, zone of flow ; jfc, interior portion whose temperature rises from the surface melting point at ff to a maximum at c ; w, threads or tongues of molten rock rising from the interior to various levels, many of these lodging within the zone of frac- ture as tongues, batholiths, etc. PP are explosive pits formed by volcanic gases derived from tongues of lava below. (After Chamberlin and Salisbury.) which the following hypothesis (abyssal injection) is based are well-founded, namely, that the crust of the earth is com- posed of acid (granitic) rock and that this is underlain by a basic (basaltic) sub- stratum, the hypothesis cannot stand, since the latter holds that the earth was formerly molten at the surface. ///. Abyssal Injection Hypothesis 2 This hypothesis is based both upon laboratory experiments and upon many obser- vations of the occurrence and relationships of igneous rocks in various parts of the world. It should be distinctly borne in mind, however, that the hypothesis is as yet unproved. 1 Chamberlin and Salisbury, Geology, 2d ed., Vol. i, p. 629. 2 Daly, R. A., The Nature of Volcanic Action: Am. Academy of Arts and Sciences, Vol. 47, June, 1911, pp. 47-122; and Igneous Rocks and their Origin. VOLCANOES AND IGNEOUS INTRUSIONS 337 It assumes that there exists, at a depth estimated at 40 kilometers (about 23 miles), a basaltic (basic) substratum which underlies an almost universal shell of acid rock (granite, etc.). Because of the pressure of the overlying rocks, this substratum of actually or potentially fused rock is so rigid as to act as a solid. It holds that all igneous action is the result of the mechanical intrusion of the substratum basalt into the over- lying crust. The intrusion of the magma is accomplished largely by " stoping"; that is, cracks in the overlying rocks are entered by the molten mass, blocks are wedged off and are ultimately absorbed in the magma. At the contact, solution takes place to some extent, but this is believed to be of secondary importance to stoping. The vent of the volcano or fissure for the last few hundred or thousand feet may have been opened by explosions or by fissuring. Once the movement of the molten magma is started, the original heat of the intrusion is maintained by chemical and exothermic reactions (the heat liberated in the formation of chemical compounds). The explosiveness of volcanoes, according to this theory, is the result of the original gases of the molten rock, as well as of the water which the magma absorbs from the intruded rocks in its ascent. The cause of the extinction of volcanoes is shown in the diagram (Fig. 332). The active vent is situated at the highest point of the injected lava, and it is in this place that the gases of the lava accumulate. The temperature of the lava in such situations is believed to be not only that of its primal heat but also to be increased by chemical re- actions and by other means connected with the pres- ence of the gas. Vents become extinct when, be- cause of the higher position of the lava in other locations, the gases which cause the fusion accumulate elsewhere. According to this theory, the composition of the lava ejected from a volcano depends upon whether it is composed entirely of the basaltic lava of the substratum, or is a mixture produced by the solution of the rock through which the basaltic lava has passed. RESUME OF PRESENT KNOWLEDGE OF VOLCANISM There is no agreement as to the origin of lava ; (i) some investiga- tors hold that a portion is derived from deeply buried sedimentary rocks which have a high temperature as a result of the rise of heat from the earth's interior, so that when the pressure of the overlying rocks is relieved, the more fusible strata liquefy; (2) some hold that it is formed as the result of heat produced by friction between great ABYSSAL INJECTION FIG. 332. Ideal section illustrating the abyssal in- jection theory. The middle vent is active because it originates at the highest point in the injected body. The other vents are extinct because of the advantage of the middle vent. The arrows show the movement of the gas ; the solid black, the crystallized portion of the injection. 338 PHYSICAL GEOLOGY blocks of the earth, when the earth's crust is yielding to strains; (3) some, that it is chiefly or entirely primal, i.e., derived from a sub- stratum of unknown thickness. Of these, the last (3) seems to be more in accord with the known facts (p. 337) than the others. The activity of a given volcano is usually independent of all others, as is shown in the history of Mauna Loa and Kilauea (p. 308), which, though forming one great mound of lava, erupt independently. On the other hand, eruptions of Pelee and Soufriere on the West Indian islands of Martinique and St. Vincent have been almost simultaneous. Origin of Volcanic Gases. The problem of the origin of gases and water vapor is to a large extent identical with that of the origin of lava. It has been proved by experiment that all rocks, even the most dense and most crystalline, contain large quantities of gas, so that a comparatively small volume of rock would be sufficient to furnish practically all of the gases and all of the water vapor given off during an eruption even of the first magnitude. It has been held, however, that water vapor, which constitutes the greater part of the emanations of dormant volcanoes, as has been stated, is derived, to a large extent at least, from either sea water or from meteoric water which has percolated down to the molten lava and been absorbed by it. Cause of the Ascension of Lava. Every theory of volcanism must account for the force which raises the lava to the surface of the earth and often throws it as fine dust thousands of feet into the air. There is general agreement that this force is to be found (i) in the tidal and other strains to which the earth is subjected ; (2) in hydrostatic pressure resulting from the weight of the overlying rock ; (3) in the enormous expansive force of the gases dissolved in the molten magma, whether original or derived from other sources ; and (4) to some degree in the expansional energy of the injected mass. Cause of Periodicity. The lava which cools in the throat of a volcano is characteristically tough. Since the cones of explosive volcanoes are built of loose ash deposits of little strength, it is evi- dent that if renewed activity were to result from an explosion alone, an opening would usually be made through the side of the mountain instead of through the crater. The latter is, however, usually the case. There must therefore be some preliminary weakening of the plug, and apparently the only cause for such weakening is to be VOLCANOES AND IGNEOUS INTRUSIONS 339 Crater found in the fluxing (Fig. 333) by intensely hot gas 1 from deep in the earth. When the plug has been formed, heat is developed be- neath it by the compression of the gas, by chemical reaction, and by gas solution. After the plug is shortened by the melting away of its lower end in this intense heat, the gas pressure may become great enough to blow out the remaining part. After the explosion, the lava in the throat of the volcano again cools and a period of inactivity ensues. The cause of extinction is discussed on P- 337- In individual cases, as for FIG. 333. Section of a dormant volcano, showing how the lava plug may be weakened by gas fluxing. The broken line shows the example in that of Stromboli, original , de P th f the so ^/l u f a , nd , the n pr , g ; ress made by the gas. (Modified after Daly.) eruptions occur when gas has accumulated under the scum of lava in the crater in sufficient volume to cause an eruption, after which quiet ensues, the surface of the lava hardens, and the gases again begin to accumulate. Influences of the Atmosphere, etc. Volcanic eruptions seem to be somewhat more prevalent when (i) the atmospheric pressure is high than when it is low, (2) after heavy rains rather than before, and (3) when tidal strains are unusually severe. None of these causes could produce an eruption, but it is probable that the increased weight of the atmosphere over a large area would aid in forcing out the lava, as . would also the weight of the water after heavy rains. Tidal strains would have a similar effect. None of these agencies could be effective unless the eruption was imminent, only a slight additional force being necessary to start it. It has long been noticed that the volcano Stromboli (p. 299) dis- charges a greater quantity of steam and bombs in stormy than in fine weather, and the fishermen make use of it as a " weatherglass " : the increase of activity indicating a falling barometer and conse- quently stormy weather; and a diminution in activity promising fair weather, 1 Such gases, called primeval gases, are believed to come directly from great depths and reach the surface for the first time. They are distinguished from resurgent gases which have a second- ary origin, that is, those which are absorbed from the intruded rock. 340 PHYSICAL GEOLOGY REFERENCES FOR VOLCANOES BONNEY, T. G., Volcanoes, their Structure and Significance. CHAMBERLIN AND SALISBURY, Geology, Vol. i. DALY, R. A., Igneous Rocks and their Origin. DANA, J. D., Manual of Geology. GEIKIE, A., Textbook of Geology. GILBERT ; G. K., Report on the Geology of the Henry Mountains. HITCHCOCK, C. H., Hawaii and its Volcanoes. HULL, E., Volcanoes: Past and Present. IDDINGS, J. P., The Problems of Volcanism. JUDD, J. W., Volcanoes. RUSSELL, I. C., Volcanoes of North America. SCHWARTZ, E. H. L., Causal Geology. SHALER, N. S., Aspects of the Earth, pp. 50-56. TOPOGRAPHIC MAPS, U. S. GEOLOGICAL SURVEY, ILLUSTRATING IGNEOUS ACTIVITY Volcanoes Laccoliths Lassen Peak, California. Henry Mts., Utah. Crater Lake National Park Sturgis, South Dakota. (Special), Oregon. Shasta, California. Laoa Plains Marysville Buttes, California. Bisuka, Idaho. San Francisco Mt., Arizona. Ellensburg, Washington. Island of Kauai, Hawaiian Islands. Lava Sills New York City and Vicinity* (Folio). Holyoke Folio, Massachusetts. New Haven, Connecticut. CHAPTER X METAMORPHISM WHEN either sedimentary or igneous rocks have been affected, either in their mineral composition or in their texture or in both, so that their original character is altered or entirely changed, they are called metamorphic rocks, and the process is known as metamorphism (Greek, metamorphoun, to transform or change). The term, as used here, will be limited to those changes which have resulted from heat or pressure, or both, whether produced locally, as for example by batholiths ; or over large areas by pressure and heat. Contact Metamorphism. The form of metamorphism most easily understood is that produced when sedimentary rocks are cut by great igneous intrusions, such as batholiths. Under such condi- IMPURE ^=1 SANDSTONE QUARTZ SCHIST 8, GNEISS INTRUSIVE GRANITE FIG. 334. Diagram showing the metamorphism produced by a great igneous intrusion upon the surrounding rock. 341 342 PHYSICAL GEOLOGY tions, it is often found that the sedimentary rocks are greatly altered near the source of the heat (Figs. 334, 335). This is shown by a change in color, in hardness, and in texture, and in some cases by the development of new minerals. Bituminous coal is changed to anthracite coal, or the even in most FIG. 335. Map showing the metamorphic zone (dotted) about an igneous intrusion. extreme stage to graphite ; limestone is metamorphosed to marble; soft sand- stone may be con- verted into hard quartzite; and shale may be metamor- phosed to dense, compact rocks, such as schist and horn- fels, a compact flint- like rock. The amount and extent of contact metamorphism depends upon the amount of heat and to an important degree upon the gase- ous emanations (mineralizers) given off by the molten rock. For example, if molten rock is intruded into a narrow fissure, the surround- ing rock will usually be little affected (Fig. 336), since the magma, having a comparatively small amount of heat, soon loses it to the neighboring rocks. Moreover, the quantity of gas present is too small to produce a marked change. In the case of great intrusions, however, such as stocks or batholiths, the country rock may be greatly altered thousands of feet away. The effect of an intrusion is naturally greatest when the supply of heat is large and long-continued. In some cases, where fragments of the surrounding rock have been inclosed in the magma, 1 black shale has been baked to a hard, red, porcelain-like rock ; granite has been more or less completely fused to dark green or black glass ; and 1 Powers, S., The Origin of the Inclusions in Dikes, Jour. Geol., Vol. 23, 1915, pp. i-io. FIG. 336. Diagram showing the greater metamorphic effect of an igneous intrusion along bedding planes. METAMORPHISM 343 occasionally the fragments have been completely absorbed. The metamorphism resulting from intrusions is more extended when the intrusion cuts across strata than when it follows bedding planes, since under the former conditions the effect of the heat is felt along the several bedding planes with which the lava comes in contact (Fig. 336). The metamorphic effect of an intrusion is greater than that of an extrusion, since in the former the heat of the magma is lost more slowly, and the neighboring rocks are consequently heated to a higher temperature and for a longer time. Moreover moisture, which is a powerful agent in metamorphism and in the production of crystal- line structure in rocks, is more likely to be present under the former conditions. It frequently happens that the rock underlying a lava flow is so little metamorphosed that no change is visible to the naked eye. The effect of great intrusions has already been discussed under Subjacent Masses (p. 327). Regional Metamorphism. Thousands of square miles of the earth's surface are underlain by metamorphic rocks. They occur over large areas in Canada, in the Adirondacks, over the greater part of New England, in the Piedmont region east of the Appalachian Mountains, in a large area south of Lake Superior, and in the Cordil- leras. Widespread or regional metamorphism may be brought about in one of two ways, (i) It may result from great igneous intrusions, such as deep-seated batholiths. (The metamorphism of the older rocks of the Laurentian region of Canada seems to have been pro- duced largely in this way.) (2) Great lateral pressure may also pro- duce sufficient heat to recrystallize the rocks affected. In regions where igneous intrusions are absent, as in New England, the meta- morphism appears to have been caused by lateral pressure alone. The fact that the rocks of some metamorphic regions are more or less highly folded and that the intensity of the metamorphism is, to some degree, in direct proportion to the intensity of the deforma- tion is offered as proof that the alteration of the rocks in such re- gions was due, either directly or indirectly, to the cause or causes which produced the folding. The indirect cause is believed to have been the pressure which produced the deformation ; the direct causes, the heat resulting from the rock mashing produced by pressure, and the presence of underground water which aided powerfully in bringing about the molecular changes which resulted in the crystalline texture. 344 PHYSICAL GEOLOGY CLASSIFICATION OF METAMORPHIC ROCKS Quartzite. A quartzite is a metamorphic sandstone. It is a com- pact rock composed of grains of quartz sand cemented by material of the same kind, that is, by silica. It can usually be distinguished from sandstone by its appearance when broken. The broken sur- face of sandstone usually has a more or less granular feel and appear- ance, since the fracture takes place in the weak cement leaving the grains outstanding. In quartzite, on the other hand, since the grains and cement are of the same material, the fracture takes place in cement and grains alike. It is often difficult to state whether a quartzite owes its character to heat and pressure or to cementation by underground water. A quartz schist is a quartzite in which a foliated structure has been developed, the planes of foliation being covered with white mica. Marble. Commercially, any calcareous rock which will take a polish is called a marble, but in a more technical sense a marble is a metamorphic limestone. It is distinguished from limestone by its granular appearance (texture) and, unlike most metamorphic rocks, if pure is not schistose. When limestone is heated where the pressure is slight, it is converted into quicklime by the escape of carbon dioxide ; but when heated under pressure, which prevents the escape of the gas, it crystallizes into marble. The clouded shadings and " veins " of marble are produced by the crystallization of impurities, with the resulting formation of colored minerals. Slate. This rock may be considered as a hardened shale or mud in which a tendency to break along parallel planes not bedding planes is developed. This con- dition is called slaty cleavage and by its means the rock splits readily into broad, thin sheets. The cause of slaty cleavage is to be found in the great lateral pressure to which such fine- VMH a fi yaVa8 p enor0cs< g rained sediments as clay or (Modified after Pirsson.) (rarely) volcanic ash are sub- jected, especially if, when compressed in one direction, they are able to expand to some extent in others. A rock is affected by such compression in three ways : (i) any particles capable of compression are flattened and correspondingly lengthened at right METAMORPHISM 345 angles to the pressure ; (2) compression also turns elongated particles into parallel positions so that they take a direction in which their longest axes are at right angles to the pressure; (3) as a result of the metamorphism accompanying compression new minerals, such as mica, are formed, and since these crystals can grow more easily in the direction in which the pressure is least along the line of least resistance they also will have their longest axes at right angles to the pres- sure. The combined effect is to produce a rock which will cleave or split much more readily in one direction than in any other. Since a bed of shale is seldom perfectly homogeneous, slate differs in the perfec- tion of its cleavage. Sandy layers, for ex- ample, are contorted and poorly cleaved, while the layers of pure clay have a perfect slaty cleavage. Slaty cleavage will be per- pendicular to the bedding if the rocks were subjected to pressure when horizontal, but may be inclined at any angle to the bedding if the beds were folded before the pressure became intense (Figs. 337, 338). The formation of slate requires much less extensive metamorphic changes than does that of schist and of gneiss (p. 346). Schist. Schists are rocks composed of thin, wavy leaves or folia in which the foliation (Latin, foliatus, leaved) or lamination is due to the abundance and parallel position of such minerals as mica, hornblende, or talc. The folia are not of uniform thickness, but are flattened lenses of the minerals, often bent and wavy, with their platy surfaces in parallel planes. The characteristic foliated structure of schists is developed when rocks have been subjected to great pres- sure. Schists are the result either of (i) the formation of new min- erals which developed at right angles to the pressure, since growth takes place more readily along the lines of least resistance ; or of (2) FIG. 338. Illustration showing the relation of slaty cleavage (nearly vertical) to bedding (dipping to the right). (Photo. L. E. Westgate.) 34-6 PHYSICAL GEOLOGY the deformation resulting from the crushing of such rocks as con- glomerates, granites, or basalts. (3) In contact metamorphism the development of minerals, especially mica, along the stratification planes of sedimentary rocks also produces a schist. Schists are given various names, depending upon their most conspicuous mineral. Mica schist is composed principally of mica and quartz and is the most common type of metamorphic rock. Mica schists are usually metamorphosed, fine-grained sandstones and shales. Hornblende schist consists largely of hornblende, and varies from green to black in color. In some cases, the characteristic needle or blade-like crystals are readily recognized, but in others the grain is so fine that the individual crystals cannot be seen. Horn- blende schists are derived from diorites, gabbros, etc., by pressure, and it is probable that impure lime- stones containing sand, clay, and iron oxides also produce hornblende schists when subjected to metamorphism. Gneiss. This is a banded, crystalline rock (Fig. 339) in which feldspar is present. It is a rock with the composition of granite, but with a banded structure. Gneiss may be con- FIG. 339. Gneiss, showing banding. (U. S. National -j i r Museum.) Sldered / or conven - ience as intermediate between an igneous rock, such as granite or diorite, and schist. It will readily be seen from the above that a gneiss may, on the one hand, so closely resemble a schist that one will be in doubt as to its classification, and on the other hand, that it may be confused with a granite. Typically, however, gneisses are easily recogniz- able and may be considered for convenience as banded granites. As in the case of schists, various qualifying adjectives are used in describing gneisses, as biotite gneiss, hornblende gneiss, garnet biotite METAMORPHISM 347 gneiss. Gneisses may be formed either (i) by the metamorphism by mashing of granite or other igneous rock; (2) by the meta- morphism of sedimentary beds; or (3) when a granite magma is intruded into sedimentary or schistose beds under pressure operating from a distance, the molten magma spreading along the sedi- mentary planes or between the folia of the schists. This intimate admixture permits of extensive mineral changes, and the two types of rock, very different in geological age, become welded into a com- posite gneiss. TABLE SHOWING METAMORPHIC CHANGES SEDIMENTS SEDIMENTARY ROCKS METAMORPHIC EQUIVALENTS Gravel Conglomerate Gneiss and various schists Sand Sandstone Quartzite and quartz schist if from pure quartz sand ; mica schist if certain impurities are present Clay Shale Slate and schists,- especially mica schist Lime deposits, such as chalk or shells Limestone Marbles IGNEOUS ROCKS METAMORPHIC ROCKS Granite, syenite, and other rocks with much feldspar Gneiss Fine-grained feldspar rocks, such as felsite and tuffs Slate and schists Diorite, basalt, and other basic rocks Hornblende schist and other schists SUMMARY OF CAUSES OF METAMORPHISM The important factors to be considered in the production of meta- morphism are heat, moisture and pressure, mechanical movements, and the nature of the material involved. Heat. The heat necessary for metamorphism may come (i) from igneous intrusions. In this way the surrounding rocks are hardened and dehydrated. The process is shown in the manufac- ture of bricks, in which the clay is dehydrated and is hardened to a rock-like mass by partial fusion. New minerals are often developed 348 PHYSICAL GEOLOGY in rocks affected by intrusions. (2) The heat developed by pressure will be discussed in a later paragraph. Moisture. When moisture is present in considerable quantity, as is the case with sedimentary rocks, the effects of heat and pressure in producing metamorphic changes are greatly increased. This is true because highly heated water, especially if alkalies are present, readily dissolves minerals which would otherwise be insoluble, and from the solution the same minerals or new ones may be formed. Water also takes part in the chemical composition of some minerals, such as mica, and is therefore necessary for their formation. The recrystallized and newly formed minerals are usually arranged with their longer axes at right angles to the pressure (p. 349), and are more stable under the new conditions than if they had not been changed. The potency of moisture is shown by the fact that rock which re- quires a temperature of 2500 F. for melting when dry, becomes pasty at 750 F. when water is present. The effect of gaseous emanations in producing metamorphism is sometimes of the greatest importance. Pressure. Simple downward pressure, such as that which results from the weight of overlying rocks, has some metamorphic effect and also tends to consolidate the sediments by bringing the grains closer together. But when the crust is under enormous lateral pressure, as a result of the contraction of the earth, the strata are folded, crushed, and mashed together, and metamorphism takes place. In this way pebbles, fossils, and crystals are flattened and elongated, or broken into fragments. By this agent alone the texture of rocks can be changed, but it is in combination with heat and moisture that the production of new minerals and the formation of highly metamorphic rock is brought about. The importance of lateral pressure in the production of regional metamorphism has been questioned by certain French geologists 1 who believe that it is brought about by heat, moisture, and vertical pressure without the aid of lateral pressure; that the sediments in the lower parts of thick geosynclines are actually fused by heat from the interior of the earth, and upon cooling become igneous rocks, capable in their turn of metamorphosing by contact the rocks which surround them. They hold that this is by far the most important element in the process of metamorphism and that dynamic action can deform but cannot transform rock; i.e., it is not competent by itself to pro- duce metamorphic changes. This theory has been generally abandoned by Ameri- can geologists and in fact by many eminent French geologists. How the Parallel Arrangement of Minerals is Produced. The conditions favorable for the production of metamorphism having 1 Haug, Traitt de Gtdogie, pp. 172-191 ; 234235. METAMORPHISM 349 been discussed, it remains to be shown why metamorphic rocks are usually cleavable. 1. Crystallization. A study of a mica or hornblende schist shows that hornblende and mica are responsible for the best rock cleavage. A microscopic examination of these rocks and of the sedimentary rocks from which they were derived shows that horn- blende and mica were built up chiefly by subsequent recrystalliza- tion from substances already in the sedimentary rocks and did not exist in them in their final form. This fact is shown by a general lack of fractures in the minerals of the cleavable rock, such as would have been developed had the rock been formed simply by crushing and by the rotation of the mineral constituents to parallel positions. Moreover, most of the mineral particles of cleavable rocks are larger than those of the same rock before the latter was metamorphosed. The gradation of shale to phyllite (a metamorphic shale) means an increase in the size of the grains. The parallel arrangement of the mineral constituents of a metamorphic rock is thus seen to be the result of crystallization, and the rotation of the original particles to a parallel position, of minor importance. 2. Granulation. Recrystallization and rotation are not the only processes instrumental in the production of easy splitting or cleavage in metamorphic rocks. In the early stages of the process the larger brittle particles are broken into small fragments or are granulated and elongated, and at the same time recrystallization builds up new minerals from the broken particles. 1 It is probable that granulation aids crystallization in that it grinds the particles into small pieces which then present a greater surface upon which the chemical process may act. Relation of Cleavage to Pressure. The proof that the planes of easy splitting of metamorphic rocks are at right angles to the pressure, or in other words parallel to the rock elongation, is seen (i) in the distortion of the pebbles of conglomerates, (2) in the distortion of fossils, the lengthening being in the plane of the cleavage, and (3) when rock is intruded by igneous masses which exert a great pressure on the walls, in the cleavage which is developed parallel to the walls. From Igneous, through Sedimentary, to Metamorphic Rocks. The history of a metamorphic rock, formed by the recrystallization of sedimentary rocks, may be briefly summarized. If a great mass of granite is exposed to the weather, it begins to decay ; its feldspar 1 Leith, C. K., Structural Geology, 1913. 350 PHYSICAL GEOLOGY and mica being disintegrated and forming simpler compounds, some of which are dissolved and carried away by the water, while the remainder is left as clay. This clay together with the insoluble quartz is transported by streams and is finally deposited in the ocean, the clay forming mud and the quartz grains, sand. The lime dissolved from the feldspars may be taken up by organisms to form lime ooze or limestone. If these sediments are laid down in a sinking geosyncline (p. 359), they may in time be buried to a depth of several thousand feet. When in the course of their burial they reach the belt of cementation (p. 61), they will be con- solidated into shales and limestones. If the sediments in the syn- cline are subjected to great lateral pressure, heat will be developed which will metamorphose them, changing the clays, sandstones, and limestones to schists, quartzites, and marbles. For a discussion of the formation of metamorphic rocks from igneous rocks, see p. 347. Weathering of Metamorphic Rocks. Metamorphic rocks usually resist weathering better than sedimentary ones, because they have been compacted by heat and pressure and have a crystalline texture. Mica schist, for example, is less easily disintegrated than the impure shale from which it was made ; hard quartzite than the less compact sandstone ; marble, however, may or may not be more resistant to weathering and erosion than the limestone from which it was derived ; the disintegration of slate is hastened by its vertical cleavage, but is hindered by its greater compactness. As a result of prolonged weathering, however, metamorphic sedimentary rocks are reduced, in time, to the same soil which their sedimentary equivalents would have made, schists weathering to clays; quartzites to sands; and marbles to calcareous clays. Because of their greater resistance to weathering, metamorphic rocks are usually associated with the scen- ery of mountains. Quartzite usually resists weathering better than any other rock on account of its small porosity, its insolubility, and its homogeneous composition. Because of the last-named character, it is little affected by changes in daily temperature (p. 31), and it is not disintegrated by the decay of a weaker constituent, as is the case with igneous rocks, such as granite. Quartzite hills are, consequently, among the last to disappear. In regions of meta- morphic rocks, where schist and marbles are involved, streams have usually cut their valleys in the softer and more soluble marbles, while the more resistant schists form hills and mountains. METAMORPHISM 351 Economic Importance. Gneisses and quartzites are often used for building stones and road material, but marble is the metamorphic rock which is the most prized, both for building purposes and for works of art. REFERENCES FOR METAMORPHISM CHAMBERLIN AND SALISBURY, Geology, 2d ed., Vol. i, pp. 426-449. COLE, G. A. J., Rocks and their Origins. LEITH, C. K., Structural Geology. PIRSSON, L. V., Rocks and Rock- Minerals. CLELAND GEOL. 23 CHAPTER XI MOUNTAINS AND PLATEAUS THE term mountain is used very loosely to indicate a conspicuous height of land. In flat regions such as southern New Jersey and the plains of Texas, heights rising more than 100 to 200 feet are dignified by the name mountain, while in mountainous regions ele- vations of 1000 or 2000 feet are often called hills. It is evident that the term is a relative one, since on plateaus a mile or two above the sea a conspicuous elevation must be still higher, and a mountain in such a situation would be at least 6000 feet above sea level. A mountain ridge or range is usually long, with a narrow crest ; when numerous ranges are associated, they constitute a mountain chain. In ancient paintings and in old geographies, the slope of mountains was usually depicted as very steep, an angle of 60 from the hori- zontal not being uncommon, but such slopes seldom occur in nature, and angles as high as 35 are rare. Mountains of Accumulation. Volcanoes are typical of this class, as they are built up by the accumulation of ash, or lava, or both. They sometimes occur singly, sometimes they are arranged along fracture lines (p. 267), and sometimes no definite order can be recog- nized. Since sand dunes (p. 52) occasionally attain a height of 600 feet and moraines (p. 159) a height of 1000 feet, they are sometimes called mountains in regions where other elevations are inconspicuous and must therefore be included under the head of accumulation mountains. Residual Mountains. These are formed when a plateau has been extensively dissected by rivers, and the ridges and pyramids, the remnants of the plateau, which have escaped erosion, stand so high above the valleys as to constitute mountains. The many " temples " in the Grand Canyon of the Colorado in Arizona (Fig. 340) show at a glance how such mountains are formed, and the Catskills of New York furnish an excellent example of residual 352 MOUNTAINS AND PLATEAUS 353 354 PHYSICAL GEOLOGY mountains in which gentle slopes are characteristic. The form of mountains of this class depends upon the nature and arrangement of the material (Fig. 340) out of which they were sculptured, and to some extent upon the climate. The Catskills owe their gentle slopes to the fact that the rocks of which they are composed do not differ greatly in hardness, and also to the smoothing effect of a moist climate. The steep-sided mesas of the southwestern United States are often the result of the erosion of lava plateaus, the hard lava forming flat- topped mountains bounded by conspicuous, vertical cliffs. Residual mountains are confined to those formed of horizontal rocks or slightly inclined rocks. The external form of complexly folded mountains (p. 356) is due to erosion, and they are in a sense residual moun- tains. They have, however, been placed in a class by themselves be- cause of the origin of the folded structure which gives them a distinct character. Fault or Block Mountains. It was shown in the study of fault- ing (p. 267, Fig. 266) that important topographic features are produced in this way, and that mountain ridges of this origin have been formed either by uplift along one side of a fault, or by sinking along one side, or by a combination of the two movements. Moun- tains formed by the elevation of wedge-shaped blocks are called horsts (p. 263, Fig. 257). In southern Utah and Oregon block or faulted mountains have been carefully studied and have been found to exhibit all the stages from young faulted mountains, in which erosion has as yet been able to accomplish little, to ancient fault mountains, in which erosion has proceeded so far that their origin can merely be conjectured. In portions of these regions block moun- tains 10 to 40 miles long and 1000 to 1200 feet high occur. The ridges are steep or clifF-like on the fault side and have a gentle slope on the opposite side. Between the faults are trough-like depressions in which lakes sometimes rest. The steep eastern slope of the Sierra Nevada Mountains marks the fault along which a great block, 500 miles in length and 70 to 100 miles broad, has been raised, the escarpment thus formed rising from 5000 to 6000 feet above the desert valleys to the eastward, and reaching a maximum height of 14,000 feet in the vicinity of Death Valley. (Russell.) Well-known examples of block mountains are the Vosges and Black Forest of Germany (p. 100, Fig. 81). Laccolith Mountains. Under the discussion of laccoliths (p. 327) it was seen that in certain localities molten material has MOUNTAINS AND PLATEAUS 355 been injected into the earth's crust in such quantity that the cover of sedimentary strata has been lifted into dome-like forms. After prolonged erosion the softer strata are partially or wholly removed, and the hard, igneous core is left as a mountain or hill. In moun- tains of this origin, the strata dip away in all directions from the center, and not uncommonly "hogbacks," with steep cliffs facing towards the center, form one or more broken rings about the moun- tain. Although mountains of this type are not abundant, a large FIG. 341. Little Sundance Dome, Sundance, Wyoming. This is a laccolith from which the overlying strata have been eroded. number are known to exist, of which those in Utah, California, Wyo- ming, South Dakota, British Columbia, and Canada might be men- tioned (Fig. 341). Domed Mountains. The Black Hills of South Dakota may be taken as a type of domed mountains. They rise 2000 to 3000 feet above the surrounding plains and about 7000 feet above sea level, and are carved from a dome-shaped uplift of the earth's crust. The length of the dome is about 100 miles and the width about 50 miles, about the size of Connecticut. As will be seen from the diagram (Fig. 342), the eroded central part is composed of crystalline rocks from which the strata dip in all directions. As a result of the more rapid erosion of a stratum of shale, a trench called the Red Valley, in many places two miles wide, entirely surrounds the center, except where it is cut through by streams. The Red Valley is separated from the flat plains surrounding the central mountain mass by a rim or " hogback," which presents a steep face towards the valley and rises several hundred feet above it. 356 PHYSICAL GEOLOGY The Uinta Mountains of Utah are formed from a flattened dome or broad arch 150 miles long and 20 to 25 miles wide, which rises about 10,000 feet above sea level. It will be seen from the diagram FIG. 342. A block diagram of a domed mountain, the Black Hills of South Dakota. The investing valleys with their steep, infacing cliffs are well shown. The central mountain mass is granite, and the three isolated mountains are intrusive masses of igneous rocks. (Fig. 343) that if all the rock which has been carried away were restored, the mountains would be three and a half miles higher than now. This does not prove that the mountains were ever as high as that, since the denudation of a mountain mass commences as soon as it begins to rise above the surrounding country, and the rate of erosion in all probability is about the same as the rate of upheaval. FIG. 343. A section across the Uinta Mountains, Utah. The range has been formed out of a single broad arch 40 miles wide, which has been greatly eroded. The original surface is indicated by the dotted line, showing that three and one half miles of rock have been removed by erosion. Complexly Folded Mountains. It is to this class that the great mountain systems of the world belong, the Appalachians, the American Cordilleras (the Rocky, Sierra Nevada, Coast, and Cascade moun- tains), the Alps, Himalayas, Pyrenees, etc. The strata which compose them may consist of a series of gentle anticlines and synclines (p. 254), or may be intricately folded and faulted. Portions of the Jura MOUNTAINS AND PLATEAUS 357 Mountains of Switzerland present a classic example of gently folded strata ; here one finds, in certain places across the system, a series FIG. 344. Section through the Juras, showing mountain ridges produced by several open folds, like great earth waves. of simple anticlines and synclines (Fig. 344). A portion of the Appalachian Mountains in Pennsylvania also presents a similar sim- ple structure (Fig. 345). In the Alps, however, the folds are much FIG. 345. Relief map of the Appalachian Mountains. (See Figs. 244 and 245, p. 255.) more pronounced and complicated, and it is often extremely difficult to determine the structure of the strata (Fig. 346). 358 PHYSICAL GEOLOGY It is evident that a series of strata, subjected to forces sufficient to produce the intense folding shown in the Alps and in the southern FIG. 346. Diagram showing a cross section of the Alps along the Simplon tunnel. The complicated structure and former extension of the strata are shown. (After Schmidt.) Appalachians (Fig. 347), will often break and fault instead of fold- ing. It is, consequently, seldom that folded strata are free from dislocations over a distance of even a few miles. The strata of folded mountains have often been so compressed that cleavage planes parallel FIG. 347. A, section across the southern Appalachians where extreme faulting has oc- curred (U. S. Geol. Surv.) ; B, section in the vicinity of Chattanooga, Tennessee. to the folds have been induced. Metamorphism is in proportion to the intensity of the compression. ORIGIN AND DEVELOPMENT OF FOLDED MOUNTAINS Four points have been established with reference to folded moun- tains : (i) they were formed from thick sediments that had accumu- lated in geosynclines ; (2) they were folded as a result of lateral pressure; (3) the rate of folding was slow; and (4) their outlines, after prolonged erosion, are determined largely by the character of the rocks and the arrangement of the strata. A discussion of these points follows. There is also reason to believe that mountains of this class are situ- MOUNTAINS AND PLATEAUS 359 ated at the junction of great earth segments or blocks where, as a result of the crowding of the latter upon each other as they are drawn toward the center of the earth, the weak strata of geosynclines are folded. 1. Geosynclines. The sedimentary strata of which folded moun- tains are formed are very thick ; in the Appalachians, the thickness is about 25,000 feet; in the Coast Ranges of California, 30,000 feet; and in the Alps, 50,000 feet. When a stratum is traced to a distance of even a few miles from the mountain chain, it is found that it rapidly becomes thinner; the strata that have a thickness of about 25,000 feet in the Appalachians, for example, are only about 2500 feet thick in the Mississippi Valley. An examination of the rocks of moun- tain masses often shows that many of them are of shallow water origin, as the occurrence of conglomerates and sandstones testifies. Ripple marks, sun cracks, and fossils afford similar evidence. * The presence of limestones, on the other hand, may indicate (p. 238) that the water in which they were deposited was deep or far from shore. The sediments that are being laid down in the seas to-day are deposited in a belt extending from the shore line to a distance usually considerably less than 50 miles (p. 237). Since there is no reason to believe that the conditions of sedimentation in the past were markedly different from those of the present, it is generally held that the strata composing the great mountain ranges were laid down near shore and, since many of them are of shallow water origin, that sink- ing accompanied and for the most part kept pace with the deposi- tion, the land rising as the geosyncline sank. Occasionally uncon- formities occur, which indicate, as has been seen (p. 270), that eleva- tion for a time interrupted the deposit of sediment. 2. Lateral Pressure. When sediments have accumulated in a geosyncline to a depth of several thousand feet, those near the bot- tom of the deposit are somewhat weakened by heat (p. 347), so that they are compressed and thrown into folds when subjected to great lateral pressure. The strata composing the Appalachian Mountains of Pennsylvania, between Harrisburg and Tyrone (Fig. 348), were compressed from a width of 81 miles to one of 66 miles; i.e., the earth's superficial crust, upon being folded, was shortened 15 miles, with a resulting mean elevation of three miles. It has been estimated that, if the folds of the Alps were smoothed out, the strata would cover an area 74 miles wider than the mountains do now, or about twice their present width. The shortening of the Front Range in 360 PHYSICAL GEOLOGY Colorado is estimated to be about 25 miles, and that of the Coast Ranges of California, 9 to 12 miles (Fig. 349). A careful study of folded regions shows that the strata are often broken and faulted, the folds frequently giving place to thrust or FIG. 348. Folds in the Appalachian Mountains between Harrisburg and Tyrone. (As restored by R. T. Chamberlin.) reverse faults, especially where the strong or competent stratum is not deeply buried. As already stated in the discussion of reverse faults (p. 263), the overriding of the strata is sometimes 10 or more miles. In fact, in the southern Appalachians thrust faults are so numerous as largely to determine the positions of the mountain ridges (Fig. 347, p. 358), and the elevation in the Scottish and FIG. 349. Profile of the Santa Cruz Mountains in the Coast Ranges of southern California. (Arnold.) Scandinavian Highlands is due, to some degree, to the fault slices piled one on top of another. Igneous rock is often associated with mountains and in some ranges is an important factor in the folding and metamorphism of the strata. It often composes the cores of mountain ranges and frequently forms their highest portions (p. 355). (i) The igneous core of mountain masses is often derived from igneous intrusions ; (2) it may be the rock of the floor of the geosynclines ; or (3) it has even been suggested that it is sometimes the lower portion of the sediments of the geosyncline which have been fused as a result of the rise of temperature (p. 348) from the interior of the earth. (Haug.) In each of these cases the igneous core is exposed only after erosion has removed a great thickness of overlying sedimentary rock. Experiments in Mountain Building. Experiments have been performed to determine whether the folds and reverse faults observed in such mountains as the Appalachians can be reproduced. In these experiments a series of layers composed of wax and other substances varying in rigidity and elasticity were prepared to represent rock strata of widely different character, such as shale, sandstone, and MOUNTAINS AND PLATEAUS 361 limestone. A load of shot, representing the weight of the overlying strata, was then placed on top of the layers. Upon the application of lateral pressure it was found that, by varying the rigidity and FIG. 350. Machine for experimenting in mountains of folded structure. (U. S. Geol. Surv.) thickness of individual layers and of the layers as a whole, the phe- nomena observed in folded regions were reproduced. A study of the apparatus (Fig. 350) gives a better idea of the conditions of the experi- ment than a written description. SURFACE FACTS AND UNDERGROUND INFERENCES ZO/V ELATIONS OF LAND AND SEA AND OF POSITION OF STRATA PRIORTO FOLDING FIG. 351. Diagrams showing the theoretical history of a folded region. The lowest figure shows the region when the present site of the mountains was a great syncline in which a load of sediment, 25,000 to 40,000 feet thick, had been laid down. The middle figure shows the region after it had yielded to great lateral pressure and had been folded and faulted. The upper figure shows the region as it is to-day. (Redrawn after Willis.) 362 PHYSICAL GEOLOGY A brief and incomplete history of a folded region is shown in Figure 351. It is incomplete because many of the important chapters of the history cannot be shown without too great confusion of detail. 3. Rate of Folding. The rate of folding must necessarily differ widely in different geosynclines and in the same geosyncline at vari- ous times. In certain cases it seems to be proved that rivers have been able to deepen their valleys as rapidly as the land surface was elevated (antecedent rivers, p. 102). It is possible that the general denudation of a region may in some cases have proceeded at about the same rate as the elevation, so that at no time was the 'surface far above sea level. This may, for example, have been true of the Appalach- ians, which now consist of comparatively low mountain ridges FIG. 352. Section across central New England, showing the uplifted peneplain and Mt. Monadnock. (Hitchcock.) although three or more miles of sediment have been removed by erosion ; and also of the folded and crumpled rocks of New Eng- land (Fig. 352). The elevation of regions of folding was not always continuous, but, as is shown by a study of the Appalachians, the folded belts were at times above sea level and suffered erosion ; upon being again depressed they received more sediment, unconformities marking the sites of the ancient erosion surfaces ; and later, they were further folded and faulted and raised above the sea. The present height of the Appalachians and Sierra Nevadas was brought about by vertical elevation and not by lateral compression. 4. To What the Topographic Features of Folded Mountains are Due. A comparison of the external form of mountains and their geo- logical structure shows that the two seldom agree. It is true that the mountain ranges in general are parallel to the strike (p. 353) of the strata, but the valleys seldom coincide with the synclines and the ridges with the anticlines. This coincidence sometimes occurs, but the reverse is as frequently seen. In the Jura Mountains, Switzerland, excellent examples of synclinal valleys may be seen MOUNTAINS AND PLATEAUS 363 (Fig. 344, p. 357), but in the Appalachians anticlinal valleys are perhaps more common than synclinal ones. This lack of coincidence is due to the fact that when strata are folded the crests of the anti- clines are stretched and consequently weakened, while the synclines are correspondingly compressed and strengthened. Moreover, when the land surface ....... emerges from the sea, the crests of the anti- clines are first attacked by erosion, and their strata may be worn through while the ^'u^-j r u . . . ric. 353. The slopes of the sides of the mountains are synclines are receiving determined largely by the dip of the rock forming them, sediment and are thus being protected. It is conceivable that a syncline may never have contained a stream, since before its surface was elevated above the sea, valleys had already been established in the anticlines. If we imagine a number of folds, the anticlines and synclines of which are exposed to erosion at the same time, it will readily be seen that erosion will develop valleys in the anticlines, as it is now doing in the Jura Mountains. In in- tensely folded mountains where overturned folds occur, as is so frequently seen in the Alps, the variable character of the strata determines the cliffs and escarp- ments of the mountains (Figs. 353, 354). The gentle slopes of mountains of this structure are most likely to be found along the dip of the strata, the cliffs MILE along the strike. FIG. 354. Section showing the effect The effect of a resistant stratum of the dip of a resistant stratum upon determining the topography the topography of a mountain. . . T of a region is well shown in por- tions of the Appalachian Mountains, where a single quartzite stratum forms long mountain ridges wherever it outcrops at the surface. The canoe valleys of the Appalachians and other folded mountains are formed by the erosion of the strata of a plunging anticline (Figs. 244, 245, p. 255). 364 PHYSICAL GEOLOGY Cycle of Erosion of Mountains. In the process of time moun- tains may be wholly reduced by erosion, and plains and plateaus be formed in their place, which will have all the structural features of folded mountains. Examples of such plateaus are to be found in the Piedmont of Virginia, in New England, and elsewhere. In moist, tropical regions the luxuriant vegetation checks erosion, with the result that the forms are less diversified than in other areas. THEORIES OF MOUNTAIN BUILDING Mountain chains are more conspicuous than plateaus because of their narrow crests and great length in proportion to their width, but when the heights of mountains and plateaus are compared, it is found that many mountain ranges are relatively low as compared with many plateaus. Portions of the Appalachian Mountains, for example, are lower than portions of the Allegheny plateau only a few miles away. The Tibet plateau is 15,000 to 16,000 feet high, being higher than many of the great mountain chains of the world. The highest of the Colorado plateaus (Aquarius) is 11,600 feet, and that at Grand Canyon, Arizona, is 6000 to 8000 feet above the sea. It is evident from the above that the cause of the elevation of the less conspicuous but more massive plateaus is as important as that of the more spectacular mountains. Cause of Lateral Pressure. In the discussion of the interior of the earth it was pointed out that the earth is composed of a hot but solid core with a cool crust. The answer to the question, "What produces lateral pressure ? " will be found to depend, to some degree, upon this relation. The explanation often given for the crumpling of the earth's crust is that, as the interior heat is lost very slowly by conduction, the crust wrinkles to accommodate itself to the smaller interior. The comparison usually made is that of an apple which has been left in a warm, dry room. Under these conditions the interior of the fruit loses water by evaporation, while the dense skin shrinks but little and is wrinkled on the contracted interior. The efficacy of this cause has been proved impossible on the ground that the shrinkage of the interior of the earth has not been sufficient to produce the lateral compression seen in the great folded tracts of the earth's surface, and in proof of this contention it is pointed out that during a single era of the earth's history (Paleozoic, p. 477) the folding of the earth's crust resulted in a shortening of between 100 and 200 miles. Since a lateral shortening of six miles of the crust is produced by one mile of radial shortening, it follows that a minimum estimate would require a radial shortening of 16 miles, and a maximum, one of 32 miles. MOUNTAINS AND PLATEAUS 365 The cause of the great deformations, such as those recorded in the Alps, the Appalachians, and other ranges, is believed to be found in the distribution of heat beneath the surface. It is thought that the heat of the interior " would be conducted from the deep interior to the outer zone 800 to 1200 miles thick, faster than from the latter outward, with the result of raising the temperature of the outer zone while that of the deep interior falls. The result of this should be a severe crowding of the outer zone upon itself, in shrinking to fit the deep interior as it loses heat and shrinks." (Chamberlin and Salisbury.) The folding of great areas therefore results, according to this theory, from the crowding of the thick outer zone on itself. The extrusion of lava from the deeper zones of the earth cooperates with the cooling of the heated interior in causing a shrinkage. The outpouring of the hundreds of thousands of square miles of lava in Oregon and neighboring states, and in the Deccan peninsula of Asia, undoubtedly contributed to the shrinkage of the interior, although the total effect was slight. The Elevation of Plateaus and Mountains. The statement is often made that great mountain ranges are formed solely as a result of lateral pressure and also, when a region only a few hundred feet above sea level is found to be underlain by much folded and meta- morphosed rocks, that "erosion has laid bare the mountains to their roots and that the ancient heights may at one time have rivaled the Alps in majesty." Such assumptions must, however, be accepted with caution. It seems safe, at least, to assume that, if great areas of the earth's surface can be raised by vertical movements to form plateaus, the elevation of a folded region may be largely due to similar vertical movements. This brings us to the modern theory of isostasy (Greek, isos, equal, and stasis, standing still). The Theory of Isostasy. 1 If oil and water are balanced in a U-tube, it is evident that, since water is the heavier, its surface will be lower than that of the lighter oil. It is upon this principle that the theory of isostasy is based. The ocean basins are believed to be underlain by heavier materials than the continents and are conse- quently lower, since they are drawn more strongly toward the center 1 Hayford, J. F., The Figure of the Earth and Isostasy from Measurements in the United States: U. S. Coast and Geodetic Surv., 1909. Hayford, J. F., The Effect of Topography and Isostatic Compensation upon the Intensity of Gravity : Special Publication 10, U. S. Coast and Geodetic Surv., 1912. Reid, H. F., Isostasy and Mountain Ranges, Bull. Am. Geog. Soc., Vol. 44, 1912, p. 354 et seq. 3 66 PHYSICAL GEOLOGY of the earth by gravity. The surface of the earth may, therefore, be considered as a mosaic of great polygonal blocks (Fig. 355), which from time to time suffer readjustment, the areas occupied by the continents being the continental segments and those by the oceans being the oceanic segments. Not only is the earth divided into these Continent FIG. 355. Diagrams representing the conception that the continents were lifted and the ocean basins sunk by movement along definite sliding planes or fault planes. The dotted lines may be taken to represent a somewhat uniform original surface, which may be looked upon as the surface before the continents and ocean basins were de- veloped. (After Salisbury.) great segments, but these in turn are made up of smaller blocks which by differential movements have produced the high plateaus and low plains of the continents, and the " deeps " of the oceans. The theory of isostasy holds that every segment of the earth, MOUNTAINS AND PLATEAUS 367 having an equal area of surface and with its apex at the center, contains the same amount of material, which it is impossible ma- terially to increase or decrease. When a large quantity of material is removed from the land by erosion and deposited in the ocean by streams, the increased weight under the ocean and the decrease under the mountains will cause the rock at a great depth to flow from the area which is more heavily weighted, to that from which the weight has been removed, and the approximate equality of material in the segments will thus be restored. As the oceanic and continental segments are drawn toward the center of the earth, the surface portions are subjected to great lateral pressure produced by the crowding of the segments against one another, and since the pressure cannot be relieved by the transfer of material by rock flowage such as is possible at great depths, it is relieved by folding and thrust faulting. Since, as has already been shown, the materials of the great mountain ranges were formed from the thick sediments of geosynclines whose basal portions were prob- ably weakened to some extent by the invasion of heat from the interior of the earth, it is clear that, if such thick but weak strata are subjected to great horizontal compression, they will be likely to be folded and faulted. According to the theory of isostasy, how- ever, the folding of strata by lateral pressure could not cause the elevation of a mountain range without the aid of the expansion of the material of which it is composed, since otherwise the quantity of material in the segment would be increased by folding and this added weight would cause a slow sinking, and material would flow from below the heavier segment to the lighter one, until the two again balanced. This theory does not tell us definitely the cause of the elevation of mountains and plateaus, but it positively states that the eleva- tion of mountains or the depression of oceanic segments must be due to an increase or decrease of density. The mountains are high because their material is light, and their elevation is due to an ex- pansion of the material in and under them ; the ocean deeps are depressed because the material under them is dense and may be sinking because this material is becoming denser. A number of examples of mountain ranges which owe their height to vertical elevation can be cited. The present altitude of the Appalachians, as has been stated, is the result of vertical movement without the aid of lateral pressure, the folding of the strata long CLELAND GEOL. 24 368 PHYSICAL GEOLOGY antedating the last elevation. The Sierra Nevadas, after folding, were peneplained and were later elevated along a great fault on the east, and their height is being increased at the present time. It is thus seen that the elevation of high mountains may be due to verti- cal movements, without the aid of folding. The Distribution of Mountains. Attention has long been called to the fact that the mountain ranges of the Pacific the Andes, western ranges of North America, etc. are situated near the edges of the continents, and the generalization has been made that moun- tains are usually located near the oceans, the higher mountains bordering the deepest basins. It is also to be noted that many exceptions exist : the Alps, Caucasus, Urals, and Himalayas are situated at considerable distances inland. The distribution of moun- tains has led to two theories as to the position of the geosynclines in which the sediments forming them were accumulated ; one hold- ing that the geosynclines existed at the edges of the continents, the other that they were between land masses. The apparent exceptions to the latter theory are attributed to the subsequent sinking of lands which formerly existed near the present shores of the oceans bordered by mountains. According to this theory, for example, the Alpine geosyncline existed between the African continent and the ancient land masses on the north ; the Appalachian geosyncline, between the Piedmont land on the east and other ancient lands on the north and west (p. 477) ; the Himalayas, between the Indian peninsula and land to the north. Permanence of Continents and Ocean Basins. It is quite generally agreed by geologists that the ocean basins and the con- tinental platforms have been very much as now for many millions of years. By this is meant that the present continents have not been covered by oceans thousands of feet deep, nor have the ocean depths been dry land over wide areas. The proof of the former lies in the fact that no deep-sea sediments have ever been found in the sedi- mentary rocks of the continents, the continents having been covered repeatedly by shallow seas (called epicontinental, p. 405), but never by any of great depth. Of the latter no positive proof has been advanced, but on the contrary the distribution of animals and plants in the past gives reason for believing that land connections once existed between South America and Africa, North America and Eu- rope, and Australia and Africa. Age of Mountains. This subject will be more fully discussed MOUNTAINS AND PLATEAUS 369 later (p. 519), but it should be noted in this connection that the time at which a region was raised above the sea was, at least, not previous to the youngest rocks of which the region is composed. For example, the Appalachian Mountains contain coal beds which show that the region was folded after their formation, i.e., after the Carboniferous (P- 477)- REFERENCES FOR MOUNTAINS DALY, R. A., Abyssal Igneous Injection as a Causal Condition and as an Effect of Mountain-building: Am. Jour. Sci., Vol. 22, 1906, pp. 195-216. DALY, R. A., Mechanics of Igneous Intrusion: Am. Jour. Sci., Vol. 15, 1903, pp. 269-298; Vol. 16, 1903, pp. 107-126. DALY, R. A., Igneous Rocks and their Origin. GEIKIE, J., Mountains; their Origin, Growth, and Decay. GEIKIE, A., Textbook of Geology, 4th ed., Vol. i, pp. 672-702. GILBERT, G. K., Report on the Geology of the Henry Mountains: U. S. Geol. and Geog. Surv. of the Rocky Mountain Region, 1877. READE, T. M., Origin of Mountain Ranges. TARR and MARTIN, College Physiography, pp. 525-581. WILLIS, B., The Mechanics of Appalachian Structure: Thirteenth Ann. Rept., U. S. Geol. Surv., Pt. 2, pp. 217-281. TOPOGRAPHIC MAP SHEETS, U. S. GEOLOGICAL SURVEY, ILLUSTRATING MOUNTAINS OF VARIOUS ORIGINS Folded Mountains Delaware Water Gap, Pennsylvania. Estillville, Kentucky. Harrisburg, Pennsylvania. Fort Payne, Alabama. Hollidaysburg, Pennsylvania. Tamalpais, California. Residual Mountains Fault Mountains Laccolith Mountains Kaaterskill, New York. Alturas, California. Henry Mts., Utah. Mt. Mitchell, North Carolina. Granite Range, Nevada. Monadnock, New Hampshire. Wausau, Wisconsin. CHAPTER XII ORE DEPOSITS ORES are concentrations in the earth's crust of economically valu- able minerals. Ores in Ready-made Cavities. A common form of deposit is the vein^ or the filling of a fissure in a rock. The contents of a fissure may consist partly or wholly of minerals, some of which may or may not be of economic value. When mineral veins contain ores, they are called lodes by miners. Fissures and other cavities are formed in several ways, as has been seen (p. 262). (i) Stretching movements of the earth's crust fracture it, producing open cracks ; (2) faulting (p. 261) forms fissures and brecciated zones ; (3) fissures are developed by shrinkage, such as occurs when igneous rocks cool or when limestone is changed to dolomite ; (4) the joints of rocks are widened ; (5) cavities are formed in limestone by solution. Cavities formed in any of these ways may contain ores. Fissure Deposits. Metalliferous veins are not composed entirely of metalliferous minerals, but on the contrary the latter often constitute a very small percent- ABCDDCBA ri rii- T-I age of the vein filling. The use- less vein material is called gangue, the common gangue minerals being quartz, calcite, and fluorite. In some veins the contents are arranged in bands parallel to the walls, the minerals and ores of one wall being represented by corresponding bands on the op- posite wall (Fig. 356). This arrangement is the result of the deposition of minerals from solution on the two walls of the fissure at the same time. Such a symmetrical arrangement, however, is not common, the layers usually being thicker on one wall than on the other, while frequently a layer on FIG. 356. Banded veins : A and Z), quartz; B y sphalerite; C, galena. ORE DEPOSITS 371 one side has no corresponding layer on the other. A banded struc- ture may also be brought about when, as a result of movements which rend the vein from one of its walls, the fissure is reopened and minerals are subsequently deposited in the cavity thus formed. Several such movements may take place and two or more bands may be formed. In some veins the filling consists wholly or in part of broken rock (Fig. 357), the spaces between which are filled with quartz or other minerals. Form and Extent of Veins. The form of veins usually FIG. 357. A section showing a fault breccia, depends Upon the shape of Such breccias have sometimes been cemented i r i i i r n together by precious minerals and are valuable the fissures which they fill, ore deposits> (After Ries and Watson.) and their width, length, and depth consequently vary greatly. Some are only a fraction of an inch wide, while others are 200 or 300 feet in width. The length is even more variable, being in some cases 50 or more miles and in others only a few feet. Some veins have been followed to a depth of more than 5000 feet, while others have disappeared a few feet beneath the surface. Source of Vein Material. It has been shown by chemical analysis that nickel, copper, tin, lead, and other metals occur in minute quantities in both sedimentary and igneous rocks, and it is generally believed that the ores which are now concentrated in veins were originally disseminated through the rocks ; that they have been dis- solved out by water, carried to fissures or other cavities and there deposited. This theory is borne out by the fact that in certain places veins are actually being formed at the present time by deposi- tion from water. For example, near Boulder, Montana, a hot spring is depositing gold-bearing quartz identical with the gold and silver-bearing quartz veins of the region. Steamboat Springs, in Nevada, are strongly alkaline and are depositing quartz in fissures and thus forming veins. Sulphides of iron, lead, mercury, and zinc are found in recently filled fissures. The water which acts as a transporting agent is either meteoric (rain water) or magmatic (the waters issuing from cooling masses of rock). The latter are believed by many geologists to be the more 372 PHYSICAL GEOLOGY effective carriers of metalliferous minerals in the majority of deposits, although meteoric waters were apparently the sole agents in some cases. Gases and vapors given off by molten magmas have also formed some deposits. The importance of igneous intrusions in the production of ore deposits is readily understood when the history of such an intrusion is considered. When a sedimentary rock, for example, is penetrated by a molten mass, it is more or less fractured. In these fractures the waters heated by the igneous mass can circulate, and if they contain minerals in solution the latter may be precipitated. Moreover, igneous rocks are often rich in metallic minerals, and the water derived from them, the magmatic waters, may be the chief source of the metal- lic minerals of the ore deposit. Cause of Precipitation. Veins exist only in the zone of fracture, that is, at a depth seldom as great as 10 or n miles, and in most cases within a mile or two of the surface. The precipitation of minerals in veins may be brought about in one of a number of ways. (i) It may be caused by the mingling of waters. This is due to the fact that having pursued different courses the waters may carry different salts which may react to cause the precipitation of metallic and other minerals. (2) Precipitation may also be brought about by the contact of solutions with rocks which contain carbon or other minerals which cause precipitation ; (3) by a decrease in temperature ; (4) by a change in pressure; and (5) by oxidation (if the solutions are brought near the surface). (6) If two rocks differing in chemical composition are in contact, as, for example, a limestone and an igneous rock, precipitation is favored at or near the plane of contact, since the waters from the two are differently mineralized. When a mineral has once formed on a fissure wall, it may act as a center of attraction and cause a further accretion of the same mineral. This process is called mass action. Replacement Deposits. Waters carrying minerals in solution sometimes attack the rocks which they penetrate, dissolving them and at the same time depositing some of their load. This is accom- plished molecule by molecule, a particle of vein material being de- posited as a particle of the rock is dissolved out. Many of the rich ore deposits are of this origin. Replacement deposits often occur along faults and near the boundary or contact of igneous with sedi- mentary rocks. The boundaries of veins are often indefinite, since the width may ORE DEPOSITS 373 depend either upon the width of the original fissure or upon the amount of the replacement of the walls, or upon both. If replace- ment has not occurred, the boundary of the vein may be distinct. Weathering and Concentration of Ores. As a metalliferous vein is eroded, it is attacked by the agents of the weather and under- ground water. The result of such action is the removal of the more soluble minerals in the surface zone, (i) If the minerals removed are worthless, the portion remaining will be richer. For example, in gold-bearing quartz veins in which the gold is contained in pyrite, the solution of the pyrite leaves the pure gold in a honeycombed, rusty quartz. It was such quartz veins which delighted the old-time prospector. (2) If the minerals removed are valu- able and are depos- ited lower in the vein by the percolating water, a rich deposit may result. In a vein containing chal- copyrite and pyrite, for example, the iron may be left in the upper part of the vein in the form of limonite. This is FIG. 358. Vein showing three zones: A, surface or weathered zone; B y oxidized or middle zone; C, unaltered or sulphide zone. The weathered zone, A, is often largely composed of iron hydroxide and is called gossan. called the gossan (chapeau de fer and eisen hut) and may be in suffi- cient quantity to be mined as iron ore (Fig. 358). Lower in the vein, in the oxidized or middle zone, the ores are in the form of oxides, carbonates, etc., and may be enriched by the addition of metallic minerals brought down from the weathered zone. In some deposits the oxidized zone is the only portion of the vein in which the mineral occurs in sufficient quantities to be ex- tracted with profit. Beneath the oxidized zone, which extends to or below the level of ground water, lies the unaltered vein material of the unoxidized zone. Here the ores occur as they were originally deposited. These three zones are not usually separated by well-defined boundaries, 374 PHYSICAL GEOLOGY the change from one to the other being sometimes so gradual that it is difficult to say where one begins and the other ends. Veins are known in which oxidized ores occur several hundred feet below the water table. Magmatic Segregation. Certain iron and nickel deposits which occur in igneous rocks were probably brought together while the rocks were in a molten condition as the result of segregation. Few workable deposits, however, have been formed in this way. Placer Gold Deposits. In the early days of gold mining in many countries the first gold was found in the gravels of stream beds. The gold of the Klondike in northwestern Canada and the majority of the early finds of Alaska were located in stream gravels, while that OUTCROP STREAM GRAVELS FIG. 359 A. Diagram showing the development of eluvial or residual placers, which may be worked like ordinary stream placers, and stream gravels. In this case the source of the gold is the quartz vein. (After Lindgren.) of Nome was found in the sands of the seashore. The gold rush to California in 1849 was due to the find- FIG. 359 B. Diagram showing ancient aurif- erous gravels (dotted) covered by a lava flow in g of Stream or placer gold, (vertical lines). In mining the gravels a tunnel The source of the nuggets or is driven as indicated. j ust m strear n gravels is evi- dently to be found in the rocks over which the streams or their tributaries now flow or formerly flowed. The way in which placer gold was transported and deposited is simple. The gold occurred either in veins or scattered through the country rock in small quantities. When these rocks were disintegrated by weathering and the fragments carried away by ORE DEPOSITS 375 the streams, the heavy gold particles quickly sank to the bed of the streams, while the lighter minerals were borne on by the current. In this way much of the gold contained in a large quantity of rock has sometimes been concentrated in a small area. It consequently happens occasionally that rich placer deposits are found in regions in which none of the rock contains gold in sufficient quantity to pay for its extraction. When conditions are favorable, gold-bearing (auriferous) gravels are worked by dredging, even when the gravel yields only twenty- five or thirty cents to the cubic yard. Ancient gravels which have been buried beneath sheets of lava are sometimes mined for their gold (Fig. 3 59 B). Sedimentary Iron Deposits. Extending in a broken belt from Nova Scotia and New York to Alabama, beds of iron ore (Clinton FIG. 360. Iron deposits in the Lake Superior region, Mesabi Range, Minnesota. (U. S. Geol. Surv.) iron ore) occur which have the same position and much the same character as other sedimentary beds, and in some cases contain ma- rine fossils. These beds of iron ore may have been precipitated from salt or from fresh water, just as iron is being deposited to-day in fresh-water ponds and lakes. The iron contained in small quan- tities in the rocks (usually igneous) of the land is leached out by per- colating waters in the form of ferrous compounds. These compounds upon exposure to the air are oxidized and ferric oxide (Fe 2 O3) is precipitated, usually in the form of limonite (2 Fe 2 O3 3 H 2 O). In this way iron accumulates in bogs and is called bog ore, and similar deposits are laid down in lakes. Another suggestion which, however, is not widely accepted, is that the Clinton iron ore has been derived from lavas rich in iron minerals which were extruded beneath the sea. The great iron deposits of the Lake Superior region (Fig. 360) are believed to have been accumulated in beds as impure iron car- bonates and silicates, too low in iron to pay for their extraction. When the deposits were uplifted to form land, they were exposed 376 PHYSICAL GEOLOGY to weathering and were enriched (i) by the removal of the impurities by solution ; (2) by the replacement of the impurities by iron oxides as the former were dissolved out; or (3) by concentration, as a result partly of the removal of impurities and partly of replacement. So wide and deep are some of these Lake Superior iron deposits that they are excavated by steam shovels. The excavation in the Mesabi region, taking into account both the removal of the ore and of the glacial drift which overlies it, is far more extensive than the work conducted at the Panama Canal. In most of the deposits underground mining methods are employed. REFERENCES FOR ORE DEPOSITS LINDGREN, W., Mineral Deposits. RIES, H., Economic Geology. U. S. Geol. Survey Bulletins, Professional Papers, and Monographs. PART II. HISTORICAL GEOLOGY CHAPTER XIII HISTORICAL GEOLOGY HISTORICAL geology deals with the evolution of the life of the past, and with the development of the continents and oceans. It traces out, as accurately as our present knowledge will permit, the changes through which the earth has passed; it endeavors to gather from the available record the history of the life of geological times and the evolutional changes which the many classes of animals and plants have undergone and, as far as possible, to determine the cause or causes of these changes. This section of geology is concerned not only with the recording of facts, but is also, to an important degree, philosophical. Human history is but a short chapter of geological history, the former being measured in thousands of years while the latter extends over millions of years. The immensity of geological time is beyond our comprehension, but some conception of it can be gained when it is remembered that the time necessary to excavate the Grand Canyon of the Colorado was, geologically, comparatively short; that a maximum thickness of sediments of not less than 40 miles has been laid down in the seas; that great mountain ranges have not only been raised but have been worn down to sea level during portions of the smaller divisions of geological history. Perhaps the most striking evidence of the length of geological time is to be seen in the evolution of life. FOSSILS A fossil is any remains or trace of an animal or plant preserved jp the rocks of the earth. _ It mav consist of the original substance of the animal, or it may be merely an impression, such as a footprint or a worm trail. Even the flint implements made by primitive man may be considered as fossils. 377 378 HISTORICAL GEOLOGY When a shell or other organic remain is buried in the mud or sand of an ocean or lake bottom, in the dune sand of a desert, in volcanic dust, in a peat bog, or in the flood plain of a river, the record of its existence may be preserved in a number of ways. (1) The Original Substance may be Preserved. In recent sediments the shells are often unchanged, even the nacreous luster being re- tained. In the ice of Siberia mammoths have been found whose flesh had been so perfectly preserved that it was eaten by dogs and wolves and possibly by the natives themselves. Insects are found in amber the fossil gum of cone-bearing trees in which they were entrapped and covered. (2) Replacement. The original substance may have been entirely replaced by some other mineral, and shells, corals, and bones are often found which, although bearing little external evidence of alteration, are composed entirely of silica or some other mineral. As alkaline water is a salient of silica the petrifaction of wood (Fig 361) is brought about when such water containing silica in solution is neutralized, since the silica is then precipitated. If then a log buried in a bed of sand or volcanic ash FIG. 361. Petrified log, Adamana, Arizona. is saturated with underground water that is slightly alkaline, the replacement of the wood by the silica will be slowly brought about as the wood decays. As each particle of wood is oxidized carbon dioxide will be formed, this acid (H 2 CO 3 ) will neu- tralize the alkali of the water and will cause the precipitation of the silica at the point where the wood decayed. By some such slow process the wood may be replaced particle by particle until the entire tree is converted into a solid cylinder of silica. Silica is not the only mineral which replaces the substance of shells, bones, and other hard parts. Pyrite, iron oxide, lime carbonate, and other minerals sometimes occur. HISTORICAL GEOLOGY 379 (3) Casts and Molds. The original substance may be carried away in solution by underground water, leaving a cavity in which only the external form is preserved ; in other words, a mold of the shell or bone is left. Often natural casts of these molds are formed by mineral matter carried into the mold or by the infiltration of mud. Molds of the interior (Fig. 362) and exterior (Fig. 363) are frequently en- countered in porous rocks. Some fine opals in Nevada have the form FIG. 362. Specimens showing the original shell (B) and a natural mold (^) of the interior of a similar speci- men from which the shell has disap- peared. (Turritella mortoni.) FIG. 363. One half of a con- cretion showing the leaf which formed the nucleus. of branches, but are in reality casts of the branches of trees, the cavities formed by the decay of the wood having been filled with silica. (4) Footprints, Trails, etc. Many animals are known from their footprints, trails, burrows, or the impressions (Fig. 380, p. 411) made by their bodies in the soft mud. Entombment of Plants and Animals. The most favorable con- ditions for the preservation of animal life are to be found on those portions of the ocean bottom which are not uncovered by tides and where sediments are accumulating. When under such conditions shellfish or other animals die, their bodies may be buried in the mud or sand and preserved. It is not unusual to find layers of rock made up largely of the remains of shells which were buried in this way. On 380 HISTORICAL GEOLOGY the surface of some slabs of rock 250 or 300 fossils may sometimes be counted. Animal and plant remains are often well preserved in lake de- posits. In deposits of this class are found leaves, branches, and flowers which were carried from the surrounding land by the streams, insects which were beaten down by the wind to the surface of the lake, and vertebrates which were floated down the streams and found a burial on the lake bottom. Some of the most beautiful fossils were made in this way, but deposits of this class are much less important than those of marine origin, both because of their smaller extent and because the contained fossils seldom afford a means of exact correlation with those of other countries. The fossils preserved in delta swamps and flood plains are often numerous, and during certain periods of the earth's history have afforded the chief record of the vertebrates of these periods. Fossils are also preserved in wind-blown sand, in peat bogs, in cav- erns, and in travertine. Imperfection of the Record. The record of ancient life must necessarily be imperfect for two reasons, (i) Only a small per- centage of the life of any one period is preserved. This can be seen best by observing the proportion of the plant and animal life of to-day that will remain as a record of the life of the twentieth century. Of the life of the sea only the animals with shells or skeletons will be preserved in large numbers ; the myriads of soft-bodied animals such as jellyfish and protozoans will not form recognizable fossils except under very exceptional conditions. The trees of the forest decay where they fall, and it is seldom that any are buried and leave a per- manent record. The same fate awaits land animals, since upon their death their bones are soon disintegrated by the agents of the atmos- phere and they crumble to dust. It is only the bones of the occasional carcass which floats downstream and is buried under favorable con- ditions that will form fossils. (2) Even after being buried, the record is not always preserved. Thousands of square miles of sediments have been metamorphosed and the contained fossils destroyed. When marine sediments have been raised to form land, they are immediately attacked by the weather and erosion and are soon carried away. We consequently find that thousands of feet of rock have been removed and the record has been completely lost. Much of the fossiliferous strata is also either buried so far beneath younger rocks as to be inaccessible HISTORICAL GEOLOGY or is under the waters of the seas and so beyond the reach of the geologist. GEOLOGICAL CHRONOLOGY Relative ages of strata are determined in two ways. (1) Order of Superposition. If a series of strata or beds is in the order in which they were laid down (Fig. 364), it is evident that the oldest will be at the bottom and the youngest at the top. It is for this reason that the strata of a geological section are always placed with the oldest at the bottom of the column. This order is conclusive proof of the rela- tive age of rocks unless they have lost their original position by faulting or folding. (2) Chronology Determined by Fossils. After the true order of a series of beds has been determined by their superposition, their contained fossils will usually make it possible to correlate them with strata which may be hundreds or even thousands of miles distant. This is rendered possible by the fact_that the inhabitants of the earth have under- gone a progressive change which has, as a whole, been gradual, but which has taken place more rapidly at certain times than at others. Certain classes became dominant for a time, and then declined but seldom entirely disappeared. As a result of this change the assemblage of animals and plants of each division of geological history differs from that of every other. The fact that life has suffered such a progressive modification is of the greatest im- portance, since, as already indicated, it furnishes a means by which the relative age of the rocks in different parts of the world can be determined. Since certain species have a short geological life (their vertical range is short), when ''such are present the relative age of the rock is readily fixed. Although fossils are the surest test of the relative age of widely separated strata it should not be con- cluded that they prove exact contemporaneity, since in favored regions an old fauna may live thousands I Q