THE LIBRARY 
 
 OF 
 
 THE UNIVERSITY 
 OF CALIFORNIA 
 
 LOS ANGELES 
 
 The RALPH D. REED LIBRARY 
 
 O 
 
 DEPARTMENT OF GEOLOGY 
 
 UNIVERSITY OF CALIFORNIA 
 
 LOS ANGELES, CALIF. 
 
 Tke RALPH D. REED LIBRARY 
 
 DEPARTMENT OK GEOLOGY 
 
 UNIVERSITY of CALIFORNIA 
 
 LOS ANGELES, CALIF. 
 
 ^Xtrf of Oil Companies of Southern Cali- 
 fornia, Alumni and Faculty of Geology Depart- 
 ment and University Library. 
 
 1940 
 
*" ' * 
 
 C> f-/ I O . 
 
 RALPH D. REED 
 
Hmerican Science Series 
 
 Physics. v 
 
 By GEORGE F. BARKER. 
 
 Chemistry. 
 
 By IRA REMSKN, President of tne Johns Hopkins University. 
 
 Astronomy. 
 
 By SIMON NK vcoMBand EDWARD S. HOLDEN. 
 
 Geology. 
 
 By THOMAS C. CHAMBKRLIN and ROLLIN D. SALISBURY, Professors 
 in the University of Chicago. 
 
 Physiography. 
 
 By ROLLIN D. SALISBURY, Professor in the University of Chicago. 
 
 Geography. 
 
 By ROLLIN D. SALISBURY, HARLANH. RAKKOWS and WALTER S. 
 TOWER, of the Department of Geography, the University of 
 .Chicago. 
 
 Central Biology. 
 
 By WILLIAM T. SEDGWICK, Professor in the Mass. Institute, and 
 EDMUND B. WILSON, Professor in Columbia University. 
 
 Botany. 
 
 By CHARLES E. BESSBY, Professor in the University of Nebraska. 
 
 Zoology. 
 
 By A. S. PACKARD, Professor in Brown University. 
 
 The Human Body. 
 
 By H. NEWELL MARTIN. 
 
 Psychology. 
 
 By WILLIAM JAMES, Professor in Harvard University. 
 
 Ethics. 
 
 By JOHN DEWKV. Professor in Columbia University, and JAMBS 
 H. Turrs, Professor in the University of Chicago. 
 
 Political Economy. 
 
 By FRANCIS A. WALKER. 
 
 Finance. 
 
 By HENRY C. ADAMS, Professor in the University of Michigan. 
 
 For full descriptions of the Advanced, Briefer, and Elemen- 
 tary Courses published under each topic, see the publishers' 
 Educational Catalog. 
 
 34 West 33d Street, N. T. 
 623S.WabaaliAve. f Chicago 
 
AMERICAN SCIENCE SERIES ADVANCED COURSE 
 
 PHYSIOGRAPHY 
 
 ROLLIN D. SALISBURY 
 
 Professor of Geographic Geology and head of the Department of Geography 
 in the University of Chicago 
 
 THIRD EDITION, REVISED 
 
 NEW YORK 
 
 HENRY HOLT AND COMPANY 
 1913 
 
COPYRIGHT, 1907 
 
 BY 
 HENRY HOLT AND COMPANY 
 
Geology 
 Library 
 
 PREFACE 
 
 THIS volume is intended for students of early college or normal- 
 school grade, who have no purpose of pursuing the study of physical 
 geography beyond its elements, but who are yet mature enough 
 for work beyond the grade appropriate for the early years of the 
 secondary schools the stage when physical geography is usually 
 studied previous to college or normal-school work. No book 
 heretofore prepared has been intended especially for this class of 
 students. The work outlined here is essentially the work covered in 
 the University of Chicago in a twelve weeks' course, taken most 
 largely by students who have but recently entered college. The 
 work outlined might appropriately be expanded to a half-year 
 course, where so much time is available. 
 
 In the preparation of the text, the effort has been to shape 
 it, where practicable, so as to lead the student into the subject 
 under discussion, rather than to tell him the conclusions which 
 have been reached by those who have made the subject their 
 special study. This point is illustrated, for example, by the treat- 
 ment of isothermal maps. This method of work has been found 
 by the author, and by numerous other teachers as well, to be 
 eminently successful in practice; but the author is far from assum- 
 ing that every teacher will approve of it, or that it is the best for 
 every teacher. That method is best for any teacher which he can 
 use most effectively. If some method other than that of this 
 text leads to better results, the teacher who uses the book should 
 be free to follow it, for text-books should be servants, not masters. 
 
 The book contains no specific suggestions to teachers. It 
 assumes that the teacher does not need them. It is the author's 
 belief that if a text were to suggest the means which any ingenious 
 teacher may devise for arousing interest and for holding it through 
 the development and the solution of problems, it would attempt 
 to do much which should be left to individual initiative. The 
 author does suggest, however, that the leading of the student (1) 
 
 8732 '17 
 
Iv PREFACE 
 
 to raise questions pertinent to the topic under discussion, (2) to 
 formulate them into definite problems, and (3) to discover the 
 means (a) of solving the problems, and (6) of testing the correct- 
 ness of his solution, must always be, in large part, the work of the 
 teacher, not of the text-book, and that no thoroughly successful 
 teaching can leave these things undone. The text-book can, at 
 best, hope merely to supply the setting for such problems and such 
 work. 
 
 The map exercises which are suggested at various places in 
 the book are essentially those which are used by the author and 
 his colleagues. These exercises may be readily extended, if time 
 permits. These particular exercises have grown up from small 
 beginnings, through the collaboration of Dr. Wallace W. Atwood 
 and Mr. Harlan H. Barrows. 
 
 Another phase of work which should not be neglected is work 
 out of doors. This must form a part of the work of every strong 
 course in this subject. Directions for local field work cannot be 
 outlined profitably in a text-book, for the work must be shaped 
 with reference to the specific locality where the subject is studied. 
 Both field work and map work should have for their aim the appli- 
 cation of the principles studied, in such a way as to make the 
 subject vital. The aim of every laboratory exercise carried out 
 in connection with this subject should be the same, and any labora- 
 tory work which does not either illustrate and enforce principles, 
 or lead to them, is not worth development. The student who 
 cannot apply what he has learned in the class-room to his out-of- 
 door surroundings, has not secured the maximum good from his 
 study of the subject. 
 
 It may seem to some teachers into whose hands this volume 
 may fall that some parts are unnecessarily simple, and especially 
 that some things are introduced which the student should know 
 before entering college. In the abstract, the author is in sympathy 
 with this view; but it is to be remembered that large numbers of 
 students enter colleges and normal schools without any knowledge 
 of this subject except that acquired in connection with general 
 geography, as studied in the elementary schools. 
 
 The writer is indebted to various colleagues for suggestions of 
 one sort and another in the preparation of this volume, but es- 
 pecially to Mr. Harlan H. Barrows, who has read the manuscript 
 with great care and intelligence, and has made many useful sug- 
 
PREFACE V 
 
 gestions. Dr. Atwood has also rendered important assistance at 
 various points. Many of the illustrations of the volume have been 
 taken from the larger work on Geology by Professor T. C. Chamber- 
 lin and the author. Many others have been taken from the publi- 
 cations of the United States Geological Survey, and a few from other 
 sources, acknowledged in the text. 
 
 UNIVERSITY OF CHICAGO, December, 1906. 
 
CONTENTS 
 
 PAGE 
 
 INTRODUCTION 3 
 
 PART I 
 
 THE LITHOSPHERE 
 CHAPTER I 
 
 RELIEF FEATURES 
 
 RELIEF FEATURES OF THE FIRST ORDER 5 
 
 The continental platforms, 10. Continuity and dis- 
 continuity of continental platforms and oceanic basins, 11. 
 Grouping of the continents, 11. Origin of relief features 
 of the first order, 12. 
 
 RELIEF FEATURES OF THE SECOND ORDER 15 
 
 Great relief features of the land, 15. Great relief 
 features of the sea bottom, 15. 
 
 Plains 16 
 
 Coastal Plains, 17. Explanation of contour map, 19. 
 Relief, 19. Drainage, 21. Culture, 21. Interior plains, 
 22. Topography of plains, 24. Extent and habitability, 
 25. 
 
 Plateaus. ' 28 
 
 Position and area of plateaus, 30. Relief of plateaus, 
 30. Other features of plateaus, 30. Origin, 31. 
 
 Mountains 33 
 
 Mountains in history, 38. Origin. 39. 
 
 SUBORDINATE TOPOGRAPHIC FEATURES 42 
 
 Land surface and ocean bottom, 42. The development 
 of minor topographic features, 42. Changes now taking 
 place on the land, 43, 
 
 vii 
 
viii CONTENTS 
 
 PAGE 
 
 THE MATERIALS OF THE LAND 45 
 
 Mantle rock, 46. Rock, 47. Classes of solid rock, 48. 
 Sedimentary rocks, 48. Igneous rocks, 51. Metamorphic 
 rocks, 53. 
 
 CHAPTER II 
 
 THE WORK OF THE ATMOSPHERE 
 
 MECHANICAL WORK. THE WORK OF THE WIND , . 55 
 
 Dust 55 
 
 Universality, 55. Sources of dust, 56. Volcanic dust, 
 56. Loess, 58. How held in the air, 61. Distribution, 
 62. Gradational effect of winds, 62. 
 
 Sand 62 
 
 Sources of sand, 62. Lodgment of wind-blown sand, 
 62. Dunes, 63. Distribution of dunes, 63. Configuration 
 of dunes, 65. Destructiveness of eolian sand, 65. Migra- 
 tion of dunes, 67. Not all eolian sand in dunes, 69. Ripple- 
 marks, 69. Gradational effects, 69. Abrasion by the 
 wind, 70. 
 
 THE CHEMICAL WORK 71 
 
 Weathering, 72. 
 CHANGES BROUGHT ABOUT UNDER THE INFLUENCE OF THE Am . . 72 
 
 Freezing and thawing, 72. Expansion and contraction 
 of rock; rock-breaking, 73. 
 SUMMARY 78- 
 
 CHAPTER III 
 
 THE WORK OF GROUND-WATER 
 
 GENERAL FACTS 80 
 
 Source of land-water, 80. The fate of rain-water, 81. 
 The existence of ground-water, 83. The source of ground- 
 water, 83. Descent of ground-water, 84. The ground- 
 water surface, 85. Amount of ground-water, 86. The 
 movement of ground-water, 86. 
 
 Springs 89 
 
 Temperature, 89. Mineral and medicinal springs, 90. 
 Geysers, 90. Artesian and flowing wells, 94. 
 
 THE WORK OF GROUND-WATER 96. 
 
 Chemical Work 95, 
 
 Solution, 96. Deposition, 99. Other changes, 105. 
 Summary, 105. 
 
CONTENTS ix 
 
 PAGE 
 
 Mechanical Work 105 
 
 Abrasion, 105. Slumping, sliding, etc., 105. 
 
 Weathering 110 
 
 Conditions att'ecting weathering, 111. 
 
 CHAPTER IV 
 
 THE WORK OF RUNNING WATER 
 
 Sources of stream water, 116. 
 
 THE EROSIVE WORK OF STREAMS 120 
 
 Load and loading, 122. Carrying, 127. Amount of 
 load, 128. Erosion defined, 129. Deposition a necessary 
 consequence of erosion, 129. 
 
 Changes Made by Rivers in their Valleys 129 
 
 The deepening of valleys, 129. Depth-limit, 131. The 
 widening of valleys, 131. Width-limit, 134. Valley flats, 
 135. The lengthening of valleys, 140. Summary, 141. 
 
 The History of a River System 141 
 
 The courses of valleys, 144. The permanent stream, 
 145. Not all valleys are grown-up gullies, 146. Growth 
 of tributaries, 147. Stages in the history of a valley, 148. 
 Cycle of erosion, 153. Peneplains, 153. 
 
 Rate of Land Degradation 154 
 
 Conditions affecting the rate of erosion, 155. 
 
 Exceptional Features Developed by Erosion 156 
 
 Canyons and gorges, 156. Bad lands, 160. Natural 
 bridges, 161. Rapids and falls, 163. Narrows, 169. 
 Rock terraces, 171. Monadnocks, rock ridges, etc., 171. 
 
 Acddents.to Streams 173 
 
 Drowning, 173. Rejuvenation, 174. Ponding, 175. 
 Piracy, 176. 
 
 Consequent and Antecedent Streams 177 
 
 DEPOSITION BY RUNNING WATER 179 
 
 Causes of Deposition 180 
 
 Loss of velocity, 180. Excess of load from tributaries, 
 181. 
 Location of Alluvial Deposits and their Topographic Forms. ... 182 
 
 At the bases of steep slopes, 182. In valley bottoms, 
 184. Flood-plain meanders, 187. Fertility of alluvial 
 plains, 191. River floods, 195. At debouchures, 198. 
 Ill-defined alluvium, 202. 
 Alluvial Terraces . . . 203 
 
x CONTENTS 
 
 CHAPTER V 
 THE WORK OF SNOW AND ICE 
 
 PAGE 
 
 MINOR FORMS OF ICE . . 207 
 
 Ice beneath the surface, 207. The ice of lakes, 207. 
 Ice on the sea, 210. Ice-foot, 212. Ice in rivers, 213. 
 Ground-ice, 214. Snow, 214. Snow-fields, 215. The 
 snow-line, 217. Ice-fields, 218. 
 
 GLACIERS 219 
 
 The Valley Glacier. . . .< 223 
 
 Its surface, 223. Waste and supply of ice, 229. Rate 
 of movement, 229. Conditions affecting rate of movement, 
 230. Nature of glacier movement, 231. 
 
 Ice-caps 234 
 
 Piedmont Glaciers 240 
 
 THE WORK OF GLACIERS 242 
 
 Erosion, 242. Materials gathered, 251. Deposition of 
 debris in transit, 253. 
 
 Deposition by Glaciers 255 
 
 The terminal moraine, 2G7. The ground moraine, 257. 
 Lateral moraines, 257. Distribution and disposition of the 
 drift, 258. Re'sume', 260. 
 
 Fluvio-glacial Deposits 265 
 
 Icebergs 269 
 
 ANCIENT GLACIERS AND ICE-SHEETS 270 
 
 Cause of Glacial Epochs : 273 
 
 CHANGES PRODUCED BY THE CONTINENTAL GLACIERS 274 
 
 Changes Produced by Erosion 274 
 
 On elevations, 274. In valleys, 274. Rock basins, 275. 
 
 Chanqes Produced by Deposition 275 
 
 General distribution of the drift, 275. Terminal 
 moraines, 276. The ground moraine, 278. Effect of drift 
 on topography, 278. Effect of drift deposits on drainage, 
 280. Stratified drift, 287. Effects of glaciation on human 
 affairs, 287. 
 
 CHAPTER VI 
 
 LAKES AND SHORES 
 
 GENERAL FACTS 292 
 
 Distribution of Lakes 293 
 
 In latitude, 293. In mountains, 293. Along rivers, 
 293. Along coasts, 294. On plateaus, 294. Other situa- 
 tions, 295. 
 
CONTENTS xi 
 
 PAGE 
 
 Area, Topographic Position, Depth, etc 29,5 
 
 Area and topographic position, 295. Depth, 296. 
 Volume, 297. Movements of lake water, 300. Changes of 
 level, 300. 
 
 Conditions Necessary for the Existence of Lakes 301 
 
 The sources of lake water, 301. 
 
 Changes now taking Place in Lakes 302 
 
 The filling of their basins, 302. The lowering of their 
 outlets, 303. Fate of lakes, 303. 
 
 The Origin of Lake Basins 303 
 
 Diastrophism. 303. Vulcanism, 305. Gradation, 305. 
 River lakes, 308. Shore lakes, 311. Glacial lakes, 311. 
 Glacial lakes an index of topographic age, 313. Lakes due 
 to slumping, 313. Solution, weathering, wind, etc., 313. 
 
 Salt Lakes 314 
 
 The Climatic Effect of Lakes 316 
 
 Economic Advantages and Disadvantages 316 
 
 THE TOPOGRAPHIC FEATURES OF SHORES 317 
 
 Gradational Changes now taking Place along Shores 317 
 
 Waves, undertow, shore currents, 318. Rivers, 329. 
 Winds, 330. Glaciers, 330. Shore ice, 333. Extinct lakes, 
 333. 
 
 CHAPTER VII 
 VULCANISM 
 
 Examples of Active Volcanoes . . . . 341 
 
 Stromboli, 341. Vesuvius, 341. Krakatoa, 348. Mont 
 Pelee and Soufriere, 350. Hawaiian volcanoes, 361. Com- 
 mon phenomena of an eruption, 366. 
 
 The Products of Volcanoes 367 
 
 Lava, 367. Cinders, ashes, etc., 368. Gases and 
 vapors, 368. 
 
 Number, Distribution, etc 368 
 
 Number, 368. Distribution, 368. Historical, 371. 
 
 Igneous Phenomena not Strictly Volcanic 371 
 
 Fissure eruptions, 371. Intrusions of lava, 374. 
 
 Causes of Vulcanism 375 
 
 Topographic Effects of Volcanic Action 378 
 
 Volcanic cones, 378. Destruction of volcanic cones, 
 382: Examples of fresh cones, 382. Examples of worn 
 cones, 382. Mt. Shasta, "382. Mt. Rainier, 383. Mt. 
 
xii CONTENTS 
 
 PACK 
 
 Hood, 383. The Marysville Buttes, 384. San Francisco 
 
 Mountain, 384. 
 
 Indirect Topographic Effects of Vulcanism 385 
 
 Volcanic necks, 385. Columnar structure, 388. 
 Mud Volcanoes . . .388 
 
 CHAPTER VIII 
 CRUSTAL MOVEMENTS. DIASTROPHISM 
 
 SECULAR CHANGES 392 
 
 Evidences of Elevation (Relative) of Land 393 
 
 Human structures, 393. Rocks, 393. Measurements, 
 393. Organic remains, 393. Raised beaches, 394. Sea 
 cliffs, 394. Sea caves, 394. 
 
 Evidences of Relative Depression 395 
 
 Human structures, 395. Submerged forests, 396. 
 Submerged valleys, 397. An Italian temple, 397. 
 
 7s it the Land or the Sea which Changes its Level ? 398 
 
 Why the Sea-level Changes 400 
 
 Sedimentation, 400. Submarine volcanic extrusions, 
 401. Diastrophism, 401. 
 
 Why the Land Changes Level 402 
 
 Changes of Level in the Interiors of Continents 402 
 
 General facts, 402. Extent, 403. Ancient changes of 
 level, 403. Future changes of level, 404. 
 
 Crustal Deformation 405 
 
 Warping and folding, 405. Faulting, 406. 
 
 EARTHQUAKES 408 
 
 Definition, 408. Strength and destructiveness, 408. 
 Examples, 412. Earthquakes starting beneath the sea, 
 420. The earthquake wave, 424. Frequency, 427. Dis- 
 tribution, 429. Causes of earthquakes, 430. Surface 
 changes caused by earthquakes, 432. 
 
 CHAPTER IX 
 ORIGIN AND HISTORY OF PHYSIOGRAPHIC FEATURES 
 
 PLAINS 435 
 
 PLATEAUS 437 
 
 MOUNTAINS 437 
 
 Distribution of mountains, 439. Heights, 440. 
 
PAGE 
 
 Oceanic mountains, 442. Changes taking place in moun- 
 tains, 443. 
 
 Origin of Mountains 445 
 
 Volcanic mountains, 445. Mountains produced by ero- 
 sion, 445. Mountains produced by intrusion and uplift, 
 447. Mountains produced by folding, 447. Mountains 
 produced by faulting, 448. Summary, 449. 
 
 Effects of Mountains on Mankind 450 
 
 Climatic effects, 450. Barriers to transportation, 452. 
 Barriers to animals and plants, 452. Ores, 452. Agricul- 
 ture in mountains, 456. Scenic effects, 456. 
 
 THE OUTLINES OF THE CONTINENTS 457 
 
 Size, 457. Position, 460. Relief, 460. Distribution of 
 various types of irregularities, 461. Agents of gradation, 
 462. Diastrophism 463. Vulcanism, 464. Historical 
 bearing 465. 
 
 ISLANDS 466 
 
 By diastrophism, 466 By vulcanism, 466. By grada- 
 tion, 466. By combinations of diastrophism, gradation, 
 and vulcanism, 467. By organic processes, 468. 
 
 CHAPTER X 
 
 TERRESTRIAL MAGNETISM 
 Declination, 477. Dip, 479. Intensity, 479. 
 
 PART II 
 
 CHAPTER XI 
 EARTH RELATIONS 
 
 Form of earth, 482 Size, 484. 
 
 Motions 484 
 
 Rotation, 484. Effect of rotation, 488. Revolution, 
 488. 
 
 Latitude, Longitude, and Time 490 
 
 Latitude 490. Longitude, 491. Longitude and time, 
 492. Lengths of degrees, 494. Inclination of axis and its 
 effects, 497. Apparent motion of the sun, 499. Latitude 
 and sun altitude, 502, 
 The Solar System 504 
 
xiv CONTENTS 
 
 PART III 
 THE ATMOSPHERE 
 
 CHAPTER XII 
 GENERAL CONCEPTION OF THE ATMOSPHERE 
 
 PAGB 
 
 Substantiality, 506. Relation to the rest of the earth, 
 507. Density and altitude, 507. Height, 508. Volume, 
 510. Mass, 510. History, 510. 
 
 CHAPTER XIII 
 
 CONSTITUTION OF THE ATMOSPHERE 
 
 Principal constituents, 512. Minor constituents, 512. 
 Impurities, 513. Relations of constituents to one another, 
 513. 
 The Functions of the Atmospheric Elements 513 
 
 Nitrogen, 513. Oxygen, 514. Carbonic-acid gas, 514. 
 Water vapor, 517. Dust, 518. 
 
 CHAPTER XIV 
 
 TEMPERATURE OF THE AIR 
 
 The thermometer, 520. 
 
 THE HEATING OF THE ATMQSPHERE 521 
 
 Sources of heat, 521. Sun heating: insolation, 522. 
 Primary distribution of heat, 523. Radiation, 526. Con- 
 duction, 526. Convection, 527. How the sun heats the 
 atmosphere, 529. 
 
 THE SEASONS . 531 
 
 Differences between summer and winter, 532. Why we 
 have summer when we do, 532. Change of seasons, 533. 
 Seasons in other latitudes, 534. Effect of varying distance 
 of the sun, 536. Effect of altitude on temperature, 537. 
 
 REPRESENTATION OF TEMPERATURE ON MAPS 539 
 
 Isotherms, 539. Isothermal charts, 539. What deter- 
 mines the positions and courses of isotherms? 540. Alti- 
 tude, 545. Isothermal surfaces, 546. 
 
 RANGE OF TEMPERATURE 553 
 
 Daily, 553. Seasonal, 559. 
 
CONTENTS xv 
 
 PAGE 
 
 ATMOSPHERIC TEMPERATURE AND ATMOSPHERIC MOVEMENT 561 
 
 Land- and sea-breezes, 561. Monsoons, 562. Moun- 
 tain and valley breezes, 562. Vertical movements and 
 temperature, 563. 
 
 CHAPTER XV 
 
 THE MOISTURE OF THE AIR 
 
 Function of atmospheric moisture, 564. Sources of 
 water vapor: evaporation, 565. Rate of evaporation, 566. 
 Function of the atmosphere in evaporation, 568. Evapora- 
 tion takes up heat, 568. Amount of water vapor in the 
 air, 568. Distribution of water vapor, 569. Atmospheric 
 moisture and atmospheric movements, 569. Saturation, 
 569. Humidity and dew-point, 570. Condensation, 572. 
 Condensation and temperature, 572. Dew and frost, 573. 
 Clouds and fog, 574. Forms of clouds, 576. Precipitation, 
 579. Rain-making, 580. Summary, 580. 
 
 CHAPTER XVI 
 
 ATMOSPHERIC PRESSURE 
 
 The barometer, 582. Air pressures unequal, 583. 
 
 Representation of Pressure on Maps and Charts 584 
 
 Isobars, 584. Isobaric surfaces, 587. The courses of 
 isobars, 588. Isobars and parallels, 589. Relation of iso- 
 bars to land and water, 589. Isobars and temperature, 
 590 Isobars and humidity, 594. High-pressure belts, 594. 
 Permanent areas of low pressure, 597. Temporary and 
 local variations of pressure, 597. 
 
 CHAPTER XVII 
 
 GENERAL CIRCULATION OF THE ATMOSPHERE 
 
 Prevailing and periodic winds, 598. The general effect 
 of unequal insolation, 598. Effect of the extra-tropical 
 belts of high pressure, 600. The high-latitude areas of low 
 pressure, 601 Direction of winds, 602. The circumpolar 
 whirl, 605. Unequal heating of land and water a disturbing 
 factor, 605. Summary, 612. Gradient, velocity, and di- 
 rections of wind, 613. 
 
xvi CONTENTS 
 
 PAGE 
 
 The General Circulation and Precipitation 614 
 
 Rainfall in the zone of the trades, 616. Rainfall in the 
 zones of the prevailing westerlies, 617. 
 
 CHAPTER XVIII 
 
 WEATHER MAPS 
 
 Aperiodic Changes of Pressure 620 
 
 Isobars, 620. Wind, 622. Cloudiness, precipitation, 
 etc., 623. Temperature, 624. Movements of cyclones and 
 anticyclones, 632. Winds incidental to cyclones and anti- 
 cyclones, 645. Origin of cyclones and anticyclones of 
 intermediate latitudes, 648. Tropical cyclones, 648. 
 Weather predictions, 656. Failure of weather predictions, 
 660. Property saved by predictions of storms, frosts, 
 floods, etc., 663. 
 
 Special Types of Storms 663 
 
 Thunder-storms, 663. Whirlwinds, 666. Tornadoes, 
 667. Waterspouts, 673. Foehn winds, Chinook winds, 
 etc., 673. 
 
 CHAPTER XIX 
 
 CLIMATE 
 
 Definition, 676. Uniformity and variability, 677. 
 
 Classification of Climates 683 
 
 Climatic Zones 684 
 
 Zones defined by latitude, 685. Zones defined by winds, 
 686. Zones defined by isotherms, 687. Oceanic climates, 
 688. Continental imates, 690. Mountain and plateau 
 climates, 691. Climatic effect of forests, 692. 
 
 The Climates of the Several Zones 693 
 
 The tropical zone, 693. Climate of intermediate zones, 
 695. Climate of the polar zones, 700. Rainfall and agri- 
 culture, etc., 701. Climate and life, 702. 
 
 Changes of Climate 703 
 
 Within historic time, 703. In geologic time, 704. 
 
CONTENTS xvil 
 
 PART IV 
 THE OCEAN 
 
 The sea-level, 707. What the physical geography of the 
 sea includes, 709. Distribution of the ocean waters, 710. 
 Depth, 711. Mass, 712. Topography of the bottom, 713. 
 
 CHAPTER XXI 
 COMPOSITION OF SEA-WATER 
 
 The mineral matter in solution, 717. Withdrawal of 
 mineral matter from the sea, 718. A suggestion as to the 
 age of the ocean, 718. Gases in sea-water, 718. Salinity, 
 density, and movement, 719. Salinity and color, 720. 
 
 CHAPTER XXII 
 THE TEMPERATURE OF THE SEA 
 
 Temperature of the surface, 721. Temperature and 
 movement, 722. Temperature beneath the surface, 722. 
 The ice of the sea, 726. 
 
 CHAPTER XXIII 
 THE MOVEMENTS OF SEA-WATER 
 
 Causes of Movement 727 
 
 Inequalities of level, 727. Wind, 728. Differential at- 
 traction of sun and moon, 728. Occasional causes, 729. 
 
 Types of Movement 729 
 
 Waves 729 
 
 Currents 730 
 
 Cause of ocean currents, 733. Climatic effects of ocean 
 currents, 733. Gradational effects of ocean currents, 734. 
 Historical suggestions, 735. 
 
 Tides 735 
 
 The periodicity and the cause of tides, 736. Solar tides, 
 741. Spring tides and neap tides, 744. Other variations 
 
xviii CONTENTS 
 
 FAQB 
 
 in the height of high tides, 744. Cotidal lines, 748. Rate 
 of movement, 748. Effects of tides on shores, 748. 
 
 CHAPTER XXIV 
 THE LIFE OF THE SEA 749 
 
 CHAPTER XXV 
 MATERIALS OF THE SEA BOTTOM 753 
 
 CHAPTER XXVI 
 RELATION OF THE SEA TO THE REST OF THE EARTH. 756 
 
PLATES 
 
 PLATE FACING PAGE 
 
 I A narrow coastal plain in Oregon 20 
 
 II A well-drained plain in Kansas 24 
 
 III An ill-drained plain in Wisconsin 25 
 
 IV Fig. 1. The Canyon of the Yellowstone River. 
 
 Fig. 2. The Grand Canyon of the Colorado River 30 
 
 V Fig. 1. Dunes on coast of New Jersey. Fig. 2. Dunes 
 along Arkansas River in Kansas. Fig. 3. Dunes in 
 
 plains of Nebraska 66 
 
 VI Limestone sinks due to solution by ground-water. Near 
 
 Pikeville, Tenn 98 
 
 VII Streams disappearing in the sand, gravel, etc., at the base 
 
 of mountains in an arid region 132 
 
 VIII A stream widening its valley by lateral planation 133 
 
 IX Fig. 1. A meandering stream. The Missouri River. 
 Fig. 2. A further stage in the development of a 
 meander. The Schell River, Missouri. Fig. 3. A 
 
 plain in old age 138 
 
 X A well-developed river flat. Valley of the Mississippi, near 
 
 Prairie du Chien, Wis 138 
 
 XI Stream flats. The Missouri and Big Sioux Rivers 139 
 
 XII Fig. 1. Youthful Valleys. Shore of Lake Michigan just 
 north of Chicago. Fig. 2. A region in a mature state 
 
 of erosion 150 
 
 XIII The Niagara Gorge 151 
 
 XIV Entrenched Meanders 174 
 
 XV A piedmont alluvial plain or compound alluvial fan in 
 
 Southern California 184 
 
 XVI The alluvial plain of the Platte Rivers in Nebraska 185 
 
 XVII Glaciers on Glacier Peak, Washington 248 
 
 XVIII A portion of the Bighorn Mountains, showing glaciated 
 
 valleys -. 249 
 
 XIX Characteristic drift topography 278 
 
 XX Fig. 1. Coastal lakes formed by the blocking of the ends 
 of drowned valleys. Fig. 2. A group of lakes on the 
 
 coastal plain of Florida 294 
 
 xix 
 
XX 
 
 PLATES 
 
 FACING PAGE 
 
 XXI The upper end of Seneca Lake, New York 295 
 
 XXII Fig. 1. A coast line developed chiefly by wave erosion. 
 
 Fig. 2. An island tied to the mainland by a "beach". . . 322 
 
 XXIII A section of the California coast, showing lands, near the 
 
 coast, which have recently emerged 394 
 
 XXIV Cushetunk and Round Mountains, New Jersey 395 
 
 XXV Dunning Mountain, Pennsylvania 438 
 
 XXVI An area southwest of Denver showing a mountain ridge 
 
 dissected by erosion 439 
 
PHYSIOGRAPHY 
 
PHYSIOGRAPHY 
 
 INTRODUCTION 
 
 Definition. The science of physiography has been variously 
 defined, and while there is still much difference of opinion as to 
 the precise limits that should be set to it, there is a strong dis- 
 position, in the school world at least, to regard physiography as 
 one with physical geography. In England, physiography is often 
 regarded as a general introduction to science, and is made to 
 include the elements of all the physical and biological sciences. In 
 some other quarters physiography is regarded as the physical 
 geography of the land. 
 
 If physiography be regarded as another name for physical 
 geography, it has to do with (1) the solid part of the earth, the 
 lithosphere, (2) the water of the earth, the hydrosphere, and (3) 
 the air or atmosphere. Physiography, however, does not deal with 
 these several spheres exhaustively. The science of the atmosphere 
 is Meteorology; the science of the ocean, which contains the larger 
 part of the water of the hydrosphere, is Oceanography; and the 
 science of waters in general is Hydrography. The complete study 
 of the lithosphere includes several subordinate sciences, all of which 
 may be considered to be parts of the broad science of Geology, 
 which has to do, to some extent, with the atmosphere and the 
 hydrosphere, as well as with the lithosphere. 
 
 Physiography may be said to deal with the atmosphere only in 
 so far as the atmosphere affects the land, the water, and life, and 
 it deals with the water primarily in its relations to the land and to 
 life. So far as concerns the lithosphere, physiography deals with 
 its surface only, though it is more than a mere description of the 
 surface; it involves a consideration of the conditions and processes 
 which have brought the surface to its present state. The processes 
 involved are largely the result of the activity of the water and 
 the air, and of the life conditioned by them; but other factors, 
 
 3 
 
4 PHYSIOGRAPHY 
 
 such as volcanic forces, and the forces which cause the slow warp- 
 ings of the outer part of the lithosphere, are also involved. In 
 other words, physiography has to do primarily with the surface of 
 the lithosphere, and with the relations of air and water to it. Its field 
 is the zone of contact of air and water with land, and of air with water. 
 
 Physiography is not sharply separated from geology. Geology 
 has to do with the history of the earth; while physiography has 
 to do with a late chapter only of that history, the history of the 
 present surface. Every period of the past has had its physiography, 
 and the history of the successive physiographies, could they be 
 fully known, would give, in large part, the history of the earth. 
 
 Physiography is also closely related to geography, but it departs 
 from that science in that it has to do primarily with the relations 
 of the lithosphere, atmosphere, and hydrosphere, and with the 
 physical results of these relations, while geography, as contrasted 
 with physical geography, concerns itself primarily with the dis- 
 tribution of life (including man) and human industries, as affected 
 by the condition of the land surface, climate, resources, etc. Physi- 
 ography may be said to be, on the one hand, a special phase of 
 geography, namely, physical geography, and, on the other, a special 
 chapter of geology, namely, the latest. Since physical geography 
 affects the distribution of life and all its activities, it is not out 
 of place, in its study, to touch again and again the biological and 
 historical bearings of the subject. 
 
 Although the lithosphere, the hydrosphere, and the atmosphere 
 seem very distinct from one another, they are in reality somewhat 
 less sharply separated than they seem, for though the larger part 
 of the hydrosphere is contained in the ocean, lakes, and rivers, a 
 not inconsiderable part has sunk into the soil and rocks, while a 
 smaller amount always exists in the form of vapor in the atmosphere. 
 The water therefore invades both the lithosphere below and the 
 atmosphere above. So, too, a part of the atmosphere penetrates 
 the soil and the rocks of the land, and is mingled with the water 
 of the ocean, lakes, and rivers. Again, solid matter from the 
 lithosphere is found in suspension in streams, lakes, etc., often 
 making them muddy, and dust is always present in the atmosphere. 
 In spite of the interpenetration of these three spheres, they remain so 
 distinct that the boundaries between them are usually well defined. 
 
 In the development of our subject, the lithosphere, the atmos- 
 phere, and the hydrosphere will be considered in order. 
 
PART I 
 THE LITHOSPHERE 
 
 CHAPTER I 
 RELIEF FEATURES 
 
 THE oceans cover nearly three-fourths of the surface of the 
 earth, while but little more than one-fourth of the lithosphere 
 rises above the level of the seas, forming land. The volume of the 
 water in the oceans is so great that if the surface of the lithosphere 
 were reduced to a common level, that is if the protuberant parts 
 were planed down and the material deposited in the depressed 
 areas, there would be no land at all, but a universal ocean nearly 
 two miles deep. The existence of land therefore results from the 
 fact that the surface of the solid part of the earth is uneven, and 
 that the water has settled in the depressions. 
 
 It would help us to get a true picture of the surface of the 
 solid part of the earth, if we could see it without the oceans; but 
 since the oceans cannot be withdrawn, some conception of its 
 surface may be gained from a relief model of the earth which does 
 not represent the water (Figs. 1 and 2); or, if such a model is not 
 available, relief maps and charts of the ocean are serviceable. 
 
 RELIEF FEATURES OF THE FIRST ORDER 
 
 The most significant feature in the surface of the lithosphere is 
 the contrast between the great depressions, which we call the 
 ocean basins, and the broad elevations, which we call the continental 
 platforms. The continental platforms and the ocean basins are 
 topographic features of the first order. The contrast between them 
 is emphasized by the fact that there is almost everywhere a rather 
 steep slope from the one to the other, a steep descent from the 
 
6 
 
 PHYSIOGRAPHY 
 
 continental platforms to the ocean basins, or, looked at from the 
 other point of view, a steep ascent from the ocean basins to the 
 continental platforms (Figs. 1, 2, and 3). 
 
 The ocean basins and the continental platforms divide the 
 surface of the earth between them. Both the basins and the 
 platforms are irregular in shape and irregular in distribution. The 
 larger part of the elevated areas is in the northern hemisphere, 
 while the depressed areas are far in excess in the southern. 
 
 FIG. l. 
 
 FJG 2. 
 
 FIG. 1. Photograph of the Jones Relief Globe, showing the North Atlantic 
 Basin depressed notably below the continents about it. The vertical 
 scale of the globe is exaggerated. 
 
 FIG. 2. Photograph of the Jones Relief Globe, showing the basin of the 
 Indian Ocean, with its distinctly marked borders. 
 
 The continental platforms are somewhat larger than the con- 
 tinents (Fig. 3), and the ocean basins are somewhat smaller than 
 the oceans. The oceanic area (more than 143,000,000 square 
 miles) is nearly three times the land area (nearly 54,000,000 square 
 miles), but the area of the ocean basins proper (about 133,000,000 
 square miles), is only about twice as great as the area of the 
 continental platforms (about 64,000,000 square miles). The 
 discrepancy between the area of the oceans and that of the ocean 
 basins results from the fact that there is more water on the earth 
 than the true ocean basins will hold, and this excess overruns the 
 rims of the basins, and spreads itself out on the low borders (the 
 continental shelves) of the continental platforms (Figs. 4 and 6). 
 
RELIEF FEATURES 7 
 
 Some 10,000,000 square miles about the borders of the continental 
 platforms are thus covered by shallow water. The result is that 
 the area of the continents falls short of the area of the continental 
 platforms by this amount, while the area of the oceans corre- 
 spondingly exceeds that of the ocean basins. The waters which 
 
 Africa 
 
 FIG. 3. A diagrammatic section of the earth about the equator, showing 
 the elevated segments (continents) and the depressed segments (ocean 
 basins). Vertical scale X 40. (Based on section in Stanford's Atlas of 
 Universal Geography.) 
 
 lie on the low borders of the continental platforms have been 
 called epicontinental (upon the continent) seas. 
 
 If all lands were graded to a common level without increasing 
 or decreasing either their area or the amount of material they 
 contain, their height above sea-level would be about 2300 feet. 
 If the bottom of the sea were graded to a common level, its area 
 remaining as now, the water would be between 12,000 and 13,000 
 
8 
 
 PHYSIOGRAPHY 
 
 feet deep everywhere. The average height of the land is therefore 
 a little less than half a mile above sea-level, while the average 
 
 / 
 
 "K 
 
 depth of the ocean bottom is but little less than two and a half 
 miles below sea-level. The difference between the average height 
 
RELIEF FEATURES 
 
 9 
 
 o_ 
 
 cf the continental platforms and the ocean basins is therefore 
 about three miles. In other words, about two-thirds of the surface 
 of the solid part of the earth is sunk about three 
 miles below the other third. Three miles is a 
 little less than y^^ of the radius of the earth. 
 
 The surfaces of both the continental plat- 
 forms and the ocean basins are uneven, and as 
 a result, the maximum unevenness, or relief, of 
 the surface of the lithosphere is much more 
 than three miles. Its lowest known point, 
 near the Fiji Islands, is nearly six miles (about 
 31,000 feet) below the level of the sea, while 
 its highest point (Mt. Everest in the Hima- 
 layas) is nearly as much (about 29,000 feet) 
 above the same plane. The maximum relief 
 of the lithosphere is therefore nearly twelve 
 miles, or about 5 | of the earth's radius. The 
 areas of those parts of the ocean basins which 
 approach a depth of six miles are, however, 
 very limited in extent, and the areas of land 
 which approach the height of six miles are 
 hardly more than points. 
 
 The following table gives some idea of the 
 relief of the lithosphere: 
 
 Approximate percent, 
 of total area of the earth. 
 
 Area of land more than 6000 feet above sea-level 2 . 3 
 Area of land between 6000 and 600 feet above 
 
 sea-level 18 . 6 
 
 Area of land between 600 feet above sea-level 
 
 and sea-level 69 
 
 Area of ocean where water is less than 600 feet 
 
 deep 7 
 
 Area of ocean where water is between 600 and 
 
 6000 feet deep 7. 
 
 Area of ocean where water is between 6000 and 
 
 12,000 feet deep 14 .8 
 
 Area of ocean where water is between 12,000 and 
 
 18,000 feet deep 39 4 
 
 Area of ocean where the depth of the water ex- 
 ceeds 18,000 feet 31 
 
 i ^r^ 
 
 This table shows that more than half the lithosphere is more 
 than a mile below sea-level. 
 
 2111*3 
 
 O S-G 
 
 
 **! 
 
 2 o-^is 
 
 J3 G 
 
10 
 
 PHYSIOGRAPHY 
 
 The following table shows the proportion of land at various 
 elevations above the sea: 
 
 Percent, of land. 
 
 Less than 600 feet about 21 .90 
 
 Between 600 and 1,500 feet about 21 .63 
 
 Between 1,500 and 3,000 feet about 21 .34 
 
 Between 3,000 and 6,000 feet about 19.51 
 
 Between 6,000 and 12,000 feet about 12.34 
 
 Between 12,000 and 18,000 feet about 2.95 
 
 Above 18,000 feet about . 33 
 
 This table shows that about two-thirds of the land has an 
 
 elevation of less than 3000 feet. About six-tenths of the land is 
 
 less than 500 meters : (1640 feet) above sea-level, and upon it 
 
 the larger part of the population of the earth lives. The facts 
 
 shown in the above table are expressed diagrammatically 
 
 Metes 
 
 8800 by Fig. 7. 
 
 The continental platforms. The continental tracts which 
 are commonly recognized are (1) the Eurasian, (2) the Afri- 
 
 FIG. 7. Diagram showing the relative areas of the lithosphere at various 
 levels above and below sea-level. Less than 10 percent, of the litho- 
 sphere is as much as 700 meters above the sea, and only 28 percent, 
 is above the sea. About half the total surface of the lithosphere is 
 more than 3500 meters below sea-level. The diagram also shows that 
 the mean surface of the lithosphere is about 2300 meters below sea-level, 
 the mean ocean depth about 3500 meters, and the mean elevation of 
 . the land above the sea-level about 700 meters. (After Wagner.) 
 
 can, (3) the North American, (4) the South American, and (5) the 
 Australian, which includes New Guinea on the north. Besides these 
 elevated segments whose summits are the recognized continents, 
 there are other lesser though still great segments not commonly 
 recognized as continental. Of these the largest is (6) Antarctica, 
 which should probably be regarded as a continent, and (7) Green- 
 land, which is universally regarded as an island. Islands, in 
 
 1 It is serviceable to remember that 1 meter = 3. 281 (approximately) feet. 
 
RELIEF FEATURES 
 
 11 
 
 general, are not to be looked upon as relief features of the first 
 order, and will be referred to in other connections. 
 
 Continuity and discontinuity of continental platforms and 
 oceanic basins. The great land areas are notably discontinuous, 
 while the sea is continuous, though its parts bear separate names, 
 as Atlantic, Pacific, etc. In contrast with the lands and the seas, 
 the continental platforms are much more nearly continuous than 
 the continental lands, while the ocean basins are less continuous 
 than the oceans. Thus the American continental protuberance is 
 connected at the northwest with the Asian protuberance, and is 
 but slightly disconnected at the northeast from the European, 
 while the elevated Eurasian platform is connected with the 
 Australian and the African. Of the continental protuberances, 
 Antarctica alone seems to be really isolated. Of the ocean basins, 
 the Arctic is measurably isolated. It is of interest to note that 
 the most isolated basin is about one pole, and the most isolated 
 protuberance at the other, so far as present knowledge of the polar 
 regions allows of generalization. Some of the smaller deep basins, 
 such as those of the Mediterranean and the Gulf of Mexico, have 
 some such measure of isolation as some of the larger islands, as, 
 for example, Greenland and New Zealand. 
 
 Grouping of the continents. The northern hemisphere con- 
 
 FIG. 8. Land and water hemispheres. 
 
 tains more than twice as much land as the southern. If the earth 
 be divided into two hemispheres having their poles in England 
 and New Zealand, respectively (Fig. 8), the first would contain 
 about f of all the land, and might be called the land hemisphere, 
 while the latter would contain only about | of the land, and might 
 
12 PHYSIOGRAPHY 
 
 be called the water hemisphere. Even in the land hemisphere, 
 however, the water would cover rather more than the surface, 
 while in the water hemisphere it would cover about \\ of it (Fig. 8). 
 Since the northern hemisphere contains of the land and a still 
 higher proportion of the economically efficient land, it has always 
 supported the larger part of the human race. 
 
 Taken together, the continents may be looked upon as forming 
 a great horseshoe-shaped protuberance of the lithosphere, ex- 
 tending around the Atlantic from Cape Horn through the 
 Americas and Europe to the Cape of Good Hope in Africa, with a 
 spur stretching to the southeast to the East Indies and Australia. 
 
 If Europe and Asia be regarded as separate, the continents, 
 except Antarctica, may be grouped in pairs. The Americas form 
 one pair, Europe and Africa another, and Asia and Australia a 
 third. Considered in this way, the longest line of each pair is in 
 a general north and south direction. The continents are often 
 said to be triangular in shape, with their broadest ends to the 
 north, and their apexes to the south. This is conspicuously true 
 of South America, and less conspicuously true of North America 
 and Africa; but it is not true of Europe and Asia, either by them- 
 selves or combined, or of Australia or Antarctica. 
 
 Origin of relief features of the first order. The origin of the 
 ocean basins and the continental platforms is not known with 
 certainty. It is not certain that they have always existed, and it 
 is not likely that the former have always been depressed as much 
 as now below the latter, though there has apparently been little 
 change for ages. The best opinion seems to favor the view that 
 the sinking of the ocean basins, rather than the elevation of 
 the continental platforms, has been the important factor in the 
 development of the topographic features of the first order. The 
 chief reason for this view is the general fact that the earth is 
 cooling, and therefore shrinking. Shrinking means that the out- 
 side is getting nearer (on the average) to the center. This must 
 result in the depression of the surface on the average, though not 
 necessarily at every point. 
 
 If the subsidence of the ocean basins be the principal factor 
 in developing the great relief features of the lithosphere, we might 
 think of the continental platforms as having been (1) wedged up 
 (Fig. 9), or warped (Fig. 10) up between the sinking parts; (2) as 
 having remained where they were before the sinking of the depressed 
 
RELIEF FEATURES 
 
 continent 
 
 13 
 
 Fig. 9 expresses diagrammatically the conception that the continents were 
 elevated and the ocean basins depressed by movement along definite 
 sliding planes or fault planes. The dotted line may be taken to rep- 
 resent a somewhat uniform original surface, which may be looked 
 upon as the hypothetical surface before continents and ocean basins 
 were developed. The diagram indicates that the continents have 
 risen above this surface, while the ocean basins have sunk below it. 
 
 Fig. 10. This diagram represents the same conception as Fig. 9, except 
 that the movement was by warping instead of faulting. 
 
 Fig. 11. This diagram represents the same conception as Fig. 9, except 
 that the continental segment is represented as not having risen. 
 
 Fig. 12. This diagram represents the same conception as Fig. 10, except 
 that the continental segment has not risen. 
 
 Fig. 13. This diagram represents the same conception as Fig. 11, except 
 that both ocean basin and continental segment are represented as 
 having sunk below the original level, the former much more than the 
 latter. 
 
14 PHYSIOGRAPHY 
 
 parts (Figs. 11 and 12); or (3) as having sunk, but as having sunk 
 less than the basins (Fig. 13). All these conceptions imply change 
 in the relative positions of continental platforms and ocean basins. 
 All may have elements of truth in them, and all may have been 
 combined, so far as now known, in the evolution of the continents. 
 Present knowledge, however, does not permit of a definite state- 
 ment of their relative value, nor does it exclude other conceptions 
 of the origin of the topographic features of the first order. It is, 
 for example, possible, or even probable, that the surface of the 
 lithosphere was never uniform, and that the topographic features 
 of the first order are not entirely the result of deformation. 
 
 One conception of the origin of ocean basins and continental 
 platforms is based on the view that the earth grew to its present 
 size from a smaller ancestral body by the ingathering of matter 
 which was once outside itself, and that this growth was not equal 
 in all places. On this view, its surface may never have been 
 smooth. Even if this conception be the true one, it is alto- 
 gether probable that movements in the outer parts of the earth 
 have set off the ocean basins and the continental platforms from 
 each other more and more sharply in the course of the long history 
 of the earth. 
 
 Even if we suppose that the ocean basins have sunk, or that the 
 continents have been upraised, the times of movement are no 
 better known than the methods; but it is probable that the move- 
 ments have been intermittent rather than constant, and that periods 
 of movement, for example periods of sinking of the ocean basins, 
 have been followed by periods of quiet. 
 
 Geological history reveals the fact that the areas of the ocean 
 and land have changed somewhat from time to time, but it is 
 not known that the relative positions of ocean basins and continental 
 platforms have changed notably. If the bottom of the sea were 
 to .sink, the ocean basins would hold more water, and some part 
 of the epicontinental sea would be drawn off the submerged parts 
 of the continental platforms, that is off the continental shelves. 
 If the bottoms of the ocean basins were to sink about 600 feet, 
 the water would be drawn off the continental shelves, and the 
 continental lands would correspond with the continental platforms. 
 If the continental tracts, on the other hand, were to sink, the 
 waters of the sea would encroach upon their borders farther than 
 now, and the area of the land would be diminished. Geology 
 
RELIEF FEATURES 15 
 
 teaches that such changes as these have taken place at various 
 times in the past, so that the lower portions of the continental 
 platforms have been alternately submerged, and converted into 
 land. 
 
 RELIEF FEATURES OF THE SECOND ORDER 
 
 The continental platforms and the ocean basins are the relief 
 features of the first order. The more strongly marked lineaments 
 of these two great divisions of the lithosphere are the relief features 
 of the second order. 
 
 Great relief features of the land. The continental platforms 
 are made up of plains, plateaus, and mountains. The plains are 
 the lowlands of the continents, and the plateaus and mountains are 
 the highlands; but no one of these great types can be defined in 
 terms of altitude alone. Most continental lands may be readily 
 classed in some one of these three divisions, but many small islands 
 do not seem clearly referable to any one of them. The difficulties 
 which they present need not, however, be considered at this point. 
 
 Great relief features of the sea bottom. The major topo- 
 graphic divisions of the land may be contrasted and compared 
 with those of the sea bottom. The continental shelves which, it 
 will be remembered, are really parts of the continental platforms, 
 are submerged plains. They are below the sea-level by an amount 
 comparable to the elevation of the land-plains above it. Some of 
 the land-plains are, however, much higher above the sea than 
 
 FIG. 14. Diagram to illustrate the relations of mountain, plateau, plain, 
 ocean basin, ocean deep, etc. 
 
 any continental shelf is below it. The great areas of ocean bottom 
 covered by water one to three and a half miles deep may be 
 compared to the higher plains and the plateaus of the land, in 
 reverse; while the very deep tracts of limited extent on the ocean 
 bottom may be compared to very high plateaus, such as Tibet, in 
 reverse. Deep holes in the sea-b3d, corresponding to mountain 
 peaks in reverse, are not known to exist. From the tables on 
 page 9, and from Fig. 7, it will be seen that the very deep areas 
 
16 
 
 PHYSIOGRAPHY 
 
 I, 
 II 
 
 of the sea, say more than 12.000 feet, are much more extensive 
 than the correspondingly high areas of the land (p. 10), while 
 the low areas of the land (plains) are 
 much more extensive than the shallow 
 (epicontinental) part of the sea. 
 
 The relation between the topography 
 of continental platforms and ocean 
 basins may be made clear in another 
 way. Extensive areas of plateaus and 
 lesser areas of mountains rise above the 
 average level of the continental plat- 
 forms, while a few relatively small 
 basins, some of them occupied by lakes, 
 sink far below it. Similarly, ridges and 
 peaks, roughly comparable to the moun- 
 tains of the land, and broad areas such iddresissipi 
 as the continental shelves, comparable 
 to the plateaus, rise well above the gen- 
 eral level of the ocean floor, while rela- 
 tively small basins (deeps) are depressed 
 far below it. These relations are ex- 
 pressed diagrammatically in Fig. 14. 
 Fig. 7 expresses the relations of the sur- o 
 face of the lithosphere to sea-level both 
 in extent and in relief. 
 
 Plains 
 
 The plains are the lowlands of the 
 earth, yet they can hardly be denned in 
 terms of altitude above sea-level, the 
 datum to which all elevations are com- 
 monly referred. They may be but a few 
 feet above sea-level, or they may be 
 thousands of feet above it. In the lat- 
 ter case, however, they are generally far 
 from the sea, and distinctly lower than 
 other lands on at least one side. Fig. 15 
 is intended to give some idea of the re- 
 lations of large plains. It will be seen that plains may be as high 
 
 
RELIEF FEATURES 17 
 
 above sea-level as low plateaus are, or even as low mountains, 
 though this is not usually the case. They are never as high as 
 plateaus or mountains in their own vicinity. 
 
 Plains differ widely among themselves, not only in height, but 
 in position, in size, in topography, in fertility, in origin, and in 
 various other ways. Various names are applied to various types 
 of plains, the names being intended to direct attention to one or 
 another distinctive feature. Considered as topographic features 
 of the second order, the most important classes of plains are 
 Coastal Plains, which border the sea, and Interior Plains, which are 
 far from the sea, or separated from it by high lands. 
 
 Coastal Plains. These plains occur on the borders of many 
 continents, as along the eastern coast of the United States south of 
 New York. They may be narrow or wide. A narrow plain is shown 
 in Fig. 16, which represents a diagrammatic plain, not an actual 
 
 FIG. 16. A narrow coastal plain. 
 
 one. It is low, and has a nearly plane surface which slopes gently 
 toward the sea. Its surface is made uneven by the shallow 
 valleys of the streams which flow across it. The inner edges of 
 coastal plains are not always so clearly defined as in this 
 illustration. 
 
 A narrow coastal plain may have originated in either of two 
 ways: (1) It may be a part of the former continental shelf from 
 which the sea has withdrawn, or (2) the sediment washed down 
 from the land may have been deposited in the shallow water of 
 an epicontinental sea, building up (aggrading) its bottom above 
 the surface of the water, and thus converting it into land. Coastal 
 plains have been made in both these ways, and both processes 
 have often been concerned in the making of a given plain. Coastal 
 
IS 
 
 PHYSIOGRAPHY 
 
RELIEF FEATURES 19 
 
 plains may also be made by the degradation of coastal lands 
 which were once high. 
 
 Plate I represents, in another way, a part of the narrow coastat 
 plain of Oregon, and Fig. 17 shows the Cdastal Plain of the Atlantic 
 and Gulf coasts of the United States. Since illustrations of the 
 sort shown in Plate I will be used frequently in the following 
 cages, the principles on which it is based must be clearly understood. 
 
 EXPLANATION OF CONTOUR MAP 
 
 "The features represented on the topographic map are of three distinct 
 kinds: (1) inequalities of surface, called relief, as plains, plateaus, valleys, 
 hills, and mountains; (2) distribution of water, called drainage, as streams, 
 lakes, and swamps; (3) the works of man, called culture, as roads, railroads, 
 boundaries, villages, and cities. 
 
 "Relief. All elevations are measured from mean sea-level. The heights 
 of many points are accurately determined, and those which are most impor- 
 tant are given on the map in figures. It is desirable, however, to give the 
 elevation of all parts of the area mapped, to delineate the horizontal outline, 
 or contour, of all slopes, and to indicate their grade or degree of steepness. 
 This is done by lines connecting points of equal elevation above mean sea- 
 level, the lines being drawn at regular vertical intervals. These lines are 
 called contours, and the uniform vertical space between each two contours 
 is called the contour interval. On the maps of the United States Geological 
 Survey the contours and elevations are printed in brown (see Plate I). 
 
 "The manner in which contours express elevation, form, and grade 
 is shown in the following sketch and corresponding contour map, Fig. 18. 
 
 "The sketch represents a river valley between two hills. In the fore- 
 ground is the sea, with a bay which is partly closed by a hooked sand-bar. 
 On each side of the valley is a terrace. From the terrace on the right a 
 hill rises gradually, while from that on the left the ground ascends steeply 
 in a precipice. Contrasted with this precipice is the gentle descent of the 
 slope at the left. In the map each of these features is indicated, directly 
 beneath its position in the sketch, by contours. The following explanation 
 may make clearer the manner in which contours delineate elevation, form, 
 and grade: 
 
 "1. A contour indicates approximately a certain height above sea-level. 
 In this illustration the contour interval is 50 feet; therefore the contours are 
 drawn at 50, 100, 150, 200 feet, and so on, above sea-level. Along the contour 
 at 250 feet lie all points of the surface 250 feet above saa; and similarly 
 with any other contour. In the space between any two contours are found 
 all elevations above the lower and below the higher contour. Thus the 
 contour at 150 feet falls just below the edge of the terrace, while that at 
 200 feet lies above the terrace; therefore all points on the terrace are shown 
 to be more than 150 but less than 200 feet above sea. The summit of the 
 higher hill is stated to be 670 feet above sea; accordingly the contour at 
 650 feet surrounds it. In this illustration nearly all the contours are num- 
 
20 
 
 PHYSIOGRAPHY 
 
 bered. Where this is not possible, certain contours say every fifth one 
 are accentuated and numbered; the heights of others may then be ascer- 
 tained by counting up or down from a numbered contour. 
 
 "2. Contours define the forms of slopes. Since contours are continuous 
 horizontal lines conforming to the surface of the ground, they wind smoothly 
 about smooth surfaces, recede into all reentrant angles of ravines, and project 
 in passing about prominences. The relations of contour curves and angles 
 to forms of the landscape can be traced in the map and sketch. 
 
 FIG. 18. Sketch and map of the same area to illustrate the representation 
 of topography by means of contour lines. (U. S. Geol. Surv.) 
 
 "3. Contours show the approximate grade of any slope. The vertical 
 space between two contours is the same, whether they lie along a cliff or on 
 a gentle slope; but to rise a given height on a gentle slope one must go farther 
 along the surface than on a steep slope, and therefore contours are far apar f 
 on gentle slopes and near together on steep ones. 
 
 "For a flat or gently undulating country a small contour interval is used; 
 for a steep or mountainous country a large interval is necessary. The smallest 
 interval used on the atlas sheets of the Geological Survey is 5 feet. This 
 is used for regions like the Mississippi delta and the Dismal Swamp. In 
 mapping great mountain masses, like those in Colorado, the interval may 
 
PLATE I 
 
 ToweHRock" " "^ '" 
 
 A narrow coastal plain in Oregon. Scale 2 miles per inch. (Port Orford 
 Sheet, U. S. Geol. Surv.) 
 
RELIEF FEATURES 21 
 
 be 250 feet. For intermediate relief contour intervals of 10, 20, 25, 50, and 
 100 feet are used. 
 
 "Drainage. Watercourses are indicated by blue lines. If the streams 
 flow the year round the line is drawn unbroken, but it the channel is dry 
 a part of the year the line is broken or dotted. Where a stream sinks and 
 reappears at the surface, the supposed underground course is shown by a 
 broken blue line. Lakes, marshes, and other bodies of water are also shown 
 in blue, by appropriate conventional signs. 
 
 "Culture. The works of man, such as roads, railroads, and towns, together 
 with boundaries of townships, counties, and states, and artificial details, 
 are printed in black." ' 
 
 CONTOUR-MAP EXERCISE 2 
 
 1. Draw a contour-line map of a conical mountain the top of which 
 is 2000 feet high, making the contour interval 200 feet. 
 
 2. Draw a contour-line map of a plain five miles square, one edge 
 of which is at sea-level and the opposite one at an elevation of 100 feet. 
 The otherwise uniform seaward slope of the land is scarred by a single 
 valley, without tributaries, which extends across the entire width of 
 the plain. Use a 10-foot contour interval, and a horizontal scale of one 
 inch to the mile. 
 
 The Coastal Plain of the eastern part of the United States has a 
 width ranging from a few to 60 miles in New Jersey, to 100 miles 
 or more in the Carolinas and Georgia (Fig. 17), and would be 
 counted a wide coastal plain. The Coastal Plain bordering the 
 Gulf of Mexico is still wider, reaching a maximum width of several 
 hundred miles in the vicinity of the Mississippi. The Coastal Plain 
 of northern Eurasia is still wider, though locally interrupted by 
 mountains, such as the Urals. 
 
 At their seaward edges coastal plains are commonly but little 
 above the sea. Their inland borders, on the other hand, especially 
 if they be wide, may be hundreds of feet above the sea. At its 
 landward edge a coastal plain may abut against a plateau or 
 against mountains by a slope somewhat steeper than that of the 
 plain itself (Fig. 6). It is this steep slope, rather than any par- 
 ticular altitude above the sea, which limits a coastal plain to 
 landward. The landward border of the Atlantic Coastal Plain 
 has an altitude ranging from 100 feet or so to several hundred 
 feet. The relatively steep slope marking its landward edge is 
 
 1 From folio preface, U. S. Geol. Surv. 
 
 2 The author's experience has been that students come to an appreciation 
 of topographic maps most readily by making them. 
 
22 PHYSIOGRAPHY 
 
 known as the Fall Line, and along it are located many impor- 
 tant cities, among them Trenton, Philadelphia, Baltimore, Wash- 
 ington, Richmond, Raleigh, Camden, Columbia, and Augusta. 
 The location of these cities was determined largely by the fact that 
 the streams were readily navigable in the Coastal Plain, but not 
 above. The position of the Fall Line was determined by the 
 inequalities of hardness of the underlying rocks. Those to the 
 west of it are much harder than those to the east. The landward 
 margin of the Coastal Plain of the Gulf of Mexico west of Alabama 
 is less well marked than that of the Atlantic border. 
 
 Such coastal plains as exist along the eastern side of North 
 America north of New Jersey, and along the western coast of the 
 continent, are narrow and discontinuous, and for considerable 
 stretches are wanting altogether. Coastal plains are therefore to 
 be looked upon as common, but not as universal, features of con- 
 tinental borders. 
 
 If the epicontinental seas were withdrawn from the submerged 
 parts of the continental platforms, the coastal plains of the present 
 land would be seen to be continuous, topographically, with the 
 continental shelves. These submerged parts of the great continental 
 protuberances of the lithosphere are therefore to be looked upon 
 as submerged coastal plains. Many of the existing coastal plains 
 of the land have emerged from the sea in very late stages of the 
 earth's history. The submerged coastal plains are more nearly 
 continuous about the continents than the coastal plains which 
 are above sea-level. 
 
 Interior plains. These plains are often higher, and sometimes 
 much higher, than coastal plains. A large part of the great area 
 between the Appalachian Mountains on the east, and the Rocky 
 Mountains on the west, is an interior plain. At the south it is 
 relatively low, and merges into the Coastal Plain bordering the 
 Gulf. At the north this Interior Plain is much higher, attaining 
 an elevation of more than 1000 feet; but its rise is so gradual 
 that it nowhere ceases to have the general effect of a great lowland. 
 At the east, also, it rises until the Appalachian Mountains are 
 approached. Along the western border of these mountains there 
 is a higher area, about 1000 feet above sea-level, often known as 
 the Cumberland, or Allegheny, Plateau (Fig. 17). This tract is 
 called a plateau, rather than a part of the plain, not more because 
 of its altitude than because it is often somewhat distinctly set 
 
RELIEF FEATURES 23 
 
 off from the lower area to the west. To the west the interior plain 
 rises gradually, and without any conspicuous increase of slope, 
 until it attains an altitude of several thousand feet. In spite of 
 this very considerable elevation, far greater than that of the 
 Cumberland Plateau, the area east of the Rocky Mountains is 
 usually called the Great Plains. The western part of this region 
 is perhaps more properly a plateau than a plain; but it is notably 
 lower than the mountains against which it abuts on the west, 
 and between its higher parts, next to the Rockies, and its lower 
 parts, adjacent to the Mississippi, there is no abrupt change of slope. 
 It is clearly a topographic unit. If the western part of this area 
 be classed as plateau, the area affords a good illustration of the 
 gradation of a plain into a plateau, for the line separating the 
 plain-part from the plateau-part would be an arbitrary one. 
 Even if the higher western part of the Great Plains be regarded 
 as plateau, it is still true that portions of the interior plain are 
 higher than many areas which are called plateaus. Areas like 
 the Great Plains are classed as plains, not primarily because of 
 their altitude above sea-level, but because they do not stand 
 conspicuously above their surroundings on any side. 
 
 The general topographic relations of the Great Plains are 
 illustrated diagrammatically by Fig. 15. If the general slope of 
 the area between the Rocky Mountains and the Mississippi River 
 had been that of the dotted line shown in the middle of this 
 figure, -Its western part would doubtless have been classed as a 
 plateau, and the line of separation between plateau and plain 
 would have been a natural one. 
 
 Here and there mountains, snch as the Black Hills of South 
 Dakota, the Ozark Mountains (plateau) of Missouri, and the 
 Ouachita (pronounced Wash'-i-ta) Mountains of Arkansas, Indian 
 Territory, and Oklahoma, rise distinctly above the general level of 
 this great Interior Plain. The Ozark and Ouachita Mountains 
 do not attain an elevation equal to that of the western margin of 
 the Great Plains, but they are so distinctively and conspicuously 
 above their immediate surroundings that they are not regarded 
 as parts of the plains, and the summit area of the Ouachita Moun- 
 tains , at least, is so limited that they cannot be regarded as a 
 plateau. The Ozark Mountain tract, on the other hand, might 
 equally well have been called the Ozark Plateau, for the character 
 of the region which bears this name is intermediate between that 
 
24 
 
 PHYSIOGRAPHY 
 
 of a well-defined mountain group and a plateau. The Black Hills 
 are higher, and more distinctly set off from the plains, than are 
 the Ozark and the Ouachita Mountains. They are mountains, in 
 spite of their name. 
 
 Interior plains have come into existence in various ways. 
 Some of them are former coastal plains, now partially shut off 
 from the sea by the development of highlands between. Some 
 of them represent areas which were once high, but which have 
 been worn down by rivers, and by the other agents which degrade 
 lands; others may have originated in other ways. 
 
 Fio. 19. A plain with little relief. Valley plain of the Cimarron River, 
 southwestern Kansas. (U*. S. Geol. Surv.) 
 
 Topography of plains. The surfaces of plains are, on the 
 whole, much less uneven than the surfaces of plateaus and moun- 
 tains. The surfaces of plains may indeed be nearly flat, though 
 more commonly they are somewhat uneven. The relief is some- 
 times slight and sometimes considerable, and in general high 
 plains are rougher than low ones. To this general statement there 
 are, however, local exceptions, for considerable areas of high plains 
 are sometimes nearly flat. 
 
PLATE II 
 
 A well-drained plain in Kansas. Scale 2 miles per inch. 
 (Anthony, Kan., Sheet, U. S. Geol. Surv.) 
 
PLATE III 
 
 An ill-drained plain in Wisconsin. Scale 1 mile per inch (Silver 
 Lake Sheet. U. S. Gecl. Surv.) 
 
RELIEF FEATURES 
 
 25 
 
 The unevennesses of surface differ in kind, as well as in amount. 
 Thus in some plains, or in some parts of plains, all depressions 
 have outlets through which the surface water flows away, while in 
 others numerous depressions have the form of basins which contain 
 ponds and lakes. Plains of the former type are well drained, if 
 the depressions are numerous, while those of the latter are ill 
 drained. Well-drained areas (PI. II) of plain prevail in the south- 
 ern part of the United States, as south of the Ohio and the 
 Missouri, while ill-drained areas (PI. Ill) abound farther north. 
 
 The topographic features of plains are relief features of the 
 third order, and will be considered later; but the points here 
 mentioned have a bearing on the topic of the next paragraph. 
 
 FIG. 20. A plain with notable relief. Iowa. (Calvin.) 
 
 Extent and habitability. Plains constitute the larger part of 
 the area of the land, and the larger part of the population of the 
 earth lives upon them (Figs. 21 and 22). They are the principal 
 theatres of human activity, partly because the climate is on the 
 whole more favorable than in higher regions, and partly because 
 there is a greater proportion of land which is nearly flat, or which 
 has but gentle slopes. As compared with higher lands, a larger 
 proportion of the surface of plains is arable, for (1) their flats and 
 gentle slopes are more generally covered with soil than the steeper 
 slopes of higher lands are, and (2) a larger proportion of their 
 surfaces is not too steep for cultivation. The larger part of the 
 agriculture of the world is therefore on plains. 
 
 When the population of the United States was about 50,000,- 
 000 l (1880), it was distributed as follows, with reference to altitude: 
 
 1 Later data on this point are not available. 
 
26 
 
 !S 
 
 "Sai 
 
RELIEF FEATURES 
 
 27 
 
28 PHYSIOGRAPHY 
 
 Less than 100 feet above sea-level 15.9% 
 
 Between 100 and 500 feet above sea-level 21 .8% 
 
 Between 500 and 1000 feet 38. 7% 
 
 Between 1000 and 1500 feet 14. 7% 
 
 Between 1500 and 3000 feet 6.2% 
 
 Above 3000 feet 2.7% 
 
 Plains also favor transportation and intercommunication, for 
 (1) the construction of roads, railways, canals, etc., is vastly easier 
 in plains than in higher and rougher regions, and (2) che streams 
 of plains are much more commonly navigable than those of moun- 
 tains and plateaus. For these reasons, and also because the larger 
 part of the raw materials used in manufacturing is grown upon 
 the plains, the larger part of the manufacturing of the world is on 
 plains. It is noteworthy that the extensive plains most favored 
 by climate and soil border the Atlantic Ocean, and, largely for this 
 reason, the borders of this ocean have been the theatres of the 
 world's culture and commerce. 
 
 Not all plains support an abundant population. Thus the north- 
 ern parts of the great Eurasian and North American plains are too 
 cold to be hospitable to varied industries or productions, and their 
 populations are likely to remain small. 
 
 Plateaus 
 
 Plateaus are tracts of land so situated as to appear high from 
 at least one side, and which have, at the same time, considerable 
 areas at or near their summit levels. Thus if a coastal plain rises 
 gradually from the sea to a height of 200 feet, and then joins by a 
 steep slope another tract of more or less level land which rises 100 
 or 200 feet higher (Fig. 6), the upper tract would commonly be 
 called a plateau, not primarily because of its altitude above sea- 
 level, but because of its distinct rise above the plain along one side 
 of it. Traced landward, the low slope of the Atlantic Coastal Plain 
 of the United States gives place to a steeper one at the Fall Line, 
 and the tract above, beyond the Coastal Plain, is the Piedmont 
 Plateau. The elevation of much of this plateau is, however, less 
 than that of much of the great interior plain of the continent. 
 
 Though plateaus are on the whole higher than plains, it may 
 be pointed out again that the distinction between them is not 
 
RELIEF FEATURES 
 
 29 
 
 more one of elevation than of relations. A tract of land is rarely 
 called a plateau unless it rises distinctly above adjacent land or 
 adjacent water on one or more sides. 
 
 In spite of the broad distinction between plateaus and plains, 
 
 these two great topographic types grade into each other so com- 
 pletely that it is often hard to say whether a given region should 
 be classed as the one or the other, and a tract which, in its sur- 
 roundings, is a plateau, might in other surroundings be a plain. 
 
30 PHYSIOGRAPHY 
 
 Distinctions in nature are often less sharp than we seem to make 
 them by arbitrary (though often necessary) definitions. 
 
 Position and area of plateaus. Plateaus often lie between 
 mountains on the one hand and plains on the other, as in the case 
 of the Piedmont and Cumberland plateaus already cited. They 
 also fre between mountains, as the plateaus of Central Asia (Fig. 24), 
 Mexico, and the western part of the United States, and they some- 
 times rise directly from the sea, as in the case of Greenland and 
 parts of Africa (Fig. 25). 
 
 JMStA 
 
 E.. 
 
 FIG. 24. Section across Asia along the 35th parallel. Vertical scale greatly 
 exaggerated. The plateau between mountains is marked Pit. (Alter 
 Heidrich.) 
 
 The aggregate area of plateaus is less than that of plains, 
 though they constitute a very considerable fraction of the land. 
 
 Relief of plateaus. The surfaces of plateaus usually have 
 greater relief than the surfaces of plains, because the valleys are 
 deeper. The plateau of the Colorado in northern Arizona has an 
 elevation of about 7000 feet, and a relief of a mile or more, for the 
 Colorado River has a valley (canyon) of that depth (Figs. 27 and 27a). 
 From the bottom of this valley, its slopes look like mountains. 
 They are indeed much higher and bolder than many mountains; 
 
 FIG. 25. Section across Africa along the parallel of 10 S. Vertical scale 
 exaggerated about fifty times. 
 
 but since there are great stretches of land about the canyon at 
 about the elevation of the tops of these slopes, the area is a plateau 
 region, rather than a mountain region. No plain has such great 
 relief as this plateau. 
 
 Other features of plateaus. Except for the greater average 
 relief of plateaus, their surfaces have much in common with the 
 surfaces of plains. There are flat plateaus, broken plateaus, roll- 
 ing plateaus, etc., and these topographic terms are often applicable 
 to different parts of the same plateau. There are plateaus which 
 
PLATE IV 
 
 FIG 1. The canyon of the Yellowstone River. Scale 2 miles per inch. 
 (Canyon, Wyo., Sheet, U. S. Geol. Surv.) 
 
 FIG. 2. The Grand Canyon of the Colorado River. Scale 4 miles per inch. 
 (TJ. S. Geol. Surv.) 
 
RELIEF FEATURES 
 
 31 
 
 are well drained, and plateaus which are ill drained; there are 
 plateaus which are relatively fertile, and plateaus which are essen- 
 tially desert. The relief features of plateaus, like those of plains, 
 are relief features of the third order. 
 
 The climate of plateaus, especially that of high ones, is distinctly 
 colder than that of plains in similar latitudes, and their precipita- 
 tion is generally less. Except in low latitudes, they are too cold to 
 be well adapted to human habitation, and their rainfall is often 
 insufficient for agriculture. Their deep valleys are barriers to trans- 
 
 FIG. 26. A valley (canyon) in a plateau. Snake River below the mouth of 
 Rattlesnake Creek. (U. S. Geol. Surv.) 
 
 portation. For these and other reasons, high plateaus are, on 
 the whole, less well adapted to human habitation than plains, and 
 the population of high plateaus is generally scanty. On the other 
 hand, the altitude of low plateaus, such as the Piedmont and the 
 Cumberland plateaus, is too slight to affect the climate adversely, 
 and such plateaus may be as fertile as plains, so far as climate is 
 concerned. Plateaus in low latitudes may have a favorable tem- 
 perature, and may be so situated as to have an adequate supply of 
 water, as illustrated by some parts of the plateau of Mexico. 
 
 Origin. Plateaus attain their height in various ways. (1) In 
 some cases their surroundings probably sank away from them. 
 If, for example, the eastern half of the Great Plains was to sink 
 a few hundred feet, while the western half did not, the latter would 
 doubtless be called a plateau (Fig. 15). (2) Some plateaus may 
 have attained their height by elevation above their surroundings, 
 
32 
 
 PHYSIOGRAPHY 
 
RELIEF FEATURES 
 
 33 
 
 while still others (3) have been built up either from plains or from 
 lower plateaus, by the outpouring of lavas. Such is the lava plateau 
 of the northwestern part of the United States (Fig. 401). 
 
 The term plateau is often applied, and properly, to small areas 
 which may owe their plateau character to other causes, such as 
 
 Fro. 27a. The Grand Canyon of the Colorado. The inner gorge in the 
 foreground, and the more distant cliffs in the background. The canyon 
 is about a mile deep. (Hull.) 
 
 isolation by the degradation of the surrounding surface. Such 
 plateaus are topographic features of a lower order, and are not 
 here considered. 
 
 Mountains 
 
 Mountains are conspicuously high lands which have but slight 
 summit areas'. Conspicuously high lands must be interpreted to 
 mean lands which are conspicuously high in their surroundings, 
 not necessarily those which have great elevation, measured in 
 feet or meters. 
 
34 
 
 PHYSIOGRAPHY 
 
 Though the tops of the highest mountains are between five and 
 six miles above the level of the sea, most mountains have not half 
 
 FIG. 28. Sierra el Late Mountains, Colo., with dissected mesa in the fore- 
 ground. (Holmes, U. S. Geol. Surv.) 
 
 this height. The highest mountains are higher than any plateaus, 
 but many mountains are not so high as the highest plateaus. Rela- 
 tively few, for example, reach the height of the Plateau of Tibet, 
 15,000 to 16,000 feet. Many elevations called mountains are not 
 even so high above sea-level as the higher parts of the higher plains. 
 Mountains differ from plateaus of similar elevation in that 
 
 FIG. 29. The Needle Mountains of Colorado. An illustration of mountain 
 topography. Taken from an elevation of about 10,700 feet. (U. S. 
 Geol. Surv.) 
 
 they have little extent of surface at the summit level. In the 
 case of mountain peaks this is indicated by the name. A moun- 
 
RELIEF FEATURES 
 
 35 
 
 tain ridge or range may be long, but, as its name implies, its crest 
 is usually narrow. The several ridges shown in Fig. 23 are ex- 
 
 FIG. 30. Lake Agnes, Canadian Pacific Railway. (Photograph by Church.) 
 
 amples. Numerous peaks or ranges are often associated, making 
 a mountain group (Fig. 28) or "a mountain chain (Fig. 23); but 
 
 FIG. 31. Cascade Pass in the Cascade Mountains. Washington. An illustra- 
 tion of mountain topography. (Willis, U. S. Geol. Surv.) 
 
 even in great mountain groups there is no great continuous area 
 of high land. Land 10,000 feet high would generally be called a 
 
36 
 
 PHYSIOGRAPHY 
 
 FIG. 32. A portion of the Elk Mountains of Colorado. 
 (Holmes, U. S. Geol. Surv.) 
 
 FIG. 33. Photograph of relief-model of Texas and surroundings. The 
 area near the coast is a part of the Coastal Plain. Inland this plain 
 gives place to a plateau tract, while at the north and west mountains 
 rise above the plateau level. The valleys are deep, and the relief is 
 greater in tl:e mountains than in the plateau, and in the plateau greater 
 than in the plain. (Hill.) 
 
RELIEF FEATURES 
 
 37 
 
 plateau if its summit area were extensive, a mountain if its summit 
 were a peak, a mountain ridge or range if its crest were a narrow 
 ridge, or a mountain area, a mountain group, or a mountain system, 
 if composed of a succession of peaks and ridges. 
 
 CALIFORNIA 
 
 / -*i.v- 
 
 FIG. 34. Topographic map of California. The State is largely mountainous, 
 but the central plain is conspicuous. (Model by Drake.) 
 
 Considered in a large way, mountains are in contrast with plains 
 and plateaus, and are the third of the three topographic types of 
 the second order, as they appear on the lands of the earth. 
 
 High mountains are on the whole the most impressive and 
 awe-inspiring features of the earth's surface. This is especially 
 
38 PHYSIOGRAPHY 
 
 the case where they rise abruptly to great heights above their 
 surroundings. In not a few cases they rise from low warm plains 
 to such heights that their summits are continually covered with 
 snow. Nowhere else are such contrasts of climate found in such 
 close proximity. 
 
 In this grouping of mountains, as the third great topographic 
 features of the lands, it must be noted that only great groups or 
 systems of mountains, such as the Appalachians, the Rockies, the 
 Sierras, the Alps, the Caucasus, the Himalayas, the Andes, and others 
 of comparable extent and magnitude are included. Since the term 
 mountain is applied to any point or ridge of such steep slopes and 
 so much above its surroundings as to be very conspicuous, if, at 
 the same time, its summit area is so small that it is not a plateau, it 
 follows that many elevations called mountains do not belong to 
 the great physiographic type which is to be brought into contrast 
 with plains and plateaus. From this category we must exclude 
 many minor and isolated elevations called mountains, especially 
 those of such small size that, in surroundings other than their* 
 own, they would not be regarded as mountains. 
 
 Mountains in history. Mountains are always more or less 
 formidable barriers, and as such have played important roles in 
 history. They have sheltered nascent civilizations from invasion, 
 and they often determine the boundaries of political states. The 
 mountains of western and southwestern Europe were an important 
 factor in producing the many small political divisions of those 
 sections, so in contrast with Russia. Mountainous highlands have 
 frequently become a refuge for weak peoples, driven by their 
 stronger enemies from the more desirable lowlands. The rela- 
 tively inaccessible highlands of Scotland, Wales, and parts of 
 India enabled such peoples to maintain their independence for 
 long periods. The Appalachian Mountains confined the English 
 settlements to the rim of the continent for nearly a century and 
 a half, and influenced their life in many ways. Later, grave 
 political dangers arose from the effectual isolation of the Ohio 
 Valley settlements from the Atlantic seaboard. 
 
 The scant soils and low temperatures of most mountains in- 
 hibit agriculture, while the difficulties of communication help to 
 restrict commerce and social intercourse. Poverty is, accordingly, 
 the common lot of the mountaineer, save in certain mining and 
 lumbering areas. Shut out from the progressive life of the plains, 
 
RELIEF FEATURES 
 
 39 
 
 mountain peoples are proverbially conservative, maintaining old 
 customs and habits, and supporting the established order of things. 
 In the Civil War the Southern Appalachians became a zone of 
 disaffection through the heart of the Confederacy, sending 100,000 
 men to the northern armies. 
 
 The most distinctive industry of the mountains is mining; 
 yet many mountains have no ores or mineral matter of commer- 
 cial value, while many ores and many mineral substances which 
 are not ores are mined in plains and plateaus. This is true, fcr 
 example, of most of the iron and coal now mined in the United 
 States. 
 
 FIG. 35. Cross-section illustrating the structure of the Appalachian Moun- 
 tains. (After Rogers.) 
 
 TI*V TRIASA^WCRJURA UP^tR JURA CRCTACCO'US' ' " TERT1AV. 
 
 FIG. 36. Section of the Alps from Saint Gothard South. (After Heim.) 
 
 FIG. 37. Cross-section of the Elk Mountain Range, Colo. 
 (Holmes, U. S. Geol. Surv.) 
 
 Pine Forest Mts. 
 
 Volcanic Mesa p uebb 
 
 Stein Mts. 
 
 Barren Val]^ 
 
 FIG. 38. Faulted Mountain (Block Mountain) structure, Nevada. 
 (Russell, U. S. Geol. Surv.) 
 
 Origin. Mountains have originated in various ways. In their 
 formation, the layers of rock of which they are composed were 
 
40 PHYSIOGRAPHY 
 
 often folded and crumpled, sometimes on a grand scale. Figs. 
 35-37 illustrate types of mountain structure common to the great 
 
 FIG. 39. Sketch of the Abert Lake, Ore. (Russell, U. S. Geol. Surv.) 
 
 ranges which belong to the topographic features of the second 
 order. 
 
 TOPOGRAPHIC MAPS 1 SHOWING GREAT PHYSIOGRAPHIC 
 
 TYPES 
 
 Note. The conventions used on the topographic maps are explained 
 on their backs. The meaning of each should be noted. 
 
 A Plain Region. Maumee Bay, Ohio Sheet. This map shows a 
 nearly level plain whose surface slopes very gently to the northeast. 
 The general flatness may be read at a glance from the fewness of the 
 contour lines (only four appear upon the entire map), and also from 
 the fact that the railroads run long distances in straight lines. Calculate 
 the average slope of the surface per mile. *.j r > ^ 
 
 A Plateau Region. Echo Cliffs, Ariz. Sheet. The numbers upon the 
 contour lines show this to be an elevated region, and the disposition of 
 the contours shows that there are considerable areas of the high land, 
 and that the region is therefore a plateau. The very deep valley of the 
 Colorado River also indicates great height of land, for such valleys 
 are found only in regions far above sea-level. Note that the Paria 
 Plateau, at the northwest, is bordered by an abrupt descent. This is 
 often true of plateaus upon at least one side. 
 
 1 These maps are topographic maps of the U. S. Geological Survey. 
 They may be had of the Director of that Survey, Washington, D. C., at 
 $3.00 per 100. Many of these maps will be referred to in the following pages. 
 See list at end of Part I. 
 
RELIEF FEATURES 4l 
 
 A Mountain Region. Hummelstown, Pa. Sheet. The massing of the 
 contours along northeast-southwest lines in the northern part of the 
 area shows a series of relatively steep slopes extending in that direction. 
 The crests between the steep slopes are narrow. The numbers on the 
 contour lines show that the elevations are of mountainous heights. 
 
 SUBORDINATE TOPOGRAPHIC FEATURES 
 
 It has already been noted that the surfaces of plains and pla- 
 teaus are often somewhat uneven, while the very name of moun- 
 tain suggests roughness of surface. In many cases the degree of 
 unevenness of surface is more or less closely related to altitude 
 above sea-level, increasing roughness going with increasing alti- 
 tude, though altitude is by no means the only factor which deter- 
 mines roughness and smoothness of surface. The minor uneven- 
 nesses of surface which affect plains, plateaus, and mountains are 
 topographic features of the third order. Some of these irregularities of 
 surface consist of elevations above the general level of their sur- 
 roundings, and some of depressions below it. Thus on the plains 
 there are ridges and hills above the general level, and valleys and 
 sometimes basins (depressions without outlets) below it, while 
 -fiats may be interspersed among the uneven tracts. The eleva- 
 tions and the depressions are bordered by slopes, which, when 
 steep, are cliffs. These subordinate features, ridges, hills, valleys, 
 basins, flats, etc., affect plateaus as well as plains; but the corre- 
 sponding features of plateaus are often more pronounced, some- 
 times so much more pronounced that they receive different names. 
 Many of the same features, often on a still larger scale, affect moun- 
 tains; but here the more or less isolated elevations, instead of 
 being merely ridges or hills, are often of mountainous size, and 
 receive individual names. And so, as terms are now used, it is 
 difficult to distinguish, in words, between mountains which are 
 topographic features of the second order and mountains which 
 are topographic features of the lower order, though in reality the 
 distinction is clear enough. Thus the Appalachian Mountains 
 are a topographic feature of the second order, but any minor ridge 
 or peak in the system, though still a mountain, is a feature of the 
 third order, and is to be compared with the hills and buttes of 
 plains and plateaus. 
 
 The depressions in the surface of plains or plateaus are of 
 different sizes, shapes, and origins, and will be the object of future 
 
42 PHYSIOGRAPHY 
 
 study. Similarly the hills and ridges of plains and plateaus, and 
 the larger, mountain-big hills of mountainous regions, are of dif- 
 ferent sizes, shapes, and origins, and their history is often intimately 
 connected with the history of the depressions with which they 
 are associated. Slopes w ? ere developed when elevations and de- 
 pressions were developed, and largely by the same means. The 
 origin of the topographic features of the third order is in general 
 well understood, for the processes which have developed them 
 are still in operation, and their results in past times may be in- 
 ferred with much confidence. These processes we shall study in 
 some detail. 
 
 Land surface and ocean bottom. Were the water removed 
 from the ocean basins, the surface of the ocean bed would appear 
 much less uneven than the surface of the land. While its aggre- 
 gate relief is a little greater than that of the land (p. 9), much 
 larger tracts of it are nearly plane, and minor irregularities, such as 
 hills and valleys with their accompanying slopes, the most wide- 
 spread of the minor irregularities of the land, are of much less 
 common occurrence; are, indeed, entirely absent from the larger 
 part of the ocean's floor. 
 
 Why this difference between land and sea bottom? Without 
 discussing this subject at this point, it may be noted that the 
 atmosphere and running water, both of which are in almost con- 
 stant motion, are always in contact with the surface of the land, 
 while the atmosphere is excluded from the ocean bottom, and the 
 water which covers it is practically motionless, except where the 
 water is very shallow. It will be seen in the sequel that the 
 differences in topography between the land and the sea bottom 
 are largely due to the contact with air and running water in the 
 one case, and with standing water in the other. 
 
 The Development of Minor Topographic Features 
 
 Since the minor topographic features of plains, plateaus, and 
 mountains have been developed in similar ways, their origin and 
 history may be considered independently of their association with 
 one or another of these great physiographic divisions. The key 
 to the history of topographic features of the third order is found 
 in the changes which the surface of the land is now undergoing, 
 or which it has undergone in such recent times that their records 
 are still clear. 
 
RELIEF FEATURES 43 
 
 Changes now taking place on the land. Certain familiar 
 changes are always taking place on the land. Some of them are 
 brought about by the atmosphere, some by water, some by ice, 
 and some by the life of the earth. The same agencies produce, 
 directly or indirectly, certain changes on the sea bottom, but the 
 changes there are not only less important than those on land, 
 but they are essentially different in their effects on the topography. 
 
 1. The air is nearly always in motion, and whenever it blows 
 over a surface on which there is dust, some of the dust is picked up 
 and blown to some other place. Even sand, the particles of which 
 are much larger than those of dust, is blown about in the same 
 way. The wind is, therefore, one of the forces which is changing 
 the surface of the land. The winds also help to distribute the 
 moisture of the atmosphere, and so influence the amount and the 
 distribution of rain and snow. Though winds do not blow at the bot- 
 tom of the sea, dust and sand blown out from the land are dropped 
 into the ocean and sink to its bottom, and the winds generate 
 water waves which work upon the shores of the seas, and affect 
 their bottoms where the water is very shallow. The winds there- 
 fore are not without their effects on the sea bottom, though these 
 effects are slight compared with those on the land. 
 
 2. On both the land and sea, rains and snows fall. The rain 
 which falls on the land disappears in various ways, but a part 
 of it runs off over the surface. When the snow of the land melts, 
 the water follows the same course. The water which runs off over 
 the land in streams is the most important single agent modifying 
 the land surface. The streams carry much sediment from the land 
 to the sea, and its deposition has its effect on the sea bottom, 
 especially near the land. 
 
 The rain- and snow-water which sinks beneath the surface 
 of the land dissolves more or less mineral matter, which appears 
 in spring water and in well water. This solution of mineral matter 
 beneath the land, and its transfer by the water to the surface, and 
 thence through streams to the sea, also help to lower the land. 
 
 While the waters which fall on the land have an indirect effect 
 on the bottom of the sea, as indicated, those which fall on the 
 sea itself have little influence on its bed. Precipitation, therefore, 
 whether in the form of rain or snow, modifies the surface of the 
 land notably, but has little influence on the ocean bottom, except 
 near its borders, where most of the sediment from the land is 
 left. 
 
44 PHYSIOGRAPHY 
 
 3. Great bodies of ice, called glaciers, move slowly over the 
 surface of the land in some places, especially on high mountains 
 and in high latitudes. Glaciers, which originate in perennial fields 
 of snow, work notable changes on their beds. They sometimes 
 push out into the sea for short distances, but they never advance 
 into deep water. At most they only affect the submerged edge 
 of the continental platform. 
 
 The winds, the streams, and the glaciers all tend to develop un- 
 evennesses of surface on the land. Since this is the case, and since 
 these agents are not in operation beneath the sea, we infer that 
 they have had much to do with developing the differences between 
 the topography of the land and that of the sea bottom. 
 
 4. The waves of the sea and of the many lakes which lie on 
 the land are continually modifying the position and the outlines 
 of their shores. The changes thus effected are slight in short periods 
 of time, but they have been very great in the course of the long 
 ages of the earth's history. They change the outlines of the 
 land rather than its relief, but they alter the relief of the sea or 
 lake bottom near shore in an important way. 
 
 The winds, rivers, glaciers, and waves are agents of gradation. 
 They degrade the surface at some points, and aggrade it (build 
 it up) at others. In general they degrade the land more than 
 they aggrade it, for much of the material moved by them finds 
 its resting-place in the sea. Conversely, they aggrade the sea 
 bottom more than they degrade it. Waves may degrade it effec- 
 tively, but only where the water is shallow. 
 
 5. Still another series of changes in the surface is being 
 brought about through the agency of life. Man, for example, 
 grades down elevations, and he grades up depressions, as along 
 railroads. He makes dams across rivers, converting portions of 
 them into ponds, or at the outlets of lakes, raising their levels; 
 he raises and changes the banks of streams, modifying their natu- 
 ral courses and their natural work; he drains marshes and lakes, 
 and, more important than all else, he clears (removes the forests) 
 and tills the land, and in so doing destroys the native vegetation 
 and stirs up the soil, thus preparing the way for the more effective 
 action of wind and running water. Man's direct influence on the 
 sea bottom is slight. 
 
 Plants and animals affect both land and sea bottom. Deposits 
 due to organisms of one sort and another, especially those due 
 
RELIEF FEATURES 45 
 
 to plants, are somewhat wide-spread in the marshes and shallow 
 lakes of the land; but they are, on the whole, of little consequence 
 compared with the deposits of shells, skeletons, and other solid 
 matter made by marine animals on the sea bottom, more especially 
 in shallow water. Organic agents are in some sense gradational, 
 chiefly aggradational; but they belong to a different category 
 from the inorganic gradational agents. 
 
 Various forms of life have a protective effect on the surface. 
 This is especially true of vegetation on the land. The forests, 
 and even the prairie vegetation, greatly restrict the erosive work 
 of wind and running water, and so decrease the rate of degrada- 
 tion which would otherwise obtain. 
 
 6. Volcanoes affect both land and sea bottom, and with approx- 
 imate equality. Volcanoes often give rise to mountain heights, 
 but the mountains to which they give rise are topographic features 
 of the third rather than the second order. The great processes of 
 vulcanism, that is the movement of liquid rock from great depths 
 up to or toward the : surf ace, affect the surface of land, and doubt- 
 less of sea bottom, in other ways, which will be mentioned in other 
 connections. 
 
 7. Tt is well known that the surface of the lithosphere seems 
 to be rising in some places and sinking in others. This has been 
 true in the past, for beds of sediment (sand, clay, etc.), containing 
 sea-shells, etc., and therefore once beneath the sea, occur at levels 
 high above it, and areas once land are now beneath the sea. Crustal 
 movements are probably responsible in large measure for the ocean 
 basins and continental platforms, and for plains, plateaus, and 
 mountains, that is for the topographic features of both the first and 
 second orders. All sorts of crustal movements, of whatever nature, 
 are grouped together under the name diastrophism. 
 
 The processes of gradation, vulcanism, and diastrophism will 
 be taken up in order, but before entering upon the study of grada- 
 tion, the materials on which the agents of gradation act must be 
 briefly reviewed. 
 
 THE MATERIALS OF THE LAND 
 
 The land is nearly everywhere covered with vegetation. In 
 some places it is dense enough to form a thick mat over the sur- 
 face, while in others it is meagre, or even wanting. The surface 
 well clothed with vegetation is the surface with which we are most 
 
46 PHYSIOGRAPHY 
 
 familiar; but there are tracts of sand on which little or nothing 
 grows, and cliffs where the rock is bare, save for scattered patches 
 of moss or lichen. In the polar regions and on lofty mountains 
 also, the land is often covered by thick beds of snow on which there 
 is no vegetation of the types with which we are familiar. 
 
 Mantle rock. Beneath the vegetation there is, in most 
 regions, a layer of loose material, composed of clay, loam, sand, 
 gravel, etc., of variable thickness. This layer of earthy matter 
 may be a few inches in thickness, or it may be scores or even 
 hundreds of feet deep. This loose material is mantle rock, because 
 it covers and conceals the solid rock which lies below. It is also 
 known by other names, among which are rock waste and regolith. 
 
 The uppermost portion of the mantle rock is commonly called 
 soil. In color the soil may be black, gray, brown, or even dull 
 red or yellow. It may be either clayey and compact, or sandy 
 and porous. In most cases it is made up largely of small particles 
 of mineral or rock. If a piece of any common sort of rock be put 
 into a mortar and ground to powder, this powder will somewhat 
 resemble soil. In general we cannot recognize the kinds of rock 
 from which the mineral particles of the soil came, for they are 
 usually very small. In addition to the mineral matter, the soil 
 contains more or less partly decayed vegetable matter. Bits of 
 roots may often be seen in it, and sometimes fragments of de- 
 cayed leaves. Both the mineral and the organic matter are neces- 
 sary parts of a good soil, but their proportions vary within wide 
 limits. The mineral matter is usually far in excess of the organic, 
 but locally, as in bogs and marshes which have been drained, the 
 organic matter is the more abundant. That part of the mantle 
 rock which is properly called soil ranges from a few inches to a few 
 feet in thickness. 
 
 The distribution and prosperity of population often bear a 
 very direct relation to the fertility of the soil. The fertile "Blue 
 Grass" region of Kentucky was the first extensive area to be settled 
 in the Ohio basin; its inhabitants have always been progressive 
 and well-to-do. Some of the hilly land to the east was slowly 
 occupied by a sparse population, condemned by a poor soil to 
 financial and intellectual poverty. The cotton and tobacco lands 
 of the Coastal Plain were partly responsible for the institution of 
 slavery. 
 
 Where the mantle rock is thicker than the soil, the soil grades 
 
RELIEF FEATURES 
 
 47 
 
 down into earthy matter of somewhat different composition, known 
 as subsoil. Between the two there is commonly no distinct separa- 
 tion, but the subsoil is often, though not always, more compact 
 than the soil, and its color is often different. Like the soil, it con- 
 tains both mineral and organic matter, though the latter is less 
 abundant than in the soil. Only the larger roots, and the roots 
 of the larger plants, penetrate the subsoil in great numbers. The 
 thickness of the subsoil is often much greater than that of the soil, 
 but, on the other hand, it is sometimes absent altogether. 
 
 Rock. Beneath the subsoil is rock. When a geologist speaks 
 of rock, he does not necessarily mean solid rock, for sand, gravel, 
 clay, etc., in large quantities and in the proper relations are in- 
 cluded under this term. The subsoil itself is a sort of rock. As 
 commonly used, however, the term rock implies solid rock, and be- 
 neath the mantle rock the larger part of the earth down to the 
 
 FIG. 40. Soil grading down into rock. Sandstone, south central Wisconsin. 
 
 (MacNeille.) 
 
 lowest accessible depths, and far beyond, is solid rock. It is 
 probable indeed that the body of the earth is solid to the core. 
 
 In many places the mantle rock grades down into the solid 
 rock in such a way as to show that the former was made by the de- 
 cay of the latter (Fig. 40). It is this fact which makes the name 
 rock waste appropriate for the mantle rock. Mantle rock of this 
 sort is local. It is made up of materials derived from the rock 
 below. In other places the plane of separation between the sub- 
 soil and the solid rock below is distinct, with no suggestion of 
 gradation (Fig. 41). In such cases the mantle rock often con- 
 
48 
 
 PHYSIOGRAPHY 
 
 tains materials which could not have been derived from the rock 
 below. They have been transported to their present position from 
 some other source. 
 
 Classes of solid rock. The solid rocks of the earth are of many 
 kinds. They differ from one another in color, in strength, in 
 
 FIG. 41. Section showing loose material (glacial drift) on solid rock. 
 Moines County, la. (la. Geol. Surv.) 
 
 Des 
 
 texture, in composition, in origin, etc.; but the common rocks may 
 be grouped into three great classes, namely, sedimentary rocks, 
 igneous rocks, and metamorphic rocks. 
 
 1. Sedimentary rocks. These rocks were once sediments not 
 unlike the muds, sands, and gravels now being deposited in rivers, 
 lakes, and seas. They are generally arranged in layers or beds, 
 varying from a few inches to several feet in thickness. Because 
 of this structure they are often called stratified rocks. The layers 
 or strata are sometimes horizontal (Fig. 42), but in other cases 
 they are tilted or inclined at various angles. 
 
 Among the common forms of stratified rock are conglomerate, 
 sandstone, and shale. Conglomerate is gravel, the pebbles and stones 
 
RELIEF FEATURES 
 
 49 
 
 of which are cemented together. Similarly, sandstone is sand, 
 the grains of which are cemented together, while shale is mud, the 
 particles of which are so compacted' or cemented that they cohere 
 into a solid mass. Various sorts of mineral matter serve as cement 
 for sedimentary rocks. In general the cementing matter was 
 
 FIG. 42. Stratified rock. Trenton Limestone, Fort Snelling, Minn. 
 
 (Calvin.) 
 
 deposited between the grains or particles of sediment, from water 
 which held it in solution, and which at some time overlay, filled, 
 or passed through the sediments. The stones of the gravel, the 
 grains of the sand, and the tiny particles of the mud, were all 
 derived from some older rock which was, in some way, broken to 
 pieces. The destruction of one kind or generation of rock there- 
 fore furnishes the material for another and younger generation of 
 rock. 
 
 Limestone is another common sort of stratified rock, but in this 
 case the mineral matter which makes the rock was chiefly de- 
 rived from the shells or other hard parts of animals which lived in 
 the sea. It is not of pebbles, sand grains, or mud particles derived 
 directly from the breaking up of older rock. Even the material 
 of the limestone, however, comes from older rock, from which it 
 was dissolved, and taken in solution to the sea. 
 
 Great layers of gravel, sand, mud, shells, etc., are being formed 
 in the ocean, in lakes, etc. We conclude, therefore, that conglom- 
 erate, sandstone, shale, and limestone were formerly beds of gravel, 
 
50 
 
 PHYSIOGRAPHY 
 
 sand, mud, shells, etc., accumulated in similar situations. Since 
 these materials, as now deposited, are arranged in nearly horizontal 
 layers, it is inferred that a nearly horizontal position is the orig- 
 inal position of the beds of sedimentary rock. 
 
 FIG. 43. Massive rock. The Upper Yosemite Falls Compare the struc- 
 ture of the rock with that shown in Fig. 42. 
 
 Stratified rocks are more wide-spread beneath the mantle rock 
 than the rocks of the other classes. They are found even in very 
 elevated mountain regions, where the strata are sometimes tilted 
 and folded in a very complicated way. Even in these high places 
 they often contain the shells or other relics of animals which once 
 lived in the sea. 
 
 From these facts the following conclusions may be drawn: 
 
RELIEF FEATURES 
 
 51 
 
 (1) The materials of which many of the rock formations of the 
 land are composed were laid down beneath the sea; and (2) these 
 deposits have been consolidated, many of them tilted out of their 
 original positions, and some of them raised to great heights, since 
 
 FIG. 44. Granitic rock, about half natural size. The white patches repre- 
 sent crystals of one or two kinds of mineral, and the dark parts represent 
 crystals of others. 
 
 their formation. Such rocks contain parts of the record of the 
 earth's history, and point to very notable changes in its sur- 
 face. 
 
 2. Igneous rocks. From volcanoes, hot liquid rock frequently 
 comes to the surface from unknown depths. This liquid rock is 
 lava. Some of the lava which rises from within the earth stops 
 before it reaches the surface, and cools where it stops, and becomes 
 solid rock. All sorts of rock formed by the solidification of lava 
 are known as igneous rocks. They do not commonly occur in 
 
52 
 
 PHYSIOGRAPHY 
 
 distinct beds or strata, and so are said to be non-stratified or massive 
 (Fig. 43). 
 
 Lavas vary much in composition. They also harden under 
 different conditions, all of which have their effect on the character 
 of the rock. The result is that there are many sorts of igneous 
 
 FIG. 45. Metamorphic rock. (Ells. Can. Geol. Survey.) 
 
 rock. One of the best known is granite. It is composed chiefly 
 of three or four minerals which have the form of imperfect crystals. 
 The minerals are sufficiently different in color and outline to be 
 readily distinguished, if the crystals are large enough to be distinctly 
 seen (Fig. 44) ; but in some igneous rocks they are so small as not 
 to be distinct. When lava cools very quickly, the mineral matter 
 of the liquid lava sometimes fails to crystallize. It then forms a 
 glassy sort of rock. 
 
 When igneous rocks decay, as all igneous rocks do, the decayed 
 particles at the surface may be blown or washed away, and may 
 
RELIEF FEATURES 
 
 53 
 
 then accumulate as sediment in proper situations. Igneous rocks 
 may therefore give rise to sedimentary rock. 
 
 3. Metamorphic rocks. This is the name given to the third class 
 of rocks, and means rock which has been notably altered from some 
 previous condition. Either sedimentary rocks or igneous rocks 
 may be changed into metamorphic rocks, especially through (1) 
 the influence of great pressure, which alters the structure of the 
 
 FIG. 46. Columnar structure in igneous rock. Sierra Nevada Mountains 
 
 rock (Fig. 45); (2) the action of water, which, by dissolving out 
 some parts and depositing new matter, changes the composition- 
 of the rock; and (3) heat, which sometimes causes the mineral 
 matter to crystallize anew in new forms. In these ways either 
 sedimentary rock or igneous rock may be greatly changed. 
 
 All large bodies of rock, whether sedimentary, igneous, or meta- 
 morphic, are traversed by cracks or joints which break them up 
 into larger or smaller masses. The joints may be vertical or in- 
 clined at any angle. Sometimes they are close together (Fig. 46) 
 and sometimes they are far apart (Fig. 42). 
 
54 PHYSIOGRAPHY 
 
 EXERCISE IN THE READING OF TOPOGRAPHIC MAPS 
 
 I. Study the following maps in preparation for conference on the maps: 1 
 
 List of Maps 2 
 
 1. Mt. Mitchell, N. C. 7. Glassboro, N. J. 
 
 2. Harrisburg, Pa. 8. Watrous, N. M. 
 
 3. Shasta Special, Cal. 9. Marsh Pass, Ariz. 
 
 4. Tooele Valley, Utah. 10. Mesa de Maya, Colo. 
 
 5. Donaldson ville, La. 11. Muskego, Wis. 
 
 6. Fargo, N. D Minn. 
 
 Note. In studying a topographic map, notice at the outset: 
 
 (a) In what part of the country the region is situated. 
 (6) The contour interval used, 
 (c) The horizontal scale used. 
 
 II. Apply the following questions to each of the above maps: 
 
 1. Is the region represented by the map a plain, a plateau, or a 
 
 mountain tract? If more than one of these great types is 
 shown, indicate the fact, and locate each definitely. 
 
 2. What range in elevation is shown on the map? 
 
 3. Is the climate of the region wet or dry? Basis for answer. 
 
 4. Is the region thickly or sparsely settled? 
 
 5. What occupations seem to be favored in this region? 
 
 REFERENCES. The following globes, models, and charts are useful in the 
 study of the topics discussed in this chapter: 
 
 1. The Jones' Reliej globe, known as "The Model of the Earth": Chicago. 
 
 2. Howell's models 0} the United States, North America, South America, 
 Eurasia, Africa, and Australia: Washington, D. C. 
 
 3. Coast Survey Charts. The illustrated catalog of these charts may be 
 had of the U. S. Coast and Geodetic Survey, Washington, D. C., and from 
 it charts may be selected intelligently. 
 
 1 The author has carried on the conference work here referred to as 
 follows: The class is divided into groups of four, and each group meets 
 the instructor for a half-hour or for an hour, as the case may be, for the 
 discussion and interpretation of the maps assigned. The maps (also relief- 
 models, photographs, etc.) are studied in advance by the students. This 
 sort of work is regarded as of the utmost importance. 
 
 2 As in the preceding and following lists, these are sheets of the topo- 
 graphic maps issued by the U. S. Geological Survey. See foot-note, page 40. 
 
CHAPTER II 
 THE WORK OF THE ATMOSPHERE 
 
 THE atmosphere is nearly everywhere in direct contact with 
 the surface of the land, and it penetrates the soil and the rock 
 beneath to considerable depths. Its effects on the soil and rock 
 are many and varied, and only a few of the more important ones 
 will be noticed here. Some of them are brought about by the 
 movements of the air, some by the chemical activity of the elements 
 of the air, while some are conditioned by the air, rather than 
 accomplished by it. 
 
 MECHANICAL WORK. THE WORK OF THE WIND 
 Dust 
 
 Universality. The atmosphere is never free from dust. On 
 windy days in dry regions the amount of dust in the air is so great 
 that it may be readily seen. Even when the air seems perfectly still, 
 dust is present. This might be inferred from the fact that even 
 on still days dust settles in houses and in enclosures of all sorts, 
 and it may be seen directly by allowing light to enter a darkened 
 room through a narrow crack or a small hole. In the light thus 
 entering, myriads of dust-motes may be seen. Dust extends 
 high up in the atmosphere, for it is found in abundance in the air 
 over even the highest mountains. It is carried far from its sources, 
 for it often falls at sea many miles from land, and it occasionally 
 settles on the decks of vessels, even in mid -ocean, in such quan- 
 ties as to be readily seen. 
 
 The universality of dust in the atmosphere may be shown in 
 another way. If rain-water which has just fallen be evaporated, 
 a slight amount of sediment remains. This sediment represents 
 the dust which was brought down by the falling drops. Similarly, 
 if fresh-fallen snow be melted and evaporated, there is a residue 
 
 55 
 
56 
 
 PHYSIOGRAPHY 
 
 of dust. This is the case even if the snow be taken from mountain 
 tops, or from such a place as Greenland, far from the cultivated 
 lands and streets which furnish much of the dust in regions which 
 are thickly settled. Since all rams and snows bring down dust, 
 we infer that dust is everywhere present in the lower part of the 
 atmosphere. 
 
 Sources of dust. All the small particles of solid matter held 
 in suspension in the atmosphere are called dust. Fine particles 
 of earthy matter caught up from the surface of the land are most 
 abundant, but the solid particles of smoke, the pollen of flo\vering 
 plants, the spores of plants which, like the puffball, do not blossom, 
 and minute organisms of other sorts are also abundant in the dust 
 of the air. Fine particles of rock blown out of volcanoes are 
 abundant in the vicinity of many active volcanoes, and a trifling 
 amount of dust reaches the earth from extra-terrestrial sources. 
 On windy days quantities of dust are gathered from streets 
 and plowed fields, and from any dry land surface which is not 
 well covered with vegetation. Where the surface is very dry, 
 as in desert regions, and the wind strong, such "clouds" or 
 "whirls" of dust are sometimes swept up by the rising currents 
 of air, so as to be seen for miles. From surfaces densely covered 
 with vegetation the air gets little dust, except pollen. Little or 
 none is gathered from surfaces which are wet, or from surfaces 
 covered with snow or ice. 
 
 Volcanic dust. Volcanoes whose eruptions are explosive often 
 
 send quantities of mineral matter, 
 broken up into fine particles, high 
 into the air. This is volcanic dust, or 
 volcanic "ash." The latter name is 
 not a good one, because the dust is 
 not the product of burning. It is 
 lava, blown into tiny bits by explosion 
 (Fig. 47). The force of explosion is 
 sometimes so great that the dust is 
 sent up high into the air, and once in 
 that position it is blown hither and 
 thither by the winds, sometimes being 
 carried great distances. 
 
 In August of 1883 a violent volcanic eruption took place on 
 the island of Krakatoa, between Java and Sumatra. Half of 
 
 FIG. 47. Particles of volcanic 
 dust. 
 
THE WORK OF THE ATMOSPHERE 
 
 57 
 
 the island was blown away, and enormous quantities of dust were 
 projected high into the air. The course of this dust in the air 
 was traced, roughly, by means of its effects upon the coloring of 
 the sunsets. In this way it was estimated to have been blown 
 completely around the earth in about fifteen days. The course 
 of most of the dust was around the earth in latitudes near the 
 equator, but from this low latitude it spread notably toward the 
 poles. It has been estimated that some of the dust was still in 
 
 FIG. 48. Thick layer of volcanic dust (5 or 6 feet) on the Richmond estate, 
 Island of St. Vincent, five miles from the crater of the Soufriere. After 
 the eruption of 1902. (Hovey, Am. Mus. Nat. Hist.) 
 
 the air three years after the eruption, and that some of it went 
 several times around the earth before settling. It is probable that 
 the dust from this single volcanic eruption found its way to nearly 
 all parts of -the earth. This example may serve to illustrate the 
 extent to which dust is carried in the upper part of the air, and 
 the length of time it may be held in suspension. Dust in the 
 lower part of the air is not usually held so long or carried so far, 
 partly because the winds are less strong, and partly because the 
 dust encounters all sorts of obstacles, such as hills, trees, etc., 
 against which it lodges. Large quantities of dust were ejected 
 
58 
 
 PHYSIOGRAPHY 
 
 from the Soufriere and from Pele, in the West Indies, in the 
 eruptions of 1902 (Fig. 48). 
 
 FIG. 49. Bluff of loess at Kansas City. (Mo. Geol. Surv.) 
 
 FIG. 50. Vertical face of loess near Huang-tu-Chai in northern Shan-si. 
 (Willis, Carnegie Institution.) 
 
 Loess. In some parts of China, in parts of Europe, and over 
 considerable areas in the Mississippi basin there are considerable 
 thicknesses of a distinctive earthy material, the particles of which 
 
THE WORK OF THE ATMOSPHERE 
 
 59 
 
 PlG. 51. A bluff of loess in China on which stands a temple. 
 (Willis, Carnegie Institution.) 
 
 Kio. 52. Facade of a group of buildings in a bluff of loess, Province o* 
 Shan-si, China. (Richthofen.) 
 
PHYSIOGRAPHY 
 
 FIG. 53. Dwellings in loess, Province of Shan-si, China. 
 (Blackwelder, Carnegie Institution.) 
 
 FIQ. 54. A roadway in China which has been deepened by the removal 
 of loess by wind and water. (Willis, Carnegie Institution.) 
 
THE WORK OF THE ATMOSPHERE 
 
 61 
 
 are smaller than sand grains, but larger than the particles of c4ay. 
 It is known as loess, and much of it, at least, was deposited by the 
 wind. From the flood plains of such rivers as the Missouri clouds 
 of dust are swept up and out over the adjacent high lands at the 
 present time, whenever the surface of the flood plain is dry and 
 the wind strong. This dust is very like loess, if, indeed, it be not 
 loess. The loess has the remarkable property of standing with 
 
 FIG. 55. Slopes of loess in China, terraced for agricultural purposes. 
 (Willis, Carnegie Institution.) 
 
 steep or vertical faces (Figs. 49-51) for long periods of time. In 
 China the loess is said to be several hundred feet thick locally, 
 but in the Mississippi basin it rarely reaches a thickness of more 
 than 30 to 50 feet. In parts of China the people have excavated 
 houses in successive tiers along the faces of the soft though steep 
 slopes of the loess (Figs. 52 and 53). 
 
 How held in the air. Though made up largely of mineral 
 matter which is much heavier than the air, dust is kept in sus- 
 pension, first, because the particles are so small that their surfaces 
 are large in proportion to their masses, so that the friction involved 
 in their descent through the air is great; and second, because 
 there are numerous upward currents in the atmosphere, and these 
 
62 PHYSIOGRAPHY 
 
 rising currents carry particles of dust upward in spite of gravity, 
 which is always tending to bring them down. As a matter of fact, 
 the dust of the atmosphere is always settling somewhere, and the 
 supply is being constantly renewed. 
 
 Distribution. In view of what is known concerning the move- 
 ments of dust in the air, it would probably involve little exagger- 
 ation to say that every square mile of the earth's surface may 
 have received dust from every other square mile which is capable 
 of furnishing dust to the atmosphere. Much of the dust of the 
 atmosphere falls into the ocean, or into other bodies of water, 
 where it is safe from further disturbance by winds; but that which 
 lodges on land may be picked up and blown about again and again. 
 
 Gradational effect of winds. Since dust is being blown con- 
 stantly from the land to the sea, and since the sea is making no 
 commensurate return to the land, the shifting of dust by the 
 wind tends, on the whole, to lower the land and to build up the 
 sea bottom. Locally, however, the wind-deposited dust aggrades 
 
 the land. 
 
 Sand 
 
 Sources of sand. Even gentle winds pick up and carry dust; 
 strong ones pick up and carry grains of sand, and even tiny pebbles. 
 Like the finer material, sand is blown about only when it is dry. 
 Abundant sand is found along many shores of seas and lakes, along 
 the bottoms of some valleys, in desert regions, and in some other 
 situations. In most of these places it is dry at times, and in some 
 of them it is dry most of the time. 
 
 Lodgment of wind-blown sand. Sand grains are not often 
 carried up to such great heights as particles of dust, nor do they 
 remain so long in the air. Because of their greater mass, they 
 drop through the atmosphere more promptly when the velocity 
 of the wind is checked. Because they are carried chiefly in the 
 lower part of the atmosphere, they are much more likely than 
 dust to be stopped by obstacles on the surface of the land. Thus 
 every tree, log, stump, building and fence, and every mound and 
 hill against which sand is blown, is likely to cause the lodgment of 
 some of it, just as they are likely to cause the lodgment of wind- 
 blown snow. It follows that sand, instead of being somewhat 
 evenly distributed, as dust is, is often accumulated in mounds and 
 ridges, which begin their growth about almost any sort of obstacle 
 on the surface. 
 
THE WORK OF THE ATMOSPHERE 63 
 
 Dunes. Mounds and ridges of wind-blown or eolian sand are 
 dunes (Fig. 56). Once started, a dune becomes an obstacle to 
 blowing sand, and the lodgment of more sand causes the dune to 
 grow. In this way, mounds and ridges of sand, scores and some- 
 times even hundreds of feet high, are built up by the wind. Small 
 dunes are much more numerous than large ones. 
 
 FIG. 56. A ripple-marked sand dune in a western valley. (U. S. Geol. Surv.) 
 
 Distribution of dunes. Dunes are found principally near the 
 sources of abundant sand. Thus they are common along the 
 Atlantic Coast of the United States south of New York. The sand 
 is here washed up on the beach by the waves, and whenever it dries 
 it may become the prey of the wind. Winds from the west blow 
 the sand into the sea; those from other directions, but especially 
 from the east, drift it up onto the land, making dunes. Dunes 
 abound along the eastern side of Lake Michigan, and some of them 
 are very large ; but they are essentially absent from the west shore 
 of the lake. This is because both the prevailing and the strongest 
 winds are from the west. Dunes are also more common on the 
 leeward than on the windward sides of valleys. Thus where west- 
 
64 
 
 PHYSIOGRAPHY 
 
 erly winds prevail, dunes are more common on the east sides of 
 valleys than on the west sides. They are on the whole more com- 
 
 Fio. 57. A group of dunes at the head of Lake Michigan. Dune Park, Ind. 
 
 (Meyers. ) 
 
 mon on the south than on the north sides of valleys, because the 
 storm winds of winter are from the north of west, rather than 
 from the south of west. Dunes abound over tracts of thousands 
 
 FIG. 58. Dunes at Longport, coast of New Jersey, showing the irregular 
 forms developed by winds which erode. 
 
 of square miles in extent in the semi-arid tracts of the Great Plains, 
 as in western Nebraska and western Kansas. The dune area be- 
 
THE WORK OF THE ATMOSPHERE 
 
 65 
 
 tween the Arkansas River and the Cimarron was the most difficult 
 portion of the famous Santa Fe trail. Dunes of great size occur 
 also in the west-central part of Wyoming. They reach their 
 greatest development in still more arid regions, such as Sahara. 
 
 Locally dunes are the most conspicuous feature of the surface. 
 They are, on the whole, more common on plains and low plateaus 
 than in mountains. 
 
 Configuration of dunes. The shapes of dunes vary widely. 
 When they take the shape of mounds, they may be round or oval, 
 or they may be very irregular in outline. When they take the form 
 of ridges they may be short or long, straight or curved. In gen- 
 eral one slope, the leeward, is steeper than the other, the wind- 
 ward. The shape of the same dune, however, shifts from time to 
 time. When dunes are in process of destruction by the wind, their 
 forms are often very irregular (Fig. 58). This is sometimes because 
 the vegetation growing on them holds the sand in which it is rooted. 
 
 Associated with the dunes there are often depressions (PL V). 
 Some of them are without outlets, while others have outlets. 
 Some of these depressions were scooped out by the wind, and some 
 of them were enclosed by the building up of dunes about thed. 
 
 FIG. 59. Lee side of a sand dune, Cape Henry, Va. The dune is advancing 
 on a forest and burying the trees. (Hitchcock.) 
 
 Destructiveness of eolian sand. The piling up of sand into 
 dunes sometimes does great damage. Narrow tracts of arable 
 
66 
 
 PHYSIOGRAPHY 
 
 FIG. 60. Sand dune showing the effect of a building on the disposition c 
 the sand. The wind reflected from the building keeps sand from ac 
 cumulating against it. Manistee, Mich. (Hitchcock.) 
 
 " 
 
 FIG. 61. Sand drifted over railway. Rowena, Wash. (Dept. of Agr.) 
 
PLATE V 
 
 FIG. 1. Dunes on coast of 
 New Jersey. Scale 1 
 miles per inch . (Cape May 
 Sheet, U. S. Geol. Surv.) 
 
 L.S.&TAT/ON 
 
 L y ff A C H 
 
 FIG. 2. Dunes along Arkan- 
 
 sas River in Kansas. Scale 
 
 2 miles per inch. (Lar- 
 nedSheet, U.S. Geol. Surv.) 
 
 Fio. 3. Dunes in plains of Nebraska. Scale 2 miles per inch. (Camp Clarke Sheet, 
 
 U. S. Geol. Surv.) 
 
THE WORK OF THE ATMOSPHERE 67 
 
 land along seacoasts have been made desolate by the formation 
 of dunes. Even forests of large trees are sometimes buried be- 
 neath them (Fig. 59). Some sorts of trees make heroic efforts 
 to maintain their life against the burying sands by throwing out 
 roots far above their original bases. In this way some of them 
 survive until they are nearly covered up; but if the sand finally 
 covers their tops, they are killed. Occasionally, too, the sand 
 buries abandoned buildings. It rarely accumulates so rapidly 
 about a building that it may not be kept away by human effort. 
 Drifting sand is sometimes the occasion of much trouble along 
 railways, as shown in Fig. 61. Many caravans have been destroyed 
 in the African desert by sand storms, and an army of Cambyses 
 numbering 50,000 is said to have been overwhelmed and buried. 
 
 Migration of dunes. Dunes are often migratory. Sand is 
 blown from their windward sides and dropped on the leeward sides. 
 The continued shifting of sand from the windward to the leeward 
 side causes the slow movement of the dune in the direction of the 
 prevalent winds. The migration of dunes, like their first de- 
 
 FIG. 62. A resurrected forest. After burying and killing the forest, the 
 sand was blown away, exposing the dead trees. (Meyers.) 
 
 velopment, often works destruction to cultivated lands, to forests, 
 to buildings, etc. 
 
 Some idea of the extent to which dunes migrate may be gained 
 both from natural phenomena and from historic records.* Thus 
 when dunes which have buried forests move on, the forests which 
 
68 
 
 PHYSIOGRAPHY 
 
 were buried and killed are again discovered. This is illustrated 
 by Fig. 62. The movement of the dune sand may discover other 
 things also. At one locality on the coast of North Carolina a sand 
 area was utilized for a cemetery. The wind has now blow T n the 
 
 FIG. 63. Migration of dune sand, exposing bones in a cemetery. 
 Hatteras Island, N. C. (Cobb.) 
 
 sand away to such an extent as to expose the bones of the bodies 
 buried (Fig. 63). Buildings buried by dunes are sometimes again 
 revealed after the dune has moved on. This is shown by Fig. 64. 
 
 Recent discoveries indicate that "there are hundreds, per- 
 haps thousands, of square miles of buried towns and cities "* in 
 Central Asia. At least a part of these cities have been buried by 
 migrating dunes. 
 
 FIG. 64. Diagram illustrating the migration of dunes. Kurische Nehrurg. 
 
 (Credner.) 
 
 So disastrous is the migration of dunes along some coasts that 
 steps are taken to prevent their movement. If a dune becomes 
 clothed with vegetation, its position is not likely to be changed 
 so long, as the vegetation remains, for the plants have the effect of 
 
 1 Nat. Geog. Mag., Vol. XVI, 1905, p. 499. 
 
THE WORK OF THE ATMOSPHERE 
 
 69 
 
 pinning the sand down. Trees, shrubs, etc., adapted to such situa- 
 tions are sometimes planted in the sand to prevent its further 
 drifting (Fig. 65). This is done at various points along the 
 western coast of Europe, where land is valuable. It has been 
 done to some extent in our own country, as at San Francisco, 
 where shrubs have been planted on the coastal slope to prevent 
 
 FIG. 65. Dune sand held by brush fences on Kurische Nehrung. 
 
 the sand of the shore from blowing over the Golden Gate Park. 
 Between 1826 and 1838 the Government spent $28,000 to fasten 
 the dune sands on the harbor shores of Provincetown, Mass. 
 Even in such cases, however, additional sand may be deposited 
 on the plant-covered dune. 
 
 Not all eolian sand in dunes. Eolian sand is not always built 
 up into dunes. It is sometimes spread somewhat evenly over 
 the surface where it lodges. Eolian sand is probably much more 
 wide-spread than dunes are. 
 
 Ripple -marks. The surface of wind-blown sand is often 
 marked by pronounced ripple-marks (Fig. 56), very like those 
 which affect the surface of sand deposited beneath water. 
 
 Gradational effects. Much sand is blown from the land into 
 the sea; but the waves wash some of the sand up on the beach 
 again. The one process reduces the volume of the land, while 
 the other increases it. The relative importance of these two pro- 
 cesses is not known. On the whole, the degradation effected by 
 the blowing of sand exceeds the aggradation effected by its deposi- 
 tion, so far as the land is concerned. The aggradational effects 
 are, however, very conspicuous locally. 
 
70 
 
 PHYSIOGRAPHY 
 
 The amount of dust and sand shifted about on the land, or 
 from the land to the sea by the wind, is very great. It has been 
 estimated that in violent dust-storms the amount of dust and 
 
 FIG. 66. A phase of wind-carving on sandstone. Wyoming. (Bastin.) 
 
 FIG. 67. A phase of wind carving near Klondike, Wyo. (Leffingwell.) 
 
 sand in the air may amount to as much as 126,000 tons per cubic 
 mile of air. The average amount of dust in the air, however, is 
 probably but a very small fraction of one percent, of this amount. 
 
THE WORK OF THE ATMOSPHERE 
 
 71 
 
 If we knew how many tons of sand and dust were blown from the 
 land to the sea each day, the figures would doubtless be most im- 
 pressive, but the amount has never been measured or even estimated. 
 Abrasion by the wind. Sand and dust blown against a sur- 
 face of rock have the effect of a sand-blast, and wear away even hard 
 rock. If the surface against which sand is driven is of unequal 
 hardness, the softer parts are worn more rapidly than the harder. 
 
 FIG. 68. Erosion Columns in Monument Park, Colo.; partly the product 
 of wind erosion. (Fairbanks.) 
 
 In regions where abundant sand is driven by the wind, projecting 
 rocks are often carved into fantastic forms (Figs. 66-68). Abra- 
 sion by wind-driven sand is of little consequence in a plain country 
 where the climate is moist, and where bare rock is rarely exposed; 
 but it is of much consequence in semi-arid regions where the topog- 
 raphy is rough, and where hills and points of bare rock are nu- 
 merous. Wind-driven dust is much less efficient than sand in the 
 wear of rock surfaces. 
 
 THE CHEMICAL WORK OF THE Am 
 
 One of the principal constituents of the atmosphere is oxygen, 
 and oxygen is a substance which is chemically active, especially in 
 
72 PHYSIOGRAPHY 
 
 the presence of moisture. This is readily seen when a piece of steel, 
 such as a knife-blade, is exposed to the air. It promptly rusts. 
 This means that both oxygen from the air, and water, have entered 
 into combination with the iron, and the iron-rust contains all 
 three substances, united into one. It is a matter of common 
 knowledge that the iron-rust scales off, and that a knife-blade will 
 soon be "eaten away" by this process, that is converted entirely 
 into rust. When oxygen unites with iron the iron is said to be 
 oxidized. If water enters into combination at the same time, as 
 it does when iron rusts, the iron oxide is said to be hydrated. Iron- 
 rust is therefore the hydrated oxide of iron. The amount (weight) of 
 oxygen and water in iron-rust is greater than the weight of the iron. 
 
 Similar changes go on in the rocks. Iron, in some combination 
 or other, is abundant in some rocks, and present in most of them, 
 and the iron in the rocks is subject to changes similar to those 
 suffered by the knife-blade; and in the rocks, as in the knife-blade, 
 the oxidation of the iron generally leads to the crumbling of the 
 rock of which it is a constituent. Other substances in the rocks 
 also are oxidized and hydrated, usually with the result of tending 
 to break up the rock. 
 
 Other constituents of the atmosphere are also active in chang- 
 ing some of the minerals of the common rocks. This is the case, 
 for example, with the carbonic acid gas (CO 2 ) of the atmosphere, 
 which enters into combination with certain constituents of the 
 rock. The union of carbonic acid gas with constituents of the 
 rock is known as carbonation. Like oxidation, carbonation usually 
 results in the crumbling of the rock affected. 
 
 Weathering. All changes which, like oxidation and carbonation, 
 lead to the breaking up of the rock are phases of the general 
 process of weathering, which includes most of the natural processes 
 by which rock at or near the' surface is caused to crumble. The 
 processes of weathering are very important. Much of the soil 
 and subsoil (mantle rock) of the earth have been produced by 
 them. Furthermore, the weathering of the rock is a necessary 
 preparation for its ready transportation by wind and water. 
 
 CHANGES BROUGHT ABOUT UNDER THE INFLUENCE OF THE AIR 
 
 The surface of the land is subject to great changes of tempera- 
 ture, and these changes are of importance in its physiography. 
 The effects of temperature changes on the rocks of the earth are 
 
THE WORK OF THE ATMOSPHERE 73 
 
 much more obvious in some regions than in others; but, directly 
 or indirectly, they are of more or less importance everywhere. 
 
 Freezing and thawing. In many regions where the surface 
 is well covered with soil, the soil freezes in winter; that is, the 
 water in the soil freezes, so that the soil becomes solid. While the 
 soil is frozen it cannot be blown or washed away. In low tem- 
 peratures, too, the precipitation falls as snow instead of rain, and 
 the snow does not immediately have the same effect as rain on 
 the land. When it is melted the water runs over the surface much 
 as rain-water would; but if the soil beneath the melted snow be 
 frozen, the effect of the running water is relatively slight. 
 
 Where the soil is thin, the waters which sink beneath the surface 
 may freeze in the cracks of the rock below. Since water ex- 
 pands about one-tenth of its volume on freezing, the ice which 
 forms in the cracks (joints) of the rock when they are nearly full 
 of water, acts like a wedge, widening the cracks and forcing the 
 parts of the rock asunder. The effect is illustrated by the frequent 
 breaking of bottles or other vessels in which water is allowed to 
 freeze. This process of rock-breaking is most important where 
 there is abundant moisture, and where the changes of temperature 
 above and below the freezing-point of water are frequent, that is 
 in middle latitudes, or in altitudes which have the temperatures 
 of middle latitudes. 
 
 Expansion and contraction of rock; rock -breaking. When 
 solid rock has little or no covering of loose material, as is often the 
 case on steep slopes, it is heated by day and cooled by night. At 
 high altitudes, and especially on slopes and cliffs exposed to the 
 noonday sun, the daily changes of temperature of the surface of the 
 rock are great. In such places the surface of the rock may become 
 very hot while the sun shines. Since rock is a poor conductor of 
 heat, only its outermost portion is heated notably. Heat expands 
 rock, and as the heated part expands it is likely to scale off from 
 the cooler, unexpanded or less expanded part beneath. As the 
 sun descends, the surface cools and contracts. The outermost 
 film of rock cools first, and most, and tends to break. The break- 
 ing of cool glass by touching it with hot water, or of hot glass by 
 touching it with cold water, involves the same principle. 
 
 The breaking of rock by heating and cooling, even when ice is 
 not formed, is a very common phenomenon. Thus on hot days 
 in summer the blocks of cement in cement walks sometimes 
 
74 
 
 PHYSIOGRAPHY 
 
 expand so much as to break (Fig. 69). The heat of the sun some- 
 times so expands the rock in the floor of a rock quarry that it is 
 similarly bowed up and broken. This has been seen repeatedly 
 in the limestone quarries about Chicago, and on the floor of the 
 Drainage Canal before the water was turned in. Many bowlders 
 which lie on the surface are seen to be "shelling off" (Fig. 70), 
 
 FIG. 69. A cement walk broken under expansion by sun-heat. 
 
 and the same process is sometimes seen on mountain tops (Fig. 71). 
 In high mountain regions where the changes of temperature of 
 the rock are great and sudden, the exposed rock is often much 
 broken. So far has this gone that the surface of many a sharp 
 mountain peak is covered with cracked and broken rock, so in- 
 secure that a touch or a step will loosen numerous pieces and start 
 them down the mountain (Fig. 72). Quantities of such debris 
 
 FIG. 70. Concentric weathering, or exfoliation of bowlder. Eastern California. 
 
 (Fairbanks.) 
 
 (called talus') bury the bases of many of the western mountains 
 many hundreds of feet (Fig. 73). The pieces of talus range in 
 size from tiny bits up to masses tons in weight. 
 
 The breaking of rock through changes of temperature is not the 
 
THE WORK OF THE ATMOSPHERE 
 
 75 
 
 FIG. 71. Exfoliation on a mountain slope. Mount Starr-King, Cal. 
 
 FIG. 72. Crumbling on a mountain top. Kearsarge Pass, Sierra Nevada 
 
 Mountains. 
 
76 
 
 PHYSIOGRAPHY 
 
 work of the atmosphere; but the atmosphere has much influence 
 on the changes of temperature, on which the process depends. 
 
 This process of rock breaking is a phase of weathering. The 
 debris loosened in this way moves from higher to lower levels 
 under the influence of gravity, if it moves at all. The general 
 effect of the process is to make high places lower, and to build up 
 the lower surfaces about the bases of steep slopes. 
 
 There are many other phases of weathering not due to the 
 atmosphere, and not altogether conditioned by it. Some of these 
 are illustrated by Figs. 74 and 75. 
 
 FIG. 73. Talus slope. 
 
 The roots of the plants penetrate the soil, loosening it, and 
 thereby make it easier for water to get below the surface. Roots 
 sometimes grow in cracks in the rock, and as they grow they act 
 like wedges (Fig. 74) . Large masses of rock are sometimes loosened 
 in this way. When a tree is uprooted, the ground is torn up, and 
 rock material to the depth of several feet is sometimes exposed 
 to the action of freezing water, air, and rain (Fig. 75). Again, 
 when plants decay, acids are formed which increase the dissolving 
 power of ground-water. Burrowing animals of all sorts loosen 
 the ground and develop channels for the entrance of water. Even 
 
THE WORK OF THE ATMOSPHERE 
 
 77 
 
 FIG. 74. Tree growing in crack in a rock, and by its growth 
 splitting the rock. 
 
 FIG. 75. Upturned tree, showing the disturbance of soil and rock. 
 (U. S. Geol. Surv.) 
 
78 PHYSIOGRAPHY 
 
 such small animals as ants and earthworms perform an important 
 work in this connection. In Massachusetts the ant has been 
 estimated to bring one-fourth of an inch of fine earth to the surface 
 each year, while Darwin estimated that the earthworm brings 
 seven to eighteen tons of material per acre to the surface each 
 year. 
 
 SUMMARY 
 
 On the whole, the tendency of the work of the atmosphere 
 and of the work which is controlled by it, is to lower the surface 
 of the land, and to loosen materials of the surface so that they 
 may be readily moved to lower levels by other agencies. The 
 most important phase of the degradational work cf the atmos- 
 phere is weathering, or the preparation of material for removal 
 by other and more powerful agents of degradation. As we shall 
 see, however, the atmosphere is not the only agent concerned in 
 weathering (see p. 110). 
 
 REFERENCES 
 
 1. CORNISH, On the Formation of Sand Dunes: Geog. Jour., Vol. IX, 
 1897, pp. 278-309. 
 
 2. UDDEN, Erosion, Transportation, and Sedimentation performed by the 
 Atmosphere: Jour, of Geol., Vol. II, pp. 318-331, and Pop. Sci. Mo., Sept., 
 1896. 
 
 3. CHAMBERLIN AND SALISBURY, Geologic Processes, Chapter II : Henry 
 Holt & Co., 1903. 
 
 4. GEIKIE, Earth Sculpture, Chapter XII: Putnam. 
 
 5. HITCHCOCK, Controlling Sand Dunes in the United States: Nat. Geog. 
 Mag., Vol. XV, pp. 43-47, 1904. 
 
 6. COBB, Nat. Geog. Mag., 1906, pp. 310-317. 
 
 7. MERRILL, Principles of Rock Weathering: Jour, of Gecl., Vol. VI, 
 pp. 704-724 and 850-871 ; also Rocks, Rock Weathering, and Soil.-t: Macmillan. 
 
 8. COWLES, The Plant Societies of Chicago and Vicinity: Bull. '2, Geog. 
 Soc. of Chicago. (Deals with relations of dunes and vegetation.) 
 
 9. SALISBURY AND ALDEN, Geography of Chicago and its Environs: Geog. 
 Soc. of Chicago, 1899, pp. 60-63. 
 
 10. All standard text-books on Geology. 
 
 11. The following Topographic Sheets of the U. S. Geol. Surv.: 
 Long Beach, N. J. North Platte. Neb. Great Bend, Kan. 
 Sandy Hook, N. J. Green Run, Md.-Va. Kinsley, Kan. 
 Barnegat, N. J. Fire Island, N. Y. Hutchinson, Kan. 
 Atlantic City, N. J. Springfield, Colo. Lakin, Kan. 
 Asbury Park, N. J. Dodge, Kan. Pratt, Kan. 
 Browns Creek, Neb. Larned, Kan. 
 
 St. Paul, Neb. Kingman, Kan. 
 
 12. The following Folios of the U. S. Geol. Surv.: 
 
 Norfolk, Va. Camp Clarke, Neb. Scotts Bluff, Neb. 
 
THE WORK OF THE ATMOSPHERE 79 
 
 TOPOGRAPHIC MAPS SHOWING EFFECTS OF WIND-WORK 
 
 I. Study the following topographic maps in preparation for the con- 
 
 ference : ' 
 
 1. Sandy Hook, N. J. N. Y. 5. Barnegat, N. J. 
 
 2. Cape May, N. J. 6. Lamed, Kan. 
 
 3. Toleston, Ind. 7. Pratt, Kan. 
 
 4. St. Paul, Neb. 8. Kinsley, Kan. 
 
 II. Geologic Folios 2 to be studied: 
 
 1. Norfolk, Va. N. C. 2. Camp Clarke, Neb. 
 
 III. Suggestions for the study of the above maps: 
 
 On these maps mounds, hills, and ridges of wind-blown sand, 
 or dunes, are represented. In studying the dunes note: 
 
 1. The various shapes, sizes, and modes of aggregation of the 
 dunes in the several regions. 
 
 2. The average height of the dunes above their surroundings 
 in the different regions. 
 
 3. The distribution of the Kansas dunes with reference to 
 the stream courses. Is any law of distribution suggested by the 
 map? If so, how is it to be accounted for? 
 
 4. What are the immediate sources of the sand which forms 
 the dunes in the different regions, so far as can be inferred from 
 the maps? 
 
 5. Note that in several of the areas depressions are numerous 
 among the dunes. May the wind be responsible for them? 
 
 6. Locate the dunes on the "Topographic Sheet" of the Camp 
 Clarke folio. Then turn to the next map of the folio, the "Areal 
 Geology Sheet." Here the dune areas are colored yellow (see 
 legend). From the topographic map form a mental picture of 
 "Jail Rock," "Smokestack Rock," and "Chimney Rock." Test 
 your picture by reference to Figs. 18, 20, and 23 in the back of 
 the folio. The wind is playing an important part in the reduction 
 of these elevations. How? 
 
 1 See foot-note, page 54. 
 
 2 The folios, like the topographic maps, are published by the U. S. Geo- 
 logical Survey. Most of them can be purchased at 25 cents each. 
 
CHAPTER III 
 THE WORK OF GROUND-WATER 
 
 GENERAL FACTS 
 
 WATER is one of the most active agents working on the land. 
 Its activity is seen on every slope during heavy falls of rain and 
 when snow is melting rapidly, and it is seen in every stream and 
 in the waves of lakes and seas. Even the water in the soil and 
 in the rocks beneath the soil is active, though its effects are less 
 obvious than those of streams and waves. 
 
 The great activity of the water, like that of the air, is due 
 primarily to its mobility; but its greater weight makes moving 
 water much more effective than moving air when it comes in con- 
 tact with the land. 
 
 Source of land-water. The waters of the land have fallen 
 from the atmosphere, which always contains some moisture in 
 the form of water vapor. This vapor is constantly passing up 
 into the air from all moist surfaces by evaporation, a process which 
 will be considered in another chapter; but the familiar fact that 
 any moist surface soon dries in the sunshine, or in any warm 
 dry place, may be taken to illustrate what is going on at all times, 
 both from moist land surfaces and from water surfaces. 
 
 Under certain conditions some of the moisture in the air is 
 condensed into drops and falls as rain; or if the temperature at 
 which the water vapor condenses is below the freezing-point of 
 water, the moisture freezes as it condenses, forming snowflakes 
 (Fig. 76) instead of rain-drops. The average amount of rain 
 and snow (snow-water) which falls on the land is something like 
 40 inches per year (10 to 12 inches of snow being counted as one 
 of water). In other words, the precipitation (the rain and snow) 
 which falls on the land each year is enough to make a layer of 
 
 80 
 
THE WORK OF GROUND-WATER 81 
 
 water rather more than three feet, or about one meter, deep over 
 the surface of the land, if it were equally distributed. Forty 
 inches of water over the land would make about 35,000 cubic 
 miles of water. Since the rivers discharge but about 6500 cubic 
 miles of water into the sea yearly, it is clear that the larger part 
 of the rainfall is not carried to the sea by rivers. 
 
 FIG. 76. Photographs of snowflakes, showing something of their diversity 
 
 of form. (Bentley.) 
 
 The fate of rain-water. The water which falls on the land 
 disappears in various ways. A part of it sinks at once beneath 
 the surface, a part forms pools or lakes upon the surface, a part 
 runs off over the surface directly, and a part of it is evaporated. 
 The proportion of the rainfall in any given place which follows 
 each of these courses depends upon several conditions, among 
 which are (1) the topography of the surface, (2) the rate of rain- 
 fall (or the rate at which snow melts), (3) the porosity of the soil 
 or rock, (4) the amount of water which the soil already contains 
 
82 PHYSIOGRAPHY 
 
 when the rain falls or the snow melts, (5) the amount of vegetation 
 on the surface, and (6) the dryness of the atmosphere. 
 
 Considering these points in order, we find (1) that the steeper 
 the slope on which rain falls or snow melts, the more rapidly the 
 water runs off, and the larger the proportion which follows this 
 course; for when it flows off rapidly there is little time for it to 
 sink in or evaporate. 
 
 (2) The more rapidly the rain falls, the less the proportion 
 which sinks in. The sinking of the water is a slow process, espe- 
 cially if the soil is compact. In sinking, the water first fills the 
 pores of the surface, and no more can enter until that already in 
 has had time to sink out of the way. If, therefore, the rain falls 
 rapidly, less sinks in and more runs off over the surface than if 
 it falls slowly. 
 
 (3) Loose or open soil, such as sand or gravel, takes in the 
 water more readily than clay or other compact material. A 
 clayey soil, therefore, causes a larger part of the rainfall to run 
 off over the surface, because it allows less to sink in in a given 
 time. Not only this, but a porous soil will hold more water, 
 because the pore space, that is the space between its constituent 
 parts, is larger. A special case of compactness arises in connection 
 with changes of temperature. When the ground is frozen, that 
 is when the water in it is frozen, the soil is rendered solid and 
 less porous, and surface water can penetrate it but slightly, even 
 if thet .surface water remains unfrozen. When the soil beneath 
 snow is frozen, the water produced by the melting of the snow does 
 not readily enter it, and so a large part is allowed to run off over 
 the surface. 
 
 (4) When the soil contains much water, less can enter, and 
 more runs off over the surface. 
 
 (5) Vegetation serves to restrain the flow of surface water, 
 and holds it longer on the surface. As a result there is more 
 time for the water to sink in, and less runs off without sinking. 
 
 (6) If the air is very dry, a larger proportion of the rainfall 
 evaporates directly, leaving less to run off or sink in. The effect 
 of dryness on evaporation is conspicuous in arid regions, where 
 the surface dries quickly after a shower, and where light snows 
 often entirely disappear in a short time by evaporation, even when 
 the temperature is constantly far below the freezing-point of 
 water. 
 
THE WORK OF GROUND-WATER 
 
 The water which sinks into the ground becomes ground-water, 
 while that which flows off over the surface without sinking is the 
 immediate run-off. Much of the ground-water ultimately reaches 
 the surface again, and some of it joins the immediate run-off in 
 the streams. All the water which the streams carry, whether 
 it has been beneath the surface or not, is the run-off. 
 
 The existence of ground-water. The soil of most regions is 
 damp at a depth of a few inches, or at any rate a few feet, even 
 when it seems dry at the surface. In deep holes, like wells, water 
 seeps or flows in from the sides, and collects in the bottom. In 
 thickly populated farming communities there are wells on almost 
 every farm, and all are supplied with water. Illinois has more 
 than 250,000 farms, and it is probable that the number of wells 
 in the state is double the number of farms. The total number of 
 wells in the United States must be several millions, and the amount 
 of water drawn out through them each day is very great; yet 
 the wells do not ordinarily go dry. Mines, too, usually encounter 
 water at no great depth from the surface. 
 
 The source of ground-water. Since rain-water and melted 
 snow are constantly sinking beneath the surface, since rain and 
 snow seem like an adequate source of supply for the ground-water, 
 and since no other source l whence it can come is known, it is 
 inferred that surface water is the source of ground-water. Other 
 phenomena point to a close connection between the two. Thus 
 in time of drought many shallow wells and some springs go dry. 
 When the drought is broken by renewed rainfall, the wells again 
 contain water, and the springs again flow. This seems to establish 
 a direct connection between precipitation from the atmosphere 
 and the supply of water beneath the surface. 
 
 While the proportion of rainfall which sinks beneath the surface 
 is determined by the conditions already suggested, the amount 
 of water beneath the surface, other things being equal, is great 
 where the rainfall is great. The amount of ground-water in any 
 region is, however, not entirely dependent on the rainfall of that 
 region, for water falling in one place may flow underground to* an- 
 other. Thus rain-water which falls in the Rocky Mountains flows 
 
 1 Water rarely sinks into the ground from lakes, rivers, etc., though 
 it does so under some circumstances. Rivers and lakes are, however, fed 
 by rain-water, either before or after it has been underground, so that ground- 
 water derived from lakes and rivers comes from atmospheric precipitation. 
 
84 PHYSIOGRAPHY 
 
 underground through porous beds of rock out under the Great 
 Plains, where it is brought to the surface through wells (Fig. 77). 
 
 FIG. 77. Diagram showing how water tailing in one place may flow under- 
 ground to another and there be brought to the surface. The layer a 
 is porous and water entering it in the mountains follows it to the plains. 
 
 Descent of ground-water. The manner in which water enters 
 the soil is readily seen. It sinks in through all the pores and cracks. 
 In the soil and subsoil pores are more common than cracks, but 
 in the solid rock beneath, cracks are common, and while pores are 
 not absent, they are often much smaller than the cracks. Since 
 the cracks in rock run in various directions, the water descends not 
 only vertically, but in oblique directions as well. Water descends 
 as long as there are cracks and pores or openings of any sort not 
 already filled with water; but the smaller these passageways 
 become, the more difficult it is for the water to pass through them. 
 If, for example, small pores, such as those which occur in compact 
 soil or rock, be diminished in size one-half, the difficulty of the 
 descent of water is much more than doubled. 
 
 Generally speaking, the rocks near the surface have more 
 and larger pores than those at greater depths. It follows that 
 as the pores get fewer and smaller with increasing depth, the 
 difficulty with which water descends increases. 
 
 It cannot be stated definitely how far down cracks and openings 
 exist; but it seems probable that all openings become very small 
 at a depth of a mile or two, and that none exist below a depth of 
 five or six miles. At this depth the rock is under the pressure of 
 a column of rock five miles high, and the weight of such a column 
 is so great that, in any ordinary sort of rock, cracks and pores 
 would be closed, if once formed. Since different sorts of rock 
 have unequal strength, pores and cracks might exist in different 
 sorts of rock at somewhat different depths, but probably in no 
 rock at a depth of more than about six miles. 
 
 For these reasons it is not probable that the water descends 
 more than five or six miles, and the amount of water below the 
 
THE WORK OF GROUND-WATER 
 
 85 
 
 depth of even one or two miles is probably far less than the amount 
 above that level. 
 
 The ground-water surface. Though the amount of water 
 beneath the surface is very great, as common phenomena show, 
 the porous rock and soil are rarely altogether full of water. This 
 is shown by the fact that it is necessary in many regions to dig 
 wells to a depth of several scores or even hundreds of feet before 
 an adequate supply of water is obtained. The surface soil is 
 rarely full of water except immediately after a heavy rain, or 
 when snow is melting. 
 
 FIG. 78. Represents a series of wells sunk in a flat tract of land. 
 
 If a series of wells be dug in a flat region, where the soil and 
 rocks are essentially uniform, they would need to be dug to about 
 the same depth in order to secure a constant supply of water. 
 This is illustrated by Fig. 78. If the well at a is dug to a given 
 depth, a well at b will need to be dug to about the same depth in 
 order to secure an equal supply of water Other wells at c and d 
 will also need to be of approximately the same depth. Under 
 these circumstances, the water in the several wells will stand at 
 about the same level. This means that the rocks and subsoil of 
 
 FIG. 79. Diagram illustrating the position of the ground-water surface 
 (the dotted line) in a region of undulating topography. 
 
 the region, below the level of the water in the several wells, 
 are full of water. The underground surface below which the rocks 
 etc., are full of water in any given region is the water surface (or 
 water table} for that region. In one region the water surface 
 may be 10 feet below the actual surface, and in another 100 feet. 
 In dry regions it may be even deeper, but where the rainfall is 
 
86 PHYSIOGRAPHY 
 
 sufficient for agricultural purposes, the water surface is rarely 
 more than a few score feet below the surface of the land. 
 
 Where the surface is uneven, the ground-water surface usually 
 undulates with it, but to a less extent, as shown in Fig. 79. 
 
 Amount of ground-water. The amount of ground-water is 
 not very definitely known, but the best estimates which have been 
 made indicate that the water in the soil, rocks, etc., of the land 
 would probably make a layer not more than 1000 feet deep, 1 if it 
 were spread out over the surface of the land. The amount of water 
 in the rocks beneath the ocean bed is probably less per square 
 mile than beneath the land surface, because the rocks there are 
 probably less porous. Even if the amount of ground-water beneath 
 the sea were as great as that in the land, square mile for square 
 mile, the total amount of ground-water would be but a fraction of 
 that in the sea. 
 
 The movement of ground-water. Ground-water is in con- 
 stant movement. This is shown in many ways. If all the water 
 is pumped out of a well, it soon fills again to its former 
 level. This shows that water flows in. The constant flow of 
 the thousands of springs shows that ground-water is in movement, 
 for only thus can the springs be supplied with water. The seepage 
 of water into mines, quarries, etc., tells the same story. 
 
 Ground Water Surface 
 
 B 
 
 -Ground Water Surface 
 
 FIG. 80. In the upper part of the figure (A) the water surface is level. 
 If a heavy rain takes place in the area at the left of that represented 
 by the figu-e the water surface at the left will be raised as indicated by 
 the lower part (B) of the figure. Movement of the ground-water will 
 follow. 
 
 The reasons why ground-water is in movement are readily 
 understood. The rainfall is not equally distributed. If there 
 be a heavy local shower in a flat region, where the water surface 
 is level or nearly so, the soil and rock in the area where the rain 
 falls are more or less completely filled with water. The result is 
 
 1 Estimates have ranged from 3000 feet to 100 feet. 
 
THE WORK OF GROUND-WATER 87 
 
 that the water surface for the region is temporarily raised, as 
 shown at c in Fig. 80. Since water is mobile, this is a condition of 
 instability, and the water from c will tend to flow off in all direc- 
 tions to places where the water surface is lower. The principle 
 involved is precisely the same as that which would be in operation 
 if a mound of water were placed on a level surface. It would 
 promptly spread. In the subsoil, or rock beneath, the water 
 tends to spread just as it would at the surface, but its movement 
 is miLch slower because of the friction of the water with the rock, 
 etc., through which it passes. 
 
 In a region of uniform rainfall, but of uneven surface, the water 
 surface is not level. Other things being equal, it is somewhat 
 higher beneath high land, and somewhat lower beneath low land 
 (Fig. 79). Where this is the case the water surface would ulti- 
 mately become level if there were no rain; but in moist regions 
 rain is so frequent that the water surface under the hills rarely 
 or never sinks down to the level of the water surface under the 
 surrounding low land, before it is raised again by additional rains. 
 As a result of inequalities of rainfall and of topography, ground- 
 water is constantly moving out from areas where the water surface 
 is higher to areas where it is lower. 
 
 While the flow of ground-water is determined primarily by the 
 ground-water surface, and while it always tends to flow from 
 higher to lower levels, it is, in some situations, forced upward. 
 Thus, if water moving down through a porous layer of rock (a, 
 Fig. 77) between beds (& and c) which do not allow it to penetrate 
 them, finds an opening through the impervious layer (6), it may 
 escape upward. It may even issue with great force, if the source 
 of supply be much higher than the point of issue. Water also 
 rises by capillary action, but not in such quantity as to give rise 
 to springs or visible seepage. 
 
 In addition to the water which comes out from beneath the sur- 
 face through wells and springs on the land, some of it flows under- 
 ground to the sea or to lakes, and issues as springs beneath them. 
 Some ground-water, too, seeps out in such small quantities as not 
 to appear to flow. In this case it does not constitute a spring. 
 
 Ground-water moves about to some extent under the influence 
 of forces other than gravity. Besides the movements resulting 
 from capillarity, some of it is taken up by the roots of plants, 
 and, passing up through the plants, escapes through their leaves 
 
88 PHYSIOGRAPHY 
 
 into the air. Still another portion of the water beneath 
 the surface is evaporated directly, without the intervention of 
 plants. Even in regions where the soil appears to be very dry 
 evaporation is constantly going on. If, for example, the water 
 surface is 500 feet down, the pores of the rock down to the water 
 surface are full of air. From the water surface below, water evap- 
 orates into the air in the rock and soil, and this air may thus 
 become more moist than the air above. Moisture is thus diffused 
 upward, and, under some circumstances, it may rise by convection. 
 
 The rise of invisible moisture from the ground may be easily 
 demonstrated in a very simple way. If a rubber blanket be spread 
 on the ground on a summer night, or if a pan be inverted on the soil, 
 the under side of the blanket or pan will often be dripping-wet in 
 the morning, before the heat of the sun affects it. Had the cool 
 blanket or the cool metal not been there to stop it, the moisture 
 from below would have escaped into the air above, unnoticed. 
 It is so escaping all day and all night, and every day and every 
 night, over all land surfaces wherever the air in the soil and below 
 it is more moist than that above. In this and other ways the 
 supply of ground-water is being constantly drawn upon. Constant 
 renewal through the descent of rain, or through underground 
 flowage from some other region, is therefore necessary to maintain 
 the supply. 
 
 It is probable that nearly all of the water which sinks beneath 
 the surface sooner or later comes up again in some one of these 
 various ways; but a small amount of it enters into combination 
 with the solid mineral matter, as in iron-rust (p. 71). So long as 
 water remains in this solid combination, it does not again escape 
 to the surface. 
 
 The rate at which ground-water moves varies greatly, and is 
 dependent chiefly on (1) the porosity. of the rock or soil, and (2) 
 the pressure of the water. The rate at which water seeps through 
 soils from irrigating ditches has been determined at various points 
 in the West. Except in very porous soils, it ranges from one to 
 eight feet per day. In very porous soils it is sometimes as much 
 as fifty feet per day. In the Potsdam standstone, a wide-spread 
 formation underlying southern Wisconsin and its surroundings to 
 the south, and the source of many artesian wells, the rate of move- 
 ment of ground-water has been estimated at half a mile a year. 
 Rain-water which enters this formation 100 miles from Chicago 
 
THE WORK OF GROUND-WATER 
 
 89 
 
 would therefore reach that city in about 200 years, if this rate is 
 correct. The water which sinks to great depths and into the very 
 small pores and cracks moves with extraordinary slowness, and 
 some of it remains entrapped within the rock for very long periods 
 of time. 
 
 Springs 
 
 All water issuing from beneath the surface is seepage. Water 
 issuing through a natural opening in such quantity as to make a 
 distinct current is a spring. Springs occur in many sorts of situa- 
 tions, but they are not located by accident. They occur where 
 there are natural passageways for the ground-water to escape 
 to the surface. Such passageways arise in various ways. Two 
 
 FIG. 81. Diagram to illustrate two types of springs as explained in text. 
 
 cases are illustrated by Fig. 81. In one, the water descends through 
 the porous bed e to the layer d, which is relatively impervious. 
 The water flows along this layer until the layer comes to the 
 surface (outcrops) and there the water flows out as a spring, s'. 
 In the other, the water moves underground through the porous 
 layer 6, under pressure, until it reaches a crack which leads 
 up to the surface. If the crack is open enough to afford 
 a passageway, the water may follow it up to the surface, 
 as at s. A spring may occur in such a situation only when the 
 opening is lower than the water surface in the rock which furnishes 
 the water. In the figure, the spring at s is lower than the .water 
 surface at w. This sort of a spring is similar to a flowing well in 
 principle, but in the latter case the opening is made by man. 
 
 Temperature. The temperature of water as it issues from 
 beneath the surface is very variable. Most springs seem cold in 
 warm weather. There is indeed a popular impression that springs 
 are cooler in summer than in winter, but this is not the case. The 
 impression arises from the fact that the water is much cooler than 
 the air in summer, and so seems cold, while in winter the water 
 is much warmer than the air, and so seems less cold than in sum- 
 mer. 
 
90 PHYSIOGRAPHY 
 
 Springs which derive their water from deep sources vary little 
 in temperature during the year, while those whose sources are 
 shallow are colder in the winter than in summer. The reason is 
 that the warmth of summer and the cold of winter are most ex- 
 treme at the surface, and become less and less with increasing depth. 
 Below the depth of 50 or 60 feet, in middle latitudes, the tempera- 
 ture does not vary sensibly with the seasons, so that springs which 
 draw their water from greater depths vary little in temperature, 
 while those which draw their supply from lesser depths vary more. 
 
 Exceptional springs are warm, and still more exceptional ones 
 are hot. Where spring water is hot it is commonly because it 
 has been in contact with hot rock. In general the water of warm 
 and hot springs probably rises from considerable depths. In 
 many cases the source of the heat is probably igneous rock (lava) 
 which has not yet become cold. It may be lava which was forced 
 upward toward but not to the surface, or it may be the deeper 
 parts of lava which flowed out on the surface. 
 
 Mineral and medicinal springs. All ground-water dissolves 
 more or less mineral matter from the rocks, and all springs there- 
 fore contain more or less mineral matter in solution ; but a spring 
 is not commonly called a mineral spring unless it contains (1) 
 much mineral matter, or (2) mineral matter which is conspicuous 
 either by reason of its color or its odor, or (3) mineral matter 
 which is unusual in spring water. 
 
 Many mineral springs are thought and sometimes justly 
 to have healing properties, and so are known as medicinal springs. 
 Many of the famous watering-places and resorts for invalids are 
 located at hot mineral springs. The Hot Springs of Arkansas, of 
 South Dakota, and of Carlsbad (Bohemia) are examples. Many 
 springs which are charged with gases are called mineral and me- 
 dicinal, especially if they have an offensive odor. In the popular 
 mind, a spring is more medicinal the worse it smells and tastes. 
 Hot water is a better solvent than cold water, so that hot springs 
 generally contain much mineral matter. 
 
 Geysers. In some parts of the world the water of hot springs 
 is forced out violently from time to time. Such springs are called 
 geysers. A geyser is therefore an intermittently eruptive hot spring. 
 Geysers are best known in the Yellowstone National Park, but 
 they are well developed also in Iceland. They exist in New Zea- 
 land, though some of the geysers of that island were destroyed 
 
THE WORK OF GROUND-WATER 
 
 91 
 
 by volcanic eruptions in 1886. In the Yellowstone Park there 
 are about 100 geysers, and more than 3000 hot springs which are 
 not eruptive. Some of the geysers send up boiling water and 
 steam to a height of 200 feet or more (Fig. 82), but this is quite 
 above the average. 
 
 From some geysers the eruptions are frequent, and from others 
 infrequent. From some they occur at regular intervals, and from 
 others they take place irregularly. One of the geysers in the 
 
 FIG. 82. Giant Geyser, Yellowstone National Park. (Wineman.) 
 
 Yellowstone Park is named "Old Faithful," because it discharges 
 its waters at nearly regular intervals of about an hour. The 
 eruptions are, however, a little less frequent and a little less regular 
 than formerly. In general, geysers which have been known for 
 long periods of time discharge their waters less and less frequently 
 as time goes on. 
 
 The features which may be seen generally at a geyser are the 
 following : 
 
 (1) An opening in the surface leading down to unknown depths. 
 Though this is sometimes called the geyser tube, it is probably not 
 
92 
 
 PHYSIOGRAPHY 
 
 always in the form of a tube. (2) A shallow basin about the 
 opening. This is sometimes, though not always, in the top of a 
 mound. In some cases there is, instead of a basin, an irregular 
 
 FIG. 83. Cone (crater) of Castle Geyser, Yellowstone National Park. 
 (Detroit Photo. Co.) 
 
 FIG. 83a. The cone of Lone Star Geyser, Yellowstone National Park. 
 
 (U. S. Geol. Surv.) 
 
 perforated mound (Figs. 83 and 83a). Both basins and mounds 
 are composed of mineral matter (commonly silica) which has been 
 deposited by the water which has issued from the geyser. (3) At 
 the time of discharge much steam, as well as liquid water, issues. 
 
THE WORK OF GROUND-WATER 93 
 
 It seems certain that steam is the force which ejects the water. 
 It is believed (1) that ground-water enters the geyser tube much 
 as it enters a well; (2) that the walls of some part of the tube are 
 of hot rock; (3) that the water in the tube is brought to the boiling 
 temperature at some point in the tube below the top of the water; 
 and (4) that when this takes place, the steam formed forces out 
 all the water above. It forces it out because water, when changed 
 to steam, expands about 1700 times. 
 
 This principle may be illustrated by experiment. If a short 
 tube of water is heated, it boils without violent discharge, especially 
 if the tube has a large diameter. But if the tube is filled with 
 sand and the sand then filled with water and heated from below, 
 the movement (convection) of the heated water in the tube is 
 greatly restricted by the sand. The result is that steam may be 
 formed abundantly below the surface, and a miniature eruption 
 may follow. 
 
 Geysers occur only in regions of relatively recent volcanic 
 activity f and the heat which is necessary for the geysers is probably 
 supplied by lava which has not yet become cool. As it is heated, 
 the geyser water is constantly cooling the hot rock, and in time 
 it will cease to be hot enough to boil the water. Geyser action 
 will then cease, unless new supplies of hot lava are forced up from 
 below. In the Yellowstone Park some geysers have died out 
 since the region became known, but little more than thirty years 
 ago. New geysers, on the other hand, have been developed in 
 the same region during this period. 
 
 The reason why the water in a geyser tube is shot out, at 
 intervals, while the water in an open kettle is not, is found in the 
 difference in the shape of the vessels holding the water. When 
 water is heated it expands. When water is heated in a kettle, 
 that at the bottom rises readily, by convection, to the top, so that 
 there is a nearly uniform temperature throughout. The geyser 
 tube is much deeper than the kettle, and in places it is probable 
 that it is small in diameter, certainly small in proportion to its 
 length. The tube is also more or less crooked. Both its small- 
 ness and its crookedness interfere with the rise of the water 
 heated below, and the result is that water below the surface is 
 brought to the boiling temperature before that at the surface is. 
 Hence steam is formed below the surface, rather than at the 
 surface, and blows out the water above. 
 
94 
 
 PHYSIOGRAPHY 
 
 If a stone or a clod of earth, or almost any other solid object, 
 be thrown into a geyser, its eruption may often be hastened a 
 little, because such things interfere with the convection of the 
 water in the tube. They help to hold the hot water down where 
 it is being heated, and so help it to reach a boiling temperature 
 at some point below the surface a little sooner than it would do 
 otherwise. Soap, especially, is supposed to hasten a geyser's 
 eruption. Its effect is probably somewhat less than is usually 
 believed. Anything which makes the water more viscous hastens 
 the eruption, because convection is less free in a thick fluid than 
 in a thin one. 
 
 Artesian and flowing wells. When the water in a well rises so 
 as to overflow, the well is said to flow. Flowing wells are not 
 
 FIG. 84. Artesian well at Woonsocket, S. D. (U. S. Geol. Surv.) 
 
 unlike springs whose waters spout up as they issue. The chief 
 difference between them is that the opening in one case is natural, 
 while in the other it was made by man. Formerly artesian wells 
 were regarded as the same as flowing wells. The name was derived 
 from Artois, France, where there was a notable well of this sort. 
 Nowadays the name artesian is often applied to deep wells, whether 
 they flow or not. 
 
THE WORK OF GROUND-WATER 
 
 95 
 
 Fig. 85 illustrates the general conditions necessary for flowing 
 wells. They are the following: (1) A porous layer or bed of rock, 
 a, which underlies an impervious layer, b, which prevents the water 
 from escaping upward until it is penetrated by the well-hole. 
 (2) The porous bed must come to the surface in a region which is 
 somewhat higher than the site of the well. (3) The rainfall where 
 
 
 FIG. 85. Diagrams illustrating the conditions favorable for artesian wells. 
 In A the porous bed a is in the form of a basin; in B it merely dips. 
 
 the porous bed comes to the surface must be sufficient to keep 
 it well filled with water. Under these conditions the water beneath 
 w, in the stratum a, is under the pressure of the water in the same 
 stratum at higher levels, and if a hole is made down to it, it will 
 gush up (Fig. 84). It is not necessary, as a rule, to take much 
 account of the layer below the water-bearing stratum a. If it is of 
 porous rock, it is generally full of water, and so prevents the 
 downward escape of that in a. 
 
 The water at the well will not rise as high as the water surface 
 in a, for in flowing through the small openings (pores and small 
 cracks) in the rock, there is loss of force by friction. An allowance 
 of about a foot to the mile must usually be made; that is, if the 
 source of supply is 100 miles away, the water surface at that point 
 should be about 100 feet higher than the top of the well, in order 
 that the water may flow out. If the water-bearing stratum, a, 
 is very porous, the allowance which must be made for friction 
 is less; if it is close-grained, the loss of force from friction is more. 
 
 Artesian wells vary much in depth. They may be but a few 
 feet deep, or they may be thousands of feet. There is an artesian 
 well in Berlin more than 4000 feet deep, one in St. Louis nearly 
 4000 feet deep, and one in Cincinnati nearly 2500 feet deep, while 
 the deepest one in Chicago is some 2700 feet deep. There are 
 numerous artesian wells in New Jersey less than 100 feet in depth. 
 
96 PHYSIOGRAPHY 
 
 The amount of water flowing from artesian wells is often great. 
 At Belle Plain, la., the water rose 77 feet above the surface when 
 a certain well was first drilled. A little later the water rose from 
 another adjacent well with such force as to greatly enlarge the 
 opening through which it rose, and with force enough to bring up 
 stones two to three pounds in weight. At first the flow from the 
 second well was estimated to be more than 5,000,000 gallons per 
 day, though it soon became less. The flow from many wells is 
 nearly constant. 
 
 Many villages and small cities get their water from artesian 
 wells. Charleston , S. C., Galveston and Fort Worth, Texas, Camden, 
 N. J., and Rockford, 111., are among the cities supplied partly or 
 wholly in this way. No great city, however, such as New York, 
 Chicago, Philadelphia, etc., is supplied with water from such 
 wells, and probably could not be. 
 
 In the semi-arid region of the Great Plains, and at various other 
 places in the West, as, for example, in some parts of California, 
 water from artesian (deep) wells is extensively used for irrigating 
 the land. 
 
 THE WORK OF GROUND-WATER 
 Chemical Work 
 
 Solution. While rock seems to be the symbol of all that L 
 stable, it is nevertheless dissolved, to some slight extent, by the 
 ground-water which passes through it, as has been stated in con- 
 nection with springs. Pure water does not dissolve mineral matter 
 readily; but the water beneath the surface is not pure. In falling 
 through the atmosphere it dissolved carbonic acid gas, oxygen, and 
 other gases, and in sinking through the soil it took up the products 
 of plant decay, so that when it became ground-water it contained 
 numerous impurities. With these impurities in solution, ground- 
 water dissolves most sorts of rock more readily than pure water 
 would. Pure water, for example, has little effect on common 
 limestone, but water with carbonic acid gas in solution dissolves 
 this rock to some appreciable extent. The descending water often 
 changes the rocks and minerals through which it passes, chemically, 
 before it effects much solution; but this is not always the case. 
 Since solution is often the result of these chemical changes, it is 
 included under the chemical, rather than under the mechanical, 
 changes produced by ground-water. 
 
THE WORK OF GROUND-WATER 
 
 97 
 
 That water does dissolve some of the material of rock is shown 
 by the character of the water which comes out of the ground. 
 When the water of wells or springs is evaporated it usually leaves 
 a little residue. This becomes noticeable in time in the inside 
 coating of boilers and kettles in which water is heated. This 
 coating is composed of mineral matter which was in solution in 
 the water, and which was left behind when the water was heated 
 and evaporated. Strictly speaking, all springs are mineral springs 
 (see p. 90), and all wells are mineral wells, for all water taken 
 out of the ground contains mineral matter. 
 
 The first work of ground-water, then, is solution, or the sub- 
 traction of material from the rocks. One result of solution is to 
 
 
 FIG. 86. Diagram to illustrate the form and relations of caverns developed 
 by solution. The black spaces represent caverns. Some limestone 
 sinks are represented at the surface where the roofs of caves have fallen in. 
 
 make the rock porous. The extreme case of porosity developed 
 in this way is found in caverns and channels beneath the surface. 
 The great caves (Fig. 86), like those of southern Indiana (Wyandotte 
 Cave and others) and Kentucky (Mammoth Cave and others), 
 are the work of ground-water. Caves of this sort occur chiefly 
 in limestone regions, for limestone is the most soluble of the com- 
 mon rocks. Even where caves and caverns are not developed, 
 small pores and cavities are often numerous. The effect of solu- 
 tion is, therefore, to weaken the rock, and finally to cause it to 
 crumble. 
 
 The roofs of underground caves sometimes fall in, leaving 
 notable sinks at the surface. These are known as limestone sinks 
 (Fig. 87). Such sinks -are characteristic of regions in which there 
 are caves. They are occasionally so numerous that the surface 
 is too much pitted to be cultivated. This is the case, for example, 
 
98 
 
 PHYSIOGRAPHY 
 
 in some parts of Kentucky and Tennessee (Plate VI). If a part 
 of the former cavern roof remains to span the depression, a 
 natural bridge is formed. 
 
 In Karst, along the east side of the Adriatic Sea, there is a 
 tract of land underlain by white limestone which is nearly free 
 from soil. Its surface is etched and eroded into fantastic forms. 
 Most of the rainfall of the region goes beneath the surface, and 
 it is the solvent action of the water before and after it sinks which 
 
 FIG. 87. A sink-hole of recent development near Meade. Kan. 
 (Johnson, U. S. Geoi. Surv.) 
 
 has developed the remarkably uneven topography of the region, 
 so bizarre that it has attracted wide attention. Numerous short 
 gullies, ravines, and valleys in the limestone terminate abruptly, 
 discharging their waters into caves or subterranean tunnels. 
 Sink-holes abound, and some of them are several hundred feet deep. 
 The slopes to the depressions, and therefore the slopes of the 
 elevations between them, are very steep, so that the surface is 
 extremely rough. Topography similar to that of this region 
 and developed in the same way is sometimes known as Karst 
 topography. 
 
 The amount of mineral matter brought to the surface through 
 wells, springs, etc., is very great. The springs of Leuk (Swit- 
 zerland) bring to the surface more than 2000 tons of gypsum 
 
PLATt ./I 
 
 Limestone sinks due to solution by ground-water. The depression contours are 
 hachured. Scale 2 miles per inch. (Pikeville, Tenn., Sheet, U. S. Geol. Surv. N , 
 
THE WORK OF GROUND-WATER 99 
 
 (a hydrated sulphate of calcium) in solution yearly. In the same 
 time the springs of Bath (England) bring up enough mineral 
 matter in solution so that, if it were taken out of the water and 
 made into a monument, it would make a column 9 feet in diam- 
 eter and 140 feet high. 
 
 Much of the water seeping out from beneath the surface finds 
 its way to rivers, and the larger part of the mineral matter in 
 solution in rivers has come from the ground-water, or seepage, 
 which has joined them. Rivers are estimated to carry nearly 
 five billion tons of mineral matter to the sea in solution each year; 
 but even this large amount does not represent all the solvent work 
 of ground-water, for much of the mineral matter which it dissolves 
 is deposited without reaching either rivers or sea, for reasons 
 which will soon appear. 
 
 The transfer of this large amount of mineral matter from the 
 land to the sea each year in solution must mean the lowering of 
 the land. It has been estimated that the land surface is lowered 
 in this way about one foot in 13,000 years, on the average. The 
 transfer of this mineral matter from the land to the sea does not 
 mean an equivalent building up of the sea bottom, for some of; the 
 mineral matter remains in solution in the sea-water. Thus salt is 
 one of the mineral substances carried by rivers to the sea; but 
 the larger part of the salt which has been carried to the sea through 
 the ages probably remains there in solution to this day. 
 
 On the other hand, much of the mineral matter which is carried 
 to the sea, especially the calcium carbonate, is used by the animals 
 and plants of the sea in the making of shells, tests, bones, etc., 
 and these are finally left on the sea bottom. 
 
 Deposition. Besides dissolving mineral matter and carrying 
 much of it away, ground- water brings about other changes in the 
 rocks of the earth. If, in the chemical laboratory, solutions of 
 various sorts are mixed in a test-tube, some of the materials in 
 solution are likely to be precipitated. The same thing takes place 
 in the rocks beneath the surface. If, for example, waters from 
 different directions enter a crack in the rock, and if these waters 
 bring different mineral matters in solution, the mingling of the 
 waters may effect a chemical change by which some of the mate- 
 rial comes out of solution, and is deposited in the crack. 
 
 It follows that while ground-water tends to make rocks porous 
 when it dissolves, mineral matter, it tends to make them compact 
 
100 PHYSIOGRAPHY 
 
 where it deposits the mineral matter which it holds in solution 
 in pores and cracks. In some places the first of these processes 
 is the more effective, and in others the second. In general, ground- 
 water probably increases the porosity of rock near the surface 
 (especially above the grcund-water surface) and increases its com- 
 pactness at greater depths. The effect of deposition is some- 
 times to cement the loose parts of rock together, making the whole 
 
 FIG. 88. The Maryland Vein, Nevada City, Cal. The vein is gold 
 bearing quartz. (U. S. Geol. Surv.) 
 
 more firm. Thus sand may be cemented into sandstone, and 
 gravel into conglomerate. 
 
 Cracks in the rock filled by mineral matter deposited from 
 solution become veins (Fig. 88), and many rocks are full of veins 
 (Fig. 89) . Ores occur in some veins, and many mines are located 
 in them. Much of the gold, silver, lead, zinc, etc., is found in such 
 positions. The ores of these metals do not usually fill the cracks, 
 but they are often associated with a much larger amount of min- 
 eral matter which is not valuable, but which must be worked over 
 in order to get out the ores which are. Mineral matter in solu- 
 tion may be deposited in caves (Fig. 90). Many of the most 
 attractive features of caves, such as stalactites, stalagmites, crys- 
 tals on the walls, etc., were formed by deposition from solution. 
 
THE WORK OF GROUND-WATER 
 
 101 
 
 FIG. 89. A piece of rock showing many veins. the white streaks. The 
 vein filling is calcite. Near Highgate Springs, Vt. (Walcott, U. S. 
 Geol. Surv.) 
 
 FIG. 90. Deposits of calcite (travertine, stalactites, and stalagmites) 
 in Wyandotte Cave, Ind. (H'ains.) 
 
102 
 
 PHYSIOGRAPHY 
 
 FIG 91. Deposit from a hot spring in Yellowstone Lake. (Fairbanks.) 
 
 * - 
 
 . 
 
 FIG. 92. Hot Springs deposit; terrace about Mammoth Springs. Yellow- 
 stone National Park. 
 
THE WORK OF GROUND-WATER 
 
 103 
 
 The deposition of mineral matter from solution is determined 
 by several conditions. The more important are the following: 
 
 (1) If water evaporates^ the mineral matter dissolved in it is left 
 behind. Surface gravels are sometimes cemented in this way. 
 
 (2) If the water which contains mineral matter is relieved of pres- 
 sure, as when it comes out to the surface, some of the mineral 
 matter may be deposited. (3) If water contains much gas in solu- 
 tion, and if the gas escapes, as it is likely to when pressure is re- 
 
 FIG. 93. Deposits about a hot spring; summit of Angel Terrace. Yellow- 
 stone National Park. (U. S. Geol. Surv.) 
 
 lieved or when it is warmed, some of the mineral matter in solu- 
 tion is likely to be deposited. (4) Some warm spring water gives 
 up its mineral matter on cooling. Some or all of these principles 
 are involved in the deposition of mineral matter about most hot 
 mineral springs (Fig. 91). (5) The mingling of solutions of dif- 
 ferent sorts, already referred to, is probably a common cause of 
 deposition. (6) In some hot springs, as in the Yellowstone Park, 
 minute plants grow in the hot waters which issue from the springs. 
 These tiny organisms, by some process not well understood, ex- 
 tract mineral matter from the water and cause it to be deposited 
 (Figs. 92 and 93). 
 
104 
 
 PHYSIOGRAPHY 
 
 These are among the simpler and more important conditions 
 under which mineral matter is deposited from solution by ground- 
 water, either while it is beneath the surface or after it issues. 
 
 Solution and deposition may be going on at the same time, 
 and perhaps at the same place; that is, the water may be dissolv- 
 ing certain substances at the same time that it is depositing others. 
 One sort of rock may thus be changed to another. A special phase 
 
 FIG. 94. Petrified tree-trunks, Yellowstone National Park. 
 (U S. Geol. Surv.) 
 
 of this process results in petrifaction. Thus the substance of a 
 buried shell, or of coral, may be changed while its form is pre- 
 served. Another illustration is afforded by petrified wood (Fig. 
 94), the substance of the wood having been replaced by mineral 
 matter. Such changes probably take place slowly, the mineral 
 matter which was in solution replacing the woody matter as it 
 decays, molecule by molecule. 
 
THE WORK OF GROUND-WATER 105 
 
 Other changes. Besides the subtraction cf mineral matter 
 from the rocks by solution, and the addition cf material to the 
 rocks of some places by deposition of mineral matter dissolved 
 elsewhere, the water works still other changes in the rocks. It 
 sometimes enters into combination with certain minerals, chang- 
 ing their character. This process (hydratiori) has already been 
 referred to in connection with the work of the air (see p. 72). 
 The moisture beneath the surface affects minerals in much the same 
 way as moisture in the air or at the surface. All changes of this 
 sort which result in the alteration of the composition of the rock, 
 or of its constituent minerals, are chemical changes. 
 
 The general result of chemical changes effected by ground- 
 water, like the result of the chemical changes effected by the ah*, 
 is to disrupt the rock. Changes of this sort are probably most 
 important near the surface, especially at and above the ground- 
 water surface. 
 
 Summary. From the preceding paragraphs it will be seen 
 that ground-water effects various changes in the rocks. The rocks 
 carrying water may, indeed, be looked upon as a sort of huge 
 chemical laboratory in which solutions are made and carried 
 from one place to another, working changes as they go. The result 
 is a slow but constant alteration in the character of the rock. Im- 
 pressed by the greatness of these changes in the long course of time, 
 an eminent geologist has said that, "given time enough, and nothing 
 in the world is more changeable than the rocks." 
 
 Mechanical Work 
 
 Abrasion. The mechanical work of ground-water is of rela- 
 tively little importance. The water is rarely concentrated into 
 considerable streams, but where it is so concentrated the under- 
 ground streams tend to enlarge their channels by erosion, somewhat 
 as surface streams do. The ground-water which flows in distinct 
 channels transports and deposits the limited amount of sediment 
 which it acquires. 
 
 Slumping, sliding, etc. Indirectly, ground-water participates 
 in changes of another sort. When the soil and earthy material 
 on a steep slope become charged with water, their weight is greatly 
 increased. At the same time the water makes them more mobile. 
 Under these circumstances the material sometimes slides down 
 slopes. Such movements are known as slumping or sliding. If 
 
106 
 
 PHYSIOGRAPHY 
 
 the movement is on a large scale, it is called a landslide. Slump- 
 ing is very common on slopes composed of unconsolidated material, 
 such as clay or accumulations of loose rock (Figs. 95 and 96). 
 Landslides give rise to a distinctive sort of topography. 
 
 Many destructive landslides have been recorded, but a few 
 facts concerning a recent one may serve to illustrate the phenomena 
 of all. On the 29th of April, 1903, there was a slide on Turtle 
 Mountain, Province of Alberta, Dominion of Canada. Here a 
 huge mass of material, nearly half a mile square and probably 
 
 FIG. 95. South face of Landslip Mountain, Colo. The protruding mass on 
 the right has slumped down. (U. S. Geol. Surv.) 
 
 400 to 500 feet deep, suddenly broke loose from the steep east 
 face of the mountain, and slid down into the valley below. It 
 covered the valley, which was half a mile wide, and even rose a 
 few hundred feet on the other side. When it came to rest it covered 
 an area of a little more than one square mile. The length of the 
 slide* was about two and a half miles, and it is estimated that the 
 time which it took was not more than 100 seconds. The heavy 
 rainfall of the preceding year had filled the rock with moisture, 
 and earthquake tremors, shortly before the slide, are believed to 
 have also hastened the catastrophe. Extensive tunnels, etc., ex- 
 
THE WORK OF GROUND-WATER 
 
 107 
 
 FIG. 96. A slope affected by slumping, creeping, etc.. Cascade Mountains, 
 Ore. (U. S. Geol. Surv.) 
 
 FIG. 97. The upper part of a mountain valley. The loose material in the 
 valley has slidden, crept, and rolled down, making what has been 
 called a talus glacier. Near Telluride, Colo. (Hole.) 
 
108 
 
 PHYSIOGRAPHY 
 
 cavated in mining at the base of the mountain may also have 
 played a part by making the under-structure less stable. Many 
 lives were lost and many buildings destroyed. 
 
 Instead of sliding down with rapid motion, the surface earth 
 sometimes moves down with extreme slowness. This sort of 
 movement is creep. It is often too slow to be seen, but it results 
 
 FIG. 98. Material settling off the face of a cliff under the influence of 
 gravity. The freezing of water in the cracks may be a factor in the 
 separation of the cliff face. (U. S. Geol. Surv.) 
 
 in the accumulation of mantle rock, especially earthy matter, at 
 the bases of slopes. In one case (Rhymney Valley, Wales) the 
 rate of creep, where it affected a railway, has been determined to 
 be 6 to 10 feet in fifty years. Movement of the same sort is now 
 constantly disturbing a railway-track a few miles from Golden, 
 Colorado. 
 
 The downward movement of loose surface material on slopes is 
 very general. It often causes the trees to incline a little down 
 
THE WORK OF GROUND-WATER 
 
 109 
 
 slope (Fig. 100). This is probably due in part to the fact that the 
 upper part of the mantle rock in which they are rooted creeps 
 down faster than the lower part. 
 
 FIG. 99. This figure shows the same phenomenon as the last, but the 
 cliff in this case is of solid rock (limestone). The open cracks are largely 
 the result of solution and weathering. East Tensleep Creek, Bighorn 
 Mountains, Wyo. (Hole.) 
 
 FIG. 100. Trees tipping down slope. North of Chicago. (Coxe.) 
 
 .In all cases of slumping, sliding, and creeping the force pro- 
 ducing the movement is gravity. Water only helps to furnish the 
 
110 PHYSIOGRAPHY 
 
 conditions for the effective action of gravity, by making the move- 
 ment easier. 
 
 Closely connected with slumping is another phase of gravity 
 work which may be mentioned here, though ground-water has little 
 to do with it. From the faces of cliffs, blocks and masses of rock 
 often settle away (Figs. 98 and 99). The freezing of ground- 
 water in the cracks may help to pry off the rock on the face of 
 the cliff, and solution may help to widen the cracks. The growth 
 of roots in the cracks acts much like freezing water, helping to pry 
 off the loosened masses from the face of the cliff. In such cases 
 the existence of cracks or joints helps ice, roots, solution, etc., 
 to become effective. 
 
 WEATHERING 
 
 Some of the processes of weathering have already been men- 
 tioned, but it may be added, by way of summary, that the chemical 
 changes (oxidation, carbonation, etc.) effected in the rock by the 
 atmosphere, the mechanical changes effected by variations of 
 temperature under the influence of the atmosphere, and the 
 chemical and mechanical changes effected by ground-water, all 
 conspire to so alter the surface of exposed rock as to cause it to 
 waste away. We have already seen (p. 74) that the surfaces of 
 the bowlders of the fields are sometimes scaling off or crumbling, and 
 they are often discolored, even when they seem firm. The upper 
 layers of stone in a quarry are frequently broken or quite different 
 in color from the lower ones. Inscriptions on old tombstones are 
 often indistinct, and they have sometimes disappeared completely 
 from stones which are but a few score years old. From the walls 
 of stone buildings, from monuments, and from other stone struc- 
 tures, flakes of stone are sometimes seen to be scaling off. 
 
 In all these cases some change has taken place in the rock 
 whereby its outer part is wasted away. All processes which pro- 
 duce this result are weathering. 
 
 The importance of rock weathering is great. Much soil is but 
 weathered rock, and without the weathering of rock much of the 
 land would be bare of soil and so of vegetation. As we have seen, 
 the weathering of the rock greatly facilitates the work of the wind, 
 not always to the advantage of man, by preparing fine material 
 which may be blown away. As we shall see in the next chapter, 
 weathering also prepares material for ready removal by running 
 
THE WORK OF GROUND-WATER 111 
 
 water. Weathering, conjointly with wind and water erosion, is 
 responsible for many striking bits of scenery (Figs. 153 and 154). 
 
 Conditions affecting weathering. There are great differences 
 in the durability of rocks. A coarse-grained rock weathers faster 
 than a fine-grained one of the same composition. Rock traversed 
 by fissures and cracks changes more rapidly than firm, impervious 
 rock. Some rocks, as limestone, are composed of relatively soluble 
 material, and some, as sandstone, of material which is compara- 
 tively insoluble. The former weather more readily than the latter, 
 so far as solution is a factor cf the weathering. A cold climate 
 favors the wedge-work of ice, but hinders the growth of vegetation 
 and chemical changes. Rock decay goes on more rapidly in warm, 
 moist regions than in cold or dry ones; but rock breaking or splinter- 
 ing, due to changes of temperature, is more effective in dry regions, 
 where daily changes of temperature are great. In deserts the 
 wear of the rock by wind-blown sand is important. In general, 
 the weathering of bare rocks probably takes place more rapidly in 
 warm, moist regions than in cold or temperate ones; yet in warm, 
 moist regions the rock is usually covered and protected against 
 some phases of weathering by a goodly layer of soil and subsoil. 
 
 The topographic position of the rocks also has an important 
 bearing on the rate at which they decay. On steep slopes, the 
 waste is commonly washed away as rapidly as formed, and the 
 bare rock is constantly exposed, whereas on plains the solid rock 
 is often deeply buried by mantle rock, and so protected from 
 some phases of weathering. 
 
 The thickness of the surface layer of weathered rock varies 
 greatly. It rarely exceeds 100 feet and is generally much less. 
 In many places the depth of soil and mantle rock represents the 
 excess of rock decay over the transportation of the decayed ma- 
 terial. 
 
 Since over the greater part of the land area there is a covering 
 of mantle rook, it follows that, in the aggregate, rock weathering 
 exceeds transportation. Since much material is carried away in 
 solution or otherwise, as rock weathers, a few feet of rock waste 
 may represent the destruction of many feet of rock. 
 
112 PHYSIOGRAPHY 
 
 MAPS ILLUSTRATING TOPOGRAPHIC EFFECTS OF 
 GROUND-WATER 
 
 Study the following maps showing limestone sinks, "sinking creeks "etc. 
 in preparation for conference. 1 
 
 1. Greenville, Tenn. N. C. 3. Bristol, Va Tenn. 
 
 2. Standingstone, Tenn. 4. Pikeville, Tenn. 
 Greenville Sheet. 
 
 Note the numerous limestone sinks in the central portion of 
 the area, and the "sinking creeks" associated with them. 
 Sketch the probable history of a "sinking creek." 
 
 Standingstone Sheet. 
 
 Note the many limestone sinks, particularly in the south- 
 western half of the map. The rocks are here essentially 
 horizontal and the sinks are therefore without definite 
 arrangement. This is in striking contrast to the condition 
 shown on the Bristol sheet. 
 
 Bristol Sheet. 
 
 The many depressions are sink-holes in limestone rocks. Note 
 the very large depressions near Adelphia, in the north- 
 western part of the map. Notice that the sink-holes 
 occur -in belts, extending northeast and southwest. This 
 is because the tilted limestone beds of the region outcrop 
 (come to the surface) along these lines (see Bristol Folio). 
 
 Pikeville Sheet. 
 
 Note the limestone sinks, especially in the northwestern part 
 of the map. 
 
 Question. What was probably the topography of each region at 
 the time the sinks developed? 
 
 REFERENCES 
 
 1. CHAMBERLIN AND SALISBURY, Geologic Processes, Chapter IV. Henry 
 Holt & Co., 1903; and all standard text-books en Geology. 
 
 2. SHALER, Chapter on Caverns, in Aspects of the Earth. Chas. Scribner's 
 Sons, 1889. 
 
 3. GEIKIE, Earth Sculpture, Chapter XIII : Putnam. 
 
 4. HOVEY, Celebrated American Caverns: Rob't Clarke Co. 
 
 5. CHAMBERLIN, Artesian Wells: Geology of Wisconsin. Vol. I, 1881, pp. 
 689-697; and 5th Ann. Rept. U. S. Geol. Surv., pp. 131-173. 
 
 6. KING, Principles and Conditions of the Movements of Ground-water: 
 19th Ann. Rept. U. S. Geol. Surv., Pt. II, pp. 59-293. 
 
 7. SLIGHTER, The Motions of Underground Water: Water Supply and 
 Irrigation Paper No. 67, U. S. Geol. Surv. 
 
 1 See foot-note, p. 54. 
 
THE WORK OF GROUND-WATER 113 
 
 8. Numerous Water Supply and Irrigation Papers, U. S. Geol. Surv. 
 
 9. WEED ; Formations of Hot Springs Deposits: 9th Ann. Kept. U. S. 
 Geol. Surv., pp. 663-676, and Amer. Jour, of Sci., Vol. XXXVII, 1889, 
 pp. 51-59. 
 
 10. CHITTENDEN, Yellowstone National Park: Rob't Clarke Co. 
 
 11. DAVIS (B. M.), The Vegetation of the Hot Springs of Yellowstone Park: 
 Science, Vol. VI, 1897, pp. 145-167 
 
CHAPTER IV 
 THE WORK OF RUNNING WATER 
 
 STREAMS are among the most wide-spread natural features of 
 the land. Only in desert regions, such as the Sahara, or in areas 
 which, like Greenland, are mainly covered with snow and ice, are 
 there extensive tracts without them. A few streams, like the 
 Mississippi and the Amazon rivers, are very large, but most of 
 them are of small size. Thousands of small creeks and brooks, 
 some of them having their source in the Rocky Mountains, some 
 in the Appalachian Mountains, and some on the plains and plateaus 
 between these mountain systems, feed the Mississippi. And so it 
 is everywhere. Every large stream receives water from many 
 small ones. 
 
 The rivers of the Mississippi basin guided the early explorers, 
 traders, and immigrants, and later became of great commercial and 
 therefore of political importance. Many of the greater streams of 
 other lands have played similar roles in history. 
 
 The flow of some streams is so gentle that they do not appear 
 to work great changes in their valleys ; but some of them wear away 
 their banks so rapidly that the changes they produce may be seen 
 from year to year, or, when the stream is in flood, from day to day, 
 or even from hour to hour. The force of streams at such times 
 is often disastrous (Figs. 101 and 102). Occasionally they sweep 
 away bridges and dams, and sometimes even buildings. The 
 strong beams and rods of the bridges, and the steel rails of railways 
 are bent almost as if they were twigs by the force of the occasional 
 torrent which follows an exceptional rain, such as a cloud-burst 
 (Fig. 103). 
 
 In the aggregate, the streams are estimated to send about 
 6500 cubic miles of water to the sea each year. The average 
 height of the land above the sea is nearly half a mile. These 6500 
 
 114 
 
THE WORK OF RUNNING WATER 
 
 115 
 
 cubic miles of water, therefore, descend, on the average, nearly half 
 a mile before they reach the sea. The energy of the water in falling 
 
 FIG. 101. The Passaic River in flood. Little Falls, N. J. 1902. 
 
 FIG. 102. A raging river. Flood of the Mississippi River, breaking through 
 
 its levees. 
 
 this distance is very great. This will be readily understood if we 
 think of this amount of water falling vertically from a height of a 
 
116 PHYSIOGRAPHY 
 
 little less than half a mile. The water in the streams has the same 
 amount of energy that it would have if it fell vertically. This 
 energy is largely expended in wearing away the materials of the 
 sides and bottoms of the valleys. Its force is therefore great, and 
 its effects on the surface of the land pronounced. 
 
 Rivers flow in mountains, in plateaus, and in plains, and wher- 
 ever they flow they modify the surface in their own peculiar way. 
 
 FIG. 103. Scene in the freight-yards of Kansas City after the flood of 1903. 
 (U. S. Weather Bureau.) 
 
 The topographic features produced by running water in mountains, 
 plateaus, and plains have much in common, and when we study 
 these features for one of these types of regions, we really study 
 them, in principle, for all. 
 
 Sources of stream water. Most streams derive a large part of 
 their water from the immediate run-off and from ground-water, 
 and many of them receive contributions from ponds, lakes, snow- 
 fields, and glaciers. The Mississippi, for example, receives water in 
 all these ways. The immediate run-off, the ground-water, and even 
 the water of the lakes and glaciers, all have their sources in the 
 rain and snow, so that rivers depend on atmospheric precipitation 
 for their supply of water. 
 
THE WORK OF RUNNING WATER 117 
 
 The direct connection between rainfall and rivers may be in- 
 ferred from various familiar phenomena. (1) Streams are more 
 numerous in regions where the rainfall is abundant (Fig. 104) 
 than in those where it is scarce (Fig. 105). (2) Multitudes of small 
 streams spring into being with each heavy fall of rain and with each 
 period of rapidly melting snow. (3) Streams are notably swollen 
 
 FIG. 104. Map showing the many streams of a humid region. Central 
 Kentucky. The area is about 225 square miles. 
 
 after rains, and most after heavy rains. (4) Many small streams 
 which flow during wet weather dry up in times of drought, while 
 others shrink. 
 
 The water of most streams which continue to flow during 
 droughts is derived largely from springs and lakes, or from the 
 melting of snow and ice about the sources of the streams. 
 
 Where water flows on the land, it is because the surface has a 
 slope. If the slope of a surface were perfectly even, the immediate 
 
118 
 
 PHYSIOGRAPHY 
 
 run-off at any given time would flow in a sheet. There are slopes 
 so smooth that their water runs off in this way; but on most slopes, 
 even those which appear to be regular, there are some uneven- 
 nesses so that, although the run-off may start as a sheet, it is 
 soon concentrated into rills and streamlets which follow the de- 
 pressions. The smallest streamlets unite to form larger ones, 
 
 FlG. 105. Map showing the few streams of an arid region. Northern Arizona. 
 The area is as great as that shown in Fig. 104. 
 
 and the little rills, after many unions with one another, reach 
 valleys which have permanent streams. These may be small (creeks 
 or brooks) or large (rivers). Streams which flow but part of the 
 time, as after a rain-storm, during wet weather, or during but a 
 part of the year, are intermittent streams. 
 
 Every permanent stream and many temporary ones flow in 
 depressions called valleys (Fig. 106). Valleys are therefore about 
 as numerous as streams. The very small depressions in which 
 water runs only after smart showers are not always called valleys. 
 
THE WORK OF RUNNING WATER 
 
 119 
 
 If they are very small they are called gullies (Fig. 107) ; or if some- 
 what larger, ravines. Gullies and ravines are but small valleys. 
 Just as the tiny streamlets unite with one another to form creeks 
 
 FIG. 106. Map showing normal drainage relations. Each stream flows in a 
 depression. The largest stream has the largest valley. Streams of 
 smaller size have smaller valleys, while the valleys of the smallest streams 
 are very small. A few miles southwest of Scio, O. (U. S. Geol. Surv.) 
 
 and these to form rivers, so the gullies in which the smallest tem- 
 porary streams flow generally unite to form wider and deeper 
 gullies (Fig. 108). These, in turn, join one another to make ravines, 
 
120 PHYSIOGRAPHY 
 
 which are but larger depressions of the same sort. Ravines lead 
 to valleys, just as gullies lead to ravines. Valleys, like streams, 
 usually end at the ocean or a lake; but in some cases, especially 
 in arid regions, they end on dry land. 
 
 There is, as a rule, some relation between the size of a valley 
 and the stream which follows it, though this relation is not one 
 
 FIG. 107. A gully developed by a single shower. (Blackwelder.) 
 
 which can be stated in mathematical terms. The large stream 
 and the large valley go together so often, however, that the com- 
 bination cannot be accidental, and leads to the inquiry whether 
 the streams make the valleys in which they flow, or whether the 
 streams flow where they do because the valleys were prepared 
 for them in advance. These are questions to which we shall seek 
 an answer in the following pages. 
 
 THE EROSIVE WORK OF STREAMS 
 
 Streams are always carrying mud, sand, etc., down their valleys. 
 This is especially well seen when rivers are in flood, for at such 
 times they are usually muddy. Besides the mud which is sus- 
 pended in the water, streams roll sand, gravel, etc., along their 
 bottoms. The movement both of the mud in the water and of 
 
THE WORK OF RUNNING WATER 
 
 121 
 
 pebbles and stones at the bottom of the water may be seen in 
 any little stream that flows along the roadside after a storm, 
 and the great Mississippi carries its load in the same way. Even 
 when streams are not in flood, they carry sediment, though its 
 amount is then less. Some of them carry so little that their 
 waters are relatively clear, while others, like the Missouri, are 
 always muddy. 
 
 Since most river-water finally reaches the sea, much of the 
 
 FIG. 108. Slope with numerous gullies, the smaller ones joining the larger 
 ones. Scott's Bluff, Neb. (U. S. Geol. Surv.) 
 
 sediment which they carry finally reaches the ocean and is deposited 
 there, chiefly near the shores. 
 
 The amount of material which certain streams carry from the 
 land to the sea has been estimated. The estimate for a given 
 river is made by calculating the volume of water discharged by 
 the river each year and then determining the average amount 
 of sediment in each unit for example, each gallon, or each cubic 
 foot, of water. In this way it has been estimated that the Missis- 
 sippi River carries to the Gulf more than 340,500,000 tons of sedi- 
 ment each year, or nearly a million tons a day. It would take 
 more than 750 daily trains of 50 cars each, each car carrying 
 
122 PHYSIOGRAPHY 
 
 25 tons, to carry an equal amount of sand and mud to the Gulf. 
 All the rivers of the earth are perhaps carrying to the sea forty 
 times as much as the Mississippi. 
 
 We have seen that ground-water dissolves rock matter, and 
 that springs bring some of this dissolved matter to the streams. 
 Streams, therefore, carry certain substances such as salt, carbonate 
 of lime, etc., in solution. These dissolved substances are generally 
 invisible, and, unlike mud and other sediment, remain in the 
 water even after it has become quiet. The presence of these 
 dissolved substances is sometimes made known by the taste; 
 but this is rarely the case in river waters. On evaporating the 
 water, however, as by boiling, the dissolved substances are left 
 behind, as on the inside of tea-kettles, boilers, etc. 
 
 The amount of matter carried to the sea in solution each 
 year by all the rivers of the earth has been estimated at nearly 
 5,000,000,000 tons. This is about one-third as much as the sedi- 
 ment carried by the rivers. 
 
 These general facts show that the rivers are constantly shifting 
 solid matter from the land to the sea. This is, indeed, their great 
 work. Even the water which falls on the land, but does not flow 
 directly to the sea, helps to make the rock decay (p. 105), and 
 so prepares it for removal by running water. It may therefore 
 be said that every drop of water which falls on the land has for 
 its mission the getting of the land into the sea. 
 
 Load and loading. The sediment moved by a stream, whether 
 in suspension or at the bottom, is its load. A stream is loaded 
 when it has all the sediment it can carry; it is but partially loaded 
 when it is carrying less than it might. 
 
 How does a stream get its load? 
 
 As the rain-water begins to flow down the slopes of the land, 
 it picks up and carries with it particles of soil, subsoil, etc.; that 
 is, particles of weathered rock. The result is that the water which 
 runs down slopes after a rain generally carries sediment to the 
 stream which it enters. This is especially true if the immediate 
 run-off flows over cultivated or abandoned fields. The water which 
 flows down freshly plowed slopes, for example, is usually very 
 muddy, while that which runs over slopes well covered with 
 vegetation, such as grass-land or forest, carries away little soil, 
 because the roots of the vegetation hold it. Gullies often develop 
 in plowed fields which lie on slopes, when adjacent fields which are 
 
THE WORK OF RUNNING WATER 
 
 123 
 
 not cultivated do not suffer in the same way. The loosening of 
 soil on hill and mountain slopes often leads to its complete removal, 
 and in some parts of France, in the southern part of our own 
 country, and elsewhere, slopes which were once productive have 
 become barren by the washing away of the soil. 
 
 The amount of sediment carried by the immediate run-off 
 from the slopes is greatest, other things being equal, where the 
 water is concentrated into streamlets, and least where it runs off 
 in sheets. It is under the former condition that little gullies are 
 made (Fig. 107). The gullies are themselves proof that erosion is 
 greater along their courses than on either side, for it was the greater 
 erosion along their courses which made them. 
 
 FIG. 109. A "boiling" or eddying stream. Woods Canyon, Alaska. 
 (Spencer, U. S. Geol. Surv.) 
 
 Much of the sediment of streams is brought to them by the 
 immediate run-off which flows down the slopes of their valleys. 
 But the stream in the valley carries away not only the sediment 
 which is brought to it by gravity and by wind, by sheet-wash 
 and temporary streamlets from the slopes above, but under favor- 
 able conditions it gathers load for itself from its bed and from its 
 banks. This is true, for example, wherever the bed of a vigorous 
 
124 PHYSIOGRAPHY 
 
 stream is composed of clay or sand, for particles of these materials 
 are easily loosened and hurried along in the current. 
 
 The stream does not pick up sediment from its bed merely by 
 the force of the forward movement of the water. We are not to 
 think of a stream as a single straightforward current. When 
 water runs through an open ditch or gutter, some of it may be 
 seen to move from the sides to the center, and some from the 
 center to the sides, while eddies are of common occurrence. These 
 subordinate motions are especially distinct where the current is 
 swift. A swift river, too, "boils" and eddies (Fig. 109), often 
 in a striking manner. In the swift Columbia, for example, eddies 
 are often so strong that it is difficult to row through them. In an 
 eddying current, objects are often "sucked under" and brought 
 up again. There are similar movements, though often less readily 
 seen, in slower streams. 
 
 All these phenomena show that there are numerous subordinate 
 currents in the main current of a river, and that they move in 
 various directions. Many of them are caused by the irregularities 
 
 FIG. 110. Diagram to illustrate the effect of irregularities, a and 6, in a 
 stream's bed, on the current striking them. 
 
 of the stream's bed (Fig. 110), from which they diverge in various 
 directions. The subordinate upward currents in the main current 
 often carry sediment up from the bottom of the stream; that is, 
 they bring it into suspension. When these subordinate currents 
 strike the sides or bottom of a stream's channel, they are often 
 effective in tearing or wearing off bits of loose matter. As we shall 
 soon see, these subordinate currents not only help to get fine 
 sediment into suspension, but they help to keep it there. 
 
 There are two reasons why a stream which is clear or nearly 
 so at the usual stage of water, becomes muddy when it is swollen. 
 One is that in time of flood there is more immediate run-off entering 
 the stream, and this usually brings abundant sediment: the other 
 is that the stream when flooded flows much more swiftly than 
 
THE WORK OF RUNNING WATER 
 
 125 
 
 at other times, and so has power to rub off and pick up much more 
 sediment for itself. 
 
 It might seem from these statements that swift streams should 
 always be muddy and slow ones always clear, but this is not the 
 case. Many a swift stream, especially in the mountains, is re- 
 markably clear, while some sluggish ones are always muddy. 
 The reason is not far to seek. Even a swift stream is clear (1) if 
 immediate run-off (slope-wash) and tributaries bring it no sediment, 
 and (2) if the materials -of its own bed are so coarse that it cannot 
 pick them up. The clearness of many swift mountain streams is 
 due to the fact that there is no mud or sand or fine material of 
 any sort in their beds or banks, while the muddiness of many 
 sluggish streams in plains, such as the Lower Missouri and the 
 Platte, is due to the fact that their bottoms and banks are of such 
 fine material that even their slow currents can get and carry it. 
 
 FIG. 111. Tools with which a river works. These cobblestones and small 
 bowlders were brought down by the stream in flood, and left where they 
 now appear. Other similar materials now in transit cause the riffles 
 in the current. Chelan River, Wash., just above its junction with the 
 Columbia. (Willis, U. S. Geol. Surv.) 
 
 Again, the stream by friction with its bed tends to drag the 
 loose sediment at its bottom along with it, somewhat as a weight 
 of any sort pulled over a surface of mud drags some of the mud 
 beneath along with it. Every stream, therefore, which is not 
 already loaded wears its bed, if it is of soft material such as 
 mud, by (1) friction of the main current, (2) impact of the sub- 
 
126 
 
 PHYSIOGRAPHY 
 
 ordinate currents, and (3) by urging or dragging along the fine 
 material of its bed. 
 
 But some river valleys are in solid rock, even in rock which is 
 very hard (Fig. 26). How are such valleys made? 
 
 In the first place, rock exposed to the water, as in a stream's 
 channel, or to the atmosphere, decays. As it decays it crumbles, 
 and the crumbled part is readily swept away. Again, the sand 
 and gravel rolled along by a stream (Fig. Ill) wear its bed, even 
 
 FIG. Ill A. Tools with which a river works. Bowlders left by the Dela- 
 ware River on its flood plain in times of flood, near the Water Gap. 
 (N. J. Geol. Surv.) 
 
 if it is of hard rock. Even the fine sediment which a stream 
 carries helps to wear its channel. The sediment which a stream 
 carries, therefore, becomes the tool, or, better, a collection of 
 tools, with which the running water works, and with these tools 
 even hard rock is worn away. 
 
 Clear water flowing over a bed of firm, hard rock effects 
 little or no mechanical wear. This is well shown in the case of 
 relatively clear streams like the Niagara. Tiny plants, like those 
 which make moist stone walls green, may often be seen growing 
 on the limestone of its bed where the water is shallow enough to 
 allow the bed to be seen. This is the case even at the brink of 
 the falls, where the current is very swift, and all the force of the 
 
THE WORK OF RUNNING WATER 
 
 127 
 
 mighty torrent is unable to sweep these tiny plants from their 
 moorings. If the stream had a moderate load of sand or mud 
 there can be no doubt that these plants would be swept away 
 with great despatch. The sediment carried by a stream is there- 
 fore a factor which influences its rate of erosion, especially where 
 the bed is of solid rock. 
 
 FIG. 112. A stream channel clogged with bowlders too big for the stream 
 to move, except in times of flood. 
 
 Carrying. It has already been stated that streams move 
 their load (sediment) (1) by rolling it along their bottoms, and (2) 
 by carrying it in suspension above the bottoms. Coarse materials, 
 such as pebbles, are generally rolled,, while fine materials, such as 
 particles of mud, are often suspended. 
 
 The material rolled on the bottom is moved directly by the 
 force of the water. Each pebble which is moved is pushed or 
 rolled along by the water which strikes against it. The principle 
 is the same as that involved in the movement of pebbles on a 
 beach, except that the stream always carries them down the valley, 
 instead of rolling them back and forth. 
 
 Mud is composed chiefly of fine particles of rock, which are 
 nearly three times as heavy as water. In spite of this, they remain 
 in suspension, often for long periods of time. The mud is kept 
 in suspension much as dust is kept in suspension in the air. Since 
 
128 PHYSIOGRAPHY 
 
 its particles are heavier than the water, they tend to sink all the 
 time. They do in fact sink; but as they sink under the influence 
 of gravity they may be caught by minor upward currents and 
 carried upward in spite of gravity. It is chiefly by means of these 
 minor upward currents in the main current that sediment is kept 
 in suspension. Because of the manner in which fine sediment 
 is carried in suspension, it helps to deepen and widen the valley 
 of the carrying stream. As the minor upward currents of a stream 
 carry sediment upward through the water, so minor downward 
 currents drive it against the bottom, and minor sideward currents 
 against the sides of the channel. In these ways even the fine 
 sediment helps the strea-m to enlarge its valley. 
 
 The particles of sediment suspended in a stream are dropped 
 and picked up again repeatedly. A particle may make a long 
 journey, but the long journey may be made up of many short 
 ones. Pai tides of mud carried from Dakota to the Gulf of Mexico 
 ordinarily make many stops in every state along the route, and 
 the time consumed in their journey is generally many times as 
 long as that consumed by the water which started them. 
 
 Amount of load. The amount of sediment a stream carries 
 depends on (1) its velocity, (2) its volume, and (3) the amount 
 and kind of sediment available. A swift, large stream can carry 
 more than a slow, small one. 
 
 The effect of velocity on the carrying power of streams may 
 be seen in most creeks and rivers whose width varies notably. 
 At the narrow places, the swift water is likely to carry away all 
 fine material, permitting only coarse pebbles and stones to remain 
 upon the bottom, while in the broader places the bottom may be 
 covered with mud. By artificially narrowing (by jetties) the 
 Mississippi near its debouchure (1875), James B. Eads not only 
 prevented further deposition of sediment there, but forced the 
 river to clear out its channel. This change permitted the larger 
 ocean vessels to reach New Orleans, and insured the commercial 
 prosperity of that city. 
 
 That fine sediment is picked up and carried more readily than 
 coarse is illustrated by the familiar fact that a stone a pound in 
 weight thrown into any common stream would sink to the bottom 
 promptly, while if a pound of fine dust were thrown into the same 
 stream, its particles would be carried forward some distance before 
 sinking to the bottom. 
 
THE WORK OF RUNNING WATER 129 
 
 A stream can carry a much greater weight of fine sediment 
 than of coarse, both because each pound of fine material carried 
 taxes the stream's energy less than a pound of coarse, and because 
 a larger part of a stream's energy can be used in carrying the 
 former than in carrying the latter. 
 
 Erosion defined. The wearing away of the land surface is 
 erosion. In general, erosion consists of three more or less distinct 
 processes. These are (1) weathering, (2) common, or the picking 
 up of the rock material loosened by weathering or by any other 
 process, and (3) transportation. The solution of rock material by 
 water is often included under corrasion. It would perhaps be well 
 to call it corrosion instead. When the running water is no longer 
 able to carry away sediment, it ceases to degrade its bed. 
 
 Deposition a necessary consequence of erosion. The sedi- 
 ment carried by rivers is deposited whenever they are unable to 
 carry it farther. The cause of deposition is most commonly loss of 
 velocity. Some of the sediment is left in the valleys, especially in 
 their lower courses; and some of it is carried to the sea, or to the 
 lake or other basin to which the river flows. Deposits of sediment 
 in valleys build up or aggrade their bottoms. Thus the Mississippi 
 is spreading sediment over the bottom of its valley for hundreds 
 of miles north of the Gulf of Mexico, and many other large streams, 
 like the Nile, the Hoang-Ho, and the Ganges, are doing the same 
 thing. The total amount of aggradation accomplished by running 
 water on land is, however, far less than the amount of degradation. 
 The deposition of sediment by streams will be more fully considered 
 in its appropriate place. 
 
 Changes Made by Rivers in their Valleys 
 
 A valley has three dimensions, depth, width, and length, and 
 each dimension is subject to change. 
 
 The deepening of valleys. Eroding streams make their valleys 
 deeper and wider. Where streams are depositing, that is, where 
 they leave more than they take away, they are making their valleys 
 shallower. In general, swift streams deepen their valleys, while 
 slow ones often make their valleys shallower. Many valleys are 
 being deepened in their upper courses where the streams are 
 swifter, and made shallow in their lower courses where the 
 streams are more sluggish. 
 
130 
 
 PHYSIOGRAPHY 
 
 Swift streams are swift because they flow in channels which 
 have relatively steep slopes; but as such streams deepen their 
 valleys, the slopes or gradients of the valley bottoms become less , 
 and the streams flow more slowly. In time every swift stream mil 
 cut its channel so low that its current will become sluggish. 
 
 There is no fixed relation between the depth of a valley on the 
 one hand and erosion or deposition in its bottom on the other. 
 Some deep valleys, like the canyon of the Colorado (Fig. 27), are 
 becoming deeper by erosion, while others which are shallow are 
 
 FIG. 113. A shallow valley becoming shallower by deposition. North Platte 
 River near the Nebraska- Wyoming line. (U. S. Geol. Surv.) 
 
 becoming shallower by deposition (Fig. 113). Some deep valleys, 
 on the other hand, are being aggraded, and some shallow ones are 
 being degraded. 
 
 The depth which a valley may attain depends primarily on the 
 height of the land in which it is cut. The higher the land, the deeper 
 the valley may become. Such valleys as the canyons of the Colo- 
 rado (Fig. 27) and the Yellowstone (Fig. 152) are never found in 
 plains (compare Figs. 140 and 169). Valleys of great depth are 
 characteristic of plateaus and mountains. With land of a given 
 height, the depth which a valley may attain depends on its dis- 
 tance from the sea by the route which the water follows. Thus, 
 if a stream flows by a direct course from a plateau 2000 feet above 
 
THE WORK OF RUNNING WATER 131 
 
 the sea and 200 miles from it, it has an average fall of 10 feet 
 per mile; but if it runs off a plateau of equal height 2000 miles 
 from the sea by the course which the water follows, the stream 
 has an average fall of but one foot per mile. If the volume of the 
 stream is the same in the two cases, the valley in the plateau 
 nearer the sea will become much deeper than the other. In other 
 words, the depth which a valley may attain depends primarily 
 on the fall (or gradient) of the water which flows through it. 
 Valleys near the borders of continents are therefore likely to be 
 deeper than those in lands of the same elevation in the interiors of 
 continents. 
 
 Depth-limit. At its lower end, a stream usually cuts its channel 
 down to, or even a little below, the level of the lake, sea, or other 
 river into which it flows. The body of water into which a river 
 flows therefore determines the depth-limit of its valley; but the valley 
 reaches this limit only at its lower end. The upper end of a river 
 valley is always above sea-level. 
 
 The lowest level to which a stream can bring its valley bottom 
 by mechanical wear is called base-level. It is to be noted, how- 
 ever, that a stream's channel may be below sea-level at and near 
 its lower end. Thus the channel of the Mississippi is below sea- 
 level for some distance above the mouth of the stream and locally 
 as much as 100 feet below. The broad valley plain of the 
 Mississippi, on the other hand, is just above sea-level in the same 
 region. 
 
 Many conditions affect the rate at which a stream erodes, and 
 everything which affects the rate of erosion affects the length 
 of time which it will take a stream to bring the bottom of its valley 
 to base-level. Other things being equal, a large stream will bring 
 its valley to base-level sooner than a small one, and any stream 
 will bring its channel to base-level in weak rock sooner than in 
 resistant rock. 
 
 FIG. 114. Diagram of a valley, the top of which is ten times the width of 
 
 the stream. 
 
 The widening of valleys. If the growth of a valley were due 
 merely to the down-cutting of the stream, the valley would be no 
 
132 
 
 PHYSIOGRAPHY 
 
 wider than the stream which flows through it (Fig. 114, see also 
 Figs. 27 and 27a). Since most valleys are very much wider 
 than their streams, other factors besides down-cutting must be 
 involved in their development. 
 
 Most valleys are much wider at their tops than at their bottoms, 
 and all valleys are being made wider all the time. The widening 
 is brought about in many ways. Some of them are the following: 
 
 (1) Sometimes a stream flows against one side of its channel 
 with such force as to under-cut the slope above (PI. VIII and 
 
 FIG. 115. River under-cutting its bank and widening its valley by planation 
 where the material is unconsolidated sand, gravel, etc. 
 
 Figs. 115 and 116). The material under-cut is likely to fall or 
 slide down, and this makes the valley wider than before. Slow 
 streams widen their valleys more rapidly than swift ones, partly 
 because they are more easily turned against their banks by any 
 sort of obstacle in the channel. 
 
 (2) Again, some of the rain falling on the slopes of a valley 
 runs down to the bottom and is likely to carry mud, sand, and 
 coarser materials with it. This also widens the valley by slowly 
 wearing back its slopes. 
 
 (3) The loose earthy matter which lies on the slopes of a valley 
 is slowly creeping downward. The movement is brought about in 
 various ways, (a) If the material is clay, it contracts when dry, 
 
PLATE VI i 
 
 Streams disappearing in the sand, gravel, etc., at the 
 base of mountains in an arid region. Scale 4 
 miles per inch. (Paradise, Nev., Sheet, U. 3, 
 Geol. Surv.) 
 
PLATE VIII 
 
 
 IS MILES SOUTHWEST OF ST. LOUIS. MISSOURI. 
 
 A stream widening its valley by lateral planation. Scale 1 miles per inch. 
 (TJ. S. Geol. Surv.) 
 
THE WORK OF RUNNING WATER 
 
 133 
 
 and as it contracts, it cracks. The gaping of the crack is due 
 chiefly to the downward movement of the clay on the down-slope 
 side (A, Fig. 117). Cracking of the same sort (Fig. 118) may be 
 seen where a pool or pond has dried up, though creep is not in- 
 volved. When the cracked clay on a slope becomes wet again, 
 as by rain, the clay swells and the cracks are closed; but the 
 swelling takes place in such a way that the cracks are closed chiefly 
 by the moving down of the clay on the upper side of the crack 
 (C, Fig. 117), not by the moving up of the clay on the lower side. 
 
 FIG. 116. The Green River, Wyo., cutting against its bank and widening iu- 
 valley by planation where the material is indurated. (Fairbanks.) 
 
 This is because gravity helps to pull the clay down, while upward 
 movement, if it took place, would have to take place against 
 gravity. (6) Again, clayey material tends to become a viscous 
 fluid when wet, and in so far as it takes on fluidity, it tends to 
 creep or flow down-slope (p. 108). Friction, the roots of plants, 
 etc., on the other hand, tend to restrain its descent. All down- 
 ward movement of this sort tends to widen the valley, for much 
 or all the material descending in this way is carried off by the 
 stream when it reaches the bottom of the valley. 
 
 (4) When the loose material of the steep valley slope is thor- 
 oughly filled with water, as after a long rain or when snow is 
 
134 
 
 PHYSIOGRAPHY 
 
 melting, it may slide or slump from higher to lower levels (Fig. 119). 
 
 Slumping is common on steep valley slopes composed of uncon- 
 solidated material like clay. Slumping 
 widens the valley at the point whence it 
 starts. Material descending the slopes in 
 this as in other ways is sooner or later 
 carried away by the rivers. 
 
 (5) Every animal which walks over the 
 slope of a valley is likely to loosen more 
 or less material if the slope is steep, and 
 if this material is moved at all, it is likely 
 to be moved downward. Burrowing ani- 
 mals of all sorts loosen the surface ma- 
 terial and prepare it to be worked down 
 the slope readily. All these processes help 
 to widen the valley. 
 
 (6) Trees which grow on the sides of 
 valleys are sometimes overturned. When- 
 ever they fall, they disturb more or less 
 earthy matter, and some of it is likely to 
 roll down if the slopes are steep. If they 
 are not, the material loosened may be car- 
 ried down by slope-wash or by other means. 
 
 (7) Fine material on the slopes of val- 
 FIG. 117. Diagram to il- 
 lustrate the effects of leys may be blown away. 
 
 Various other processes are also in oper- 
 ation, helping to loosen rock or soil on the 
 
 cracking open. Tn B s i O p es anc j & [\ processes which loosen the 
 
 the process has gone . 
 
 b material m this position prepare it for de- 
 scent, and the descent or removal of mat- 
 ter from the slopes of a valley always 
 increases its width. All valleys, therefore, 
 are being widened all the time. In most 
 
 by the moving down of processe s of widening, the stream itself is 
 
 a rather than the mov- ... 
 
 ing up of b. an important factor, for it carries away 
 
 much of the material which descends the 
 
 slopes. Along the bases of the slopes of many valleys there is 
 much debris (talus) waiting to be carried away (Fig. 121). 
 
 Width-limit. As a result of all the processes which wear back 
 their slopes, adjacent valleys may be widened until the divide 
 
 drying and wetting on 
 a clay slope. In A the 
 clay is drying and 
 
 further, and it is 
 which has moved down, 
 while a remains where 
 it was in A. C repre- 
 sents the same after it 
 has been wet again and 
 the crack closed, chiefly 
 
THE WORK OF RUNNING WATER 
 
 135 
 
 between them is worn away (Figs. 122 and 122a). More commonly, 
 however, the divide between valleys becomes low without dis- 
 appearing altogether (Fig. 123). 
 
 FIG. 118. Sun-cracks in the flood plain of the Missouri. (Chamber lin.) 
 
 Valley flats. As already implied, streams, after they have 
 cut their channels down to low gradients, develop flats, or flood 
 
 1 
 
 FIG. 119. Slumping in the side of a valley, two miles southeast of Trout 
 Lake, near Telluride, Colo. (Hole.) 
 
 plains, in the bottoms of their valleys. These flats are always below 
 the level of the surface in which the valley lies. Thus the Missis- 
 
136 
 
 PHYSIOGRAPHY 
 
 sippi River at Dubuque has a flat between one and two miles wide, 
 about 300 feet below its surroundings, and about 600 feet above 
 sea-level. Near St. Louis the flat is 10 miles wide, about 150 
 
 FIG. 120. Slumping on the slope of Monte Crist o Cieek, Alaska. 
 (U. S. Geol. Surv.) 
 
 FIG. 121. Talus at base of valley slope, ready to be can led off by the 
 stream. Little Canyon looking south into Snake River. (U. S. 
 Geol. Surv.) 
 
 feet below its surroundings, and about 400 feet above sea-level. 
 At Memphis it is about 35 miles wide and but 220 feet above sea- 
 level. At Vicksburg it has a similar width and a height of but 
 
THE WORK OF RUNNING WATER 
 
 137 
 
 90 feet. Though increasing width of flat down-stream is char- 
 acteristic of valleys in general, it must not be understood that the 
 
 FIG 
 
 . 122. Diagram showing streams in adjacent valleys, under-cutting the 
 divide between them. They may, in time, destroy the divide by lateral 
 planation. 
 
 m 
 
 A 
 
 FIG. 122a. Diagram to show how the divide between streams may be 
 done away with by lateral planation. In A the stream at the left is 
 represented as under-cutting the divide between the two valleys. Later, 
 by shifting of its channel, the stream in the other valley might under- 
 cut the other slope of the divide, as shown in B. In C both streams are 
 represented as under-cutting the divide between them, and in D the 
 divide has been done away with. 
 
 increase of width is uniform. Narrower portions (often where the 
 rock is more resistant) often alternate with wider ones (often 
 where the rock is less resistant). 
 
138 
 
 PHYSIOGRAPHY 
 
 Combining these facts with a generalization previously made, 
 we may say (1) that rivers tend constantly to get the material of 
 the land into the sea; (2) that in working to this end they develop 
 
 FIG. 123. Diagram to illustrate the leveling of the surface by valley erosion. 
 The ground profile represented at the top shows two young valleys, 
 1 and 1, in an otherwise flat surface. In time these valleys will develop 
 the cross-sections represented by 2, 2, and later those represented by 3, 3, 
 4,4, etc. The divide between them may finally reach 5, when the 
 surface is nearly flat. 
 
 flats below the general level of the surface in which the valleys lie; 
 and (3) that these flats are, in general, wider and lower near the 
 sea, and narrower and higher far from it. Plates VIII to X and 
 Figs. 124 and 125 show valley flats in various sorts of regions. 
 
 FIG. 124. A valley flat in an early stage of development. Monte Cristo 
 Creek, Alaska. (U. S. Geol. Surv.) 
 
 Most valley flats are developed chiefly by the side-cutting of 
 the streams (PI. VIII) after they have become sluggish. The 
 streams which flow through flats generally meander, that is, they 
 have very winding courses (Pis. IX, X, and XI). 
 
 The valley flat is a sort of base-level, though the first flat developed 
 by a stream is not necessarily the lowest level to which it may 
 
PLATE IX 
 
 FIG. 1. A meandering stream. The Mis- 
 souri River. Scale 2 miles per inch. 
 (Marshall, Mo., Sheet, U. S. Geol. Surv.) 
 
 FIG. 2. A further stage in the 
 development of a meander. 
 The Schell River, Missouri. 
 Scale 2 miles per inch. 
 (Butler, Mo., Sheet, U. S. 
 Geol. Surv.) 
 
 FIG. 3. A plain in old age. Scale 2 miles per inch. (Abilene, Kan., Sheet, 
 
 U. S. Geol. Surv.) 
 
PLATE X 
 
 
 A well-developed river flat. Valley of the Mississippi, near Prairie du Chien, 
 Wis. Scale 2 miles per inch. (Waukon, la. Wis. Sheet, U. S. Geol 
 Surv.) 
 
PLATE XI 
 
 Stream flats. The Missouri and Big Sioux rivers. Scale 2 miles per inch 
 (Elk Point, S. Dak. -la. Neb. Sheet, U. S. Geol. Surv.) 
 
THE WORK OF RUNNING WATER 
 
 139 
 
 bring its valley bottom. It is the lowest level to which the stream 
 can bring its valley under the conditions which exist when the flat 
 is developed. It is therefore a temporary base-level, and serves as 
 
 FIG. 125. A wide valley flat. Milk River near Pendant d'Oreille, Canada. 
 
 (U. S. Geol. Surv.) 
 
 FIG. 126. Trout Creek, Yellowstone Park. (U. S. Geol. Surv.) 
 
 the limit below which tributary streams may not cut. Later, 
 under changed conditions, the stream may sink its channel well 
 below its first flat, and when this is done by a main stream, all its 
 tributaries may do the same. 
 
140 
 
 PHYSIOGRAPHY 
 
 The lengthening of valleys. Valleys are lengthened, too, 
 in various ways. Illustration of one way in which they are made 
 longer is furnished by the gullies developed on hillsides during 
 heavy rains. The gully made during one rain-storm is often 
 
 FIG. 12V. 
 
 FIG. 128. 
 
 FIG. 127. Two young valleys heading toward each other. 
 
 FIG. 128. Valleys of Fig. 127 developed headward until their respective 
 
 heads have met and the divide has been lowered a little at the point 
 
 of meeting. 
 
 lengthened at its upper end (headward) during the next, by the 
 water which flows in at its head. The process of lengthening 
 may sometimes be seen even during the progress of a single storm. 
 The heads of valleys often have the characteristics of ravines or 
 gullies. Valleys are, indeed, in some cases no more than growing 
 ravines which are working their heads inland, after the manner 
 of hillside gullies. 
 
 By this process the head of a valley may advance until a 
 permanent divide is established. Thus in Fig. 127 the heads of 
 the valleys, a and 6, may be worn back farther into the upland; 
 but when the heads of the valleys reach the points shown in Fig. 
 
 FIG. 129. Diagram to illustrate the lowering of a divide without shifting 
 it. The crest of the divide is at a, b, and c successively. If the erosion 
 was unequal on the two sides, the divide would be shifted. 
 
 128, neither can advance farther, if the rates of erosion are the 
 same on both sides of the divide. The divide is then permanent, 
 for though continued rainfall may lower it, it cannot shift its posi- 
 tion (Fig. 129). 
 
THE WORK OF RUNNING WATER 
 
 141 
 
 It is not to be understood that all valleys are being lengthened 
 at their heads in this way. Thus the head of the St. Lawrence 
 River is at the foot of Lake Ontario, and will remain there as long 
 as the lake shore remains where 
 it now is. 
 
 In its growth in length, the 
 head of one valley may reach 
 another valley, when the two 
 become one. This is illustrated 
 by Fig. 130. Streams are 
 sometimes lengthened at their 
 lower ends. This is the case 
 where the sediment which they 
 deposit at their debouchures 
 (lower ends) builds the land 
 out into the sea. The streams 
 then find their way across the 
 new-made land. Across such 
 lands the streams have chan- 
 nels, but never valleys of much 
 depth. There are various other 
 ways in which valleys become 
 longer, but they will not be 
 ' considered at this point. 
 
 FIG. 130. Diagram to illustrate one 
 mode of valley lengthening. In A 
 there are two small valleys, a and 
 b, and the former ends at the base 
 of the steep slope. In B the valley 
 b is represented as having been 
 lengthened so as to join a, and the 
 two have become one. 
 
 Summary. All valleys are 
 being made deeper in at least 
 some part of their courses all 
 the time; all valleys are being 
 
 made wider all the time; and some valleys are growing longer. All 
 streams sooner or later develop flats in their valleys, and these flats 
 may increase in width till the divides between them are worn away. 
 Where the divides between streams are not worn away by the 
 lateral planation of the streams, they may become so low as to 
 be inconspicuous. In either case the area affected becomes nearly 
 flat, at a level as low as running water can cut it. The land is 
 then base-leveled. 
 
 The History of a River System 
 
 Since valleys grow deeper, wider, and longer year by year, 
 they must formerly have been smaller than now. If, in imagination, 
 
142 
 
 PHYSIOGRAPHY 
 
 we trace them backward in their history, we may think of a time 
 when the large valleys of the present day were small, when the 
 small valleys were only ravines, when the ravines were only gullies, 
 and when the present gullies did not exist. Or, going still fur- 
 ther back, we may imagine a time when even the large valleys 
 had a beginning. 
 
 A principal method of valley birth and growth is illustrated 
 by the development of a gully. The rain-water which falls on the 
 
 FIG. 131. Gullies developing on easily eroded soil. Clear Lake, Cal. Every 
 shower will cause them to grow headward. (Fairbanks.) 
 
 surface tends to gather in such depressions as exist, and to flow 
 through them down the slopes. The water concentrated in the 
 depressions flows faster than that not so concentrated, and wears 
 the surface there more than elsewhere, and so starts a gully. The 
 gully started during one shower is made deeper, wider, and longer 
 by the next. Year by year, as the result of repeated show r ers and 
 repeated meltings of snows, the gully may grow to be a ravine, and 
 still later, by the same processes, it may become a valley. A 
 hillside gully is essentially like a river valley except in size, and 
 many valleys are but gullies grown big. 
 
 Not all gullies, however, become valleys, and not all valleys 
 
THE WORK OF RUNNING WATER 
 
 143 
 
 start as gullies. On a steep slope numerous gullies may start 
 (Figs. 107, 131, 132, and 133); but as they grow, some are so 
 widened as to take in others (Fig. 134), and the number is reduced. 
 
 FIG. 132. Gullies on slope above a valley flat. (Montana.) 
 
 Relatively few gullies become even ravines, fewer still become 
 small valleys, and very few ever attain great size. As valleys 
 
 FIG. 133. Surface much furrowed by the development of erosion gullies. 
 Montana. (George. ) 
 
 develop from gullies, the heads of some work back faster than others, 
 with the result that many valleys are arrested in their develop- 
 ment early, and so are dwarfed (Fig. 135). For example, g, 
 
144 
 
 PHYSIOGRAPHY 
 
 Fig. 135A, will grow in length little more, because the water 
 which falls on the land above its head flows off by some 
 
 other route to the sea. Later 
 stages in the development of 
 these valleys are illustrated 
 by Fig. 135, B and C. The 
 
 FIG. 134. Diagram illustrating how one contest among gullies and val- 
 gully takes another as a result of leys resu lting in the survival 
 lateral erosion. J ' 
 
 of a few and the exceptional 
 
 development of a very small number, may be called a struggle for 
 existence. 
 
 FIG. 135. Diagrams illustrating successive stages in the struggle for existence 
 and dominion among streams. 
 
 The courses of valleys. The headward growth of a gully is 
 due chiefly to the erosion of the water which flows into its upper 
 end. If the material about the upper end of a gully is of uniform 
 hardness, the head of the gully works back in the direction from 
 
THE WORK OF RUNNING WATER 145 
 
 which the greatest volume of water enters. Unevenness of the 
 surface about the head of the gully may concentrate the inflo\ving 
 water now at this point and now at that, as the head of the gully 
 advances. Consequently the head of the gully is rarely worn back 
 in a straight line. It turns to the right (b f , Fig. 136) where more 
 
 FIG. 136. Diagram to illustrate the direction of lengthening of a valley. 
 At 1 the valley is straight. If at this stage more water comes in from 
 the direction b than from the direction a, the wear is greater toward b 
 than toward a, and the head turns as shown in 2. If at this stage more 
 water comes in from the direction c than from any other direction, the 
 head turns in this direction, as shown in 3. 
 
 water comes in from that side, and to the left, 6", where there is 
 more inflow and wear on that side. 
 
 If the material about the head of the gully is less hard at one 
 point than at another, the head of the gully will work back on 
 the material which is most easily worn, even though the amount 
 of water flowing in from that direction is no greater than elsewhere. 
 Inequalities of slope or material, therefore, cause the gully's head 
 to turn now to one side and now to the other, and where the gully's 
 head goes, there the valley which develops from it follows, if the 
 gully reaches valley-hood. The crookedness of many valleys is 
 thus explained. 
 
 The permanent stream. Water commonly flows in a gully 
 only when it rains and when the snow is melting, and for a short 
 time afterward; but many valleys were developed from gullies, 
 and sooner or later most valleys have permanent streams. Where 
 does the water for the permanent stream come from? 
 
 The answer to this question may be readily inferred. When 
 a valley has been deepened so that its bottom is well below the 
 ground-water surface, the ground-water seeps or flows out into 
 the valley, and once in the valley in sufficient quantity, it becomes 
 
146 PHYSIOGRAPHY 
 
 a stream (Fig. 137). The valley whose cross-section is shown by 
 1, Fig. 137, would not have a stream; the valley whose cross-section 
 is represented by 2 would have a stream in wet weather, when 
 the ground- water level is at a; while the valley 3 would have a 
 permanent stream because it is well below the ground-water level, 
 b, of dry times. In regions where the ground-water surface is deep, 
 
 \2> 
 
 FIG. 137. Diagram showing ground-water surface: a the ground-water sur- 
 face at ordinary times, and b in times of drought. When a valley has 
 been cut below a there will be a stream in wet weather, but it will go 
 dry in time of drought. When the valley is down to 3 below the ground- 
 water surface of dry weather the stream will be permanent. 
 
 the valley must be deep to get a stream. In regions where the 
 ground- water surface is near the land surface, even shallow valleys 
 may have permanent streams. 
 
 Streams which are fed by lakes and streams which have their 
 sources in snow- and ice-fields which persist from year to year, 
 are not immediately dependent on ground-water, though they 
 often receive it. 
 
 Not all valleys are grown-up gullies. Not all valleys were 
 formed by the growth of gullies. A great area in the northern 
 part of North America, for example, was once covered by a great 
 sheet of snow and ice. When it finally melted, large parts of 
 the surface were left without well-defined valleys, but with 
 numerous lakes. The rainfall of the region was enough to make 
 many of these lakes overflow. When a lake overflows, the out- 
 going water follows the lowest line accessible to it, so long as there 
 is a line of descent. In this case, the running water will start to 
 cut a valley all the way from the lake which furnishes the water, 
 to the end of the stream, at the same time. No part of such a 
 valley is much older than another. Valleys developed in this way 
 may have permanent streams at the outset, since they are not 
 dependent on ground-water. The course of a valley developed 
 in this way was not determined by the direction in which the 
 head of the valley grew, but by the direction which the water took 
 at the outset, that is, by the course of the lowest descending slope. 
 
THE WORK OF RUNNING WATER 
 
 147 
 
 Growth of tributaries. Most valleys are joined by many 
 smaller tributary valleys. The reason may be easily understood 
 by the study of a gully. 
 
 If the slopes of a gully were worn back everywhere at the same 
 rate, tributaries would not develop; but the sides are rarely or 
 
 FIG. 138. Diagram showing tributaries in an early stage of development. 
 
 never worn back equally. Either the material is softer at some 
 places than at others, or the water flowing down the slopes is 
 concentrated along some lines more than along others. In either 
 case the erosion of the side sbpes is greater at some points than 
 at others, and where the erosion on the slope of a main gully is 
 greater than at adjacent points, a tributary gully is started (Fig. 
 138). Tributary gullies are therefore developed in the same way, 
 and for the same reason, as the larger ones from which they grow. 
 The tributary gully grows in length, width, and depth as its main 
 
 FIG. 139. Diagrammatic representation of a surface much dissected by the 
 development of numerous tributaries. 
 
 did, and in time it may become a valley and acquire a permanent 
 stream. Tributaries to the tributaries are developed in turn, until 
 a network of watercourses affects the surface. Figs. 139, 140, and 
 141 show a surface in this condition. A valley developed by 
 outflow from a lake develops tributaries in the same way as one 
 
148 
 
 PHYSIOGRAPHY 
 
 developed from a gully. Such a valley might also get tributaries 
 by the inflow of water from other lakes. 
 
 A valley and its tributaries constitute a valley system. A 
 stream and its tributaries constitute a drainage system, and the 
 area drained by a river system through a valley system is a 
 
 '- - 
 ^^ 
 
 FIG. 140. Photograph of the model of an area in northwestern Connecticut, 
 showing a surface much dissscted by erosion. (Model by Howell.) 
 
 drainage basin. From the conditions under which a valley system 
 develops, the outline of a drainage basin often comes to be rudely 
 pear-shaped (Fig. 142). 
 
 Stages in the history of a valley. We have seen that valleys 
 normally grow as they advance in years. When a valley is young, 
 it is narrow, and its slopes are steep. If the land is high, it has 
 a high gradient (unless far from the sea) and soon becomes deep. 
 
THE WORK OF RUNNING WATER 
 
 149 
 
 Its cross-section is then somewhat V-shapsd (Fig. 143), and its 
 tributaries are short. The mature valley is wider (Fig. 144), its 
 slopes are often gentler, and its tributaries are longer and older. 
 
 FIG 141. Contour map of the area shown in Fig. 104, representing the same 
 type of surface shown in Figs. 139 and 140. 
 
 An old valley is wide, has a broad flat or flood plain and a low 
 gradient. 
 
 A stream also, as well as its valley, passes from youth to ma- 
 
150 PHYSIOGRAPHY 
 
 turity, and from maturity to old age. In its youth it is likely 
 to be swift and impetuous, unless it flows through low land. 
 In maturity it is much steadier in its flow, and when it reaches 
 old age it meanders through its wide plain. Even an old 
 stream, however, may take on the vigor of youth when it is 
 flooded. 
 
 The terms youth, maturity, and old age are also applied to 
 river systems. Every river system, aided by weathering, has 
 
 FIG. 142. Map of the principal streams of southern New Jersey, and outlines 
 of their basins, shown in dotted lines. 
 
 entered upon the task of carrying to the sea all the land of its basin 
 which is above base-level. So long as the river system has the 
 larger part of its task before it, it is young (Fig. 1, PL XII). In 
 youth the land is often ill drained and may have many ponds 
 and lakes (PL III). When the main valleys have become wide 
 and deep, and the areas of upland have been well cut up (dissected) 
 
PLATE XII 
 
 . 1. Youthful valleys. Shore of Lake Michigan just north of Chicago. Scale 
 1 mile per inch. (Highwood Sheet, U. S. Geol. Surv.) 
 
 FIG. 2. A region in a mature stage of erosion. Scale 2 miles per inch. 
 (Kentucky, U. S. Geol. Surv.) 
 
The Niagara Gorge. Scale 1 mile per inch. (Niagara Falls Sheet, U. S. Geol. Surv.) 
 
THE WORK OF RUNNING WATER 
 
 151 
 
 FIG. 143. A young V-shaped valley, the Stehekin River, Wash. 
 (U. S. Geol. Surv.) 
 
 FIG. 144. A valley much older than that shown in Fig. 143, Gray Copper 
 Gulch, southwestern Colorado. (U. S. Geol. Surv.) 
 
152 PHYSIOGRAPHY 
 
 by valleys, the river system is said to have reached maturity (PI. 
 XII, Fig. 2). The land is then well drained. When the task of 
 base-leveling its drainage basin is nearing completion, the river 
 system has reached old age, Fig. 3 (PL IX). The master stream 
 of a drainage system attains the characteristics of maturity and 
 age sooner than its tributaries, and in its lower course sooner 
 than in its upper. 
 
 The topography of a drainage basin is youthful when its river 
 system is youthful, mature when its river system is mature, and 
 old when its drainage is old. In an area of youthful topography 
 much of the surface has not yet been much affected by erosion 
 (Fig. 1, PL XII); in an area of mature topography the surface 
 has been largely reduced to slopes by erosion (Fig. 2, PL XII); 
 while an area of old topography is one which has been brought 
 down to general flatness by erosion (Fig. 3, PL IX). Some 
 parts of a drainage basin, especially those parts near the 
 master stream, may take on the characteristics of age, while 
 other parts farther from the trunk stream may not be advanced 
 beyond maturity or even youth. 
 
 MAP EXERCISE 
 
 Topographic Maps Showing Erosion Topography in Various Stages of 
 
 Development 
 
 I. Study the following maps in preparation for the conference : 
 
 1. Emporia, Kan. 7. Fredonia, Kan. 
 
 2. Kanawha, W. Va. 8. Mt. Guyot, Tenn. N. C. 
 
 3. Prince Frederick, Md. 9. Casselton, N. D. 
 
 4. Ridgeway, N. Y. 10. Princeton, Ind. 111. 
 
 5. Canyon, Wyo. 11. Bright Angel, Ariz. 
 
 6. Yosemite, Cal. 
 
 II. Suggestions for the study of each map : 
 
 1. Does the map represent a plain, plateau, or mountain region? 
 
 If more than one of these great types appears, note location 
 of each. 
 
 2. What is the age of the topography, in terms of erosion, and how 
 
 is it shown? (If different parts of an area are in different 
 stages, note the fact.) 
 
THE WORK OF RUNNING WATER 153 
 
 3. What inferences may be made from the map as to the climate of 
 
 the region represented? The evidence on which the inference 
 is based? 
 
 Note. Does the evidence (1) merely suggest the inference, 
 or (2) make the inference probable, or (3) make the inference 
 altogether trustworthy? 
 
 4. Are there topographic features which cannot be accounted for 
 
 by the erosion of running water? If so, where? 
 
 Cycle of erosion. The time necessary for the development of 
 a base-level throughout a drainage basin is a cycle of erosion. 
 This period of time is very long. While land is high and the 
 streams swift, erosion is rapid; but the nearer the land approaches 
 base-level, the slower the processes of erosion. The last part 
 of the process of base-leveling is therefore the slowest of all. 
 
 Peneplains. It is doubtful whether any extensive land area 
 was ever worn down to a perfect base-level ; but great areas have 
 been worn down almost to that level. In such cases low hills or 
 
 FIG. 145. A peneplain near Camp Douglas, Wis. (Atwood.) 
 
 ridges remain between the valleys, and hard bodies of rock may 
 rise abruptly above the general level of the plain of degradation. 
 A region in this condition is called a peneplain (an almost-plain, 
 Fig. 145). It has a surface which has been brought nearly, but 
 not quite, to base-level. If conspicuous elevations of slight extent 
 remain above it, they are monadnocks. The name was derived 
 from Mount Monadnock (N. H.), because that mountain was formed 
 in this way. 
 
154 PHYSIOGRAPHY 
 
 Rate of Land Degradation 
 
 Since all lands are being cut down by running water, it is a 
 matter of interest to know how fast they are being brought low. 
 It is also of interest to know whether the lands are to be destroyed 
 altogether, and if so, how long they are to last. 
 
 It has been estimated recently by the United States Geological 
 Survey that the Mississippi River carries to the sea yearly about 
 .340,500,000 tons of sediment in suspension, and about 136,400,000 
 tons in solution. Of these amounts, the Ohio and the Missouri 
 contribute more than half, and the Missouri more than twice as 
 much as the Ohio. The Colorado River is estimated to remove 
 387 tons, on the average, from each square mile of its basin. The 
 same Survey estimates that the total amount of sediment carried 
 from the United States to the sea in a year is about 513,000,000 
 tons, and the amount in solution about 270,000,000 tons. On the 
 basis of these figures, it is estimated that the land is being degrad- 
 ed at the average rate of about 1 foot in 9,120 years. It is clear 
 that some parts of the country are being cut down much faster 
 than this, and other parts much more slowly. 
 
 If this rate were to be continued without interruption, and 
 if nothing occurred to counteract it, the North American conti- 
 nent would be reduced to sea-level in about 18,000,000 years, for 
 its average height is about 2000 feet. But as already pointed 
 out, however, the present rate of down-cutting cannot continue, 
 for as the land becomes lower the rate of erosion must diminish, 
 since the water must then move more slowly. As a matter of fact, 
 mechanical erosion by running water would cease when the sur- 
 face was brought to base-level, though solution would still go on. 
 
 Other changes, to be discussed later, are likely to occur to pre- 
 vent the land from being worn down to base-level. The continent 
 is therefore likely to endure not only much longer than 18,000,000 
 years, but probably indefinitely. Nevertheless, these figures serve 
 a useful purpose in indicating the rate of change which the land 
 is undergoing as the result of the fall of rain and snow upon it. 
 
 Conditions affecting the rate of erosion. Some of the condi- 
 tions affecting the rate of erosion by running water have been 
 stated or implied in the preceding pages. By way of summary 
 they may be brought together at this point. 
 
 The rate at which running water wears down the surface over 
 
THE WORK OF RUNNING WATER 
 
 155 
 
 which it flows depends largely on (1) the volume of the water, 
 (2) its velocity, (3) the character of the material over which it 
 flows, and (4) the amount and character of the load it carries. 
 
 (1) The volume of water flowing over the land outside of streams 
 depends chiefly on the rainfall. The volume of a stream depends 
 chiefly on (a) the area which it drains, and (6) the amount of pre- 
 cipitation within its basin. The larger the area and the greater 
 the amount of the precipitation the larger the stream. 
 
 (2) The velocity of running water depends on (a) its gradient or 
 slope, (6) its volume, especially its depth, (c) its load, and (d) the 
 shape and configuration of its channel. The higher the gradient, 
 the greater the volume, the less the load, and the smoother and 
 narrower its channel, the faster the flow. 
 
 The effect of slope on velocity needs no explanation. That 
 increase of volume increases the rate of flow is shown by the 
 familiar fact that a stream in flood runs faster than at other times. 
 The erosive force of a flooded stream has already been referred 
 to (Figs. 101 and 102). The carrying of sediment, in whatever 
 form, is a tax on the stream's energy, and the more the load the 
 greater the tax. The energy used in carrying is taken from the 
 energy which would otherwise be available for flowing. A smooth 
 channel offers less friction than a rough one, and so favors high 
 velocity. But, apart from smoothness, the channel which favors 
 
 great velocity is that which offers 
 least area of contact with the water. 
 Thus a broad shallow channel 
 (Fig. 146) has a greater surface of 
 contact with the water than a 
 deeper, narrower channel (Fig. 
 147). The water in the former has 
 nore friction with its bed, and fric- 
 tion retards the current. Nearly 
 FIG. 147. A deeper and nanower all streams which flow now in a 
 channel than that shown in Fig. narrow channel and now in a wide 
 146, with the same gradient. A , , . , 
 
 stream in a channel such as is one, have greater velocity where 
 
 FIG. 146.- 
 
 -A broad, shallow river 
 channel. 
 
 represented in Fig. 147 will 
 flow faster than one in such a 
 channel as that shown in Fig. 
 146. 
 
 their channels are narrowed. 
 
 (3) The character of the surface 
 
 of its basin, and especially the 
 character of the material in its channel, also influences the rate of 
 a stream's erosion. If the surface of the land on which the rain 
 
156 
 
 PHYSIOGRAPHY 
 
 FIG. 148. Oneonta Gorge, Canyon of the Columbia, Ore. (Fairbanks.) 
 
 FIG. 149. Grand Canyon of the Colorado. (Peabody.) 
 
THE WORK OF RUNNING WATER 157 
 
 falls is bare solid rock, the immediate run-off brings little sedi- 
 ment to the stream, and if the bed of the stream is bare solid rock, 
 the stream wears it less than if it is of mud or sand. 
 
 (4) To work most effectively, the stream must carry load (tools) 
 enough to enable it to cut rapidly, but not so much as to make 
 it flow so slowly that it cannot use its tools effectively. 
 
 Exceptional Features Developed by Erosion 
 
 Canyons and gorges. When valleys are so narrow and deep 
 as to be striking, they are called gorges or canyons. In general 
 canyons are larger than gorges, though there is no sharp distinction 
 between them. The sides of small gorges and young canyons 
 are sometimes nearly vertical (Fig. 148), but the sides of the large 
 canyons are rarely so (Fig. 149). The distinction between a canyon 
 and a valley which is not a canyon is not a very sharp one, and, in 
 regions where canyons abound, the term is often applied to all 
 valleys. 
 
 The Colorado canyon (Figs. 27, 27a, and 149) is the greatest 
 canyon known. Its maximum depth is about a mile, but where 
 it has this depth, it is often 8 to 10 miles wide from rim to rim, 
 though very narrow at the bottom. With a depth of one mile and 
 a width of 8 miles, the slope, if uniform, would have an angle of 
 less than 15. The cross-section of such a valley is shown in Fig. 
 
 150. But the slopes of the 
 canyon are not uniform, as 
 shown by Fig. 151. The ir- 
 
 FIG. 150. Diagram showing the proper- regularities of slope are caused 
 tions of a valley the width of which . i i 
 
 is eight times the depth. These are by inequalities in the hardness 
 approximately the proportions of the o f t h e roc k o f t h e canyon 
 Colorado Canyon. 
 
 walls. 
 
 The Yellowstone River also has a notable canyon 1000 feet or 
 so deep (Fig. 152 and Fig. 1 cf PI. IV). Its wiith is less in pro- 
 portion to its depth than that of the canyon of the Colorado. 
 
 Narrow valleys mean that the processes of valley deepening 
 have outrun the processes of valley widening. This in turn means 
 that the stream which made the gorge or the canyon was swift, 
 or that the processes of valley widening (p. 131) were slow, or 
 both. 
 
158 PHYSIOGRAPHY 
 
 Valleys are deepened rapidly when their gradients are high 
 and the streams strong. They are widened slowly when the climate 
 is arid so that there is little slope-wash, when the stream is so 
 swift that it does not meander, and when the material of the sides 
 is such that it will stand with steep slopes. Solid rock, for ex- 
 ample, will stand with steeper slopes than loose sand. We con- 
 clude that (1) great altitude, (2) arid climate, (3) strong streams, 
 and (4) a rock structure which will stand in steep slopes, favor 
 the development of valleys of the canyon type. In other words, 
 youthful valleys in plateaus and mountains are likely to be canyons, 
 
 ^^!M&'^^'-'^^ 
 
 FIG. 151. Cross-section of the Colorado Canyon. 
 (After Gilbert and Brigham.) 
 
 if climate and rock structure favor. The plateaus of the western 
 part of the United States furnish these conditions, and canyons 
 are there common. This is true not only of main streams but 
 of their tributaries as well. 
 
 A strong stream in an arid region is possible when the valley 
 is supplied with abundant water from a humid region above. 
 The Colorado River is an example. 
 
 Since gorges often occur in humid regions, it is clear that all 
 the conditions favoring the development of canyons need not 
 be present in order to develop gorges. Thus the Niagara River 
 has a gorge or canyon below the falls (PI. XIII). Here the 
 down-cutting is so rapid that the processes of valley widening 
 have not kept pace with it, in spite of the fact that the region is 
 humid. 
 
 The deeper canyons of the west constitute well-nigh impassable 
 barriers to travel athwart their courses, while their rivers rarely 
 serve the needs of commerce or irrigation. Considerations of 
 defence doubtless led the cliff-dwellers to make their homes in the 
 almost inaccessible canyon walls. 
 
 Canyons must ultimately develop into valleys of another 
 type, for the stream of the canyon will ultimately cut to 
 base-level. The valley will then cease to become deeper, but 
 
THE WORK OF RUNNING WATER 
 
 159 
 
 the processes of valley widening will still go on, and the 
 narrow valley will become wider and wider until it ceases to 
 be a canyon. 
 
 FIG. 152. The canyon of the Yellowstone below the falls. 
 Yellowstone National Park. 
 
 Bad lands. To a type of topography developed in early 
 maturity in certain high regions where the rock is but slightly, 
 though unequally, resistant, a special name, bad land, is some- 
 times given. Some idea of bad-land topography is gained from 
 
160 
 
 PHYSIOGRAPHY 
 
 FIG. 153. Bad-land topography north of Scott's Bluff, Neb. 
 (U. S. Geol. Surv.) 
 
 FIG. 154. Bad-land topography southwest foot of Mesa Verde, Colo. 
 (U. S. Geol. Surv.) 
 
THE WORK OF RUNNING WATER 
 
 161 
 
 Figs. 153 and 154. Bad-land topography is found in various 
 localities in the West, conspicuously in western Nebraska, in 
 Wyoming and the western parts of the Dakotas. The formations 
 here are often beds of sandstone or shale, alternating with un- 
 indurated beds of clay. Climatic factors are also concerned in 
 the development of bad-land topography. A semi-arid climate, 
 where the precipitation is much concentrated, seems to be most 
 favorable for its development. 
 
 Natural bridges. If a stream flowing over jointed rock has 
 falls, the conditions are sometimes afforded for the development of 
 
 fe^=^ 
 
 - J -eq 
 
 _^ 
 1 
 
 ~r 
 
 r H 
 
 ( 
 
 ^r 
 
 
 ^3 
 
 ? ?' <'v\ 
 
 ^T^ 
 
 2 
 
 ; 
 
 5d 
 
 '.r^.A/j 
 
 
 U^-U 
 
 FIG. 155. Diagram to illustrate the initial stage in the development of a 
 natural bridge. Longitudinal section at the left, cross-section at the 
 right. 
 
 an exceptional and striking scenic feature. If above a waterfall 
 there were an open joint in the bed of the stream (as at 6, Fig. 155), 
 some portion of the water would descend through it. After reach- 
 ing a lower level it might find or make a passage through 
 the rock to the river below the falls. If even a little water 
 
 FIG. 156. A stage later than that shown in Fig. 155. 
 
 takes such a course, the flow will enlarge its channel, making a 
 passageway from the joint through which the water descends to 
 the valley below the fall (bcde, Fig. 155). This passageway 
 may become large enough to accommodate all the water of the 
 river. In this case, the entire fall would be transferred from the 
 position which it previously occupied (/) to the position of the 
 enlarged joint (&). The fall would then recede. The under- 
 ground channel between the old falls and the new would be bridged 
 by rock (&/" and /", Fig. 156), making a natural bridge. A bridge 
 of this sort is now in process of development in Two Medicine 
 
162 
 
 PHYSIOGRAPHY 
 
 FIG. 157. A partially developed natural bridge in Two Medicine River, 
 Mont. (Whitney.) 
 
 FIG. 158. The Natural Bridge of Virginia. (U. S. Geol. Surv.) 
 
THE WORK OF RUNNING WATER 163 
 
 River in northwestern Montana (Fig. 157). Once in existence, 
 a natural bridge will slowly weather away. The natural bridge 
 near Lexington, Va. (Fig. 158), almost 200 feet above the stream 
 which flows beneath it, is believed to have been developed in this 
 way. It is not to be understood, however, that all natural bridges 
 have had this history (see p. 98). 
 
 Effects of Inequalities of Hardness of Rock 
 
 Rapids and falls. The bed of a stream is often steeper at 
 some point than at others (Fig. 159), and there the stream flows 
 more rapidly. In such a case as that illustrated by Fig. 159 the 
 quickened flow constitutes a rapid. If the water in a stream's 
 
 FIG. 159. Chandlar Rapids in river of the same name in Alaska. 
 (U. S. Geol. Surv.) 
 
 bed drops over a cliff, it makes a waterfall (Figs. 160, 161, and 162). 
 Between a waterfall and a rapid there are all gradations (Fig. 163). 
 Steep rapids are often called falls, and both are sometimes called 
 cascades. 
 
 Falls and rapids occur in many places and under many con- 
 ditions, but they are most common where the material of the valley 
 
164 
 
 PHYSIOGRAPHY 
 
 FIG. 160. Niagara Falls. (U. S. Geol. Surv.) 
 
 FIG. 161. The lower falls of the Yellowstone. 
 
THE WORK OF RUNNING WATER 
 
 165 
 
 FIG. 162. Twin Falls, Snake River. (U. S. Geol. Surv.) 
 
 FIG. 163. Rustic Falls. A succession of slight falls in the Yellow- 
 stone Park. (U. S. Geol. Surv.) 
 
166 
 
 PHYSIOGRAPHY 
 
 bottom is of unequal hardness. They are commonly located where 
 the river passes from more resistant rock to that which is less 
 resistant. 
 
 The falls and the rapids of many rivers add greatly to their 
 beauty, and sometimes enhance their value to mankind by afford- 
 ing abundant water-power. Niagara Falls affords about 4,000,000 
 horse-power, so much of which has been or seems likely to be 
 granted to manufacturing companies, that a movement has been 
 begun to "save the falls." The Falls of St. Anthony did much 
 to make Minneapolis the greatest flour manufacturing city of the 
 world. Some of the great manufacturing cities of New England 
 also grew up about low falls and rapids. Advantageous as falls 
 are as a source of power, they are enemies of navigation. The 
 falls (really rapids) of the Ohio necessitated the breaking of bulk 
 at that point, and so determined the location and early growth 
 of Louisville. A canal designed to evade the obstacle was com- 
 pleted in 1830. 
 
 Falls and rapids are undergoing constant change, although 
 the change is usually very slow. The falls of the Niagara are 
 receding up-stream, because the falling water undermines the 
 hard layer of rock over which it is precipitated (Fig. 164). As a 
 fall recedes, it generally becomes lower. In such cases it is clear 
 
 that the fall will disappear if it 
 recedes far enough. If the hard 
 rock over w r hich the water drops 
 be in the position shown in Fig. 
 165, the fall will not recede, 
 though it will become lower and 
 will disappear when the stream 
 cuts down to base-level where 
 the fall is. Rapids and falls are 
 therefore temporary features of 
 Like canyons, they 
 are marks of youth, for they 
 show that the stream is well above base-level. In time, all existing 
 rapids and waterfalls will disappear, for they can no longer exist 
 after rivers have reached base-level, the goal of every stream. 
 
 From a waterfall we may reason backward in time as well 
 as forward. If existing falls are to disappear, was there a time 
 before thev existed? 
 
 FIG. 164. Diagram illustrating the streams 
 conditions at Niagara. (Gilbert.) 
 
 " marks 
 
THE WORK OF RUNNING WATER 
 
 167 
 
 Suppose the material along the line followed by vigorous 
 drainage to be of unequal hardness. The less resistant part will be 
 
 FIG. 165. Diagram illustrating a condition where a fall will not recede. 
 
 worn more rapidly than the more resistant part farther up the 
 stream, with the result shown in Fig. 166. The continued wear 
 
 FIG. 166. Diagram illustrating the development of a fall where the hard 
 layer dips up-stream. 
 
 of the water in such a case will cause the rapids at a to become 
 more rapid, and the process of steepening the bed of the descend- 
 ing water will go on until the rapids become a fall. In this case, 
 the rapid.s and falls depend on inequalities of hardness discovered 
 by the stream in the excavation of its valley. This is perhaps the 
 commonest way in which falls and rapids originate. Falls originat- 
 ing in this way are developed gradually. Such falls may be called 
 subsequent falls, since they do not depend on the original shape 
 of the surface. 
 
 In other cases the surface run-off, in following its course to 
 the sea, may reach a cliff and plunge over it. In this case, the 
 steep descent of surface existed before the stream found it, and the 
 falls began when the river came. Since such falls result from the 
 irregularities of surface over which the river began to flow, they 
 may be called consequent falls. A good example of such a fall is 
 that of the Niagara, formed when the outflow from Lake Erie 
 reached and fell over a cliff on its way to Lake Ontario. Since 
 the fall began it has receded some seven miles. 
 
 Falls are formed in still other ways. A landslide or a lava- 
 flow may form a dam, over which the water falls or flows in rapids. 
 In such cases, especially the former, the dams, and therefore the 
 rapids and falls, are often temporary. 
 
168 
 
 PHYSIOGRAPHY 
 
 At the bottoms of falls pot-holes (Fig. 167) are sometimes de- 
 veloped. The start is made as a result of slight inequalities in the 
 
 FIG. 167. Pot-holes in granite. Upper Tuolumne River, Cal. 
 
 surface of the rock. The holes reach their conspicuous size as 
 the result of wear by stones kept in motion in them by the eddies of 
 the falling water. 
 
 FIG. 168. Diagram showing a narrow place in a valley where the stream 
 crosses a hard layer of rock. 
 
 Narrows. When a stream cuts through a bed of hard rock, 
 it not only develops rapids and falls, but the hard rock affects the 
 valley in other ways. The resistant rock weathers less rapidly 
 
THE WORK OF RUNNING WATER 
 
 169 
 
 than the weak rock, and hence the valley is narrower where the 
 rock is resistant than where it is weak. Such a constriction of 
 the valley is a narrows (Fig. 168) or a water-gap. The Delaware 
 Water Gap through the Kittatinny Mountain is a well-known 
 
 FIG. 169. The Lower Narrows of the Baraboo River, Wisconsin. 
 (At wood.) 
 
 example. The narrows of the Baraboo River in Wisconsin (Fig. 169) 
 is another good example. Unlike falls, narrows are not most 
 conspicuous in the youth of the stream, but at a later time, after 
 the valley has been much widened in the weak rock adjacent to 
 
 FIG. 170. Rock terraces, due to resistant layers of rock. 
 
 that which is resistant. Falls are common in horizontal or nearly 
 horizontal beds, but narrows are commonly developed in stratified 
 rock only where the beds are tilted. 
 
 Narrows sometimes serve as gateways through mountains, 
 and so control lines of travel and transportation. The narrows 
 of Wills Creek in Wills Mountain, Maryland, may serve as an 
 example. From Fort Cumberland (site of Cumberland), built by 
 the Ohio Company to guard the important passageway, Nemacolin's 
 Path and Washington's and Braddock's roads ran west through it, 
 
170 
 
 PHYSIOGRAPHY 
 
 and the Cumberland National Road and an important railway 
 now pass through it. 
 
 FIG. 171. A monadnock: a mass of igneous rock isolated by erosion and 
 remaining because of its superior hardness. Matteo Tepee, Wyo. 
 (Detroit Photo. Co.) 
 
 FIG. 172. Hogbacks, due to the erosion of tilted beds of unequal resistance. 
 The harder layers stand up as ridges and constitute the "hogbacks." 
 (Powell.) 
 
 Rock terraces. Again, if the hard layer through which a 
 stream cuts is horizontal, the resistant rock weathers less rapidly 
 
THE WORK OF RUNNING WATER 
 
 171 
 
 than the weaker rock above and below, giving rise to rock terraces, 
 as shown in Fig. 170. 
 
 Monadnocks, rock ridges, etc. Elsewhere than in valleys, 
 
 FIG. 173. A butte. A characteristic feature of the arid plateau region of 
 the West. The butte is really a monadnock. (U. S. Geol. Surv.) 
 
 too, rock of more than average resistance makes itself felt in the 
 topography, for rain-wash, wind, and most phases of weathering 
 
 FIG. 174. The Encnanted Mesa. A striking butte in New Mexico. The 
 name mesa is not commonly applied to elevations of such small summit 
 area. (R. T. Chamberlin.) 
 
 affect resistant rock less than weak rock. The result is that hard 
 rock often remains as hills, or even as mountains (monadnocks, 
 
172 
 
 PHYSIOGRAPHY 
 
 p. 153), after the weaker rock about it has been reduced nearly 
 to base-level. Fig. 171 is an example. An elongate narrow ridge, 
 
 FIG. 175. A canoe-shaped valley bordered by a > ridge formed by 
 the outcrop of a hard layer. (Willis.) 
 
 due to the isolation of a tilted layer of resistant rock, is sometimes 
 called a " hogback" (Fig. 172). * 
 
 FIG. 176. Diagram to illustrate the effects of erosion on a fold or anticline, 
 both ends of which dip down (or plunge). (Willis.) 
 
 In the West similar elevations are often called buttes (Figs. 171 
 and 174). A hard stratum of rock, such as a lava-bed, overlying 
 
THE WORK OF RUNNING WATER 
 
 173 
 
 less resistant formations, such as clay or soft shale, often gives 
 rise to buttes. If such an elevation has a considerable expanse of 
 surface at its top, it is a mesa (Fig. 28), though this term is also 
 applied to wide terraces, especially if high. 
 
 The elevations due to the isolation of outcrops of hard rock 
 by the removal of their less resistant surroundings often take on 
 peculiar forms, dependent on the structure of the rock. Elongate 
 ridges are common where the strata are folded. Where the tops 
 of the original folds were not horizontal, erosion gives the ridges 
 which result from the isolating of the outcrops of hard rock 
 peculiar forms, as shown in Figs. 175 and 176. Such forms are 
 not uncommon in the Appalachian Mountains. 
 
 Accidents to Streams 
 
 Drowning. Streams are subject to many accidents. If the land 
 through which they flow sinks so as to decrease their gradients, 
 
 FIG. 177. 
 
 FIG. 178. 
 
 FIG. 177. Chesapeake Bay and its surroundings. The bay is a drowned 
 river valley, and the lower ends of its tributary valleys are also drowned. 
 
 FIG. 178. The drainage of the region about Chesapeake Bay as it would 
 .have been but for drowning. 
 
 they flow less rapidly, or may even cease to flow. If the lower 
 end of a valley sinks below sea-level, sea-water enters and forms 
 
174 
 
 PHYSIOGRAPHY 
 
 an estuary. In such cases the lower end of the river and its valley 
 are drowned. If along a coast the streams end in bays, the inference 
 is that the coast has sunk, and that the rivers and valleys have been 
 drowned. The Atlantic coast between New York and the Carolinas is 
 a good example (Fig. 177). Delaware Bay, Chesapeake Bay, and 
 numerous other smaller bays mark the sites of drowned rivers. With- 
 out the drowning, the drainage of this region would be somewhat 
 as shown in Fig. 178. By comparison of these figures, it is apparent 
 that drowning has the effect of isolating the parts of a river system. 
 Rejuvenation. If the basin of an old stream is raised, so that 
 the gradient of the stream is increased, its velocity is increased, 
 and it again takes on the characters of youth. Such a stream is 
 said to be rejuvenated (Fig. 179). The rejuvenation of a stream 
 means the beginning of a new cycle of erosion, even though the 
 preceding one was incomplete. 
 
 If the old stream was meandering in its valley, as old streams 
 often do, the quickened stream cuts its meanders deeper. The 
 meanders are thus entrenched. Where a stream has entrenched 
 meanders, there is a strong presumption that it has been reju- 
 venated. Entrenched meanders are 
 :;hown by many streams. Plate 
 XIV shows the entrenched meanders 
 of the Conodoguinet River in Penn- 
 sylvania. Young valleys in the bot- 
 toms of old ones, and entrenched 
 meanders, are among the more com- 
 mon marks of a second cycle of 
 erosion. 
 
 Ponding. If a portion of the 
 stream's bed is warped upward, the 
 gradient above the point of up-warp 
 is lessened, the flow is checked, and 
 the stream widened. Streams above 
 such an obstruction are ponded, that 
 !is, the waters accumulate in a pond 
 FIG. 179. Diagram to illustrate or lake. If the up-warp is great 
 an ideal case of rejuvenation enoug h ft may completely dam the 
 as the result of uplift. The 
 black area at the bottom repre- stream. Streams are also some- 
 
 sents the sea. times ponded by lava-flows, by land- 
 
 slides, etc., and by darm made by man. The mill-ponds along 
 
PLATE XIV 
 
 "7" 
 
 Entrenched meanders. Scale 1 mile per inch. (Harrisburg, Pa., Sheet, U. S. 
 
 Geol. Surv.) 
 
THE WORK OF RUNNING WATER 
 
 175 
 
 numerous creeks are illustrations of streams ponded in the last of 
 these ways. If the ponded waters flow out over the dam, they will 
 ultimately cut it down. If the dam is sufficiently high, the water 
 
 FIG. 180. A rejuvenated stream, the Wisconsin River. The stream is large 
 and swift, and flows through a young, narrow gorge, cut in the plain 
 shown in Fig. 145. 
 
 may be forced out of its valley altogether and find a new course. 
 
 Piracy. One stream may steal another. One way in which 
 this is done is suggested by Figs. 181 and 182. The head of a 
 
 FIG. 181. 
 
 FIG. 182. 
 
 Diagrams to illustrate a phase of piracy. By the headward growth of a, 
 Fig. 181, it reaches 6, and carries off its upper waters, a, Fig. 182, is 
 a pirate; b, Fig. 182, has been diverted, and c has been beheaded. 
 
 valley as at a, Fig. 181, may work back until it reaches the channel 
 of another stream, such as b. It then carries off the water coming 
 down to b (Fig. 182). The stealing of one stream by another is 
 stream piracy. The stream which steals is a pirate. The stream 
 
176 
 
 PHYSIOGRAPHY 
 
 stolen is diverted, and the stream which has lost its upper waters 
 is beheaded. When a stream is diverted from a narrows or water- 
 gap, the latter becomes a wind-gap. Such gaps are common in most 
 mountain regions. 
 
 The numerous wind-gaps of the Blue Ridge Mountains figured 
 prominently in the westward movement and in the strategy of the 
 Virginia campaigns of the Civil War. Farther south, the Cumber- 
 land Gap afforded the early emigrant the most available route 
 across the mountains, and during the last quarter of the eighteenth 
 century probably more than 300,000 people passed through it to 
 settle in Kentucky and Tennessee. 
 
 THE 
 KITTATINNY 
 
 PLAIN 
 
 THE 
 
 5HENANDOAH 
 PLAIN 
 
 FIG. 183. FIG. 184. 
 
 The capture of the head of Beaverdam Creek by the Shenandoah River. 
 Virginia-West Virginia. (After Willis.) 
 
 Piracy has been much more common among rivers than is 
 generally known. In the Appalachian region, for example, where 
 the conditions for piracy have been favorable, there are few large 
 streams which have not either increased their waters by piracy, 
 or suffered loss by the piracy of others. Figs. 183 and 184 afford 
 one illustration. Piracy is favored by inequalities of hardness, 
 for the streams which do not cross hard rock deepen their channels 
 more readily than those which do. 
 
THE WORK OF RUNNING WATER 
 
 177 
 
 Consequent and Antecedent Streams 
 
 When streams develop on a land surface in harmony with its 
 slope, they are said to be consequent (Fig. 185), that is, consequent 
 on the slope of the surface in which they developed. After streams 
 
 FIG. 185. A consequent stream whose course is in harmony with that oi 
 the slope of the area it drains. 
 
 have established their courses, the land surface which they drain 
 may be warped or deformed, but the deformation may go on so 
 slowly that the streams are able to hold their courses established 
 before it began (Fig. 186). The streams then have courses which 
 they would not have taken had the deformation taken place 
 before they were established. Such streams, with courses ante- 
 dating the present general slope of the surface and out of harmony 
 
 FIG. 186. An antecedent stream. The stream and its valley are conceived 
 to have developed as consequent stream and valley. An up-warp athwart 
 the valley then took place, but so slowly that the stream cut down its bed 
 as fast as the up-warp raised it, so that the stream held its old course, 
 not now in harmony with the slope of the area drained. 
 
 with it, are antecedent streams. They may at the outset have 
 been consequent, but they ceased to be consequent when the 
 deformation took place. 
 
178 PHYSIOGRAPHY 
 
 MAP EXERCISE 
 
 Maps Showing the Topographic Effects of (1) Inequalities of Hardness, 
 (2) Piracy, (3) Cycles of Erosion, etc. 
 
 I. Study the following maps in preparation for the conference: 
 
 1. Niagara, N. Y. 7. Harpers Ferry, Va. W. Va. Md 
 
 2. London, Ky. 8. Monterey, Va. W. Va. 
 
 3. Charleston, W. Va. 9. Kaaterskill, X. Y. 
 
 4. Tuscumbia, Mo. 10. Pawpaw, Md. Pa. 
 
 5. Lancaster, Wis. la. 111. 11. Relay, Md. 
 
 6. Independence, Kan. 12. Fredericksburg, Va. Md. 
 
 II. Apply each of the following questions to each of the above maps: 
 
 1. The age of the topography in terms of erosion? Reasons? 
 
 Are different parts of the area represented in different stages 
 of topographic development? 
 
 2. Is more than one cycle of erosion shown? If so, the evidence? 
 
 3. Is there any indication of inequalities of hardness in the under- 
 
 lying rock? If so, what? 
 
 4. Is there anything in the topography to indicate the position of 
 
 the strata beneath the region? If so, what? 
 
 5. What inferences can be made from the map as to (1) the 
 
 climate, (2) the density of the population, and (3) the occu- 
 pations of the people? 
 
 Note. In making inferences from the topographic map, 
 the measure of certainty or uncertainty should be carefully 
 recognized. Some inferences may be certain, some almost 
 certain, some probable, some possible but not very probable, 
 etc. The student should note to which of these classes each 
 inference belongs. 
 
 III. Questions on individual maps: 
 Monterey and Charleston Sheets. 
 
 Compare and contrast (1) the topography and (2) the drainage 
 
 of the two areas represented by these sheets. 
 Lancaster Sheet. 
 
 1. Account for the depression which extends between Sageville 
 
 and Dubuque, in the southern part of the map. 
 
 2. What is the probable explanation of Sinsinawa and Sherrill 
 
 mounds, in the southern part of the region? 
 Kaaterskill Sheet. 
 
 1. Find a case of piracy shown on the map. 
 
 2. Is there any chance for future piracy in this region? 
 
THE WORK OF RUNNING WATER 179 
 
 3. The possible explanations of the steep slope in the eastern 
 
 part of the area? The probable explanation? 
 Pawpaw Sheet. 
 
 1. How are the great meanders of the Potomac to be accounted 
 
 for, in view of the fact that the valley is narrow? 
 Harpers Ferry Sheet. 
 
 1. What is the origin of Snickers Gap in the Blue Ridge? 
 London Sheet. 
 
 1. Account for the large depressions near Lincoln, in the south- 
 west, and elsewhere. 
 Relay Sheet. 
 
 1. What indication is there of change of level in this region? 
 Fredericksburg Sheet. 
 
 1. Account for the peculiar character of the lower portions of 
 Aquia Creek, Potomac Creek, Rappahannock River, etc. 
 
 DEPOSITION BY RUNNING WATEK 
 
 We have seen that rivers carry mud, sand, gravel, etc., from 
 land to sea, and that their goal is the degradation of the land 
 nearly to the level of the sea. We have also seen that rivers do 
 not always carry the sediment derived from the land directly 
 to the sea. It is often dropped for a time on the land, perhaps 
 to be picked up and carried on again when the conditions for its 
 transportation are more favorable. We have now to inquire into 
 (1) the causes which make running water drop some of its load, 
 temporarily, at least; (2) the places where the material is aban- 
 doned; (3) the topographic features developed by the deposition 
 of sediment; (4) the effect of deposition on the stream depositing 
 it; and (5) the advantages and disadvantages of stream deposition 
 to mankind. 
 
 Causes of Deposition 
 
 When running water drops its load, or any part of it, it is 
 generally because the current has lost something of its velocity. 
 We have already seen (p. 155) that gradient and volume are the 
 most important factors in determining the velocity of a small 
 stream. 
 
 1. Loss of velocity. The commonest cause of loss of velocity 
 is decrease of slope or gradient. Running water may lose velocity 
 (1) suddenly, as when it passes from a steep slope, whether of hill 
 
180 
 
 PHYSIOGRAPHY 
 
 or mountain, to a gentle one, or to a body of standing water, or 
 (2) slowly, as in descending a valley the gradient of which becomes 
 gradually less. We therefore look to the places where these changes 
 in velocity occur for the principal deposits of running water. 
 Streams also become slower wherever their channels become wider, 
 if volumes and gradient remain constant. 
 
 A less common cause of decrease of velocity in a stream is 
 
 FIG. 187 The lower end of the Mississippi, showing its distributaries. 
 
 (C. & G. Surv.) 
 
 decrease of volume. Streams generally increase in size with in- 
 creasing distance from their sources, but to this general rule there are 
 exceptions. (1) If a stream flows through a very dry region, it 
 may receive few tributaries and few springs. Evaporation, on 
 the other hand, is great, and some of the water may be absorbed 
 by the thirsty soil and rock through which it flows. This is 
 
THE WORK OF RUNNING WATER 181 
 
 especially the case if the ground-water surface (p. 85) of the 
 region is below the level of the stream. In a dry region there- 
 fore a stream may diminish as it flows, and may even disappear 
 altogether (Pis. VII and XV). (2) A stream sometimes breaks 
 up into several streams (Fig. 187). The volume of each is less 
 than that of the original stream. (3) Still again, many streams, 
 especially in semi-arid regions, have much of their water withdrawn 
 for purposes of irrigation. Many streams in the West are made 
 smaller in this way. (4) Streams decrease in volume as their floods 
 decline. 
 
 Increase of load makes running water flow more slowly; but 
 a stream which is increasing its load by its own action is an eroding 
 not a depositing stream. A stream may deposit coarse sediment 
 and pick up fine in its stead, but in this case the amount of fine 
 material w r hich it picks up is usually greater than the amount 
 of coarse which it leaves. Erosion is therefore greater than depo- 
 sition, and a stream which erodes more than it deposits is not a 
 depositing stream, as the term is commonly used. 
 
 2. Excess of load from tributaries. Tributary streams with 
 high gradients may bring to their mains more sediment than the 
 latter can carry away. This is an occasional cause of deposition 
 in the channel of the main stream, especially where mountain 
 torrents with high gradients join older streams which have reduced 
 their channels to much lower gradients. 
 
 Location of Alluvial Deposits and their Topographic Forms 
 
 The deposits made by running water are found principally 
 in those situations where the flow of the water is checked or 
 stopped. 
 
 i. At the bases of steep slopes. Every shower washes fine 
 sediment down the slopes of the hills, and much of it is left at their 
 bases. Fences in such situations are often buried, little by little, 
 by the mud thus lodged. Temporary streams, bred of showers, 
 sometimes flow down steep slopes, and are suddenly checked at 
 their bases. Such streams gather much debris in their headlong 
 courses down the slopes, but abandon it where their velocity is 
 suddenly checked. Thus, at the lower end of every new-made 
 gully on the hillside there is a mass of debris which was washed 
 out of the gully itself (Figs. 107 and 188). Material in such 
 
182 PHYSIOGRAPHY 
 
 positions accur ulates in the form of a partial cone, known as an 
 alluvial cone. Alluvial cones have much in common with cones 
 of talus; but in the former, gravity brings the material down by 
 the help of water, while in the latter gravity brings the material 
 down without the aid of water, or. with but little help from it. 
 Between talus cones and alluvial cones there are however all gra- 
 dations. 
 
 Conspicuous alluvial cones are rather more common in semi- 
 arid regions than elsewhere, if steep slopes are present; for in 
 such regions the rainfall is fitful, and the occasional heavy showers, 
 which give rise to temporary and powerful torrents, favor the de- 
 
 FIG. 188. An alluvial cone. (U. S. Geol. Surv.) 
 
 veloprnent of cones of great size. Talus cones often have great 
 development in the same regions. At the bases of the mountain 
 ranges in the Great Basin the talus and alluvial cones from the 
 mountains are sometimes 2000 or 3000 feet high. 
 
 An alluvial fan is the same as an alluvial cone, except that it 
 has a lower angle of slope. The term fan is indeed more appro- 
 priate than cone for most alluvial accumulations at the bases 
 of slopes. The lower angle of the fan may be due to the less 
 abrupt change of slope where it is developed, to the larger 
 quantity of water concerned in its deposition, to the smaller 
 amount of detritus, or to its greater fineness. Less change 
 of slope, more water, and less and finer material, all favor the 
 wider distribution of the sediment, and so the development of 
 fans rather than cones. Nearly all young rivers descending 
 
THE WORK OF RUNNING WATER 183 
 
 from mountains build fans where they leave the mountains. 
 Thus, the rivers descending from the Sierras to the great 
 valley of California build great fans at the base of the mountain 
 range. Most of the rivers descending from the Rockies to the 
 plains to the east do the same thing. The fans of streams descend- 
 ing from the mountains are often many miles across. The fan of 
 the Merced River in California, for example, has a radius of about 
 40 miles. 
 
 The fans made by neighboring streams often grow laterally until 
 they merge. The union of several such fans makes a compound 
 alluvial fan, or a piedmont alluvial plain (PI. XV). Such plains 
 exist at the bases of most considerable mountain ranges. The 
 depth of alluvial material in such situations is often scores and 
 sometimes hundreds of feet. 
 
 Alluvial cones and fans react on the course of the water which 
 makes them. The loose debris of the cones and fans is capable of 
 absorbing much water, and the water of even a considerable stream 
 may sink into its fan (PI. XV). Before it disappears, the stream 
 is often divided into several smaller ones. This is because the 
 sediment deposited by the stream in its channel makes the channel 
 too small to hold all the water. Some of it therefore runs over 
 (out of the channel) and makes a new channel for itself. The 
 deposits which clog the channel may be the result of (1) dimin- 
 ished slope, ana so diminished activity, or (2) diminished volume, 
 due to absorption of water. The distributaries thus formed, 
 being small, are likely to be slower than the stream from which 
 they sprang, and so more likely to choke themselves. They there- 
 fore give rise to other and smaller distributaries. Thus the water 
 of the main stream is likely to be spread about over its cone or fan, 
 and the stream sometimes disappears. 
 
 Aside from well-developed fans and cones there is much sedi- 
 ment at the bases of slopes which are not steep. In such positions, 
 however, the alluvium is often without distinct topographic form. 
 Such accumulations at the bases of slopes are almost as widespread 
 as the bases of slopes themselves. 
 
 Alluvial fans and piedmont alluvial plains are often valuable 
 for agricultural purposes. In some parts of California, for ex- 
 ample, the alluvial lands are so valuable that holdings are generally 
 small and highly improved. Even in semi-arid regions they are 
 often extensively cultivated, the water being supplied (1) by wells, 
 
184 
 
 PHYSIOGRAPHY 
 
 through which the debris of the fan is made to yield up the wate. 
 it has absorbed, or (2) by irrigation ditches, which connect 
 
 FIG. 189. A branching stream. Junction of the Cooper and Yukon rivers, 
 Alaska. Shows also bars, etc, (U. S. Geol. Surv.) 
 
 the stream farther up the valley, and lead the water out of its 
 natural channel over the fan or plain (Fig. 200). 
 
 Scale 
 
 FIG. 190. A braided river, Dawson Co.. Neb. (U. S. Geol. Surv.) 
 
 2. In valley bottoms. A stream which makes deposits in its 
 channel reduces the size of the channel. In time it may become 
 
PLATE XV 
 
 ./. / 
 
 fe* 
 
 ~x \ 
 
 i I 
 
 FARTS OF LOS ANCELES AND 
 
 SAN BERHASOINO COUNTIES, CALIFORNIA. 
 
 
 A. piedmont alluvial plain or compound alluvial fan in Southern California. Scale 1 
 mile per inch. (Cucamonga Sheet, U. S. Geol. Surv.) 
 
PLATE XVI 
 
 The alluvial plain of the Platte rivers in Nebraska. The South Platte 
 is braided and the North Platte shows bars. The map also shows 
 irrigating canals leading out from the river. Scale 2 miles per 
 inch. (Paxton Sheet, U. S. Geol. Surv.) 
 
THE WORK OF RUNNL\ 7 G WATER 
 
 185 
 
 too small to hold all the water. A part then breaks out, and 
 follows a new course in the valley flat. This process may be re- 
 peated again and again (Figs. 189 and 190). The diverging stream 
 may or may not return to the main. Those which do not return 
 are called distributary streams. This term is sometimes applied 
 also to all diverging streams, without reference to their return. 
 The breaking up of a stream into parts may go so far, especially 
 when the water is low, that there can hardly be said to be a main 
 channel. The stream then becomes a network of minor streams, 
 or a braided stream. The Platte River in Nebraska is an excellent 
 example (Fig. 190). This condition exists only at low water. 
 
 FIG. 191. Bars in river. The Yellowstone River, 34 miles south of 
 
 Livingston, Mont. 
 
 At high water the entire flat through which the minor streams 
 shown in Fig. 190 flow is covered by water, and becomes the bed 
 of a single river (PI. XVI). 
 
 Streams sometimes deposit sand-bars in their channels (Figs. 
 189 and 191), especially in low water, even when they do not be- 
 come braided. These bars are obstacles to navigation, and are a 
 constant source of embarrassment to river traffic in the low stages 
 of many navigable streams. The bars deposited in low water 
 are often swept away in times of flood, when the velocity of the 
 
186 
 
 PHYSIOGRAPHY 
 
 stream is greatly increased. Occasionally bars become more or 
 less permanent islands. If they become covered with forests 
 they are less easily eroded by the swift waters of floods, since the 
 roots have a strong protective influence. 
 
 The profiles of most valleys are curves, the curvature becom- 
 ing less and less steep as the lower end of the stream is approached 
 (Fig. 192). It therefore happens that as a stream descends its 
 
 FIG. 192. Profile of a normal valley. 
 
 valley it generally reaches a point where its reduced gradient so 
 diminishes its velocity that it must abandon some of its load. In 
 this way sediment is distributed for long distances along valley 
 bottoms. It is left in the channels of the streams and spread over 
 their flood plains, aggrading them and making them alluvial 
 plains. Deposition in a valley which has no flat tends to develop 
 one (Fig. 193). 
 
 Deposition on valley flats has but little effect on their topog- 
 
 V ::':/T 
 
 FIG. 193. Flat developed by aggradation diagrammatic. 
 
 raphy; but a few minor features deserve mention. Among them 
 are natural levees. This term is applied to the low ridges on stream 
 flats along the banks of the channel (Fig. 194). They are built 
 
 FIG. 194. Levees of the Mississippi in cross-section, four miles north of 
 Donaldsonville, La. Vertical scale X 50. The horizontal line repre- 
 sents sea-level. The bottom of the channel is far below sea-level at this 
 point. 
 
 in times of flood. At such times the current in the main channel 
 is swift; but as the water escapes its channel and spreads over the 
 adjacent flat, its velocity is checked promptly, because its depth 
 
THE WORK OF RUNNING WATER 
 
 187 
 
 suddenly becomes less. It must therefore abandon much of its 
 load then and there. Repeated deposition in this position gives 
 rise to the levees. Natural levees are sometimes high enough 
 and continuous enough to turn the courses of tributary streams. 
 This is well illustrated by the Yazoo River of Mississippi, which 
 flows some 200 miles in the flat of the Mississippi before being 
 able to join it. Near Vicksburg the Mississippi swings over 
 to the east side of the valley, and thus receives its tributary, 
 which the levees have shut off. The early population of Louisiana 
 
 FIG. 195. 
 
 FIG. 196. 
 
 FIG. 195. Diagram illustrating an early stage in the development of river 
 meanders. The dotted area represents the area over which the stream 
 has' worked. 
 
 FIG. 196. A later stage in the development of meanders. 
 
 and Mississippi was largely distributed in narrow belts along the 
 levees of the Mississippi and its tributaries and distributaries. 
 Here was the highest, driest land, of great fertility, fronting ready- 
 made highways. 
 
 Flood-plain meanders. A stream with an alluvial plain is 
 likely to meander widely (Pis. IX, X, and XI). In general terms 
 this may be said to be the result of low velocity, which allows 
 it to be turned aside easily. Were the course of such a stream 
 made straight, it would soon become crooked again. The manner 
 of change is illustrated by Figs. 195 and 196. If the banks be 
 
188 
 
 PHYSIOGRAPHY 
 
 less resistant at some points than at others, as is always the case, 
 the stream will cut in at those points. If the configuration of the 
 channel is such as to direct a current against a given point, 6 (Fig. 
 195), the result is the same, even without inequality of material. 
 Once a curve in the bank is started, it is increased by the current 
 which is directed into it. Furthermore, as the current issues from 
 the curve, it impinges against the opposite bank and develops a 
 
 curve at that point. The 
 water issuing from this curve 
 develops another, and so on. 
 Once started, the curves 
 or meanders tend to become 
 more and more pronounced 
 (Fig. 196). In the case rep- 
 resented by Fig. 1, PI. IX, 
 the narrow neck of land be- 
 tween curves is almost cut 
 through. When this is ac- 
 complished, the stream will 
 abandon its wide curve. A 
 later stage in the process 
 is shown in Fig. 2, PI. IX 
 (the Osage River near Schell, 
 Mo.). 
 
 When the stream has cut 
 off a meander, the abandoned 
 part of the channel often 
 remains unfilled with sedi- 
 ment. If it contains standing 
 
 ., , ,, water, as it often does, it be- 
 
 FIG. 197. Meanders and cut-offs in the 
 
 Mississippi Valley below Vicksburg. comes the Site of a lake (Fig. 
 The figure shows the migration of the 197)> Such lakes sometimes 
 meanders down-stream and their ten- 
 dency to increase in size. have the form of an ox-bow. 
 
 and so are called ox-bow lakes 
 (Pis. IX and X). They are also known as bayous. 
 
 In meandering, a stream sometimes reaches and undermines the 
 valley bluff, thus widening its valley. 
 
 By the shifting of their courses, as the result of deposition 
 and meandering, streams have affected human interests in many 
 ways. Villages built on the banks of a stream because of the 
 
THE WORK OF RUNNING WATER 
 
 189 
 
 river traffic which the situation favored have sometimes been 
 abandoned by changes in the stream's course. Such villages 
 usually decay when the stream has withdrawn its patronage. 
 Some have been destroyed, while others have been preserved at 
 great expense. Kaskaskia, the capital of Illinois until 1819, was 
 situated on the flood plain of the Mississippi. In 1881 a change in 
 the channel of the river converted the larger part of the village site 
 into an island, the last vestige of which was washed away in 1899. 
 Large sums have been expended by the National Government and 
 by the Chicago and Alton Railway to keep the Missouri in its course 
 
 FIG. 198. A cement-lined canal prepared for irrigation. Truckee-Carson 
 project, Nev. The cement lining prevents free seepage. (U. S. Geol. 
 Surv.) 
 
 at Glasgow, Mo. Again, streams are sometimes the boundaries 
 between counties, and even states. In such cases the shifting of 
 the stream would transfer territory from one state to another. To 
 prevent this, complicated legal devices and complicated definitions 
 of boundaries are sometimes resorted to. The case is still more 
 serious where a river forms an international boundary. Thus the 
 shifting of the Rio Grande makes that river an unsatisfactory 
 boundary between the United States and Mexico. 
 
190 
 
 PHYSIOGRAPHY 
 
 FlG. 199. An irrigating canal not cemented, before the water is turned in. 
 Salt River Valley, Ariz. (U. S. Geol. Surv.) 
 
 FIG. 200. A diversion dam where the water of the stream is raised and 
 turned into the canal. Truckee-Carson project, Nev. (U. S. Geol. Surv.) 
 
THE WORK OF RUNNING WATER 
 
 191 
 
 Fertility of alluvial plains. Alluvial plains are often very 
 fertile and are among the tracts most prized for agricultural pur- 
 poses. This was as true in ancient times as now, for the valleys 
 of the Nile, the Po, and of several of the rivers of southern Asia 
 were the garden spots of ancient civilizations. The frequent de- 
 posits of silt and mud on such plains continually renew the soil 
 and render it fertile. So strictly were the earlier civilizations 
 confined to valley plains that the period antedating 800 B.C. has 
 been called, with some propriety, the "fluvial period" of history. 
 In America, valleys have been sought out for habitation from the 
 
 FIG. 201. An irrigating canal filled with water. Salt River Valley, Ariz. 
 
 (U. S. Geol. Surv.) 
 
 earliest times. In Virginia and Maryland early settlements were 
 made in the valleys of the James and the Potomac ; and in Pennsyl- 
 vania, in the valleys of the Delaware, the Schuylkill, and the Sus- 
 quehanna. In New York the principal settlements were long 
 confined to the valleys of the Hudson and the Mohawk; and when 
 the early settlements of Massachusetts began to spread beyond 
 the coast, they occupied the Connecticut Valley. 
 
 Valley flats, as well as alluvial fans, are favorably situated for 
 irrigation. Figs. 198 and 199 show irrigation canals or large ditches, 
 Fig. 200 the beginning (head) of a canal, and Fig. 201 a canal filled 
 
192 
 
 PHYSIOGRAPHY 
 
 FIG. 202. Fields prepared for irrigation by methods of squares. Las Cruces, 
 N. M. (Photograph by Fairbanks.) 
 
 n,. '^?n' 
 /!5^A/'^?^^^r 
 
 --f^*' - o -~^r * * T 
 >"7 4 AK *' i D A K * 
 
 '-^^^ 
 
 
 \ r* 
 
 \ < 
 
 t^fi i 
 
 rv 
 
 X /rr-ftt/e area 
 
 . Projects untltr 
 <D cor.strvcfion. 
 
 (J) Projects approved 
 
 FIG. 203. Map showing irrigation projects completed and under construction; 
 spring, 1906. (Blanchard.) 
 
THE WORK OF RUNNING WATER 
 
 193 
 
 FIG. 204. A type of the arid lands of the West before irrigation. 
 (U. S. Geol. Surv.) 
 
 FIG. 205. The same type of land shown in Fig. 204, after irrigation. 
 River Valley, Ariz. (U. S. Geol. Surv.) 
 
 Salt 
 
194 
 
 PHYSIOGRAPHY 
 
 with water. Fig. 202 shows a field prepared by ditching for irrigation. 
 Water is drawn, as required, from the canals into the small ditches of 
 the field. Great progress has already been made in the utilization 
 of the arid lands in the western part of the United States. The 
 lands thus utilized are largely in valleys and on plains adja- 
 cent to mountains. The general distribution of the irrigated and 
 irrigable lands is shown in Fig. 203. The Government has under- 
 taken the construction of many reservoirs in favorable sites in 
 the mountains to hold the waters of the wet seasons, so that they 
 may be drawn out and used on the lands below during the growing 
 season. The sites selected for dams are usually narrow places in 
 the valleys (Fig. 206). 
 
 FIG. 206. Roosevelt dam site. Salt River project, Ariz. 
 (U. S. Geol. Surv.) 
 
 River floods. Alluvial plains are, however, not without their 
 drawbacks as agricultural regions, for the floods to which they 
 are subject are often disastrous both to life and to property. 
 
 Terrible illustrations are afforded by the valleys of many great 
 rivers. Thus in the spring of 1897 many thousand square miles 
 of the flood plain of the Lower Mississippi were covered with 
 water. It was estimated that 50,000 to 60,000 people suffered 
 serious loss. In 1881 and 1882 the floods of the same stream and 
 
THE WORK OF RUNNING WATER 
 
 195 
 
 of the Ohio are estimated to have caused a loss of $15,000,000 and 
 138 lives. The losses occasioned by the floods of the Ohio alone 
 were estimated at $10,000,000 in 1884, and at $40,000,000 in 1903. 
 There was a disastrous flood in the valley of the W abash and 
 another in the valley of the Susquehanna in 1904, each causing 
 the destruction of property to the extent of nearly $10,000,000. 
 
 FIG. 207. Diagram illustrating changes in the course of the Yellow River. 
 The shaded area represents the area subject to flooding by the main 
 stream and its tributaries. (Richthofen.) 
 
 Cities built on flood plains are also subject to great injury from 
 floods. An exceptional flood of the Passaic River (N. J.) in 1902 
 is estimated to have destroyed millions of dollars' worth of prop- 
 erty in the city of Paterson alone. 
 
 Disastrous floods occur from time to time in most great valleys. 
 In 1885 a heavy rainfall of about 24 inches of water over an area 
 of about 1000 square miles in the valley of the Ganges caused a 
 disastrous flood. The volume of the river was greatly swollen, and 
 
196 
 
 PHYSIOGRAPHY 
 
 the water rushed down the valley with terrible velocity, under- 
 mining the banks, cutting new channels in the valley plain, sweeping 
 
 FIG. 208. Delta of Lake St Clair. (Lake Survey Chart.) 
 
 FIG. 209. A general view of the lower part of the delta of the Mississippi. 
 
 away roads, ditches, bridges, aqueducts, retaining-walls, and even 
 villages. 
 
THE WORK OF RUNNING WATER 197 
 
 The most disastrous river floods recorded are those of the 
 Hoang-ho or Yellow River of China. Previous to 1892, this river 
 flowed into the Yellow Sea south of the Shan-tung promontory. In 
 that year it shifted its course when in flood, and formed a new 
 channel leading northwest into the Gulf of Pechili, 300 miles to 
 the north (Fig. 207). Such changes in a stream's course are of 
 much consequence to commerce. 
 
 The alluvial plains of some valleys are protected against flood by 
 levees, or dykes. In such cases the natural levees are built higher by 
 man, and the gaps in them are filled. They then protect the flat 
 outside in ordinary floods; but extraordinary floods sometimes 
 burst through the dykes, working great disaster. Some parts 
 of the rich flood plain of the Mississippi which are used for agri- 
 culture are so subject to flood that all buildings connected with 
 the farms are placed above the flat. 
 
 3. At debouchures. Where a swift stream flows into the sea or 
 a lake, its current is promptly checked and soon destroyed alto- 
 
 FIG. 210. Diagrammatic profile and section of an alluvial fan. 
 
 gether. Its load is accordingly dropped. If not washed away 
 by waves, etc., the deposits of river-borne sediment in such places 
 make deZtas (Figs. 208 and 209). 
 
 The delta has some points in common with the alluvial fan. 
 In both cases the principal deposit is concentrated at the point 
 where the velocity is suddenly checked. In the case of the delta, 
 however, the current is checked more completely, and the debris 
 accumulates (at the outset) below the surface of the standing 
 water. In form, the delta differs from the alluvial fan in that its 
 edge has a steep slope (compare Figs. 210 and 211). 
 
 Once a delta is started below water, deposition takes place 
 upon its surface, which may be built up to, and even above, the 
 water-level. That part of the delta above the surface of the water 
 in which it is built is like a flat alluvial fan. 
 
198 
 
 PHYSIOGRAPHY 
 
 Waves, currents, etc., may prevent the building of a delta, but 
 otherwise all sediment-bearing streams make deltas at their de- 
 bouchures. Deltas are sometimes built where one stream flows into 
 
 FIG. 211. Diagrammatic profile and section of a delta. 
 
 another (Fig. 212). This is especially the case where a swift, 
 debris-laden stream joins a slow one. Deltas built into rivers 
 are usually of slight extent. 
 
 Much land has been made by delta-building. Thus the Colorado 
 River has built a great delta many square miles (above water) in 
 
 FIG. 212. Delta of the Chelan River built into the Columbia River. Wash. 
 (Willis, U. S. Geol. Surv.) 
 
 area at the head of the Gulf of California (Fig. 213). The delta 
 has been built quite across the gulf near its upper end, shutting 
 off the head. In the arid climate of the region this shut-off head 
 has become a nearly dry basin, the lowest part of which is about 
 300 feet below sea-level. The Skagit River, in Washington, has 
 built out its delta so as to surround what were high islands in Puget 
 Sound, thus joining them to the mainland. The deltas of the 
 
THE WORK OF RUNNING WATER 
 
 199 
 
 Mississippi (Fig. 209), the Nile (Fig. 214), and the Hoang-ho rivers 
 are among the large and well-known deltas. The united delta of the 
 Ganges and Brahmaputra is also a great one, having an area (above 
 water) of some 50,000 square miles. The Po has built a delta 14 
 
 FIG. 213. Relief map of an area about the head of the Gulf of California, 
 showing the delta of the Colorado River, outlined in a general way by 
 dotted lines. (U. S. Reclamation Service.) 
 
 miles beyond the former port of Adria, which gave its name to the 
 Adriatic Sea. The Rhone River (France) has advanced its delta 
 (Fig. 215) some 15 miles in as many centuries. 
 
 The borders of a delta are often difficult of determination. 
 A delta is sometimes said to be limited up-stream by the point 
 where the distributaries begin to be given off. This definition is 
 
200 
 
 PHYSIOGRAPHY 
 
 convenient, but arbitrary. It is less definite, but perhaps truer, to 
 regard the up-stream border of the land reclaimed from the sea 
 
 Jfemjim 
 FIG. 214. The delta of the Nile. (Prestwich.) 
 
 or lake by the river deposits, as the head of the delta. This defini- 
 tion would in many cases make the areas of deltas much greater 
 than the other. On this basis, the head of the delta of the Missis- 
 sippi, for example, would be near the mouth of the Ohio. 
 
 FIG. 215. Delta of the Rhone River. (Prestwich.) 
 
 The effect of delta-building is to increase the area of the land; 
 but it is to be noted that the processes which lead to delta-building 
 
THE WORK OF RUNNING WATER 
 
 201 
 
 reduce the volume of the land-masses, even though they increase 
 their area. 
 
 The outline of some deltas is determined by the surroundings 
 in which they are built. When, for example, a delta is built into a 
 bay, the form of the bay-head determines the shape of the delta. 
 The normal form of a delta built on an open coast is somewhat 
 semicircular, though there is often a fringe of delta fingers which 
 together have some resemblance to the Greek letter J, which 
 gave these terminal deposits of streams their names. 
 
 The silting up of river mouths is sometimes disastrous to cities 
 whose commerce is based on river trade. Thus the silting up 
 
 FIG. 216. Delta of the Danube. (Prestwich.) 
 
 of the Tapti River was largely responsible for the decline of Surat, 
 once the leading commercial centre of India. Between 1797 and 
 1847 its population declined from 800,000 to 80,000. 
 
 The surfaces of deltas are usually nearly plane, and the streams 
 which cross them often give off distributaries, as the preceding 
 figures show. These distributaries are subject to great and sudden 
 changes of course, as well as to minor shiftings which are in prog- 
 ress all the time. These changes sometimes affect commerce in a 
 vital way. Thus the site of Kasimbazar, in India, described as 
 the "chief emporium of the Gangetic trade" early in the eighteenth 
 century, is now a swamp as a result of a sudden change in the 
 course of the Bhagirathi River (a distributary of the Ganges), on 
 the banks of which it stood. 
 
202 PHYSIOGRAPHY 
 
 Many deltas are cultivated, and some of them, like that of the 
 Hoang-ho, support dense populations. Delta lands are, however, 
 
 FIG. 217. Terraces on the Fraser River at Lilloet, B. C. 
 (Photograph by Calvin.) 
 
 subject to disastrous floods. It is estimated that the flood of the 
 Hoang-ho River in September, 1887, drowned at least a million 
 people who lived upon its delta, and caused the death of many 
 
 FIG. 218. Diagram to illustrate the development of river terraces. 
 
 more by disease and famine afterward. Many villages were com- 
 pletely destroyed, and hundreds more were temDorarilv submerged. 
 
THE WORK OF RUNNING WATER 203 
 
 Ill-defined alluvium. Alluvial deposits as a whole are wide- 
 spread. A large part of the surface of the land is covered with a 
 little alluvial material, though relatively small areas are deeply 
 covered. Alluvial material is so disposed as to tend to even up 
 slopes. Thus alluvial fans and cones tend to bring the steep slope 
 above and the gentle slope below into harmony. 
 
 Alluvial Terraces 
 
 When a river which has an alluvial flat is rejuvenated, the 
 stream sinks its channel below the level of the flat (Fig. 179). 
 The remnants of the old flood plain then constitute alluvial terraces 
 (Figs. 217 and 218). Such terraces are also formed in other ways. 
 Thus if a stream is temporarily supplied with an excess of load, 
 it aggrades its valley (Fig. 193). If, later, the source of the ex- 
 cessive load is removed, the stream sets to work to remove that 
 which was temporarily laid aside in its flood plain, even without 
 rejuvenation. The more conspicuous alluvial terraces arise in some 
 such ways. Many cities, such as Dubuque, la., Peoria, 111., and 
 Harrisburg, Pa., were begun on stream terraces, though they have 
 now spread above them. 
 
 MAP EXERCISES 
 
 Topographic Maps and Coast Survey Charts Showing Stream Deposition, 
 
 Terraces, etc. 
 
 List of Maps 
 
 1. Cucamonga, Cal. 6. Sacramento, Cal. 
 
 2. Marshall, Mo. 7. Savanna, la. 111. 
 
 3. Marseilles, 111. 8. Mississippi River Chart No. 14. 1 
 
 4. Donaldsonville, La. 9. Coast Survey Chart No. 19. 2 
 
 5. Tacoma, Wash. 
 
 Cucamonga Sheet. 
 
 1. Of what kind of material are the uplands probably composed? 
 
 2. Of what kind of material are the lowlands composed? 
 
 '. The Mississippi River charts can be had of the Mississippi River Commis- 
 sion, St. Louis. Catalogs and prices furnished on application to the Com- 
 mission. 
 
 2 The Coast Survey charts are published by the U. S. Coast and Geodetic 
 Survey, Washington, D. C. Catalogs and prices furnished on application to 
 the Director of that Survey. 
 
204 PHYSIOGRAPHY 
 
 3. Why do the contour lines in general curve out from the upland 
 
 along the drainage lines? 
 
 4. Why are the outward curves of the contours in some instances 
 
 notched backward toward the mountains along the immediate 
 drainage lines? 
 
 5. Why are the streams intermittent on the lowlands? 
 
 6. Explain the peculiar manner in which the streams of waste 
 
 divide. 
 
 7. What possible water-supply is there for the dense population of 
 
 the lowland? 
 Marshall Sheet. 
 
 1. How is the extent to which the Missouri River has shifted its 
 
 course in recent times shown? 
 
 2. What are the evidences that deposition is now in progress? 
 
 3. What probably determined the immediate location of Miami, 
 
 Dewitt, and Brunswick? 
 Marseilles Sheet. 
 
 1. Interpret the flats upon which the town of Seneca and the Black 
 
 Ash Swamp are situated. 
 Donaldsonville Sheet. 
 
 1. Explain the general distribution of the higher land. 
 
 2. What is the explanation of the alluvial deposits northeast of 
 
 Colomb Park? 
 
 3. Note the relation of the minor streams to the main stream. 
 
 4. Note the location of common roads, railroads, and settlements. 
 Tacoma Sheet. 
 
 1. What phase of river work is the Puyallup River now performing? 
 
 2. What was the probable origin of the low marshy ground at the 
 
 head of Commencement Bay? 
 
 3. Possible reasons why Wapato and Hylebo creeks flow independ- 
 
 ently to Commencement Bay instead of into Puyallup 
 
 River? 
 Sacramento Sheet. 
 
 1. What was the probable origin of the plain covering the western 
 
 half of this area? 
 Savanna Sheet. 
 
 1. Explain the lakes and swamps in the lowland. 
 
 2. What is the probable meaning of the abrupt slope east of Dyson 
 
 Lake? 
 
 3. Is there any discrepancy (in stage of development) between the 
 
 valley of the Mississippi and its tributaries? 
 
 4. Note the distribution of the common roads in the northern part 
 
 of the area, and suggest the explanation. 
 
THE WORK OF RUNNING WATER 205 
 
 Mississippi River Chart No. 14. 
 
 (Read carefully the note printed in red.) 
 
 1. What was the origin of Lake Chi cot? 
 
 2. What changes have taken place at Bachelor's Bend since 1883? 
 
 3. Can you find evidence on the map that the meanders of the Mis- 
 
 sissippi are working down the valley? 
 
 REFERENCES 
 
 " 1. CHAMBERLIN AND SALISBURY, Geologic Processes, Chapter III: Henry 
 Holt & Co., 1903, and other standard text-books on Geology. 
 
 2. RUSSELL, Rivers of North America: G. P. Putnam's Sons, 1898. 
 
 3. SHALER, Aspects of the Earth, Chapters III and IV. Chas. Scribner's 
 Sons, 1889. 
 
 4. GILBERT, Geology of the Henry Mountains, Chapter on Land Sculpture: 
 Government Publication, 1877; Chapter on Niagara Falls, in Physiography 
 of the United States: Am. Bk. Co., 1896; and Physical History of Niagara 
 River: Am. Geol., Vol. XXV11, pp. 375-377. 
 
 5. CAMPBELL, Drainage Modifications and their Interpretation: Jour, of 
 Geol., Vol. IV, 1896, pp. 567-581 and 657-678. 
 
 6. WILLIS, The Northern Appalachians, in Physiography of the United 
 States: Am. Bk. Co., 1896. 
 
 7. HAYES, The Southern Appalachians, in Physiography of the United 
 States: Am. Bk. Co., 1896; and Physiography of the Chattanooga District: 
 19th Ann. Rept. U. S. Geol. Surv., Pt. II, pp. 1-58. 
 
 8. DAVIS, The Seine, the Meuse, and the Moselle: Nat. Geog. Mag., Vol. 
 VII, pp. 181-202 and 228-238; Rivers and Valleys of Pennsylvania: Nat. 
 Geog. Mag., Vol. I, especially pp. 203-219; Stream Contest along the Blue 
 Ridge: Bull. Geog. Soc. of Philadelphia, Vol. HI, pp. 213-244; Geographic 
 Cycle in an Arid Climate: Jour. Geol., Vol. XIII, p. 381; and Development 
 of River Meanders: Geol. Mag., Vol. X, 1903. 
 
 9. DALY, Accordance of Summit Levels among Alpine Mountains: Jour. 
 Geol., Vol. XII, p. 105. 
 
 10. WALCOTT, Natural Bridge of Virginia: Nat. Geog. Mag., Vol. V, 
 1893, pp. 59-62. 
 
 11. CLELAND, North American Natural Biidges, in Pop. Sci. Mo., May, 
 1911, and Bull. Geol. Soc. Am., Vol. XXI, pp. 313-338. 
 
 12. MERRILL, Principles of Rock Weathering: Jour, of Geol., Vol. IV, 
 1896, pp. 704-724 and 850-872; also Rocks, Rock Weathering, and Soils, 
 Parts III, IV, and V: The Macmillan Co. 
 
 13. DUTTON, Tertiary History of the Grand Canyon District, especially 
 Chapter III: Mono. II, U. S. Geol. Surv., 1882. 
 
 14. JOHNSON (L. C.), The Nita Crevasse: Bull. Geol. Soc. of Am., Vol. 
 II, pp. 20-25, 1891. 
 
 15. GANNET, The Flood of April, 1897, in the Lower Mississippi. Scot. 
 Geog. Mag., Vol. XIII, 1897, p. 419. 
 
206 PHYSIOGRAPHY 
 
 16. JEFFERSON, Limiting Width of Meander Belts: Nat. Geog. Mag., 
 Vol. XIII, pp. 373-384. 
 
 17. TOWER, Development of Cut-off Meanders: Bull. Am. Geog. Soc., 
 Vol. XXXVI, pp. 589-599. 
 
 18. DAVIS, River Terraces in New England: Bull. Mus. Comp. Zool., 
 Geol. Ser., Vol. V, pp. 282-346, 1902. 
 
 19. DODGE, The Geographical Development of Alluvial River Terraces: 
 Proc. Boston Soc. Nat. Hist., Vol. XXVI, pp. 257-273. 
 
 20. All standard text-books on geology. 
 
CHAPTER V 
 THE WORK OF SNOW AND ICE 
 
 WE have seen that the atmosphere, the ground-water, and the 
 waters on the surface of the land bring about important changes 
 in its configuration. We are now to study the work of water in 
 its solid form. 
 
 Ice beneath the surface. The wedge- work of ice in the 
 crevices of rock has already been mentioned (p. 73). When 
 the great areas where water freezes during some part of the year 
 are considered, it appears that the aggregate effect of the freezing 
 of water in the pores and crevices must be great in long periods 
 of time. The water which freezes in the soil also has some effect 
 on the surface. This is shown by the disturbance of the walls of 
 buildings if their foundations do not go below the depth of freez- 
 ing, and by the working up of stones and bowlders through the 
 soil, etc. The frozen water in the soil makes it solid temporarily, 
 and so retards or prevents surface erosion, thus having a pro- 
 tective effect. The moisture rising from the soil, either by evap- 
 oration or by capillary action, sometimes freezes as it reaches 
 the surface. There may be continued additions, from below, 
 to the frost (ice) thus formed, resulting in the upward growth of 
 ice, as shown in Fig. 219. 
 
 Snow, the most familiar form of ice, is more wide-spread than 
 any other. Besides snow, the more familiar forms of ice appear 
 on lakes, rivers, and the seas of high latitudes, and on the lands 
 of high mountains or of high latitudes. 
 
 The ice of lakes. To understand the formation of ice on ponds 
 and lakes, we may follow the changes which take place as the cold 
 of winter comes on. 
 
 Fresh water is densest at a temperature of about 39 F. 
 (about 4 C.). The surface water of ponds and lakes in middle 
 
 207 
 
208 
 
 PHYSIOGRAPHY 
 
 latitudes is usually much warmer than 39 in summer. The 
 waters below the surface are generally cooler, but for some dis- 
 tance at least, and often to the bottom, they are warmer than 
 39 F. As the surface water cools in autumn and winter, it becomes 
 heavier than the warmer water beneath, and slowly sinks. This 
 process goes on, or tends to, until all the water from top to bottom 
 has a temperature of about 39. With further cooling the upper- 
 most water expands slightly, and so becomes lighter and remains 
 
 FIG. 219. Ice-crystals forming in the upper part of the soil grow by the 
 addition of moisture rising from below. The ice added below pushes 
 up the ice already formed. Columns of ice two or three inches in height 
 are formed in this way. often raising small stones. (Photo, by Roberts.) 
 
 at the surface. When cooled to 32 F. (0 C.), it freezes. In 
 freezing it expands about one-tenth of its volume. 
 
 Deep lakes in middle latitudes, such as the Great Lakes of the 
 United States, do not freeze over even in the coldest winters, for 
 the body of the water of such lakes is not cooled to 39 F., and so 
 long as their deeper parts have a temperature above 39, the sur- 
 face water sinks as it is cooled, and so does not reach the freezing 
 temperature. Such lakes, therefore, freeze over only about their 
 borders, where the water is so shallow that its temperature from 
 top to bottom is reduced to the temperature of greatest density. 
 Theoretically, this colder water near shore should spread to the 
 
THE WORK OF SNOW AND ICE 
 
 209 
 
 greater depths farther from shore; and it actually does move in 
 this direction whenever it is heavier than that of greater depths, 
 but the movement is often too slow to prevent the freezing. 
 
 FIG. 220. Ice crowding on shore. Lake Mendota, Wis. 
 (Buckley, Wis. Geol. Surv.) 
 
 FIG. 221. Shore of Wall Lake, Iowa. (Photo, by Calvin.) 
 
 Like most other solids, ice contracts as its temperature is re- 
 duced. If the temperature falls notably after the lake or pond is 
 frozen over, the ice, in contracting, pulls away from the shore 
 
210 
 
 PHYSIOGRAPHY 
 
 or cracks open, sometimes with loud, pistol-like reports. Water 
 rises between the ice and the shore, or into the opened cracks, 
 
 
 FIG. 222. A low terrace of gravel and sand formed by ice. Shore of Oco- 
 nomowoc Lake, Wis. (Fenneman, Wis. Geol. Surv.) 
 
 and freezes, and the ice-cover again fits the lake. When the cold 
 "spell" is over, the temperature of the ice rises, and the ice ex- 
 
 FIG. 223. The shove of ice on the shore of Lake Mendota, Wis. 
 (Photo, by Buckley.) 
 
 pands. The expanding ice may be crowded up on the shores, 
 especially if their slopes be very gentle (Fig. 220), or bowed up 
 
THE WORK OF SNOW AND ICE 
 
 211 
 
 away from the shores. In the former case, sand, gravel, and bowl- 
 ders frozen into the bottom of the ice are pushed up with it. Many 
 
 FIG. 224. Shove of shore ice where the shore is marshy. The ice of the 
 marsh is pushed up into ridges. (Buckley, Wis. Geol. Surv.) 
 
 Marsh 
 
 Lak 
 
 FIG. 225. Diagram representing the effects of ice-shove on a marsh adjoin- 
 ing a lake, and on a high steep bank. It is to be remembered that 
 the ground is frozen when the shove takes place, and therefore more 
 resistant than when not frozen. The thrust must therefore be strong 
 to produce the observed result. 
 
 "walled lakes" (Fig. 221), that is, lakes with accumulations of 
 bowlders resembling walls about their shores, owe their peculiar 
 
212 
 
 PHYSIOGRAPHY 
 
 features to the shove of shore ice. Low terraces along shores, and 
 low ridges are also made by the landward shove of the ice (Fig. 222). 
 When the shore of the lake is steep and of loose earthy matter, 
 the expanding ice sometimes crowds in under the soil, even over- 
 turning trees near the shore (Fig. 223). 
 
 The ice of a lake may be continuous laterally with the ice of 
 the soil (Fig. 224), and in this case the shove of the ice, on ex- 
 panding, may thrust up the frozen soil, making conspicuous ridges 
 of it. Fig. 224 shows a ridge formed in this way on the shores of 
 Lake Mendota near Madison, Wis., in the winter of 1898-9. The 
 ridge here is chiefly of the ice of the marsh which bordered the 
 lake at this point. 
 
 Ice on the sea. In high latitudes ice is formed along the sea- 
 shore. Unlike fresh water, sea-water condenses till it freezes, at a 
 temperature of 26 to 28 F. The variation in the temperature is 
 due to the varying salinity of the water. 
 
 The ice crystals formed from sea-water are individually with- 
 out salt; but a mass of ice formed from sea- water contains iri- 
 
 FIG. 226. Floe-ice on the shore of Greenland. 
 
 elusions of crystallized salt or of brine, excluded from the salt 
 water as it froze. If large quantities of such ice be melted, the 
 resulting water is more or less salty, though apart from these 
 inclusions the ice is fresh. In polar regions the sea ice attains a 
 depth of several feet, at least as much as eight or ten. Floating 
 ice of much greater thickness is sometimes seen, but it is doubtful 
 if these great thicknesses represent the ice formed by the freezing 
 
THE WORK OF SNOW AND ICE 213 
 
 of undisturbed sea-water. At any rate, the ice formed in winter 
 is often broken up in the summer inj;o floating pieces, floe-ice (Fig. 
 226), and the floe-ice is sometimes crowded together in ice-packs, 
 the separate pieces being so jammed together that some of them 
 are ended up and stand high above the water. If the ice-pack 
 of one summer is still far enough north at the end of the warm 
 season, it is frozen together, and its aggregate thickness, made up 
 as it is of blocks of ice some of which are on edge, is far beyond that 
 of normal sea ice. 
 
 Ice-foot. In high latitudes the snows along shore begin to 
 accumulate in the autumn, before the sea-water freezes, and the 
 water dashed up in storms freezes on and in the snow, converting 
 it into ice. The first sea ice may be forced up on the land some- 
 what above normal sea-level by waves and tides, and it is thickened 
 by the snow which lodges on it. In these ways the ice on the 
 shore sometimes becomes very thick, with its upper edge many 
 feet above sea-level. Such shore ice is known as an ice-foot. On 
 the ice-foot, rock fragments broken off from cliffs above often 
 gather in quantity. This protects the ice beneath from melting, 
 and remnants of it may endure through the summer. 
 
 Ice in rivers. Rivers also freeze over in cold climates, and 
 when the ice breaks up in the spring, stones and bowlders to which 
 it was frozen in the banks are sometimes floated miles down the 
 river. Not only are bowlders frozen into the ice floated away, 
 but huge pieces are occasionally torn from points of rock which 
 project into the river. At Montreal stone buildings 30 to 50 feet 
 square, projecting so as to have river ice form about them, have 
 been moved by the ice of the St. Lawrence. 
 
 When the river ice breaks up, masses of it may be carried 
 down-stream, and may accumulate in vast fields or "jams " behind 
 obstructions in the river. Where they are formed above bridges, 
 the bridges are likely to be swept away. The jams also occasion 
 disastrous floods above their sites, and when they break, the waters 
 accumulated above may sweep down the valleys with destructive 
 violence. 
 
 Northward-flowing rivers in the northern continents are espe- 
 cially subject to such floods. The snows of their upper basins 
 often melt while the lower parts of the streams are still frozen over. 
 The free discharge of the upper waters is thus prevented, and 
 freshets result. 
 
214 PHYSIOGRAPHY 
 
 When frozen over, many rivers of northern latitudes serve as 
 roadways. 
 
 Ground-ice. Tee sometimes forms on the bottoms of stony 
 rivers where the current is swift. It ultimately freezes around the 
 stones and bowlders on the bed, and when enough of it freezes to 
 them, they may be raised from the bottom and floated away. Ice 
 sometimes forms in quantities on the bottom (or below the sur- 
 face) of shallow seas, such as the Gulf of St. Lawrence and the 
 Baltic Sea. Ice thus formed is called ground-ice or sometimes 
 anchor-ice. Small vessels are said to be occasionally surrounded 
 and entrapped by the sudden appearance of large quantities of 
 this ice at the surface. 
 
 Ground-ice in rivers seems to be due (1) sometimes to the fact 
 that the bed of the stream is frozen, and the water in contact with 
 it freezes; and (2) sometimes to the fact that, though the tempera- 
 ture of the river as a whole is slightly below 32 F., the greater 
 motion of the upper and swifter part keeps the water there from 
 freezing, while the quieter water below congeals. 
 
 The cause of the development of ice on the bottom of shallow 
 seas is not clear. The suggestions made above as to the cause 
 of river-bottom ice do not seem applicable. It may be that fresh- 
 water springs issuing into sea-water which is below 32 F., but 
 above the freezing temperature of salt water, sometimes freeze 
 before being thoroughly mixed with the salt water. The ice formed 
 about anchors is probably sometimes the result of the low tempera- 
 ture of the anchor before it was lowered. Ice due to this cause 
 would not, however, endure long. 
 
 Snow. When the moisture in the air condenses at a tempera- 
 ture of less than 32 F., it commonly takes the form of snow- 
 flakes (Fig. 76). Snowflakes are not frozen rain-drops; they are 
 formed instead of rain-drops when the temperature at which the 
 water vapor condenses in the air is below the freezing-point. 
 
 Snow falls in high latitudes during much of the year, and in 
 middle latitudes during the winter season. Except on high moun- 
 tains, little falls in low latitudes, and the little that does fall gen- 
 erally melts quickly. The period of snowfall and the duration of 
 the period when snow lies on the surface increase both with in- 
 creasing altitude and latitude, so that above the polar circles 
 most of the precipitation falls as snow, and snow lies on most of 
 the land surface all the time, even at low levels. The same is 
 
THE WORK OF SNOW AND ICE 
 
 215 
 
 true in very high altitudes, even in tropical latitudes. In such 
 situations, indeed, the snowfall of the cold summer is often much 
 greater than that of the winter. 
 
 While snow lies on the surface, it serves to protect it. It shields 
 the vegetation beneath from excessive changes of temperature, 
 and especially from the repeated thawings (by day) and freezings 
 (by night) which are injurious to many plants, and it keeps the 
 dust and sand beneath from being blown about by the wind. The 
 conditions which preserve the snow also prevent the effective 
 wear of the surface by running water, so long as the snow is on the 
 ground. 
 
 FIG 227. Mt. Hood, a snow-capped mountain. (By permission of 
 Lipman, Wolfe & Co.) 
 
 Snow-fields. Where snow endures from year to year over any 
 considerable area, it constitutes a snow-field. Snow-fields are 
 widely distributed. Stated in general terms, they occur in moun- 
 tains in nearly all latitudes, but the altitude which is neces- 
 sary in the equatorial region is great (15,000 to 18,000 feet), that in 
 the temperate region less, and that in the polar regions slight. 
 In the polar regions, indeed, snow-fields occur even down to sea- 
 level. Stated in other terms, snow-fields occur in sufficiently high 
 altitudes in any latitude, and at any altitude in sufficiently high 
 latitudes. 
 
216 PHYSIOGRAPHY 
 
 Snow-fields are by no means rare in the United States. They 
 occur in the high mountains of California, Colorado, and Utah 
 (rare), and in the high mountains of all the states farther north 
 (Fig. 227). The snow-fields of the more northerly states are more 
 numerous and on the whole larger than those farther south. In 
 the mountains north of the United States they are still more ex- 
 tensive, and in Alaska some of them attain considerable size (Fig. 
 228). 
 
 Small snow-fields also occur in the high mountains of Mexico 
 and South America. They occur in the Alps, the Pyrenees, the 
 Caucasus, and the Scandinavian mountains of Europe, and in the 
 
 FIG. 228. Snow-fields in the Skolai Range of Alaska. Chisana glacier in 
 the foreground. (U. S. Geol. Surv.) 
 
 Himalayas and the higher mountains of the regions farther north 
 and northeast in Asia. They occur also in Africa, even very near 
 the equator, though they are small and limited to very high moun- 
 tains. Besides these and other small fields of snow and ice, there 
 are two great fields in Greenland and Antarctica. The snow- 
 and-ice field of Greenland contains much more snow and ice than 
 all the mountain snow-fields mentioned above, and that of Antarc- 
 tica contains probably several times as much as that of all other 
 fields together. 
 
 It is not improbable that there are as much as a million cubic 
 miles of snow and ice now on the land. If this amount of ice 
 were all melted and returned to the sea. it would raise its level 
 about 30 feet. 
 
 The snow-line. The line above which the snows of winter 
 are not all melted is the snow-line. The snow- line may fluctuate 
 
THE WORK OF SNOW AND ICE 217 
 
 somewhat from year to year, but during any given period, his- 
 torically speaking, it is relatively constant. Its average position 
 for a period of years should perhaps be regarded as the snow-line 
 for that period. 
 
 1. The position of the snow-line is influenced by temperature. 
 This is shown by the general fact that it is higher in lower (warmer) 
 latitudes and lower in higher (colder); but in various mountains, 
 the Himalayas for example, the snow-line is much higher on the 
 north side than on the south, although the temperature on the 
 south side is higher than that on the north. It is therefore evident 
 that something besides temperature is involved in the position o] the 
 snow-line. 
 
 2. An addition al factor is the amount of snowfall. The southerly 
 winds blowing over the Himalayas carry much more moisture than 
 the northerly ones. The result is that the fall of snow on the 
 southern slope is much heavier than that on the northern. The 
 same is true in the mountains of Switzerland. The position of the 
 snow-line is therefore influenced by the amount of snowfall as well 
 as by the temperature. Six inches of snow on the colder north 
 slope of mountains (northern hemisphere) may disappear in the 
 fewer melting days of summer, while as many feet of snow on the 
 warmer south slope may not disappear during the longer period of 
 melting in that position. 
 
 3. Again, snow does not disappear entirely by melting. Some 
 of it evaporates, and aridity favors evaporation. A snow-field 
 in a dry region is therefore wasted more by evaporation than 
 one in a humid region. Wind increases evaporation if the air is 
 dry. 
 
 4. Topographic relations also affect the position of the snow- 
 line in a given place, for some situations favor accumulation and 
 afford protection against the sun. 
 
 (1) Temperature and (2) amount of snowfall are the principal 
 factors which determine the position of the snow-line, and (3) 
 humidity (or aridity) and (4) topographic relations are minor 
 factors. Since these factors vary from place to place, no particu- 
 lar altitude in any particular latitude can be specified as the one 
 necessary for the existence of perennial snow. 
 
 The following table shows the position of the snow-line at a 
 few points: 
 
218 
 
 PHYSIOGRAPHY 
 
 Bolivian Andes, west side, 
 
 Bolivian Andes ; east side, 
 
 Chilean Andes, 
 
 Mexico 
 
 Teneriffe, 
 
 Himalayas, north side, 
 
 Himalayas, south side, 
 
 Caucasus Mountains, 
 
 Pyrenees Mountains, 
 
 Alps, 
 
 Norway, 
 
 Lapland, 
 
 Alaska, 
 
 Greenland. 
 
 Near equator. 
 Near equator, 
 Lat. 33 S., 
 
 Lat. 33 N., 
 Lat. about 28 N., 
 Lat. about 28 C N., 
 Lat. 40 + N., 
 Lat. 40 + N., 
 Lat. about 46 J N. 
 
 Lat. 70 N., 
 Lat. 60-70 N., 
 
 About 18,500 feet. 
 About 16,000 feet. 
 About 12,800 feet. 
 About 14,800 feet. 
 About 13,000 feet. 
 About 16,700 feet. 
 About 13,000 feet. 
 About 8300 to 14,000 feet 
 About 6500 feet. 
 About 9000 feet. 
 About 5000 feet. 
 About 3000 feet. 
 About 5500 feet. 
 About 2200 feet. 
 
 Ice-fields. Every considerable snow-field is also an ice-field, 
 for where snow accumulates to great depths and lies long upon the 
 surface, the greater part of it is converted into ice. The beginning 
 of this change is seen in the snow which has lain for a few days at 
 the surface. It loses its flaky character and becomes coarse-grained, 
 so that it is harsh to the touch. The change is still more con- 
 spicuous in the last banks of snow in the spring. The snow of such 
 banks is made up of coarse granules, often of considerable size. 
 The change of the flakes into granules is a process which is due, 
 in part, to the melting of the surface snow and the sinking and 
 re-freezing of the water below the surface; but since the change 
 appears to go on even where there is no melting, melting and re- 
 freezing are probably only a part of the process of change. 
 
 While this transformation is going on, the snow becomes more 
 compact. As it lies on the surface, its own weight tends to com- 
 press it. The sinking water which re-freezes below the surface 
 tends to bind the granules to one another, and as a result of the 
 compression, of the transformation of the flakes into granules, and 
 of the binding together of the granules themselves by the freezing 
 of water between them, the whole mass tends to become solid. 
 Just how solid and how dense snow must become before it is to be 
 called ice, cannot be stated; but every great snow-field is really 
 an ice-field, scarcely more than frosted over with snow. The 
 last snow-banks of spring are often essentially ice. 
 
THE WORK OF SNOW AND ICE 219 
 
 GLACIERS. 
 
 If the body of ice developed from snow becomes great enough, 
 it begins to spread or creep out from its place of accumulation. 
 Ice thus moving is glacier ice. Not all snow-fields give origin to 
 glaciers, but nearly all glaciers have their sources in snow-fields. 
 The distribution of glaciers is therefore much the same as the dis- 
 tribution of snow-fields. 
 
 Types of glaciers. Glaciers assume various shapes, depending 
 chiefly on the amount of ice and on the configuration of the sur- 
 face on which it lies. If the originating snow-field lies on the 
 slope of a mountain, the ice moves down the slope; and if a valley 
 
 FlG. 229. Summit of the Nizina-Tanana glacier, Alaska. 
 (Rohn, U. S. Geol Surv.) 
 
 leads out from the area of the snow-field, the movement of the ice 
 is chiefly concentrated in the valley. If the ice lies on a flat sur- 
 face, it spreads in all directions from its centre. 
 
 Glaciers which occupy valleys are called valley glaciers. In 
 common speech, "a glacier" is usually understood to be a valley 
 glacier. All valley glaciers are sometimes called alpine glaciers, 
 because they belong to the same general class as those of the Alps ; 
 but the valley glaciers of high latitudes differ in some ways, espe- 
 cially in their steeper sides and ends, from those of the Alps. For 
 this reason, valley glaciers may be classed as alpine (Fig. 230) 
 and high-latitude (Fig. 231) glaciers. 
 
 In high latitudes glacier ice sometimes lies on plains or plateaus. 
 In such positions glaciers may be nearly circular in outline, and may 
 
220 
 
 PHYSIOGRAPHY 
 
 FIG. 230 The Rhone glacier. (Photo, by Reid.) 
 
 FIG. 231. The end of the Bryant glacier, a high-latitude glacier of North 
 Greenland. (Photo, by Chamberlin.) 
 
THE WORK OF SNOW AND ICE 
 
 221 
 
 spread radially from their centres. Such glaciers are ice-caps or 
 ice-sheets. Ice-caps may be large or small. The main ice-caps 
 of Antarctica and Greenland (Figs. 252 and 259) are large, but 
 
 30 Miles 
 
 FIG. 232. A small ice-cap in the northwestern part of Iceland. 
 (After Thoraddsen.) 
 
 FIG. 233. A cliff glacier, coast of North Greenland. The height of the 
 cliff is perhaps 2000 feet. The water in the foreground is the sea. 
 
 small ones of the same type are found on various promontories 
 along the coast of Greenland, on Iceland (Fig. 232), and on some 
 Arctic islands. 
 
222 PHYSIOGRAPHY 
 
 Glaciers sometimes occur at the bases of mountains, being 
 formed by the union of the spreading ends of valley glaciers. 
 Such glaciers are piedmont glaciers (Fig. 260). Again, many snow- 
 fields nestled in the depressions of mountain cliffs give origin to 
 small glaciers which never descend to a valley. Such ill-formed 
 and poorly developed glaciers are cliff glaciers (Fig. 233). Cliff 
 glaciers grade into valley glaciers (Figs. 234 and 235). Ice broken 
 
 FIG. 234. Glaciers intermediate in type between a cliff glacier and a valley 
 glacier. Cascade Mts., Wash. (Willis, U. S. Geol. Surv.) 
 
 off from ice-sheets, or from valley glaciers which reach a cliff, may 
 accumulate below, freeze together again, and assume movement. 
 Such a glacier is sometimes called a reconstructed glacier. 
 
 Of these types, valley glaciers are most common and familiar, 
 but ice-caps contain far more ice. The leading characteristics of 
 glaciers may be studied in connection with the most familiar 
 form. 
 
THE WORK OF SNOW AND ICE 
 
 223 
 
 The Valley Glacier 
 
 The general form of a valley glacier (Fig. 236) is determined 
 by the shape of the valley in which it lies. If the valley is crooked, 
 the glacier curves to match it, and if the bottom of the valley is 
 uneven, the surface of the ice is more or less uneven, in keeping 
 with it. The valley glacier has sometimes been called a "river of 
 
 FIG. 235. Dana glacier, Mt. Dana, Cal. A glacier of the same type as 
 that shown in Fig. 234. 
 
 ice," but the differences between a glacier and a river are so much 
 greater than their likenesses that this definition is misleading. 
 
 The surface. The upper end of a valley glacier is in the snow- 
 field, and is covered with snow all the time, while the lower end 
 may be covered during the winter. Some glaciers carry so much 
 stony and earthy debris on their surfaces as to conceal the ice in 
 some places, especially near the lower end. 
 
 The centre of a valley glacier is usually higher than its sides, 
 so that its upper surface is generally somewhat convex in cross- 
 section. The profile of the surface of a glacier corresponds some- 
 what to the profile of the bottom of the valley in which it lies 
 (Figs. 230 and 236), as already noted, but its slope is sometimes 
 
224 
 
 PHYSIOGRAPHY 
 
 notably increased near its lower end, because of the steep slope of 
 
 the upper surface of the ice. 
 
 The surface of the glacier 
 is often uneven. In many 
 cases it is cracked, and the 
 cracks or crevasses frequently 
 gape. A principal cause of 
 the crevasses is the movement 
 of the brittle ice over an un- 
 even bed (Fig. 238). When 
 the slope of a glacier bed in- 
 creases suddenly, an ice cas- 
 cade is developed (Figs. 230 
 and 237) ; but an ice cascade 
 has little in common with the 
 rapids or falls of rivers. 
 Crevasses formed by the pas- 
 sage of ice over a steep place 
 in its bed are usually trans- 
 verse to the glacier. Cre- 
 vasses are sometimes parallel 
 to the sides of the glacier and 
 oblique to them, and such 
 crevasses are due to other 
 
 FIG. 236. Aletsch glacier, Switzerland, causes. The breaking of the 
 
 ice as it moves is one of the 
 
 many features wherein a glacier differs from a river. 
 
 PIG. 237. Diagrammatic and longitudinal sections of glaciers. (After Heim.) 
 
 As the ice moves forward, the crevasses sometimes tend to 
 close, though they rarely heal in such a way as to leave the surface 
 
THE WORK OF SNOW AND ICE 
 
 225 
 
 of the ice smooth. So long as a crevasse is open, the sun's rays 
 and the sun-warmed air enter it and melt the ice. The effect of 
 the melting is to widen the crevasse, especially its upper part. 
 
 FIG. 238. Crevassed glacier, the cracking due to change in grade of bed. 
 
 North Greenland. 
 
 The result is that when the movement tends to close the crevasse, 
 the opposing faces rarely fit together. This is illustrated by Figs. 
 
 FIG. 239. Crevassing in the upper part of a glacier on Mt. Hood, Ore. 
 
 (Meyers.) 
 
 240 and 241. The crevassing and the subsequent melting are 
 therefore a cause of unevenness of surface. 
 
226 
 
 PHYSIOGRAPHY 
 
 Another cause of surface irregularity is the drainage from the 
 surface. The valley glacier often extends far below the snow 
 line, and is within the region of active melting during the summer 
 season. Some of the surface water sinks beneath the surface, 
 
 FIG. 240. Diagram to illustrate one reason why ice crevasses fail to heal 
 as explained in text. 
 
 but some of it runs in little streams on the ice until it reaches a 
 crevasse or the edge of the glacier. These surface streams wear 
 notable channels (valleys) in the ice (Fig. 242), which, though 
 rarely deep, help to destroy the smoothness of the surface. 
 
 The stony and earthy debris which many valley glaciers carry 
 on their surfaces also gives rise to irregularities. The large stones 
 
 FIG. 241. Seracs of glaciers. (Photo, by Reid.) 
 
 protect the ice beneath from melting, and therefore come to stand 
 on pedestals of ice, as the unprotected ice about them is melted 
 away. Considerable aggregations of debris of any sort have the 
 same effect, by protecting the ice beneath from melting, thus 
 giving rise to mounds or ridges of ice covered with debris 
 
THE WORK OF SNOW AND ICE 
 
 227 
 
 (Fig. 243). Small or thin stones on the surface of the ice affect its 
 
 topography in the opposite way. 
 Rock absorbs heat better than the 
 ice does, and thin pieces of rock 
 are warmed through. They may 
 then melt their way into the ice 
 more rapidly than the sun melts 
 down the surface about them, thus 
 making depressions in the ice. 
 Patches of dust blown on the ice 
 have the same effect. The de- 
 pressions to which they give rise are 
 known as "dust-wells" (Fig. 245). 
 Dust-wells are sometimes so close 
 together that one must watch his 
 
 FIG. 242. Valley of a superglaciai steps in walking over the ice. Their 
 
 depths depend upon their diameters 
 and the angle of the sun's rays (Fig. 
 
 246). Their bottoms do not descend below the plane where the 
 
 sun's rays strike the heat-absorbing sediment. Dust-wells are 
 
 stream in the Bighorn Mts. 
 (Photo, by Blackwelder.) 
 
 FIG. 243. Bowlders on ice pinnacles. Forno glacier, Switzerland. 
 (Photo, by Reid.) 
 
 usually full of water at the end of a warm (melting temperature) 
 
228 
 
 PHYSIOGRAPHY 
 
 FIG. 2436. 
 
 FIG. 244. 
 
 FIG. 243a. Ice columns capped by slabs of rock, on Parker Creek glacier, 
 
 California. (U. S. Gepl. Surv.) 
 FIG. 243Z>. An ice pyramid on Mt. Lyell glacier, California. The protecting 
 
 stone has fallen from the column, which has since melted into the 
 
 pyramidal form. (~U. S. Geol. Surv.) 
 FIG. 244. Diagram to show how debris on ice gives rise to prominences. 
 
 (Gilbert.) 
 
 
 I * 
 
 I 
 
 *; ,* y-f* 
 
 .. -* 
 
 ^ * ^* * * . - 
 
 * . . .' , :- ! . 
 
 /** 
 
 %-r/ : 
 ^ 
 
 -.. 
 
 FIG. 245. Dust-wells, North Greenland. (Photo, by Chamberlin.) 
 
THE WORK OF SNOW AND ICE 
 
 229 
 
 day, but the water usually drains out at night. This drainage 
 shows that the glacier ice is, on the whole, very leaky. Depressions 
 resembling dust-wells, and of the same origin, sometimes develop 
 on the compact surface of snow which has lain for some time on 
 the ground. 
 
 Movement 
 
 Waste and supply of ice. The ice of a glacier is continually 
 wasting. The waste is due partly to surface melting, especially 
 in summer, partly to melting below the surface, for much of the 
 
 * P 
 
 IS 
 
 FIG. 246. 
 
 FIG. 247. 
 
 FIG. 246. Diagram to illustrate the fact that wells of larger diameter 
 may be deeper than those of smaller diameter. The slanting lilies 
 represent the direction of the sun's rays when the sun is highest. 
 
 FIG. 247. Diagram illustrating certain features of glacier motion. The 
 figure at the left represents a vertical section, and the top as moving 
 faster than the bottom. The figure at the right represents a part of 
 the surface, and the central part as moving faster than the sides. 
 
 subsurface ice of most glaciers is at the melting temperature much 
 of the time, and partly to evaporation. 
 
 In spite of the rapid waste of glaciers, particularly at their 
 lower ends and in summer, they often remain nearly constant 
 in size for long periods of time. This shows that there must be a 
 source of supply to replace the waste. This source is found in 
 the snow-fields. From them the ice creeps down the valleys 
 until it reaches an altitude so low and so warm that the waste 
 (chiefly melting) at its end balances its forward motion. 
 
 The fact of movement was first established by noting (1) that 
 the ends of glaciers were sometimes farther down the valleys than 
 they had been at earlier times, and (2) that familiar objects at the 
 ends of glaciers were overturned and pushed forward by the ice. 
 
 Rate of movement. Once the fact of movement was estab- 
 lished, various means were devised for measuring its rate. Rows 
 
230 PHYSIOGRAPHY 
 
 of stakes were set across a glacier in a straight line, and their 
 positions with reference to fixed points on the sides of the valley 
 marked. After a time they were found to have moved down the 
 valley. In most cases it appears that those in the central part of 
 the glacier have moved faster than others, as shown by Fig. 247. 
 
 In this and in other similar ways the rate of movement of 
 numerous glaciers has been determined. It ranges from an amount 
 so small as to be measured with difficulty, to several feet per day. 
 One very large glacier in north Greenland has been estimated to 
 move 100 feet per day, but this is certainly far beyond the rate for 
 any of the more accessible and better-known glaciers. Few of the 
 better-known mountain-valley glaciers move more than a foot or 
 two a day. 
 
 Conditions affecting rate of movement. The rate of glacier 
 movement appears to depend chiefly on (1) the depth of the moving 
 ice; (2) the slope of the surface over which it moves; (3) the slope 
 of the upper surface of the ice; (4) the topography of the bed over 
 which it passes; (5) the temperature; (6) the amount of water 
 in the ice, including that which falls upon it or is carried to it by 
 the drainage of its surroundings, as well as that produced by the 
 melting of the glacier itself; and (7) the amount of load (debris) 
 which the ice carries, especially in its bottom. Great thickness, 
 a steep slope, smoothness of bed, a high (for ice) temperature, and 
 much water favor rapid movement. Since some of these conditions, 
 notably temperature and amount of water, vary with the season, 
 the rate of movement for any given glacier is not constant through- 
 out the year, and is generally greater in summer than in winter. 
 Other conditions, especially the first of those mentioned above, vary 
 through longer periods of time, and occasion periodic variations 
 in the rate of movement. 
 
 MAP EXERCISE 
 
 Topographic Maps Showing Glaciers 
 
 Study the following maps showing existing glaciers, in preparation 
 for conference: 
 
 1. Shasta, Cal. 
 
 2. Mt. Lyell, Cal. 
 
 3. Mt. Stuart, Wash, 
 
 4. Glacier Peak, Wash. 
 
 5. Cloud Peak, Wyo. 
 
THE WORK OF SNOW AND ICE 231 
 
 In each note (a) the altitude of the glaciers, (6) their size, and 
 (c) their exposure (on east, west, north, or south slopes). 
 
 Is there any relation between the heights of (a) the lower, and (6) 
 the upper ends of the glaciers, and the exposures? 
 
 Compare the heights of the lower ends of the glaciers in California, 
 Washington, and Wyoming. Why the differences? 
 
 Note the peculiar shape of the upper ends of many of the valleys 
 shown on the Cloud Peak Sheet, for example that of the South Fork of 
 Clear Creek. Such valley heads are cirques, and cirques generally 
 indicate the former existence of glaciers. 
 
 Nature of glacier movement. Glacier movement has been 
 much discussed, but no general agreement concerning its nature 
 
 FIG. 248. The spreading end of a glacier, North Greenland. 
 
 has been reached. From the fact that the ice moves down the 
 valley, conforming to it somewhat as a river does to its valley, it 
 has been thought that a glacier flows like a stiff liquid. This 
 idea seemed at first to be supported by the fact that when a glacier 
 moves out from its mountain valley to the plain beyond, it gen- 
 erally spreads (Fig. 248), somewhat as a stiff liquid might. It 
 is to be noted, however, that the spreading or deploying end cracks 
 open. In further support of this explanation of glacier motion, 
 various experiments have been performed upon ice. They show 
 that a bar of ice may be bent or moulded into almost any desired 
 
232 
 
 PHYSIOGRAPHY 
 
 form, if it be subjected to sufficient pressure, applied slowly enough 
 through long periods of time. 
 
 +. +.* .** + .+ 
 
 FIG. 249. Diagram to show relations of a high-latitude glacier to its 
 
 valley walls. 
 
 But in spite of the apparent mobility of ice, and in spite of the 
 fact that in so many ways its motion seems to resemble that of a 
 
 stiff liquid, it is very doubtful if its 
 real motion is one of flowage, as 
 that term is ordinarily understood. 
 It has already been stated that the 
 ice often cracks open when it passes 
 over irregularities of bed, as well as 
 under some other circumstances. 
 Cracking open is not a characteristic 
 of liquids. Many glaciers of high 
 latitudes do not rest against the 
 sides of the valleys in which they 
 lie (Figs. 249 and 250). Such gla- 
 ciers are often crevassed longitu- 
 dinally, and the crevasses sometimes 
 have great length. If the ice flows, 
 therefore, it must be supposed to 
 flow until it cracks open. It is 
 not evident that a fluid, however 
 viscous, would do this. These and 
 many other considerations, which 
 will not be detailed here, have led 
 to the view that the resemblance 
 between glacier motion and the 
 motion of a stiff liquid is more 
 seeming than real. 
 
 The fundamental element in 
 glacier motion probably consists in the melting and re-freezing of 
 its substance. The process is a very complex one, and cannot 
 be fully analyzed here, but some of its elements may be stated. 
 When water from the surface sinks into the glacier and freezes 
 
 FIG. 250. A part of the verti- 
 cal side of a North Green- 
 land glacier. The vertical or 
 even overhanging faces are 
 sometimes more than 100 feet 
 high. 
 
THE WORK OF SNOW AND ICE 233 
 
 again it expands, and the ice where the freezing takes place is 
 subject to great stress. The force of the stress which freezing 
 water exerts is illustrated by the familiar fact that vessels of very 
 considerable strength are broken when water freezes in them. 
 The freezing of the water which has descended must have the effect 
 of moving the ice, and the movement must be chiefly down the 
 valley, for in this direction gravity helps, while in the opposite 
 direction it hinders. Furthermore, the water before re-freezing 
 moves, and always downward, not only toward the bottom of the 
 ice, but often, at least, toward the lower end of the valley as well. 
 The flow of the water is therefore a way of transferring the ice of the 
 glacier down-valley. 
 
 There are causes of melting and re-freezing other than those 
 which are dependent upon the direct influence of the sun or the 
 heat of the interior of the earth. These are bound up with the 
 movement of the ice itself, and, without attempting here to ex- 
 plain the principle involved, it may be stated that it is now be- 
 lieved that an important part of glacier motion is to be explained 
 by the melting which results from the pressure involved in motion, 
 and in the re-freezing of the water thus produced. It is believed 
 therefore that though the aggregate motion of a glacier is in a 
 way comparable to the motion of a viscous body, the actual motion 
 is probably that of a solid, small parts of which frequently pass 
 from a solid to a liquid condition for brief spaces of time. 
 
 Another element of glacier motion is sliding, for parts of a 
 glacier sometimes slide or shear over other parts (Fig. 251). The 
 motion of the lower portion of a glacier which carries much debris 
 is greatly retarded by its load. The relatively clean ice above 
 the bottom moves less slowly, and appears very often to be thrust 
 forward or sheared over the debris-laden portion below. This 
 phase of motion is probably much more common and of much 
 more consequence than was formerly supposed. It is best seen 
 in the glaciers of high latitudes, where the vertical edges and 
 ends allow the structure of the ice to be well seen. Under some 
 conditions a glacier may probably slide over its bed. Sliding is 
 not, however, believed to be a principal element in glacier motion. 
 
 Size. There are in the Alps nearly 2000 glaciers. The longest 
 of them is about 10 miles long. Less than 40 are as much as five 
 miles long, and the great majority are less than one mile in length. 
 Some of them are but a few hundred feet wide, and few of them 
 
234 
 
 PHYSIOGRAPHY 
 
 are so much as a mile wide. The thickness of ice is rarely known, 
 except at the lower ends, but it is generally to be measured by 
 hundreds of feet, rather than by denominations of a higher order. 
 Larger alpine glaciers occur in the Caucasus Mountains of 
 Europe and in Alaska. Seward Glacier in Alaska is more than 
 50 miles long, and 3 miles wide at the narrowest part. The glaciers 
 of the western mountains of the United States (south of Alaska) 
 are mostly shorter than the longer glaciers of the Alps. Many 
 
 FIG. 251. Shearing planes in ice, well defined. 
 
 of them are indeed cliff glaciers, or intermediate in type between 
 valley glaciers and cliff glaciers (Figs. 234 and 235). 
 
 Ice-caps 
 
 Ice-caps lie on plains or plateaus instead of occupying moun- 
 tain valleys. As already stated, they may be large or small. 
 Large ones may cover valleys and hills alike. Very large ones 
 are sometimes called continental glaciers. At the present time 
 the ice-caps of Greenland and Antarctica are the only ones which 
 attain great size. 
 
THE WORK OF SNOW AND ICE 
 
 235 
 
 The area of Greenland has been variously estimated at from 
 400,000 to 600,000 square miles, and all of it, except its borders, 
 is covered with a vast field of snow and ice (Fig. 252). Near 
 its margin, occasional mountain tops project above the snow, 
 and here its surface carries some debris; but, except about its 
 
 FIG. 252. Map showing the ice-cap of Greenland, 
 the island are free of ice. 
 
 Only the borders of 
 
 edge, nothing is visible, so far as known, through the entire 
 island, save one vast plateau of snow-covered ice. The sur- 
 face snow is frequently driven by the wind into rolling billows. 
 The snow- and ice-covered plateau rises gradually toward the 
 centre of the island, where it attains an elevation of 8000 or 9000 
 feet. The thickness of the ice is not known, but where thickest 
 it is probably some thousands of feet. 
 
236 
 
 PHYSIOGRAPHY 
 
 The ice of this great field is creeping slowly outward. The 
 rate of movement has not been determined and is probably not 
 
 FIG. 253. Edge of the Greenland ice-sheet. 
 
 the same at all points; but it has been estimated not to exceed a 
 foot a week. Near its margin the ice-cap is much crevassed; 
 
 Fia. 254. A mountain projecting up through the ice, North Greenland. 
 
 but the interior is comparatively smooth and unbroken, so far 
 as now known. 
 
 The ice-cap of Greenland is, in one sense, more of a desert 
 
THE WORK OF SNOW AND ICE 
 
 237 
 
 FIG. 255. A nunatak projecting up through a Greenland glacier. 
 
 FIG. 256. Three small glaciers descending to the sea, North Greenland, 
 
238 
 
 PHYSIOGRAPHY 
 
 FIG. 257. Front of Miles glacier, Alaska, where it reaches Copper River. 
 (U. S. Geol. Surv.) 
 
 FIG. 257a. Glacier and Icebergs. 
 
 FIG. 258. Iceberg, coast of Greenland. 
 
THE WORK OF SNOW AND ICE 
 
 239 
 
 than the Sahara, since it is inhabited, even less than that desert, 
 by plants and animals. There are, it is true, tiny red plants upon 
 it at various points about its border. Taken singly, they are too 
 small to be readily noticed, but they sometimes occur in such 
 multitudes as to give the snow a distinctly red color, known as 
 "red snow." Occasional small animals, especially the larvas of 
 
 FIG. 259. Map of Antarctica. The dotted line represents the approximate 
 limit of abundant floating ice. (After Bartholomew.) 
 
 certain insects, are also found on the snow some little distance at 
 least back from its margin. 
 
 Where the edge of the Greenland ice-cap lies back a few miles 
 from the coast, the rock plateau outside it is affected by numerous 
 valleys which lead down to the sea. Where the edge of the ice- 
 cap reaches the heads of these valleys, ice moves down them in 
 advance of the edge of the ice-cap, making valley glaciers. Many 
 of the valley glaciers move down to the sea, where their ends are 
 
240 
 
 PHYSIOGRAPHY 
 
 broken off (Fig. 256) and floated away as icebergs. Many of these 
 glaciers are far larger than any of the Swiss glaciers, and some 
 are even larger than the great Seward Glacier of Alaska. While 
 their number is very large, the total amount of ice in them is 
 trivial compared with the amount in the one great ice-cap glacier 
 from which they originate. 
 
 The Antarctic snow- and ice-cap is far more extensive than 
 that of Greenland, but its area is not even approximately known 
 (Fig. 259). It appears to be some millions of square miles in 
 extent, but mountains project up through it, as determined by 
 the Schackleton expedition. At many points the ice descends to 
 sea, where huge blocks of it are broken off and floated away as 
 icebergs. Whether this Antarctic ice-cap rests on a continuous 
 land-mass which if free from ice would be called a continent , or 
 whether it rests on numerous islands which but for the ice would 
 be separated by shallow water, is not known. 
 
 FIG. 260. Malaspina glacier, a piedmont glacier in Alaska. (After Russell.) 
 
 Piedmont Glaciers. 
 
 In Alaska a number of large alpine glaciers emerge from adja- 
 cent valleys of the St. Elias range upon a low plain, where their 
 ends spread and unite to form a vast plateau of ice, 70 miles long 
 
THE WORK OF SNOW AND ICE 
 
 241 
 
 and 20 to 25 miles wide. This peculiar body of ice is the Malas- 
 pina Glacier (Fig. 260). Its area is considerably more than that 
 of the state of Delaware. Its central portion is free from rock 
 debris, but is interrupted by thousands of crevasses. On warm 
 summer days hundreds of rivulets flow in channels of clear ice 
 until they lose themselves in yawning crevasses. The deep roar 
 of some stream in its tunnel far below the surface is frequently 
 heard. Nearer the margin, where the ice is not so broken, there 
 are many small ponds with high walls of ice. A belt along the 
 margin five miles or less in width is covered by earthy matter and is 
 
 FIG. 261. Forest on the southern border of Malaspina glacier. 
 (Photo, by Russell.) 
 
 densely forested (Fig. 261). The undergrowth is here so thick 
 that travelers have to cut their paths, and on the edge of the ice 
 there are trees three feet in diameter. The forest extends four or 
 five miles from the edge of the ice. The ice beneath the surface 
 debris is probably 1000 feet thick. Another large but unexplored 
 glacier of the same type lies a few miles west of the Malaspina. 
 Others occur about north Greenland. 
 
 Piedmont glaciers are of slight importance, from a quantitative 
 point of view, but they constitute an interesting type. 
 
242 
 
 PHYSIOGRAPHY 
 
 THE WORK OF GLACIERS. 
 
 Glaciers do a twofold work They wear or erode the surface 
 over which they pass, and they carry away and ultimately deposit 
 the material which they acquire by erosion, as well as all that falls 
 or blows upon them. 
 
 Erosion. As the snow-field develops, it often lies upon a sur- 
 face which is uneven and more or less covered with loose pieces 
 of rock (Figs. 72 and 274). As the snow accumulates, all projecting 
 
 FIG. 262. Surface of rock rounded and smoothed by ice. Bronx Park 
 New York City. (U. S. Geol. Surv.) 
 
 stones are covered and enclosed by it, and, when the snow becomes 
 ice and begins to move, these masses of rock are carried along in 
 its bottom. The ice therefore has some load when it starts. 
 
 Where the snow and ice accumulate about projecting points of 
 bed-rock, the ice tends to break them off when it moves. If they 
 are too strong to be broken off bodily, their surfaces are worn by 
 the passage of the ice carrying rock debris in its bottom. Again, 
 as a glacier creeps out over surfaces covered with soil or other 
 mantle rock, the ice freezes to the soil, etc.; that is, the ice above 
 the ground becomes continuous with the ice in the soil. This 
 union is brought 'about, in part at least, by the freezing of descend- 
 ing water. When this has been done, further movement causes 
 more or less of the soil to be moved along. 
 
 The first effects of the glacier therefore are (1) to clean off 
 the loose debris from the surface, and (2) to break or wear off 
 
THE WORK OF SNOW AND ICE 
 
 243 
 
 projecting points of the bed-rock over which it passes. The 
 general effect of the movement of the ice may be comoared to 
 
 FIG. 263. Ice-worn rock, Bell's Island, Lake Huron. (Bell.) 
 
 the effect of a flexible rasp which fits itself, though sometimes with 
 difficulty, to the irregularities of the surface over which it moves. 
 
 FIG. 264. Diagram representing a hill unworn by ice, and the irregular 
 contact of soil and rock. 
 
 FIG. 265. Diagram showing the effect of glacial wear on a hill such as is 
 shown in Fig. 264. 
 
 Clean ice, moving over smooth, solid rock, would erode little; but 
 rock-shod ice wears the surface over which it moves, even where 
 that surface is smooth, solid rock. 
 
244 
 
 PHYSIOGRAPHY 
 
 An ice-sheet glacier is generally much thicker than a valley 
 glacier, and it generally moves over a surface which has less slope. 
 
 FIG. 266. A mountain valley which has been strongly glaciated, 
 Wasatch Mountains. (Photo, by Church.) 
 
 An ice-sheet of great thickness may move over considerable hills 
 and valleys, without being notably turned from its course by 
 
 FIG. 267. A mountain valley in the same range as the last, but not 
 glaciated. (Photo, by Church.) 
 
 them. The hills and projecting points of rock overridden are worn 
 down and smoothed off (compare Figs. 264 and 265), the wear being 
 greatest on the side of the hill against which the ice moves. The 
 
THE WORK OF SNOW AND ICE 
 
 245 
 
 result is that glacial erosion sometimes so shapes the rock hills 
 that their forms indicate the direction of movement. 
 
 FIG. 268. A normally eroded mountain mass not affected by glaciatioa 
 
 (Davis.) 
 
 FIG. 269. The same mountain mass shown in Fig. 268 affected by glaciers 
 which still occupy its valleys. (Davis.) 
 
 Valleys tnrough which glaciers pass are widened and deepened 
 and their walls made smoother. Valley glaciers tend to trans- 
 
246 
 
 PHYSIOGRAPHY 
 
 form Y-shaped valleys into U-shaped ones, a result often con- 
 spicuous in mountain regions (Figs. 266 and 271). Where a 
 
 FIG. 270. The same mountain mass shown in the two preceding figures 
 after the ice has melted. (Davis.) 
 
 glacier deepens a mountain valley notably, it brings about a lack 
 of adjustment between the valley which is deepened, and its tribu- 
 
 FIG. 271. A hanging valley near Lake Kootenay. (Photo, by Atwood.) 
 
 taries which are not so deepened. The effect is illustrated by 
 Figs. 270 and 271. The lower ends of the tributary valleys are 
 well above the bottoms of their mains. Such valleys are called 
 
THE WORK OF SNOW AND ICE 
 
 247 
 
 hanging valleys. Hanging valleys abound in the mountains of 
 the West, where glaciers were formerly much more extensive than 
 
 FIG. 272. Figure showing contrast between glaciated rock surface below 
 and non-glaciatqd crests above. Kearsarge Pinnacles, Bubbs Creek 
 Canyon, Cal. 
 
 FIG. 273. A glacial cirque with a small glacier in its head. Bighorn Mts. 
 Wyo. (Photo, by Blackwelder.) 
 
 now. Valley glaciers descending to the sea sometimes deepen the 
 the lower ends of their valleys so that they become narrow bays or 
 
248 
 
 PHYSIOGRAPHY 
 
 fjords, when the ice melts. Glacier erosion is not, however, the 
 only factor in fjord-making (p. 173). 
 
 FIG. 274. A glacial cirque north of Grizzly Peak, Colo. (Photo, by Hole.) 
 
 Fig. 237 shows that there is often a steep descent of a moun- 
 tain glacier near its head. This steep slope, and especially its lower 
 part, is the site of great erosion, which carries the head of the 
 
 FIG. 275. Striated rock surface. Kingston, Des Moines Co., la. 
 (U. S. Geol. Surv.) 
 
 valley back farther and farther into the mountain, and at the 
 same time gives it steep slopes at sides and head (Fig. 273 and PI. 
 XVIII). The big, blunt, steep-sided heads of valleys developed 
 by the erosion of valley glaciers are known as cirques. Cirques 
 have remarkable development in the Uintas, the Bighorns, the 
 
PLATE XVfl 
 
 C^/'f' ^y 
 
 ,oV^ - 
 aX& ; y 
 
 WASHINGTON. - f^C^ 
 
 Glaciers on Glacier Peak, Washington. Scale 2 miles per inch. (Glacier Peak 
 Sheet, U. S. Geol. Surv.) 
 
PLATE XVIII 
 
 A portion of the Bighorn Mountains, showing glaciated valleys, the heads of which 
 are in many cases cirques. Scale 2 miles per inch. (Cloud Peak, Wyo., Sheet, 
 U. S. Geol. Surv.) 
 
THE WORK OF SNOW AND ICE 
 
 249 
 
 Sierras, and many other mountains of the West. There are often 
 basins excavated in the solid rock in the cirques, and such basins 
 are the sites of some of the numerous little lakes which add so 
 much to the beauty of scenery in mountains which have been 
 recently affected by ice. 
 
 The effect of ice-sheets on valleys is less obvious than that of 
 
 FIG. 276. Rock grooved by glaciation. The gorge was probably formed by 
 a stream under the ice, and then worn by the ice. Kelley's Island, 
 Lake Erie. (U. S. Geol. Surv.) 
 
 valley glaciers, because in this case the ice affected divides as well 
 as valleys. It is probable that the valleys through which the 
 ice of a large ice-cap moves are deepened more than the neigh- 
 boring hilltops are cut down. If so, glacier erosion increases the relief 
 of the rock surface. At the same time, it probably reduces the 
 roughness of the surface by reducing the steepness of slopes, and 
 by obliterating many minor irregularities of hill and valley slopes. 
 
250 
 
 PHYSIOGRAPHY 
 
 If the ice of an ice-sheet crosses valleys, as the ice of great 
 ice-caps often does, the valleys are not deepened notably, though 
 their upper slopes may be much worn. 
 
 FIG. 277. Small protuberances of rock showing the effect of ice wear. 
 The movement was from left to right. Near Darlington, Ind. (U. S. 
 Geol. Surv.) 
 
 As the ice wears the surface, it makes distinct scratches, called 
 strice (Fig. 275), on the bed-rock over which it passes. Grooves 
 (Fig. 276) may be developed instead of striae under favorable con- 
 ditions. The striae are made by the stones carried in the bottom 
 of the ice. The grooves are developed where the bed-rock is softer, 
 
 FIG. 278. Diagram showing, by the wear in the depressions, the direction 
 of ice movement, left to right. 
 
 or where great bowlders are held firmly in the bottom of the ice, 
 and urged along under great pressure. Fine clayey material 
 in the bottom of the ice polishes the rock below. The polish, 
 the striae, and the grooves left on the surface of the rock after the 
 ice has melted are among the most distinctive marks of the former 
 existence of glaciers. In any limited area, these striae are gen- 
 erally nearly parallel to one another and show the direction, or 
 one of two directions, in which the ice moved. Between these 
 two directions it is usually possible to decide by the help of the 
 little irregularities of surface, as shown in Figs. 277 and 278. 
 
 The stones in the bottom part of the ice are rubbed against 
 one another, as well as against the bed of the glacier, and are 
 
THE WORK OF SNOW AND ICE 
 
 251 
 
 scratched much as the bed-rock is (Figs. 279 and 280). Since 
 the stones in the ice shift their positions from time to time as 
 the ice moves, they are frequently striated on two or more sides. 
 As the materials carried by the ice rub against one another 
 and against the bed over which they are carried, they become 
 finer and finer. The finest products of the grinding constitute 
 "rock flour," while coarser parts have the size of sand grains, 
 pebbles, or even large stones. Thus it happens that the materials 
 gathered and shaped by the ice are of all grades of coarseness, 
 from huge masses many feet in diameter down to the finest earth 
 
 FIG. 279. Stones striated by glacial wear. 
 
 (Fig. 281). The larger masses of rock are bowlders; the smaller 
 pieces are cobble-stones, pebbles, etc., while the finer materials are 
 sand and ground-up rock (rock flour), popularly called clay. 
 
 Materials gathered. From its mode of erosion it will readily 
 be seen that the bottom of a glacier may be charged with various 
 sorts of material. There may be (1) bowlders which the ice has 
 picked up from the surface, or which it has broken off from pro- 
 jecting points of rock over which it has passed; (2) smaller pieces 
 of rock picked up in the same way; (3) the fine products (rock 
 flour) produced by the grinding of the debris in the ice on the 
 rock-bed over which it passes, and similar products resulting 
 from the rubbing of stones in the ice against one another; and 
 (4) sand, clay, soil, vegetation, etc., derived from the surface 
 overridden. Thus the materials which the ice carries are of all 
 grades of coarseness and fineness, from large bowlders to fine 
 
252 
 
 PHYSIOGRAPHY 
 
 FIG. 280. Stones in the drift striated and beveled by glacial wear. 
 
 FIG. 281. Section of drift showing its heterogeneity. 
 
THE WORK OF SNOW AND ICE 253 
 
 clay. The coarser material may be angular or round at the 
 outset, and its form may be changed and its surface striated as 
 it is moved forward. Whether one sort of material or another 
 predominates depends primarily on the nature of the surface 
 overridden. 
 
 Disposition of debris in transit. The larger part of the 
 material carried by a glacier is carried in its basal portion; but 
 some is carried in the body of the ice, well above its bottom, and 
 in the case of most glaciers, some is carried on the surface of the 
 ice. 
 
 The position of the material in the base of the ice is readily 
 understood from the manner in which the debris is gathered. 
 The material above the base reaches its position in various ways. 
 
 FIG. 282. Diagram illustrating one way in which a glacier gets 
 englacial material. 
 
 Sometimes the ice passes over a considerable elevation. In this 
 case material may be torn from its top and carried along at a 
 level corresponding somewhat to that from which it was derived. 
 This is illustrated by Fig. 282. Under some circumstances, too, 
 ice moves from the bottom of the glacier upward (Fig. 283), and 
 carries debris with it, and in this way debris which was once at 
 the bottom may later find itself in a higher position. 
 
 The material on the surface of a glacier reaches its position in 
 various ways. Where the slopes above the glacier are steep, rock 
 material may fall or slide down to the surface of the ice. Great 
 masses of snow (avalanches) sometimes slide down upon a glacier 
 from the steep slopes above, bringing quantities of debris, and 
 dust is blown upon the ice. 
 
 The material falling or sliding down to a glacier from the 
 cliffs above tends to accumulate near the margin of the ice, and 
 as it lies on the surface of the glacier, constitutes lateral moraines 
 (Fig. 284). If two glaciers unite, as is sometimes the case, the 
 
254 
 
 PHYSIOGRAPHY 
 
 FIG. 283. End of a North Greenland glacier, showing the upturning of 
 the layers of ice at the end. At one point a few stones are seen on the 
 surface of the ice, where an upturned layer comes to the surface. This 
 structure is common in North Greenland. 
 
 FIG. 284. Figure showing the union of glaciers and the development of 
 medial moraines by the union; also the position of lateral moraines. 
 (After Tyndall.) 
 
THE WORK OF SNOW AND ICE 255 
 
 two lateral moraines of the adjacent sides may unite, forming a 
 single medial moraine (Fig. 284). 
 
 Both medial and lateral moraines arise in other ways. If a 
 glacier passes over an elevation which reaches well up into the 
 ice, material torn from the elevation is at first carried along in 
 the ice; but as the lower end of the glacier is approached, surface 
 melting may bring the surface of the ice down to the level of the 
 debris, when it appears at the surface as a medial moraine (Fig. 
 282). In other cases lateral and medial moraines arise by the 
 upturning of the layers of ice, as shown in Fig. 285. Moraines 
 arising in this way are common on the glaciers of North Greenland. 
 
 FIG. 285. Diagram illustrating a way in which lateral and medial moraines 
 are formed in many of the North Greenland glaciers. The horizontal 
 line at the base represents sea-level. 
 
 Deposition by Glaciers. 
 
 While the ice is in motion it is depositing more or less debris 
 beneath itself, against the hills and projecting bosses of rock. 
 Again, the bottom of the ice may become heavily loaded in passing 
 over a region which yields load readily, and a part of it may be 
 deposited farther on, where changed conditions of movement have 
 rendered the load excessive. Debris deposited at one time may 
 be taken up at another. Thus beneath the moving ice there is 
 more or less deposition in progress all the time, though much of 
 it is not permanent. In this respect, deposition by glaciers is 
 somewhat like the deposition of streams. 
 
 The position of the end of the glacier is determined by the 
 relation between ice waste and forward movement. When the 
 ice advances as much as it is melted back, its end or edge is con- 
 stant in position. But it is to be remembered that even when 
 the end of a glacier is constant in position, the ice itself is in con- 
 tinual movement. Under these circumstances, material is con- 
 tinually brought to the end or edge of the glacier and left there. 
 The result is that if the end of a valley glacier or the edge of an 
 ice-cap be constant or nearly constant in position for a long period 
 of time, a thick body of drift is accumulated beneath its margin 
 (Figs. 286 and 287). 
 
256 
 
 PHYSIOGRAPHY 
 
 FIG. 286. Thick accumulation of drift under the end of a glacier. The 
 end has probably been in about the same place for a long time. McCor- 
 mick Bay, North Greenland. 
 
 FIG. 287. An accumulation similar to that shown in Fig. 286, after the ice 
 has melted away; near the last. 
 
THE WORK OF SNOW AND ICE 257 
 
 The terminal moraine. The belt of thick drift accumulated 
 beneath the end of a valley glacier or beneath the edge of an 
 ice-cap is a terminal moraine. The terminal moraine becomes 
 massive only when the end of the glacier remains nearly constant 
 in position for a long time. The same term is sometimes applied 
 to the debris on the end of a valley glacier, or on the edge of an 
 ice-sheet. The material of this super-glacial terminal moraine is 
 added to the sub-marginal terminal moraine when the ice melts. 
 In general, the sub-marginal accumulation is much greater than 
 that which is let down from the top when the ice melts. 
 
 The ground moraine. When a glacier melts, all the debris 
 which it carried is deposited. When the ice is gone, therefore, the 
 whole surface which it covered is likely to be strewn with its debris. 
 All the rock debris left by the ice is drift. The drift deposited 
 by the ice but not aggregated into thick belts at its edge is ground 
 moraine. The area of the ground moraine is much more exten- 
 sive than the area of the terminal moraine. 
 
 Many spots once covered by the ice are left without drift 
 when the ice melts, for the ice does not always carry debris at 
 every point in its bottom. Areas of bare rock are therefore found, 
 and sometimes commonly, in the area from which glacier ice has 
 melted. 
 
 Lateral moraines. The term lateral moraine is applied not 
 only to certain aggregations of drift on valley glaciers (Fig. 284) , 
 but also to certain aggregations of drift which are left by a 
 valley glacier after melting. When a valley glacier is melted, 
 the lateral moraines of its surface are left in the valleys which it 
 occupied; but these lateral moraines are not commonly massive 
 enough to be conspicuous after the ice is gone. On the other 
 hand, the lateral moraines which remain when a valley glacier has 
 disappeared are often large (Fig. 288). In many cases, indeed, 
 they are the most conspicuous deposits left. They are often hun- 
 dreds of feet high, and in some cases even more than a thousand. 
 
 The making of these huge lateral moraines is a somewhat com- 
 plex process. They are made up in small part of the lateral moraines 
 which were on the ice, and in much larger part of material accumu- 
 lated beneath the lateral margin of the glacier. Their explanation 
 seems to be found in the fact that a valley glacier moves not only 
 down the valley, but also sidewise from the center toward either 
 side. Spreading sidewise from the center, the ice is constantly 
 
258 
 
 PHYSIOGRAPHY 
 
 shifting debris from the axis of the valley to the edge of the ice on 
 either side. The lateral moraine left after the ice is gone is there- 
 fore of the nature of a terminal moraine beneath the lateral margin 
 of the ice. 
 
 Lateral moraines of valley glaciers are much more likely to 
 remain after the ice is gone than the corresponding terminal 
 moraines, for the latter are more likely to be washed away by the 
 waters issuing from the ice or flowing down subsequently through 
 the valley. 
 
 Ice-caps do not develop lateral moraines, because they have no 
 lateral margins. 
 
 FIG. 288. A lateral moraine left by a former glacier in the Bighorn Moun- 
 tains of Wyoming. (Photo, by Blackwelder.) 
 
 Distribution and Disposition of Drift Glaciers distribute their 
 debris unevenly over the surface. The drift of terminal moraines 
 is generally much thicker than that of the ground moraine near 
 at hand, and the drift in the lateral moraines of valley glaciers 
 is often very thick. The drift therefore modifies the topography 
 in some notable measure. 
 
 In valleys, the terminal moraines often constitute dams, and 
 so pond the waters of the streams above, and make lakes (Fig. 
 289), though not all the lakes of glaciated valleys are due to 
 moraine-dams (p. 311). 
 
 The drift of ice-sheets is more likely to be abundant in valleys 
 
260 
 
 PHYSIOGRAPHY 
 
 and other low places than on hills and ridges. The drift of ice- 
 sheets, therefore, usually diminishes rather than increases the relief 
 of the surface (Fig. 290). 
 
 FIG. 290. Diagram to illustrate how drift may decrease the 
 relief of the surface. 
 
 Glacier drift is often irregularly disposed, so that its own sur- 
 face is somewhat rough, even where it diminishes surface relief. 
 Its surface is marked in many places by hillocks, mounds, etc., of 
 drift, and by basin-like depressions (Figs. 291 and 292). In some 
 
 FIG. 291. Sketch of drift (terminal moraine) topography near 
 Hackettstown, N. J. (N. J. Geol. Surv.) 
 
 of the basins water stands, making lakes, ponds, or marshes. 
 The surface of drift is therefore very unlike the surface developed 
 by the erosion of running water, for in the latter the depressions 
 have outlets, and the hills and ridges stand in a very definite re- 
 lation to the valleys (compare Figs. 106, 139, and 292). 
 
 R6sum6. The more distinct marks which a valley glacier leaves 
 behind it are the following: (1) A U-shaped valley, often with its 
 tributary valleys hanging (Fig. 271), and with its head in the 
 form of a cirque (Fig. 273) ; (2) the upper end of the valley which 
 
262 
 
 PHYSIOGRAPHY 
 
 it occupied well cleaned out (Fig. 293), the loose rock-debris hav- 
 ing been carried down the valley; (3) the rock of the valley 
 smoothed and striated (Fig. 294); (4) rock basins in the bottom 
 of the valley, especially near its head; (5) a body of drift com- 
 posed of coarse and fine material, often without trace of stratifi- 
 cation or orderly arrangement; (6) the stones of the drift are 
 often worn, but not rounded as streams or waves round them, and 
 
 FIG. 293. Portion of the upper part of a valley cleaned out by ice. The 
 figure shows also the contrast between glaciated topography below 
 and the non-glaciated above. Needle Mountains, Colo. (U. S. Geol. 
 Surv.) 
 
 they often have planed and striated faces; (7) the drift is disposed 
 as no other transporting agent disposes the material which it leaves 
 (Fig. 295). The singular lateral-moraine ridges of valley glaciers, 
 and the terminal moraines which often partially or wholly ob- 
 struct valleys, giving rise to lakes, ponds, and marshes, are among 
 the distinctive deposits of valley glaciers. (8) Another distinc- 
 tive though less common mark of glacier deposits is the huge 
 bowlders in delicately balanced positions (perched bowlders) (Fig. 
 
 296). 
 
 Ice-sheets likewise leave (1) bodies of drift very much like that 
 
THE WORK OF SNOW AND ICE 
 
 2G3 
 
 FIG. 294. Striae, grooves, etc., in a canyon tributary to the Big Cotton- 
 wood Canyon. Wasatch Mts., Utah. (Photo, by Church.) 
 
 FIG. 295. The moraines about the lower end of a glaciated mountain 
 valley. Bloody Canyon, Cal. (U. S. Geol. Surv.) 
 
264 
 
 PHYSIOGRAPHY 
 
 FIG 296. A perched bowlder, size 12X8X8 feet. East of Englewood, N. J. 
 
 (N. J. Geol. Surv.) 
 
 FIG. 297. Topography of drift shown in contours; an area near Minneapolis, 
 Minn. Scale about one inch to the mile. (U. S. Geol. Surv.) 
 
THE WORK OF SNOW AND ICE 
 
 266 
 
 of mountain glaciers, though often less coarse. This drift is in the 
 form of terminal moraines. (Figs. 291 and 297) and ground moraines 
 (Fig. 298), the surfaces of which are marked by (2) numerous lakes, 
 ponds, and marshes, which fill the kettle-like or saucer-like de- 
 pressions in the surface of the drift. (3) Ice-sheets also smooth, 
 striate, and groove the surface of the rock over which they move. 
 
 FIG. 298. One phase of ground moraine topography. Elongated hills of 
 drift of the type shown here are called drumlins. Southeastern Wis- 
 consin. (U. S. Geol. Surv.) 
 
 Fluvio-glacial Deposits. 
 
 Even while glaciers are growing, their ice is melting to some 
 extent, and when they disappear, it is primarily by melting. Drain- 
 age is vigorous at the edge of an ice-sheet, and below the ends of 
 valley glaciers much of the time, and in the summer, when the 
 ice is melting rapidly, the streams which carry off the water are 
 greatly swollen. Water-work, therefore, accompanies ice-work 
 in all cases, and since there is on the whole about as much water 
 as ice, and since the water has the last chance at the material left 
 by the ice, it follows that some of the drift, as deposited by the 
 glacier, is more or less modified by water subsequently. The 
 streams which flow from glaciers carry away much debris de- 
 rived from the ice. At the outset, this consists of both coarse 
 
266 
 
 PHYSIOGRAPHY 
 
 and fine material, but the gravel and small bowlders are soon 
 dropped and only the finer material is carried far. Many such 
 streams carry so much silt in suspension that the water is turbid. 
 If the silt is whitish, as it often is, the streams are said to be 
 "milky." 
 
 Profile of Valley Train 
 
 FIG. 299. Diagram to illustrate the profile of a valley train, and its rela- 
 tions to the terminal moraine in which it heads. 
 
 By the deposition of the gravel, sand, and silt, the valleys below 
 glaciers are often aggraded to some extent by the fluvio-glacial 
 debris. Deposits of this sort are stratified, and so are in contrast 
 with the deposits made by the ice. Furthermore, the surface of 
 the stream deposits is generally plane, and therefore in contrast 
 with the topography of the drift deposited by the ice. 
 
 FIG. 300. The outwash plain and the terminal moraine near Baraboo, Wis. 
 (Photo, by Atwood.) 
 
 The material deposited by the stream in the valley below a 
 glacier is a valley train. It is simply an alluvial plain developed 
 under special circumstances. Valley trains are best developed 
 just outside terminal moraines (Fig. 299). 
 
 In the case of an ice-cap, the water which issues from the ice 
 often fails to find a valley. Each issuing stream thus tends to 
 develop an alluvial fan. By growth, these fans may merge, 
 making an alluvial plain. Such a plain, composed of material 
 washed out from the ice is an outwash plain (Fig. 300), which is 
 
THE WORK OF SNOW AND ICE 
 
 267 
 
 often wider than long. It is of coarse material next the ice and 
 of finer material farther away. Outwash plains, like valley trains, 
 are best developed just outside the terminal moraines of ice-sheets, 
 and their materials are stratified. 
 
 A glacier may obstruct surface drainage. If in its forward 
 movement the ice obstructs the lower end of a valley, the water 
 above accumulates and constitutes a lake. Drainage from the 
 melting ice often builds deltas in lakes, just as other streams 
 build deltas in the bodies of standing water into which they flow. 
 
 FIG. 301. An esker in Finland. 
 
 In addition to the drainage outside the ice, there is running 
 water in the ice and under it. The streams beneath the ice (sub- 
 glacial streams) sometimes deposit gravel in their channels. These 
 channels may be so built up that when the ice melts, the old bed 
 of the stream appears as a low but narrow ridge, called an esker, 
 composed chiefly of gravel and sand (Fig. 301). The sub-ice 
 channels sometimes have the effect of tubes through which the 
 water is forced with considerable velocity. As it issues from 
 beneath the ice, its velocity is checked, and it sometimes makes 
 extensive deposits of gravel and sand at the margin of the ice. 
 These deposits are stratified, but the stratification is often irregular. 
 
268 
 
 PHYSIOGRAPHY 
 
 They are often left against the edge of the ice, and when the 
 edge melts, the deposits appear as mounds and ridges, called 
 kames (Fig. 302). 
 
 In warm weather, there are many small streams on the surfaces 
 of glaciers (Fig. 242), especially if the surface of the ice is not 
 much crevassed; but these surface streams rarely make deposits 
 of consequence. On ice-caps there is little debris on the ice 
 
 FIG. 302. A group of kames near Connecticut Farms, N. J. 
 (N. J. Geol. Surv.) 
 
 except at its immediate edge, so that the surface streams have 
 access to little debris, and they are commonly clear. 
 
 As the ice melts away, the waters produced by the melting 
 flow over the surface of the drift which the ice had already depos- 
 ited, and modify its surface to some extent by eroding in some 
 places and depositing in others. 
 
THE WORK OF SNOW AND ICE 269 
 
 As a result of all these phases of water-work, much of the drift 
 is stratified. The stratified drift is sometimes above the unstrati- 
 fied, sometimes below it, and sometimes interbedded with it. In 
 some cases, too, it lies beyond the limit which the ice reached. 
 
 Icebergs. 
 
 Where glaciers move down into the sea, their ends may be 
 broken off and floated away as icebergs. The breaking-off is 
 brought about in various ways. As the ice pushes out into 
 deepening water, it may reach a point where the depth of water 
 is so great that its buoyant effect breaks off the ice, which is 
 lighter than water. The ice is often already partly broken by 
 crevasses before it reaches the sea. 
 
 Bergs derived from Greenland are found as far south as New- 
 foundland in considerable numbers. Occasionally they reach still 
 lower latitudes, but by the time they have moved so far from 
 their source, they have usually become small by melting. The 
 bergs from Greenland are rarely 200 feet out of water, and most 
 of them are not more than 100 feet, even near their sources. They 
 are sometimes a mile or more across. In the South Polar regions 
 bergs are still larger in area, though higher ones are perhaps not 
 common. A berg 200 feet out of water, disregarding projecting 
 points, may be 1000 to 1500 feet thick. Though river or lake ice 
 is about nine-tenths as heavy as water, glacier ice is less heavy, 
 unless loaded down with rock debris. This is because the snow- 
 ice does not become so compact as the ice formed on rivers and 
 lakes. 
 
 As the icebergs sail away, they carry more or less of the debris 
 which was in the bottom of the glacier. In the water, melting 
 takes place, and the debris which was held by the melted parts 
 falls to the bottom. If, as is often the case, the iceberg capsizes as 
 it sets sail, the bottom debris of the glacier may appear on the 
 sides or top of the iceberg, if it does not at once slide off and sink. 
 This debris absorbs more of the sun's heat than the ice does, and 
 is soon melted out of the ice; or more strictly, the ice is melted 
 away from around it. If it was on the side, it drops out into the 
 sea. 
 
 Icebergs frequently turn, or cant, because of (1) the cutting 
 of the waves, (2) the splitting off of pieces of ice, (3) unequal 
 
270 PHYSIOGRAPHY 
 
 melting, etc., all of which tend to shift their centers of gravity 
 and so disturb their equilibrium. 
 
 Observations on northern icebergs indicate that they do not 
 carry much debris far. The average berg is probably free of 
 debris before it has floated 100 miles. The common notion that 
 the banks of Newfoundland were made largely by berg deposits 
 probably has no foundation in fact. 
 
 Where Icebergs ground in large numbers, as on the shores of 
 Labrador and Newfoundland, they may erode the bottom to come 
 slight extent. 
 
 Icebergs in the North Atlantic occasionally reach the tract of 
 transatlantic commerce. Since they are sometimes surrounded 
 by fog, they may be a menace to shipping and travel. 
 
 ANCIENT GLACIERS AND ICE-SHEETS. 
 
 There have been times in the earth's history when glaciers were 
 much more extensive than now. The latest of these periods is 
 known as the glacial period. During this period the glaciers of the 
 western mountains were very much larger than now, and glaciers 
 were numerous in many mountains where there are none now. 
 Small ones existed even in the mountains of New Mexico, Arizona, 
 and Nevada. The amount of ice in the glaciers of Utah or Colo- 
 rado at that time was far in excess of all that now exists in the 
 United States south of Alaska. The glaciers in the western moun- 
 tains north of the United States also were correspondingly larger 
 than now, while east of the mountains an area some 4,000,000 
 square miles in extent (Fig. 303), lying partly in Canada and partly 
 in the United States, was covered with an ice-sheet or continental 
 glacier. 
 
 The ice-sheet of North America seems to have originated in 
 two principal centers, one on either side of Hudson Bay. The 
 beginning of each was doubtless a great snow-field. The snow- 
 and ice-fields grew by the fall of snow, and later by the spread 
 of the ice to which the snow gave rise. The two ice-sheets finally 
 became one by growth (Fig. 303). It is to be noted that the 
 great continental glacier did not originate in mountains, but on 
 high plains. In addition to the large valley glaciers of the western 
 mountains, bodies of ice of the ice-sheet type were developed in 
 favorable situations in these mountains, though their continuity 
 
THE WORK OF SNOW AND ICE 
 
 271 
 
 was much interrupted by the crests and peaks. The valley glaciers 
 often merged on the plains below, where piedmont glaciers of great 
 size were developed. 
 
 At the time of its greatest extent, the ice-sheet of North America 
 
 FIG. 303. Sketch-map showing the area in North America covered by ice 
 at the maximum stage of glaciation. (Chamberlin.) 
 
 extended south so as to cover all of New England, the northern 
 parts of New Jersey and Pennsylvania, and much of Ohio and 
 Indiana. Its edge crossed the Ohio River where Cincinnati now 
 stands, and advanced a few miles into Kentucky. Farther west 
 it reached almost to the southern end of Illinois. Its edge crossed 
 the Mississippi near St. Louis, and followed, in a general way, 
 
272 PHYSIOGRAPHY 
 
 the course of the Missouri River to Montana. Most of the con- 
 tinent north of this line was covered with snow and ice, but there 
 was an area of eight or ten thousand square miles, mainly in 
 southwestern Wisconsin, which the ice did not cover. Because 
 of the absence of drift in this region, it is known as the Driftless 
 Area. 
 
 The conditions for extensive glaciation existed in Europe at 
 about the same time. The glaciers of the Alps, for example, were 
 many times as large as those of the present time. To the south 
 they extended quite through the mountain valleys and spread 
 themselves out on the plains of northern Italy. On other sides 
 also the glaciers were correspondingly larger than now. This 
 great extension of the glaciers is known from the moraines, and 
 from the striated rock, etc., which the ice left where it melted. 
 Similar conditions existed in the other mountains of Europe where 
 glaciers now exist, and in some where glaciers are not now present. 
 
 In northern Europe, as in the northern part of North America, 
 there was an extensive ice-sheet, but its area was only about half 
 that of the ice-sheet of North America. The center from which 
 the ice-sheet radiated was the high mountains of Scandinavia, 
 with perhaps subordinate centers in the highlands of Scotland, 
 and in the Urals. At the time of its greatest extension, this ice- 
 sheet covered all but the southernmost part of Great Britain, all 
 of northern Germany, and much of Russia (Fig. 304). 
 
 Great ice-sheets are not known to have developed in other 
 continents, but their mountain glaciers were greatly enlarged. 
 
 In both Europe and North America the history of the con- 
 tinental glaciers was most complex. In each continent there 
 were several successive ice-sheets, separated from one another by 
 considerable intervals of time. The sequence of events in North 
 America was somewhat as follows: After the development of the 
 first great ice-sheet, it shrank to small proportions, or disappeared 
 altogether, probably because of a change of climate. The dwin- 
 dling of the first ice-sheet was followed by a relatively warm period, 
 during which plants and animals took possession of the region 
 abandoned by the ice. Another continental ice-sheet then de- 
 veloped, overspreading the region from which the first had with- 
 drawn, and extending still farther south. As it advanced, the 
 second ice-sheet occasionally buried the soil which had formed on 
 the top of the drift deposited by the ice of the first epoch. Such 
 
THE WORK OF SNOW AND ICE 
 
 273 
 
 soils, sometimes with the remains of plants which can be identified. 
 lying between a sheet of drift below and another above, are one of 
 the means by which it is known that there was more than one 
 continental glacier. By this and other means, a third, fourth, 
 and fifth ice-sheet, each somewhat smaller than its predecessor, 
 developed and disappeared. In other words, there were at least 
 five epochs when ice-sheets were extensive, separated by epochs 
 
 FIG. 304. Sketch-map showing the area of Europe covered by the con- 
 tinental glacier at the time of its maximum development. (After 
 Jas. Geikie.) 
 
 when the ice was greatly diminished, or when it disappeared alto- 
 gether. The ice-sheets of Europe had a similar history. 
 
 Cause of the Glacial Epochs. 
 
 The cause of the development of the great ice-sheets was doubt- 
 less climatic, the chief factor being a reduction of temperature. 
 The cause of this cold climate is not certainly known. Various 
 hypotheses have been proposed to explain it, but to most of them 
 there seem to be fatal objections. This subject will not be dis- 
 cussed here, but it may be stated that the only hypothesis which 
 
274 PHYSIOGRAPHY 
 
 seems not to be discredited is that which refers the change of 
 climate to a change in the constitution of the atmosphere. It 
 appears that an increase in the amount of carbonic-acid gas and 
 water vapor would result in an amelioration of climate, while a 
 decrease in these elements would result in a reduction of tempera- 
 ture. This hypothesis cannot be elaborated here, but it may be 
 stated that plausible reasons have been suggested for fluctuations 
 in the amounts of these substances in the atmosphere, and also 
 for the relatively heavy precipitation (which is as necessary as 
 low temperature for glaciation) in the regions where the ice-sheets 
 developed. 
 
 CHANGES PRODUCED BY THE CONTINENTAL GLACIERS. 
 
 The ice-sheets of North America modified the surface which 
 they covered to some notable extent. A brief resume" of the 
 changes they produced will serve to review and emphasize the 
 work of ice-sheets. The changes wrought by the ice-sheet fall 
 into two classes: (1) those brought about by the erosion of the ice, 
 and (2) those brought about by the deposition of the drift. 
 
 It is important to remember that the continental glacier of 
 North America developed on the surface of a rather high plain, 
 the topography of which had been shaped in large measure by 
 rain and river erosion. This is inferred from the topography of 
 the area not covered by the ice. 
 
 Changes Produced by Erosion, 
 
 1. On elevations. The ice was thick enough to pass over the 
 hills and low mountains, such as those of New England and north- 
 ern New York, within the area shown in Fig. 303. As it over- 
 spread these and lesser elevations, it wore off their tops. It re- 
 duced all points which stood up above the general surface, and so 
 tended to make the surface less rough. The general effect on 
 elevations is shown by Figs. 264 and 265. 
 
 2. In valleys. The ice also deepened the valleys through which 
 it moved. In many cases it deepened them as much as it lowered 
 the hills, or even more. In the latter case, the relief of the surface 
 was increased; but even where this was true, the roughness of the 
 surface was often diminished, for roughness depends on the fre- 
 
THE WORK OF SNOW AND ICE 275 
 
 quency with which elevations and depressions, such as hills and 
 valleys, succeed one another, and on the steepness of their slopes, 
 quite as much as on the amount of relief (Fig. 305). Where the 
 edge of an ice-sheet was differentiated into valley glaciers which 
 moved down to the sea, the ice sometimes gouged out the valleys 
 far below sea-level, giving rise to narrow bays, or fiords, after the 
 ice melted. 
 
 3. Rock basins. Another effect of ice erosion w r as to gouge 
 out hollows where the underlying rock was relatively weak. The 
 result was the formation of basins in the surface of the rock. Such 
 rock basins are probably less common in the area of the con- 
 tinental ice-sheet than in mountain valleys affected by glaciers. 
 
 FIG. 305. Diagram to show that roughness of surface and amount of relief 
 are not necessarily the same. A represents greater relief, but B might 
 be regarded as a rougher surface. 
 
 The ice also polished, striated, and grooved the surface of the rock 
 over which it moved, though these effects are not important 
 topographically. 
 
 Changes Produced by Deposition. 
 
 Sooner or later the ice deposited all of the material which it 
 eroded from the surface over which it passed. Had the drift 
 been equally thick everywhere, its effect would have been to 
 raise the surface without altering its topography; but it is dis- 
 tributed with great inequality, and this inequality modified the 
 topography. 
 
 1. General distribution of the drift. The tendency of the mov- 
 ing ice was always to transfer its drift from the point where it 
 was picked up toward the margin of the ice. In general, there 
 fore, the drift left by the continental glaciers is thicker toward 
 their former margins, and thinner toward their centers. It is very 
 thick, for example, in a belt extending from western New York 
 through Ohio, Indiana, Illinois, Wisconsin, Minnesota, and Iowa, 
 
276 
 
 PHYSIOGRAPHY 
 
 to Dakota and Montana. In considerable tracts north of the 
 boundary of the United States, on the other hand, toward the 
 center of the ice-fields, little drift was left. 
 
 FIG. 306. Terminal moraine topography near Oconomowoc. Wis. 
 (Wis. Geol. Surv.) 
 
 FIG. 307. Map showing the position of some of the principal terminal 
 moraines of the United States. 
 
 Terminal moraines. The last ice-sheet, especially, developed 
 stout terminal moraines. The position of some of them is shown 
 
THE WORK OF SNOW AND ICE 277 
 
 in Fig. 307. It will be seen that they lie well north of the southern- 
 most margin of the drift, because the ice-sheet which made them 
 did not advance so far to the south as some of its predecessors had 
 done. 
 
 As the edge of the ice was melted back, it sometimes halted for 
 a time far back from the position of its maximum advance. Beneath 
 
 FIG. 308. Bowlders on the terminal moraine of the Okanagan glacier, 
 Wash. (U. S. Geol. Surv.) 
 
 the edge in such positions, terminal moraines were made. Such 
 terminal moraines are sometimes called recessional moraines. 
 They are terminal to the ice at the time they were made, but not 
 terminal to the drift sheet as a whole. This explains why one 
 ice-sheet came to develop several terminal moraines. 
 
 The terminal moraine of an ice-sheet is not always, and per- 
 
 FIG. 309. A single bowlder in the area shown in Fig. 308. 
 (Willis, U. S. Geol. Surv.) 
 
 haps not usually, a conspicuous ridge, though it is often conspicu- 
 ous in a region of slight relief. Its topography is much more dis- 
 tinctive than its size. Its surface is often marked by hillocks, 
 mounds, ridges, etc., associated with depressions of similar shapes 
 (Figs. 291, 292 and 306). While this sort of topography is so wide- 
 spread as to be characteristic, it is not pronounced in all terminal 
 moraines. The depressions often contain ponds, lakes, or marshes. 
 
278 
 
 PHYSIOGRAPHY 
 
 Bowlders frequently abound on a surface of terminal moraines 
 (Fig. 308). 
 
 The ground moraine. The area of the ground moraine is 
 much more extensive than that of the terminal moraines, and its 
 topography is, in general, less rough. The hills and hollows are 
 
 FIG. 310. "Pilot Rock.' 
 
 A glacial bowlder near Coiile" City, Wash. 
 (Carrey.) 
 
 less steep-sided, and the curves of the surface broader (Fig. 311, 
 and PI. XIX). Portions of the ground moraine are sometimes in 
 the form of elongate or oval hills, called drumlins. Drumlins occur 
 in many places, some of the best known being in Wisconsin and 
 New York (Figs. 298 and 312-314). At the battle of Bunker Hill, 
 the Americans occupied and fortified a drumlin. 
 
 Fio. 311. Ground moraine topography. (At wood.) 
 
 Effect of drift on topography. The drift is sometimes so 
 disposed as to increase the relief of the surface (Fig. 315), but 
 oftener so as to decrease it (Fig. 290), because more drift, on 
 the whole, was left in the low places than on the high ones. On 
 the other hand, the drift was sometimes left in such a way as to 
 
PLATE XIX 
 
 / T.. Cr. :^ 
 
 Characteristic drift topography. Scale 1 mile per inch. (Eagle, Wis. 
 Sheet, U. S. Geol. Surv.) 
 
THE WORK OF SNOW AND ICE 
 
 279 
 
 FIG. 312. Drumlins in contour, near Clyde, N. Y. (U. S. Geol. Surv.) 
 
 FlQ. 313. A Wisconsin drumlin seen, from the side. Two miles north of 
 Sullivan. (Alden, tT. S. Geol. Surv.) 
 
280 
 
 PHYSIOGRAPHY 
 
 make the surface rougher than the surface of the rock below, 
 even where the relief was decreased. > 
 
 Effect of drift deposits on drainage. The drift left by the 
 ice sometimes filled valleys at some points, but not at others. 
 Drift fillings in valleys make dams, above which water is likely 
 
 FIG. 314. The same drumlin shown in Fig. 313 seen from the end. 
 (U. S. Geol. Surv.) 
 
 to accumulate, making lakes. If a valley was filled in two places, 
 as sometimes happened, the unfilled place between became a 
 basin fit for a lake. The number of lakes developed in this way 
 is very large. The finger lakes, of New York and Devil's Lake, 
 Wisconsin (Fig. 316), are good examples. 
 
 FIG. 315. Diagram to show how drift may be so disposed as to increase 
 the relief of the surface. 
 
 Rock basins have already been referred to; but it often hap- 
 pened tk&t'T&wms the bottoms of which are in rock were made 
 deeper by the deposition of drift about their rims. The Great Lakes 
 probably occupy rock basins, but their margins were built up by 
 drift, making them deeper. 
 
 The ice-sheets gave rise to lakes and ponds in other ways'a'lso. 
 
THE WORK OF SNOW AND ICE 
 
 281 
 
 Many of them occupy depressions in the surface of the drift. Such 
 lakes are especially numerous in terminal moraines, but they are 
 not rare in the ground moraine, and are by no means unknown 
 in the stratified drift. Glaciation affords the explanation of 
 the numerous lakes of North America, nearly all of which are in 
 the area which was covered by the ice-sheet or by mountain 
 
 
 FIG. 316. Sketch showing a lake in a former river valley, held in by drift 
 dams. The dotted areas are terminal moraines. 
 
 glaciers. They are most numerous in the area covered by the 
 ice of the last glacial epoch (Fig. 303), as in North Dakota, Minne- 
 sota, Wisconsin, Michigan, New York, and New England. Except 
 in special situations, where they are of wholly different types, 
 lakes do not occur south of the drift. 
 
 Some lakes developed by the ice had but a temporary exist- 
 ence. Some of them came into existence along the margin of 
 
282 
 
 PHYSIOGRAPHY 
 
 the ice-sheets, the ice itself often forming one border of the lake. 
 Such lakes disappeared when the ice melted. 
 
 One of the largest of the marginal lakes (Lake Agassiz) lay in 
 the valley of the Red River of the North (Fig. 317). When this 
 lake was largest, its length was about 700 miles, and its maximum 
 width about 250 miles. Its area was about 110,000 square miles, 
 
 ' .= ;^S 9ca / 
 
 < jff R 'l 
 
 J 
 
 FIG. 317. Map of the extinct Lake Agassiz, and a few other glacial lakes. 
 Lake Winnipeg occupies a part of the old basin of Lake Agassiz. (Upham , 
 U. S. Geol. Surv.) 
 
 or nearly one-fifth more than the combined area of all the Great 
 Lakes, but the water was not very deep. It came into existence 
 when the ice at the north obstructed drainage in that direction. 
 The water rose in the basin until it overflowed to the south. When 
 the ice at the north melted, a new and lower outlet was opened 
 in that direction, and the lake was drained. Lake Winnipeg and 
 several smaller lakes may be looked upon as remnants of this great 
 lake, for they occupy the deepest depressions in the old basin. 
 
 The borders of the former lake are marked by old beaches, 
 and locally by deltas. The silt-covered bottom of the Jake is one 
 of the most important wheat-producing areas in the United States. 
 
THE WORK OF SNOW AND ICE 
 
 283 
 
 FIG. 318. The beginning of the Great Lakes. The ice still occupied the 
 larger parts of the present lake basins. (After Taylor and Leverett, 
 U. S. Geol. Surv.) 
 
 FIG. 319. A later stage in the development of Lakes Chicago and Maumee. 
 The ice has retreated, and the outlet of Lake Maumee has been shifted. 
 (After Leverett and Taylor, U. S. Geol. Surv.) 
 
284 
 
 PHYSIOGRAPHY 
 
 The Great Lakes of the present day were greatly expanded 
 while the ice blocked their present outlets. A part of their history, 
 
 FIG. 320. The Great Lakes at the Algonquin-Iroquois stage. 
 (After Taylor.) 
 
 FIG. 321. A still later stage of the Great Lakes. The sea is thought to 
 have covered the area shaded by lines at the east. (After Taylor.) 
 
 dating from the time of the last ice occupancy, is suggested by 
 Figs. 318 to 321. The basins of these lakes did not exist, so far 
 
THE WORK OF SNOW AND ICE 
 
 285 
 
 as known, before the glacial period, but considerable rivers may 
 have flowed along the lines of their axes. From these river val- 
 leys, lake basins appear to have been developed as a result of (1) 
 the deepening of portions of the valleys by ice erosion, (2) the 
 building up of the rims of the basins by the deposition of drift, and 
 (3) perhaps the down-warping of the sites of the basins. 
 
 FIG. 322. Sketch-map showing the drainage of the upper Ohio basin as it 
 is believed to have been before the glaciation of the region. (Tight, 
 U. S. Geol. Surv.) 
 
 The irregular disposition of drift also deranged the rivers. 
 After the ice melted, the surface drainage followed the lowest 
 lines open to it, but these lines did not always correspond with 
 the former valleys, for some of them had been filled, and most of 
 them were blocked up in some places. After the ice melted, there- 
 fore, the surface waters followed former valleys in some cases, 
 and in others flowed across areas where there had been no valleys. 
 In choosing their new courses, the streams sometimes fell over 
 
PHYSIOGRAPHY 
 
 cliffs or ran down steep slopes. Thus arose falls and rapids which, 
 on the whole, are rather common in the streams of the glaciated 
 area. Figs. 322 and 323 indicate something of the changes 
 effected by the deposition of the drift in the basin of the upper 
 Ohio. 
 
 We have already seen that rapids and falls are marks of young 
 streams. Most lakes also are marks of youth. Rivers are, on 
 the whole, hostile to lakes, for outflowing waters cut down their out- 
 lets, and inflowing waters bring in sediment which, when deposited 
 in the basins, tends to fill them up. Many small lakes have already 
 
 FIG. 323. Present drainage of the area shown in Fig. 322. 
 
 become extinct in these ways, and many others have been made 
 sensibly smaller. The fact that so many falls, rapids, lakes, etc., 
 still exist within the glaciated area shows that the time since the 
 melting of the last ice-sheet has not been long enough for these 
 features to be destroyed. 
 
 Marshes also abound within the glaciated area. In some cases 
 they represent the beds of former lakes and ponds, while in others 
 they are simply basins too shallow to hold bodies cf water suffi- 
 ciently deep to prevent the growth of plants. 
 
THE WORK OF SNOW AND ICE 287 
 
 Lakes, ponds, marshes, falls, rapids, etc., are much more 
 abundant in the area covered by the last ice-sheet than in the area 
 of drift outside of the last ice-sheet. This is largely because the 
 southernmost part of the drift, as now exposed, is older, and 
 has been subject to rain and river erosion long enough for sur- 
 face drainage to destroy most of the lakes. The oldest drift 
 of the glacial period is believed to be many times (probably as 
 many as twenty-five) as old as the youngest. 
 
 Stratified drift. Valley trains, outwash plains (p. 266), deltas 
 (p. 198), etc., were developed by the continental glaciers, but 
 only those of the last ice-sheet are well preserved. Some of 
 the valley trains are long, and in some the deposits are deep. Thus 
 the Rock River in southern Wisconsin filled its valley with gravel 
 and sand to a depth of 300 to 400 feet just below the terminal 
 moraine of the last glacial epoch, while the ice-water flowed through 
 it. The Columbia River, swollen by the waters from the melting 
 ice, filled its valley, locally, to the depth of 700 feet with material 
 washed out from the ice. 
 
 Since the ice melted, most of the valley trains have been par- 
 tially carried away, and the remnants of the old plains of aggrada- 
 tion are now terraces (Fig. 217). 
 
 Outwash plains (p. 266) also are extensive. Thus much of the 
 southern part of Long Island is an extensive outwash plain, spread- 
 ing southward from the terminal moraine which makes the back- 
 bone of the island. Just southeast of Brooklyn the outwash plain, 
 in preference to the hills of the adjacent moraine, has been occupied 
 by several suburbs. 
 
 Kames and eskers (p. 267) also diversify the surface of the 
 drift at numerous points, the former being far more numerous than 
 the latter (Figs. 301 and 302). 
 
 Effects of glaciation on human affairs. The effects of glacia- 
 tion have had much influence on the industrial history of the region 
 which the ice covered. 
 
 The increase of mantle rock in the United States, as a result of 
 glaciation, is of significance. This increase is of value in regions 
 where slopes were considerable, for where such slopes are found 
 in driftless regions, the soil is often very thin or absent, and the 
 area of arable land is thereby restricted. Abundant soil is much 
 more likely to be found on similar slopes in the glaciated area. 
 Furthermore, since the general effect of glaciation was to reduce 
 
288 PHYSIOGRAPHY 
 
 slopes, it tended to reduce the areas where the slopes were too 
 steep to be cultivated. 
 
 Again, the quality of the soil was improved in many places by 
 glaciation, but this is not true everywhere. It is worth noting 
 that most of the wheat and hay grown in the United States east 
 of the Rocky Mountains, are within the area which was glaciated. 
 This is probably not altogether because of the drift, but partly 
 because of the climate. 
 
 The reduction of roughness and the smoothing of slopes effected 
 by glaciation made the construction of roads easier, and so, on the 
 whole, has facilitated transportation. Locally, however, the sur- 
 face was made rougher, with disadvantageous results. 
 
 The falls, rapids, and lakes which resulted from glaciation 
 have increased the water-power, and the lakes, ponds, and marshes 
 which serve as reservoirs have tended to equalize the flow of the 
 streams throughout the year. The flow of streams from lakes is 
 much steadier than the flow of streams which have no permanent 
 reservoirs to draw upon. The drift is much thicker, on the whole, 
 than the mantle rock of other regions. This greater thickness of 
 loose material on the surface tends to hold back the rain-water 
 after it falls, the porous drift itself serving as a sort of reservoir 
 which yields up the water slowly. 
 
 The economic significance of lakes is noted elsewhere (p. 316). 
 
 The drift materials are somewhat extensively utilized. Thus 
 much of the drift clay (rock-flour) is used for the manufacture of 
 brick, tile, etc., and the gravel is used for road-making, and in the 
 manufacture of various sorts of cements. 
 
 Such are some of the beneficent results of glaciation. There 
 are also some considerations on the other side. 
 
 In some places the quality of the soil has been injured, for 
 in many areas the drift is stony, and great labor is necessary to put 
 it in workable condition. In some places, too, it is too sandy 
 or gravelly to make good soil, and in other places its surface is 
 too rough to allow of successful tilling. In still other situations, 
 as in much of New England, the ice left a thin stony mantle of 
 drift covering a rough hilly surface. This, combined with a some- 
 what unfavorable climate, made agriculture unprofitable in much of 
 this region, and so favored the early development of the fisheries, 
 and, together with abundant water-power, has made New England 
 a manufacturing rather than an agricultural region. 
 
THE WORK OF SNOW AND ICE 289 
 
 In spite of these adverse considerations, it seems probable, on 
 the whole, that the glaciated area of the United States was con- 
 siderably benefited by the work of the ice. 
 
 MAP EXERCISE. 
 Maps /or the Study of the Topographic Effects of Glaciation. 
 
 I. The maps to be studied in preparation for conference: 
 
 1. Lancaster, Wis. la. 8. Passaic, N. J. 
 
 2. Eagle, Wis. 9. Palmyra, N. Y. 
 
 3. Whitewater, Wis. 10. Canada Lake, N. Y. 
 
 4. Muskego, Wis. 11. Paradox Lake, N. Y. 
 
 5. Geneva, Wis. 12. Leadville, Colo. 
 
 6. St. Croix Dalles, Wis. Minn. 13. Hayden Peak, Utah. 
 
 7. Brooklyn, N. Y. 
 
 II. Questions to be answered in writing: 
 
 1. Contrast the Lancaster Sheet (which represents a part of the 
 "Driftless Area") with the Eagle Sheet (a glaciated area), and indi- 
 cate the essential ways in which the two areas differ topographically. 
 
 2. Locate (a) terminal moraine belts and (6) outwash plains on at 
 least two maps. 
 
 3. Interpret the peculiar topography shown on the Palmyra, N. Y., 
 Sheet. How were the numerous hills formed, what are they called, and 
 why do they all trend in the same direction? 
 
 4. Select three maps showing regions whose present topography 
 is controlled largely by the drift; three where the topography is con- 
 trolled chiefly by the underlying rock. 
 
 5. Under what conditions does drift control present topography? 
 Under what conditions does the underlying rock control topography 
 within the glaciated area? 
 
 6. The probable origin of the lakes on 
 
 (a) The Paradox Lake Sheet. 
 
 (6) The Eagle and Geneva sheets. 
 
 (c) The Leadville, Colo., Sheet, especially Twin Lakes. 
 
 7. What altitude appears to have been necessary to develop glaciers 
 in the area of the Leadville and Hayden Peak sheets? 
 
 REFERENCES. 
 A. EXISTING GLACIERS. 
 
 1. CHAMBERLIN AND SALISBURY, Geologic Processes, Vol. I, Chapter V: 
 Henry Holt & Co., 1903; and other standard text-books on Geology. 
 
 2. RUSSELL, Glaciers of North America: Ginn & Co., 1897. 
 
 3. RUSSELL, Glaciers of Mount Rainier: 18th Ann. Rept. U. S. Geol. 
 Surv., Pt. II, 1896-1897, pp. 349-415. 
 
290 PHYSIOGRAPHY 
 
 4. Alaskan Glaciers. REID, 16th Ann. Kept. U. S. Geol. Surv., Pt. I, 
 1894-1895, pp. 421-459, and Nat. Geog. Mag., Vol. IV, 1891, pp. 19-55. 
 RUSSELL, Nat. Geog. Mag., Vol. Ill, 1891, pp. 176-188; Jour, of Geol., Vol. I, 
 pp. 219-245; and 13th Ann. Rept. U. S. Geol. Surv., Pt. II, 1891-1892, pp. 
 7-91. GILBERT, Glaciers, Harriman Alaskan Expedition: Doubleday, Page 
 & Co. 
 
 5. Glaciers of Greenland. CHAMBERLIN, Jour, of Geol., Vol. II, 1894, 
 pp. 649-666 and 768-788; Vol. Ill, 1895, pp. 61-69, 198-218, 469-480, 
 565-582, 668-681, and 833-843; and Vol. IV, 1896, pp. 582-592. SALISBURY, 
 Jour, of Geol., Vol. IV, 1896, pp. 769-810. 
 
 6. SHALER AND DAVIS, Glaciers: James R. Osgood & Co., 1881. 
 
 7. Variations of Glaciers. REID, Jour, of Geol., Vol. Ill, 1895, pp. 278- 
 288; Vol. V, 1897, pp. 378-383; Vol. VI, 1898, pp. 473-476; Vol. VII, 1899, 
 pp. 217-225; Vol. IX, 1901, pp. 250-254; Vol. X, 1902, pp. 313-328; Vol. 
 XI, 1903, pp. 285-288; Vol. XII, 1904-, pp. 252-263; and Vol. XIII, 1905, 
 pp. 313-318. GILBERT, Bull. Sierra Club, Vol. V, pp. 20-25. 
 
 8. TYNDALL, The Glaciers of the Alps: Murray, 1860. 
 
 B. GLACIER MOTION. 
 
 9. CHAMBERLIN AND SALISBURY, Geologic Processes, Vol. I, pp. 308-323: 
 Henry Holt & Co., 1903. 
 
 10. REID, Mechanics of Glacier Motion: Jour, of Geol., Vol. IV, p. 912. 
 
 11. AITKIN, Am. Jour. Sci., Vol. V, 1873, p. 305; Vol. XXXIV, 1887, 
 p. 149; and Nature, Vol. XXXIX, 1888, p. 203. 
 
 12. RUSSELL, The Influence of Debris on the Flow of Glaciers: Jour, of 
 Geol., Vol. Ill, p. 823. 
 
 C. RESULTS OF GLACIATION. 
 
 13. CHAMBERLIN AND SALISBURY, Earth History, Vol. Ill, pp. 327-446: 
 Henry Holt & Co., 1906. 
 
 14. SALISBURY, The Drift: Jour, of Geol., Vol. II, pp. 708-724 and 
 837-851, and Vol. Ill, pp. 70-97. 
 
 15. CHAMBERLIN, Genetic Classification of the Drift: Jour, of Geol., Vol. II, 
 pp. 517-538. 
 
 16. CHAMBERLIN, A Preliminary Paper on the Terminal Moraines of 
 the Second Glacial Epoch: 3d Ann. Rept. U. S. Geol. Surv., 1881-1882, 
 pp. 295-401. 
 
 17. J. GEIKIE, The Great Ice Age, 3d Ed.: D. Appleton & Co., 1895, and 
 Earth Sculpture, Chapters X and XI: Putnams. 
 
 18. A. GEIKIE, Scenery of Scotland, chapters on Glacial Action: Macmillan 
 Co., 1887. 
 
 19. Folios of the U. S. Geol. Surv., especially those of the high mountains 
 of the West. 
 
 20. LEVERETT, Illinois Glacial Lobe: Mono. XXXVIII, U. S. Geol. 
 Surv.; Glacial Formations, etc., of the Erie and Ohio Basins: Mono. XLI, U.. 
 S. Geol. Surv. STONE, Glacial Gravels of Maine: Mono. XXXIV, U. S. Geol. 
 Surv. 
 
THE WORK OF SNOW AND ICE 291 
 
 21. SALISBURY, Glacial Geology of New Jersey: N. J. Geol. Surv., Vol. V, 
 1902. 
 
 22. CHAMBERLIN, Rock Scorings of the Great Ice Age: 7(h Ann. Kept. U. S. 
 Geol. Surv., pp. 155-248, 1885-1886. 
 
 D. EXPLORATION IN POLAR REGIONS. 
 
 23. PEARY, Northward over the Great Ice: Fred. A. Stokes Co., 1898. 
 
 24. NANSEN, First Crossing of Greenland: Longmans, Green & Co., 1890; 
 and Farthest North: Harper & Bros., 1897. 
 
 25. Most other books on the Arctic and Antarctic regions give some 
 account of the ice. 
 
 E. GLACIAL LAKES. 
 
 26. TAYLOR, Short History of the Great Lakes, in Studies in Indiana 
 Geography: Inland Publishing Co., Terre Haute, Ind., 1897. See also 20 above. 
 
 27. CHAMBERLIN AND SALISBURY, Earth History, Vol. Ill, pp. 394-403: 
 Henry Holt & Co., 1906. 
 
 28. UPHAM, Glacial Lake Agassiz: Mono. XXV, U. S. Geol. Surv. 
 
 29. SALISBURY AND KUMMEL, Lake Passaic, an Extinct Glacial Lake: 
 Geol. Surv. of N. J.; Ann. Rept. of the State Geologist for 1893. 
 
 30. FAIRCHILD, Glacial Genesee Lakes: Bull. Geol. Soc. Am., Vol. VII, 
 pp. 423-452. 
 
CHAPTER VI 
 LAKES AND SHORES 
 
 GENERAL FACTS 
 
 Definition. In general, a lake is an inland body of standing 
 water larger than a pool or a pond; but the term is sometimes 
 applied to the widened parts of rivers (Fig. 324), and sometimes 
 
 i 
 
 FIG. 324. Lake Pepin, a widened part of the Mississippi River between 
 Wisconsin and Minnesota. Maximum width about 2A miles. The 
 widening of the river is apparently due to the detritus brought down 
 by the Chippevva River and deposited in the Mississippi. (Miss. Riv. Com. ) 
 
 to bodies of water which lie along coasts, even when they are at 
 sea-level, and sometimes when they are in direct connection with 
 the sea (PI. XX). 
 
 The distinction between lakes and similar bodies of water 
 
 292 
 
LAKES AND SHORES 293 
 
 which are not lakes is rather arbitrary. The amount of widening 
 which a river must undergo before it is called a lake is as arbitrary 
 as the size which the body of standing water must attain before 
 this name may be applied to it. In the interior of the United 
 States, a pond is usually understood to mean a body of water 
 smaller or shallower than a lake; but this usage is not universal, 
 for some beautiful lakes (for example, Green Pond in New Jersey) 
 are called ponds. Ponds and lakes differ from inland seas, bays, 
 and lagoons (1) in being more completely (in most cases altogether) 
 shut off from the ocean, and (2) in being for the most part at a 
 level above very rarely below that of the sea; but between bays 
 and lagoons which are nearly enclosed, and coastal lakes, there 
 are all possible gradations. 
 
 Most lakes are fresh, but a few, like Great Salt Lake and the 
 Dead Sea, are much more salt than the sea itself. 
 
 Distribution of Lakes 
 
 1. In latitude. Lakes occur in most latitudes, but they are 
 more abundant in high latitudes than in low. They do not abound, 
 however, in all high latitudes. Northern Asia, for example, has 
 relatively few. This distribution of lakes is connected with 
 former glaciation, a connection which has already been pointed 
 out. 
 
 2. In mountains. Lakes are abundant in some mountain regions 
 but not in all. They are numerous in the western mountains of 
 the United States, especially at the North, but they are essentially 
 absent from the Appalachian Mountains south of northern Penn- 
 sylvania. They are, in general, more abundant in high moun- 
 tains than in low ones, and if they occur in low mountains at 
 all, it is likely to be in high latitudes only. In other words 
 lakes are common in mountains which have been recently 
 glaciated. 
 
 3. Along rivers. Another situation where lakes occur, though 
 less commonly, is along rivers; but they do not occur along all 
 rivers. Outside of high latitudes and high mountains (glaciated 
 regions), lakes are common only along streams which have low 
 gradients and wide flats. There are numerous lakes, for example, 
 on the alluvial plain of the Mississippi (Fig. 197), and on the 
 flats of some of its tributaries, such as the Missouri and the Red 
 
294 
 
 PHYSIOGRAPHY 
 
 River of Louisiana (Fig. 325). In many of these cases the origin 
 of the lakes is clearly connected with the changing of the river 
 channel (Fig. 197). 
 
 4. Along coasts. Another situation where lakes are of rather 
 common occurrence is along coasts (PI. XX and Fig. 326) , though 
 many coasts are without them. Coastal lakes stand in no 
 apparent relation to latitude and are always at low altitudes. 
 
 . : L:Cruss .. '' 
 
 . ''.'.''. . ShrevQport 
 
 FIG. 325. Lakes along the Red River of Louisiana. The lakes are at the 
 lower ends of the tributary streams. 
 
 The level of the water in them is often nearly or quite the same 
 as that of the sea. 
 
 5. On coastal plains. Low lands in proximity to the sea, 
 though back from the coast, are sometimes affected by lakes, 
 especially if the climate is moist. This is illustrated by Florida 
 (Fig. 2, PI. XX), where the number and abundance of lakes is 
 comparable to that of equal areas in the northern part of the 
 United States. 
 
 6. Glaciated plains and plateaus. Plains and plateaus which 
 have been recently glaciated are likely to abound in lakes, as 
 indicated in the preceding chapter. 
 
 7. On plateaus. Lakes sometimes occur on plateaus even 
 where there has been no glaciation. The most considerable 
 
PLATE XX 
 
 FIG. 1. Coastal lakes formed by the blocking of the ends of drowned valleys. 
 Scale 1 mile per inch. (Marthas Vineyard, Mass., Sheet, U. S. Geol. 
 SunO 
 
 Fio. 2. A group of lakes on the coastal plain of Florida. Scale 1 mile 
 oer inch. (Williston Sheet, U. S. Geol. Surv.) 
 
PLATE XXI 
 
 The upper end of Seneca Lake, New York. The flat between Montour Falls 
 and Watkins is a delta which has been built out into the lake by the 
 inflowing creek. Scale 1 mile per inch. (Watkin's Sheet, U. S 
 Geol. Surv.) 
 
LAKES AND SHORES 
 
 295 
 
 examples are the great lakes of central and southern Africa; but 
 there are not a few lakes, mostly shallow, on the Great Plains (pla- 
 teaus) of the United States, even where the rainfall is not great. 
 
 8. Other situations. Lakes occur in a few other situations, as 
 in the tops of some volcanic mountains, and on plains which have 
 not been glaciated and which are far from the coast. 
 
 FIG. 326. Lakes on the coaet of New Jersey. 
 
 Area, Topographic Position, Depth, etc. 
 
 Area and topographic position. Lake Superior, Lake Huron, 
 Lake Michigan, Lake Erie, and Lake Ontario are examples of 
 great lakes. These lakes, indeed, constitute the greatest chain 
 of lakes in the world, and have an aggregate area of nearly 95,000 
 
296 PHYSIOGRAPHY 
 
 square miles. Five of the great lakes of the Dominion of Canada 
 have a combined area of more than 32,000 square miles. All these 
 large lakes lie at relatively low altitudes. 
 
 Lakes have a great range in altitude as well as in size. The 
 Yellowstone Lake is the highest lake of much size in the United 
 States. It is 7738 feet above sea-level, and its area 140 square 
 miles. Lake Titicaca, the largest lake (except Lake Maracaibo) 
 in South America, is both higher (12,500 feet) and larger (3200 
 square miles). The surfaces of a few great lakes are below sea- 
 level. This is true of the Caspian Sea ( 85 feet), the Dead Sea 
 (-1268 feet), and the Sea of Tiberias (-682 feet). 
 
 Depth. Most lakes are rather shallow. The number in which 
 the water is less than 50 feet deep probably far exceeds the num- 
 ber in which the depth is greater; but some lakes are exceedingly 
 deep. It need hardly be said that the popular notion that many 
 lakes are bottomless, -is without foundation. Many of the lakes 
 which are locally believed to be without bottom are, indeed, 
 shallow. 
 
 Lake Superior has a maximum depth of about 1000 feet, and 
 Lakes Michigan, Huron, and Ontario all have depths exceeding 
 700 feet. Lake Erie, on the other hand, is much shallower, its 
 maximum depth being only about 200 feet. 
 
 A few lakes have much greater depths. The deepest, so far as 
 known, is Lake Baikal, in Siberia, which is stated to have a maxi- 
 mum depth of about 4700 feet, or about one-seventh the depth 
 of the deepest part of the ocean. The Caspian Sea, really a lake, 
 is next deepest, and has a maximum depth of about 3200 feet. 
 Other lakes of great depth are Crater Lake, Oregon, about 2000 
 feet; Lake Tahoe, California, 1645 feet; Lake Chelan, Washington, 
 about 1500 feet; and lakes Maggiore, Como, and de Garda, in 
 northern Italy, and the Dead Sea, all of which have depths of 
 more than 1000 feet. 
 
 The bottoms of most lakes are well above sea-level; but in 
 exceptional cases their bottoms are far below. The lowest point 
 in the bottom of the Caspian Sea is a little more than 3000 feet 
 below sea-level, and the lowest point in the basin of Lake Baikal 
 is nearly as far down. The lowest point in the bottom of Lake 
 Ontario is about 500 feet below sea-level, in Superior about 400 feet, 
 and in Lake Chelan more than 400 feet. Except in lakes along 
 the coast, the bottoms of the small lakes are rarely so low as sea- 
 
LAKES AND SHORES 
 
 297 
 
 level; but the bottoms of the three north-Italian lakes mentioned 
 are all several hundred feet below the ocean. 
 
 Various facts concerning the lakes which exceed 10,000 square 
 miles in area, and concerning a few others, are given in the 
 following table, though few lakes have such remarkable dimen- 
 sions: 
 
 Name. 
 
 Approximate 
 area in square 
 miles. 
 
 Approximate 
 altitude of 
 surface. 1 
 
 Approximate 
 maximum depth 
 in feet. 
 
 Caspian 
 
 170,000 
 
 -85 
 
 3,200 
 
 Superior. . .... 
 
 31,200 
 
 602 
 
 1,008 
 
 Victoria Nyanza 
 
 26,000 
 
 3,800 
 
 240 
 
 Aral . 
 
 25,050 
 
 160 
 
 1,200 
 
 Michigan 
 
 22,500 
 
 581 
 
 870 
 
 Huron 
 
 22,320 
 
 581 
 
 700 
 
 Nyassa. . . . . . . -. . . 
 
 14,200 2 
 
 1,500 
 
 2,300 
 
 Baikal 
 
 13,000 
 
 1,700 
 
 4,700 3 
 
 Tanganyika 
 
 12,000 
 
 2,700 
 
 2,100 
 
 dreat Bear 
 
 11,200 
 
 390 
 
 270 
 
 Erie 
 
 9,960 
 
 573 
 
 200 
 
 Winnipeg. . 
 
 9,400 
 
 710 
 
 70 
 
 Balkash 
 
 8600 
 
 900 
 
 80 
 
 Ontario 
 
 7,240 
 
 247 
 
 738 
 
 Chad 
 
 6 OOO 4 to 40,000 
 
 900 
 
 8 to 20 
 
 Titicaca 
 
 3,200 
 
 12,500 
 
 700 
 
 Dead Sea 
 
 360 
 
 -1,268 
 
 1,300 s 
 
 'Garda 
 
 189 
 
 215 
 
 1,135 
 
 Chelan 
 
 85 
 
 1,079 
 
 1,500 
 
 Como 
 
 60 
 
 650 
 
 1,340 
 
 Crater. . 
 
 25 
 
 6,239 
 
 2,000 
 
 
 
 
 
 Great as is the depth of the water in some of the lakes, the 
 ^shape of their basins is often very different from that which might 
 be imagined from the mere statement of the depths. Fig. 327 
 represents the cross-sections of the basins of some of the Great 
 Lakes, drawn to scale. The basins of many smaller lakes (Fig. 
 328) are much more striking in cross-section. 
 
 Volume. No accurate estimate of the volume of water in 
 lakes has ever been made, but their combined volume is insignifi- 
 cant when compared with the sea. If the water of all lakes were 
 added to the ocean, it would probably not raise its surface two 
 feet. 
 
 1 The negative sign means below sea-level. 
 
 2 Sometimes given as low as 10,200. 
 
 * Depth of 5,618 feet recently reported. 
 4 Range between wet and dry seasons. 
 6 Sometimes given as low as 1,171 feet. 
 
298 
 
 PHYSIOGRAPHY 
 
 
 a 
 
 a 
 
 O 
 
 ' 
 
 43 
 1 
 
 c3 
 
 
 1 
 
 
 1 
 
 
 X 
 9 
 
 
 Q 
 
 
 "3 
 
 o 
 
 
 -3 
 g 
 
 
 2 
 
 (4 
 
 
 
 
 s 
 
 
 & 
 
 O 
 
 p. 
 
 o 
 
 OQ 
 
 I 
 
 
 e 
 
 
 1 
 
 
 
 '53 
 
 tt 
 
 Hi 
 
 2 
 
 V 
 
 
 o 
 
 ,8 
 
 
 HJ 
 
 
 O 
 
 
 
 
 
 "o 
 
 
 .0 
 
 
 1 
 
 
 
 CO 
 
LAKES AND SHORES 
 
 299 
 
 I 
 
 to 
 
 sni-3.5 
 
 73 '~ c a 
 
 ^2|^ 
 +==0.2 Sis 
 
 c X s C 
 
 c73 H O 
 
 - 
 
 eg 
 
 *- 3 (N 
 
 1 a 
 
PHYSIOGRAPHY 
 
 Movements of lake water. The waters of all lakes are af- 
 fected by waves, and the waters of many lakes by movements of 
 other sorts. In some lakes there is a more or less well-defined 
 system of currents (Fig. 329) or drifts. A sudden change of 
 atmospheric pressure on one part of a large lake causes changes 
 of level everywhere. If the pressure is increased in one place, 
 the surface of the water there is lowered and the surface else- 
 where correspondingly raised. If the change is one which les- 
 sens the pressure locally, the water surface beneath the lessened 
 
 FIG. 329. Diagram showing the currents in the Great Lakes. 
 (U. S. Weather Bureau.) 
 
 pressure rises, while it falls elsewhere. Once these changes are 
 set up, there is some pulsation of the water-level before equilibrium 
 is again established. These movements are called seiches. Seiches 
 have been much studied in the Swiss lakes. In very large lakes 
 tides may be observed, though they are not usually detected except 
 by instruments devised to record them. Slumping about the 
 shores of a -lake, earthquakes, etc., also cause movements of its 
 waters. 
 
 Changes of level. The levels of lakes without surface outlets 
 change notably from time to time, according to the amount of 
 
LAKES AND SHORES 301 
 
 precipitation on their surroundings. Many small lakes rise several 
 feet in wet weather, and fall correspondingly in drought. 
 
 Conditions Necessary for the Existence of Lakes 
 
 The conditions necessary for the existence of lakes are (1) 
 depressions without outlets, and (2) a sufficient supply of water. 
 
 Depressions without outlets must not be understood to mean 
 that lakes have no outlets. It means that, below the level of the 
 lake-outlet, there is, in each case, a depression which has no outlet. 
 It is this depression without an outlet which holds the water which 
 makes the lake. When the water reaches the top of this depres- 
 sion it overflows. 
 
 A sufficient supply of water means enough so that water con- 
 stantly 1 remains in the depression. If the bottom of a basin be of 
 porous material, such as gravel, more water may be necessary, in 
 order that water may stand in the depression, than if the bottom 
 be of compact material like clay. If, however, the ground-water 
 surface in the surroundings of the basin is above the level of the 
 bottom of the basin (Fig. 330), the water will not escape from the 
 
 FIG. 330. If the water table about a lake is above the lake level, there will 
 be no leakage from the lake, even if its basin be of porous material. 
 
 basin, even if the latter be of porous material. The humidity of the 
 atmosphere also affects the amount of water necessary for a lake. 
 In moist regions, most considerable depressions without outlets 
 contain lakes, while those of arid regions are often lakeless. 
 
 The sources of lake water. The sources of lake water are 
 rain, melting snow and ice, springs and rivers, and immediate 
 run-off. Since springs and rivers are dependent upon rain and 
 snow, the source of lake water may be said to be atmospheric 
 precipitation. 
 
 1 There are temporary lakes, in the basins of which water is not always 
 present. 
 
302 PHYSIOGRAPHY 
 
 Changes now taking Place in Lakes 
 
 Various changes are now in progress in all lakes, and a study 
 of these changes throws light on the past and the future of lakes. 
 
 The filling of their basins. Lake basins are being filled con- 
 stantly. In many cases, the filling will in time obliterate the 
 basin, and the lake will then disappear. The filling takes place 
 in various ways. 
 
 1. In the first place, all streams and other surface waters which 
 flow into lakes bring sediment, and essentially all this sediment 
 is left in their basins. This is shown by the fact that the streams 
 which flow from lakes are usually clear, even when those which 
 enter bear much sediment. 
 
 In some lakes deltas are being built, and a delta reduces the 
 area of the lake in which it is built. In rare cases, deltas have been 
 built across the middle parts of narrow lakes, separating the one 
 lake into two, as at Interlaken, Switzerland. Deltas have been 
 built at the ends of some of the Finger Lakes of New York, as 
 shown in PI. XXI, shortening them sensibly. Some of the impor- 
 tant cities about these lakes are on deltas. Ithaca is an example. 
 Sheet-floods and all rain-wash which enters a lake, even though 
 not organized into streams, bring detritus to the basin and help 
 to fill it. 
 
 2. Lake basins are also being filled by the work of the waves. 
 The waves of lakes are cutting into their shores at some points 
 most of the time, and most of the material thus worn from the 
 land is deposited in the lake basin. The lake may extend its area 
 by wave-cutting, but most of the material worn away from the 
 shore is deposited in the lake basin. 
 
 Rivers and waves are the principal agents which are filling 
 lake basins and diminishing the volume of their waters, but they 
 are not the only ones. 
 
 3. Numerous shell-bearing animals live in lakes. The ma- 
 terial for their shells is extracted from the water, and the shells 
 are left on the bottom when the animals die, thus helping to fill 
 the basins. 
 
 4. Plants grow in lakes, especially in their shallow borders, 
 and. the organic matter helps to fill the basins when the plants 
 die. Some lake plants secrete lime carbonate and others silica, 
 
LAKES AND SHORES 303 
 
 and these materials help to fill the basin the same as the shells of 
 animals. 
 
 5. Winds blow dust and sand from the land into lakes, and 
 thus help to fill their basins. 
 
 In all these ways the lake basins are being gradually filled. 
 
 The lowering of their outlets. Most lake basins are being 
 affected in another way. The water flowing out of a lake cuts down 
 the level of the outlet, and, as this is lowered, the depth of the 
 basin below the outlet is diminished. The limit to which the out- 
 flowing water may cut the outlet of the lake is base-level. 
 
 Fate of lakes. A lake may be destroyed by the lowering of 
 its outlet, if its bottom is sufficiently high; but where the bottom 
 is below sea-level, river erosion could never cut the outlet down 
 to it. In such cases, filling and cutting may accomplish what 
 cutting alone could not. As a result of these processes, all existing 
 lakes must ultimately become extinct. In their destruction, rivers 
 are probably the most important agents. Their relation to lakes is 
 such as to have led to the epigram: "Rivers are the mortal enemies 
 of lakes." This has especial reference to lakes which are not on 
 valley flats. 
 
 Lakes are occasionally destroyed by drying up. This some- 
 times results from a change of climate, but it may also result from 
 a diversion of inflowing waters. 
 
 The Origin of Lake Basins 
 
 Lake basins originate in many different ways. Most of them 
 are the result of gradational processes, but some of them are due 
 to vulcanism and some to diastrophism, and, while these latter 
 topics have not yet been studied, we may anticipate their con- 
 sideration sufficiently to note the ways in which they produce 
 depressions without outlets in the surface of the land. 
 
 Diastrophism. This term includes all crustal movements, 
 whether up or down. Movements of the earth's crust give rise to 
 basins in various ways. There are often basin-like depressions 
 beneath the shallow water over the continental shelves. If such 
 areas were converted into land, either by their own rise or by the 
 withdrawal of the sea from them, the basins would appear on the 
 surface of the new land. Newly emerged portions of the sea 
 bottom are therefore regions where lake basins sometimes occur. 
 
304 PHYSIOGRAPHY 
 
 Some of the lake basins of ' Florida- '"and of the plains of Siberia 
 perhaps arose in this way. At the outset, lakes in such basins 
 would be salt, but they might become fresh (p. 314). 
 
 A lake basin may arise by crustal warpinj within land areas. 
 Thus, if a portion of a flat area were warped downward, while its 
 surroundings were not, there would be a basin, and. under proper 
 climatic conditions, it might become the site of a lake. 
 
 Lake basins may originate by the warping of a river valley, 
 where the warping leaves one part of the valley higher than a part 
 farther up-stream. The up-warp constitutes a dam, and the waters 
 above are ponded (p. 174) and a lake produced. This origin has 
 been assigned to the basin of Lake Geneva in Switzerland. Lakes 
 produced in this way are likely to be but short-lived, for in most 
 cases the outflowing water would soon cut down the obstruction. 
 
 A portion of a valley may sink through faulting, thus giving 
 rise to a basin. Such an origin has been ascribed to the basin of 
 the Dead Sea and to certain lakes in Oregon (Figs. 39 and 331). 
 
 FIG. 331. Section showing the structure of the rock about Abert and 
 Warner Lake, Oregon. (U. S. Geol. Surv.) 
 
 According to one interpretation, Lakes Stefanie, Rudolf, Albert. 
 Tanganyika, Leopold, and Nyassa, in Africa, all lie in a great rift 
 (or sunken) villey. 
 
 Lakes have probably originated at various times in the past 
 when mountains have been made by the folding of rock strata. 
 Wherever two parallel folds are developed in the surface of the 
 lithosphere, a trough is formed between them. Such troughs 
 have probably sometimes been deeper in the middle than at either 
 end, and lakes have probably resulted. Lakes made in this way 
 would be short-lived, as a rule, since they are situated favorably 
 for receiving abundant drainage, and the outflowing water would 
 soon reduce the outlets so as to drain them. It is sometimes dif- 
 ficult to distinguish between basins produced by faulting and 
 those produced by warping, especially if the solid rock is covered 
 by a heavy mantle of drift or soil. Thus lake basins are known 
 to have originated during earthquakes, sometimes it may be by 
 warping, but perhaps more commonly by faulting. In 1811 and 
 
LAKES AND SHORES 
 
 305 
 
 1812 a considerable area in the Lower Mississippi Valley sank dur- 
 ing a time of earthquakes. One of the areas most depressed be- 
 came the site of Reelf oot Lake, which lies in the 
 flat of the Mississippi, partly in Tennessee and 
 partly in Kentucky. 
 
 Vulcanism. In the tops of some extinct 
 volcanic mountains there are basins known as g 
 craters, some of which are occupied by lakes, g 
 Such lakes occur near Rome (Lake Nemi), | 
 Naples (Lake Averno), and in France. Ponds a 
 or small lakes are known in craters in many | 
 other places, even in such dry regions as Nevada Ji 
 and northern Arizona. Streams of lava some- 
 times obstruct river valleys, giving rise to lakes. 
 Snag Lake, in California, and Tiberias Lake, in 
 the valley of the Jordan, are illustrations. 
 Others occur in France and in other regions of 
 recent volcanoes. Crater Lake in Oregon (Figs. 
 333 and 334), a lake five or six miles in 
 diameter and 2000 feet deep (p. 297), occupies 
 a basin or caldera made by the sinking of the 
 top of a volcanic mountain. While it lies in a 
 depression in the top of an extinct volcano, its 
 basin is really due to diastrophism rather than 
 vulcanism. This lake is of such extraordinary 
 interest that the area about it has been set off 
 as a National Park. The general conception of 
 its history is represented by the hypothetical 
 Figs. 335 and 336, the former representing the 
 volcanic mountain as it is supposed to have 
 been before the top sank in, and the latter the 
 present basin, free from water. The island in 
 Figs. 333 and 334 is a small volcanic cone 
 developed since the sinking of the top. Basins 
 .which become the sites of lakes occasionally 
 Idevelop on the surfaces of lava-flows, perhaps 
 as the result of the changes of surface incident 
 ' to cooling. 
 
 Gradation. Various agents of gradation produce lake basins, 
 and some of them produce them in several different wavs. 
 
306 
 
 PHYSIOGRAPHY 
 
 FIG. 333. Map of Crater Lake, Ore. Contour interval 200 feet. Sound- 
 ings in feet. Lake surface 6239 feet above the sea-level. 
 (U. S. Geol. Surv.) 
 
 FIG. 334. Western border of Crater Lake from Victor Rock to Llao Rock. 
 
 (U. S. Geol. Surv.) 
 
LAKES AND SHORES 
 
 307 
 
 FIG. 335. Mount Mazama (the name given to the former mountain where 
 Crater Lake now is), as it is conceived to have been before the collapse 
 which gave rise to the lake basin. (U. S. Geol. Surv.) 
 
 FIG. 336. The r?m of Crater Lake. (U. S. Geol. Surv.) 
 
308 PHYSIOGRAPHY 
 
 i. River lakes. Reference has already been made (p. 293) to 
 lakes on the flood plains of streams, formed by the meandering and 
 later the cutting off of the streams; but rivers give rise to lakes 
 in other ways. If a tributary brings to its main more sediment 
 than the latter can carry away, the excess is deposited as an obstruc- 
 tion, and ponds the water above (Fig. 324). If a main stream 
 aggrades its channel, it tends to obstruct the inflow of its tribu- 
 taries, giving rise to lakes along them. This has been the com- 
 monly accepted explanation of the lakes along the Red River of 
 Louisiana (Fig. 325); but it now appears that the obstructions 
 to the tributaries are due to organic accumulations, rather than 
 to sediment in the ordinary sense of the term. 1 
 
 It is well known that "rafts" sometimes form in streams. The 
 "rafts" are jams formed of timber which falls into the river as the 
 result of the caving in of the forested banks, due to lateral plana- 
 tion of the meandering stream. The trees thus floated down the 
 stream lodge against the banks at favoring points, and once this 
 lodgment is started, the jam continues to grow. The branches 
 of the trees greatly aid in the growth of the raft by helping to 
 catch and hold the floating trees. 
 
 The Red River is known to have been the site of a great raft. 
 It commenced to form somewhere below Alexandria (Fig. 336a), 
 and its head had reached the vicinity of Alexandria by the latter 
 part of the sixteenth century. The raft was really a series of 
 more or less disconnected jams, each completely filling the river. 
 The effect of the early jams was to pond the water of the river 
 above, and to force it out of its old channel through low places 
 in its banks. The whole river was thus diverted to a new course 
 below Alexandria (Fig. 336&). Driftwood accumulated about the 
 new outlet, forming another jam, and in this way the raft gradu- 
 ally grew until it extended itself up the river about 160 miles. 
 Between 1820 and 1872, its average rate of growth was about 
 four-fifths of a mile per year, but in two instances accumulations 
 of over five miles are recorded during freshets. As the raft grew 
 up-stream,' -it obstructed tributaries, and developed lakes along 
 them. 
 
 There is no good record of the lakes along the lower part of 
 the raft-ridden section. At the time of the early settlement, its 
 
 1 Veatch, A. C., Professional Pap3r 46, U. S. Gcol. Surv., 1906. 
 
LAKES AND SHORES 
 
 509 
 
 lower end was near Natchitoches, and the location of this city 
 was largely determined by the fact that the foot of the raft was 
 the head of ordinary navigation Record of the lakes along the 
 upper portion of the raft is much fuller. The group of lakes near 
 Shreveport was formed near the end of the eighteenth century. 
 Before 1873, when the raft was finally removed, it had advanced 
 almost to the Arkansas line, forming Poston Lake, the most north- 
 erly of the series (Fig. 336a). 
 
 FIG. 336a. The lakes of Red River Valley, La., at their fullest recorded 
 development. (Veatch, U. S. Geol. Surv.) 
 
 Since the removal of the raft, the river has lowered its channel 
 15 feet at a point 15 miles above Shreveport, and 3 feet at Shreve- 
 port. As a result of the deepening of the channel, tributaries have 
 lowered their valleys in the attempt to adjust themselves to their 
 main stream, and the lakes are being drained. In many places 
 land which was formerly covered by lake water is now under cul- 
 tivation. When the topographic adjustment of the tributaries has 
 
310 
 
 PHYSIOGRAPHY 
 
 been completed, other areas still submerged will also be available 
 for cultivation. 
 
 Rivers are partly or wholly responsible for a class of lakes 
 which may be called delta lakes. Lake Pontchartrain in Louisiana 
 is an example (Fig. 209). Here detritus brought down by the river 
 
 FIG. 3366. Map showing the diversion of the Red River below Alexandria. 
 The shaded areas are subject to overflow. (Veatch, U. S. Geol. Surv.) 
 
 was deposited around an area of shallow water, converting the latter 
 into a basin. Marshes and ponds and lakes are sometimes made by 
 the building of alluvial cones or fans across a valley. Lake Tulare 
 in California owes its basin to an alluvial fan made by a stream 
 (King River) descending from the Sierras. 
 
LAKES. AND SHORES 311 
 
 2. Waves and shore currents. Waves and shore currents give 
 rise to lakes by shutting in the drowned ends of valleys or other 
 bays. Illustrations are numerous along many coasts (PI. XX and 
 Fig. 337). 
 
 3. Glacial lakes. The relation between the distribution of lakes 
 and the distribution of ice in former times is so close that it cannot 
 be looked upon as accidental, and the study of lakes has shown 
 that many of their basins are due to glaciation. Glaciers give rise 
 
 FIG. 337. Maps showing lakes (ponds) along the shore of Lake Ontario, 
 shut off from the main lake by sand-bars. (U. S. Geol. Surv.) 
 
 to lake basins in many ways, some of which have already been 
 mentioned. 
 
 a. The mountain glacier descending over a steep slope often 
 digs out a basin at the base of that slope (Fig. 338). Hundreds 
 of lake basins in the valleys of the western mountains of the United 
 States, and many hi similar positions in other parts of the world, 
 were formed in this way. Such lakes are usually small. They 
 occupy rock basins, and often lie in cirques (p. 247, and PI. XVIII). 
 
 b. Where glaciers pass over rocks of unequal hardness, they 
 sometimes erode the weaker more than the stronger, thus gouging 
 out hollows. Lake basins formed in this way are common in 
 
312 
 
 PHYSIOGRAPHY 
 
 mountain valleys, and are not unknown within 'the Krea covered 
 by the ice-sheets. 
 
 Lake basins of the above types are due to glacial erosion. 
 
 c. A glacier descending a mountain valley may obstruct the 
 lower end of a tributary "valley, giving rise to a lake above. Such 
 a basin has been called an ice-barrier basin. The Marjelen See 
 in Switzerland is an example of such a lake. Many former lakes 
 due to ice barriers, for example Lake Agassiz, have become extinct. 
 
 d. By far the larger number of lake basins due to glaciation 
 arose through the disposition of the debris which the ice carried 
 
 FIG. 338. Shadow Lake, in a rock basin of glacial origin, near the head of 
 San Joaquin Valley, Sierra Nevada Mountains, Cal. (Fairbanks.) 
 
 and left on the surface when it melted. Of such basins there are 
 many varieties: (1) The terminal moraines of mountain glaciers 
 often cross and obstruct valleys, giving rise to basins and so to 
 lakes (Fig. 289). (2) In many other cases, drift is so deposited 
 with relation to a rock slope as to leave a depression between the 
 main body of drift and the rock. Basins of this sort are enclosed 
 partly by solid rock and partly by drift. Illustrations are furnished 
 by many of the lakes of the United States and Europe. (3) Drift 
 may fill a valley at tw r o points, leaving the intermediate portion 
 unfilled. The intermediate part becomes a basin and, under 
 proper conditions of water-supply, a lake. (4) Other lake basins 
 owe their origin to the unequal disposition of the drift itself. Prob- 
 ably the larger number of lake basins in the northern part of North 
 America and northern Europe are merely depressions in the surface 
 
LAKES AND SHORES 313 
 
 of the drift. Lakes whose basins are of this type are not among 
 the larger or the deeper lakes. 
 
 In some of the states within the glaciated area of North 
 America, ponds and lakes are numbered by the thousand. The 
 area of the lakes in Minnesota alone has been estimated at more 
 than 5000 square miles. 
 
 Many glacial lakes owe their origin to a combination of the 
 above conditions and relations. Here belong the Great Lakes. 
 As already indicated, these basins are probably due (1) partly to 
 glacial excavation, which gouged them out to considerable depths; 
 (2) partly to the piling up of the debris thus eroded about the 
 rims of the basins; and perhaps (3) partly to the downward warping 
 of the surface beneath the water. 
 
 Lakes due to slumping. Valleys are sometimes obstructed 
 by landslides, thus giving rise to basins which become the sites 
 of lakes. Such a lake, five miles long and more than seven hun- 
 dred feet deep, was formed on the Upper Ganges in 1892. Two 
 years later, the dam which held back the water broke, and the 
 resulting flood wrought great destruction in the valley below. 
 
 Solution, weathering, wind, etc. Basins suitable for ponds 
 and lakes sometimes originate by solution of the underlying rock. 
 Limestone sinks (p. 97) may become the sites of ponds and per- 
 haps of lakes, but considerable basins of this origin are not known. 
 Basins are sometimes formed by solution of the surface rock. 
 Some of the lake basins of Florida probably arose in this w T ay. 
 
 The surface of rock weathers (p. 110) unequally. If the weather- 
 ing of one area is greater than that of its surroundings, the weathered 
 material may be blown away, leaving a depression suitable for 
 holding water. Wind-driven sand sometimes scours out small de- 
 pressions which hold pools of water, though basins of such size 
 as to become the sites of lakes are not known to have originated in 
 this way. Eolian sand is sometimes piled up about low places, 
 enclosing them, thus giving rise to marshes, ponds, and even 
 lakes. 
 
 Glacial lakes an index of topographic age. Since rivers are 
 antagonistic to lakes, and since rivers are always active, it follows 
 that a region of abundant lakes is a region in its topographic youth, 
 unless the lakes are in valley flats. Lakes at high altitudes are of 
 relatively recent origin. 
 
314 PHYSIOGRAPHY 
 
 SALT LAKES' 
 
 While most lakes are fresh, a few, such as Great Salt Lake, the 
 Caspian, 'Aral, and Dead seas, are salt. Salt lakes are found chiefly 
 in arid climates. 
 
 Fresh lakes may becrme salt, and salt lakes may become fresh. 
 These changes are usually the result of changes of climate. If 
 aridity increases so that evaporation from a fresh lake exceeds 
 the intake (precipitation and inflow), a fresh lake may become 
 salt. If evaporation from a salt lake becomes less than the intake 
 of fresh water, the lake will be freshened, and, under proper con- 
 ditions, may become altogether fresh. The best-known illustra- 
 tion of these changes is furnished by Great Salt Lake and the 
 earlier lake which preceded it in the same region. 
 
 1. The first lake which occupied this basin appears to have 
 been fresh. A climatic change seems then to have taken place, 
 a relatively humid climate giving place to an arid one. When 
 this change had been accomplished, evaporation from the sur- 
 face of the lake exceeded the intake of fresh water, and the level 
 of the lake was lowered. As the water evaporated, the mineral 
 matter which it held in solution was left behind. Salt was one 
 of these substances, and as more and more water evaporated, the 
 salinity of that which remained increased, and the lake became 
 salt. 
 
 2. Another change of climate, this time in the direction of in- 
 creased humidity, ensued. The intake of fresh water then ex- 
 ceeded evaporation from the lake, and the saltness of the water 
 was diminished by dilution. At the same time, the level of the 
 lake rose until it overflowed, finding an outlet by way of Snake 
 River to the Columbia. The continued inflow of fresh water and 
 the continued outflow of the diluted salt water resulted in the 
 progressive freshening of the lake, and it finally became fresh. 
 The expanded lake of former times in this basin is known as Lake 
 Bonnevilk (Fig. 339), which, at its maximum, covered an area 
 of about 17,000 square miles. Its surface was about 1000 feet 
 higher than that of Great Salt Lake. 
 
 Another change of climate, this time in the direction of aridity, 
 reduced Lake Bonneville. Its surface sank below its outlet, and, 
 as this happened, its waters gradually became saline. Evapora- 
 
LAKES AND SHORES 
 
 315 
 
 ,tion in excess of intake continuing, the former great lake was in 
 time reduced to the relatively small Great Salt Lake of the present, 
 with an area of about 2000 square miles and an average depth of 
 only about 15 feet. Its waters are saturated with salt, and much 
 salt has been deposited. 
 
 FIG. 339. Former lakes of the Great Basin. (U. S. Geol. Surv.) 
 
 Farther west a similar series of changes is recorded by the former 
 Lake Lahontan (Fig. 339), and by Mono Lake. 
 
 Salt Lake, and the sites of some extinct salt lakes, yield salt 
 in commercial quantities. Great Salt Lake was estimated, a few 
 years ago, to contain 400,000,000 tons of common salt, besides 
 large quantities of other mineral matter. Much of the mineral 
 matter formerly held in solution by this lake has already been 
 deposited. Utah produced more than 400,000 barrels of salt in 
 1902, and 253,829 (value $321,301) in 1904. Deposits of salt made 
 
316 PHYSIOGRAPHY 
 
 by lakes and inland seas which are now extinct, are the chief sources 
 of sa'.t. 
 
 Accessible salt deposits and "salt-licks" determined, or helped 
 to determine, the location of numerous early trans- Allegheny settle- 
 ments, as in the Blue Grass region of Kentucky. 
 
 The Climatic Effects of Lakes 
 
 The great number of lakes in the northern parts of the United 
 States and Europe have some influence upon the climate of the 
 regions in which they occur. They increase its humidity to some 
 slight extent at least, and, since water is heated less readily than 
 the land and gives up its heat less readily, the lakes have the 
 effect of tempering the climate. Until they freeze over, they tend 
 to keep the temperature of their surroundings a little higher than 
 it would otherwise be in the autumn and early winter, and to reduce 
 the temperature of spring. The temperature effects of lakes are 
 felt chiefly on the sides toward which the prevailing winds blow. 
 
 Economic Advantages and Disadvantages 
 
 The question as to whether lakes are beneficial or harmful to 
 mankind may be looked at from various points of view. 
 
 1. The Great Lakes serve as highways and make cheap trans- 
 portation possible. In this way they serve a good purpose, as shown 
 by their extensive commerce. 2. Many cities, like Chicago, draw 
 their water-supply from lakes. A city located as Chicago is could 
 not readily get an adequate supply from any other source, except 
 at far greater cost. 3. Lakes furnish a certain amount of food 
 material, especially fish. 4. By tempering the climate, they modify, 
 to some slight extent at least, agricultural pursuits. Thus the 
 prevailing westerly winds temper the climate of the east shore of 
 Lake Michigan in such a way as to make it favorable for fruit- 
 growing, while the west side of the lake, affected by winds not 
 tempered by the lake, is not favorable for this industry. In these 
 and other ways the lakes seem to serve mankind. 
 
 On the other hand, it is to be remembered that such a body of 
 water as Lake Michigan occupies some 22,450 square miles of sur- 
 face, much of which would, presumably, have been good farming 
 land, if the lake basin had not been made. The value of such an 
 
LAKES AND SHORES 317 
 
 area of good farming land might offset, or more than offset, the 
 economic advantages of the lake. 
 
 Small lakes are of little consequence as highways, and the modi- 
 fications of climate which they effect are slight. The same may 
 be said of marshes. There can be no doubt that if all the area 
 occupied by lakes were cultivated land instead, the returns would 
 be much greater than those which now accrue from any uses to 
 which the lakes are put. But lakes have a value not to be esti- 
 mated in dollars and cents. They beautify the landscape and 
 afford the means for rest and recreation which could not well be 
 spared. The actual value of such considerations is not easy of 
 definite estimate. 
 
 The advantages which a primitive people may derive from a 
 location upon the shores or islands of a lake are suggested by the 
 fact that the earliest European civilization arose about the lakes 
 of Switzerland, while the lakes of Mexico and Peru were the seat 
 of the ancient civilizations of those countries. 
 
 THE TOPOGRAPHIC FEATURES OF SHORES 
 
 In connection with the discussion concerning changes now tak- 
 ing place in lakes, reference was made to certain topographic fea- 
 tures of lake shores, because their development has a bearing on 
 the history of lake basins, and so on the life of the lakes them- 
 selves; but the topic is of such importance that it merits more than 
 incidental mention. 
 
 The topographic features developed by lakes along their shores 
 are similar to those developed by the sea along its coasts, except 
 that the latter are on a larger scale. A discussion of the develop- 
 ment of the topographic features of lake shores is therefore appli- 
 cable, in most of its details, to the shores of the sea. 
 
 Gradation is affecting the shores of oceans and lakes every- 
 where; diastrophism is affecting them in many places, though 
 not universally, at least not to such an extent as to be sensible 
 from year to year, while the effects of vulcanism on shores are 
 very limited and of little consequence in this connection. 
 
 Gradational Changes now taking Place along Shores 
 
 Waves, currents, rivers, winds, glaciers, ice formed along the 
 shore, and various other agencies are working on the shores of 
 
318 PHYSIOGRAPHY 
 
 seas and lakes, and each has some effect on the coast-line. Of these, 
 the waves and the movements of the water to which the waves 
 give rise, are the most important. 
 
 1. Waves, Undertow, Shore Currents. The top of a wave is its 
 crest, and the depression between two adjacent crests is the 
 trough. The vertical distance between the crest and the bottom 
 cf the trough is the height of the wave, and the horizontal 
 distance between two adjacent crests is the length. The time which 
 it takes one crest or one trough to travel the length of the wave 
 is the period of the wave. In the open sea, storm waves often 
 have a height of 20 to 30 feet, and in rare cases even 50 feet. On 
 the shore their heights may be much greater, as we shall see. The 
 length of great waves may be as much as 1500 feet, and the velocity 
 as much as 60 miles per hour. Such lengths and velocities are, 
 however, far beyond the average. 
 
 In the open sea wave motion does not involve the forward 
 movement of the water. Each particle of water describes a curve, 
 and theoretically comes to rest at the point whence it started, 
 though the wave form moves on. Some conception of the motion 
 involved may be gained from a waving field of grain or grass, where 
 each moving stem is fixed to the ground, though wave after wave 
 crosses the field; or from a long piece of rope one end of which is 
 fixed, while the other end is shaken up and down. Successive waves 
 travel from the shaken end to the other end. 
 
 Fig. 340 gives some idea of the nature of the movement of the 
 water in the waves of the open sea. 
 
 FIG. 340. Diagram to illustrate the movement of water in waves. The 
 small circles represent the movement of water particles. 
 
 If the water in a wave moved forward at the velocity at which 
 the wave form travels, the sea would hardly be navigable. 
 
 When the wind is very strong, the top of a wave may be blown 
 forward, that is, the wave "breaks," and so has a motion independ- 
 ent of the true wave motion. Even when the waves do not break, 
 the surface water is slipped along to some extent by the moving 
 air. 
 
 High waves in the ocean are often called "seas," and when a 
 sailor says that there is a "high sea," he means that there are high 
 
LAKES AND SHORES 319 
 
 waves. The destructiveness of waves in the open sea depends 
 quite as much on their length as on their height. With a given 
 height, the longer the wave, the less its destruction. 
 
 Waves generated by a storm often run far beyond the place 
 where they were started. They diminish in height, but keep 
 their velocity and their length if the water is deep and the waves 
 are unobstructed by islands, etc. Waves which have outrun the 
 storm which started them constitute the swell or the ground-swell. 
 In the case of great hurricanes, destructive waves are sometimes 
 felt a thousand miles from the storm. This was the case on the 
 coast of New Jersey in 1889, when the storm was to the south. 
 As a result of storms in different places, the open sea is never 
 altogether quiet. 
 
 As a wave advances from the open sea into shallow water, it 
 undergoes notable changes. Where the water is so shallow that 
 wave motion is sensible down to its bottom, the wave "drags" 
 bottom. The velocity and the length of the wave are then 
 diminished, and its height increased. The top then pitches for- 
 ward as surf. 
 
 In strong winds and in shallow water, therefore, there is a 
 distinct forward movement of some of the water of a wave. Waves 
 in which there is pronounced forward movement are sometimes 
 called waves of translation. 
 
 The water thrown against the shore in the wave runs back 
 again, and this from-shore motion is the undertow. The undertow 
 tends to run down the steepest slope, but it is often directed 
 obliquely by incoming waves. Its movement is checked, too, by 
 every incoming crest. 
 
 Where waves strike a shore obliquely, the water moves more 
 or less along shore, and the sum of these movements along shore 
 gives rise to a shore or littoral current. 
 
 The waves, the undertow, and the shore currents all affect 
 the shore. The waves erode the shore-line in some places, and 
 all of these movements erode the bottom in some places. All the 
 sediment acquired by erosion is deposited sooner or later. Whether 
 the movements of water along shores erode or deposit, they affect 
 the outline of the coast, and often its vertical configuration. 
 
 The amount of motion in waves diminishes rapidly downward, 
 and is insensible below a few hundred feet. Submarine structures, 
 such as piers, etc., are rarely disturbed below 30 feet. 
 
s- 
 
 320 PHYSIOGRAPHY 
 
 The erosive work of waves. The force of the wave as it is 
 hurled against the shore is often great. The surf is sometimes 
 thrown up to heights of more than 100 feet with force enough to 
 destroy lighthouses and even cliffs of rock. Windows of the Dunnet 
 Head lighthouse on the coast of Scotland are said to have been 
 broken at heights of 300 feet above sea-level during severe gales. 
 In some cases the bursting of doors and windows during such 
 storms appears to be due to the explosive action of the air within 
 the buildings, as the surf dashed against them falls back. It has 
 been estimated that, in exceptional storms, the strength of waves 
 on the exposed coast of Britain has been as much as three tons per 
 square foot, and that the average force of winter waves is about 
 one ton per square foot. Such waves would move masses of rock 
 tons in w r eight. It is clear, therefore, that the force of waves is 
 adequate for powerful erosion. 
 
 If a coast-line were regular, but composed of rock of unequal 
 hardness, it would not be likely to remain regular, so far as wave 
 erosion is concerned, for the waves would 
 wear the weaker rock more and the stronger 
 rock less. The result would be the develop- 
 ment of reentrants on the weaker rock, 
 while the stronger rock would remain as 
 projections of land into the sea (Fig. 341). 
 Under these circumstances the irregularities 
 of the coast would go on increasing, so far 
 as wave erosion is concerned, until the 
 
 reentrants had become so deep that the Fl ; 341. Diagram to 
 
 . , illustrate the effect of 
 
 diminished force of the waves in them would wave erosion on rocks 
 
 wear the weaker rock at those points no f unequal hardness. 
 
 btarting with astraight 
 faster than the stronger \vaves wear the line, indicated by the 
 
 harder rock of the projecting points be- dotted line, the erosion 
 
 of the waves would de- 
 tween. When this stage is reached, the velop some such out line 
 
 shape of the coast-line is stable, so far as as shown :TF, weak rock, 
 
 and S, resistant rock. 
 wave erosion is concerned. Since coast-lines 
 
 are made of stronger and weaker rock structures, irregularities of 
 this sort are constantly in process of development. They become 
 greater along seacoasts than along lake shores, because the waves 
 of seas are stronger than those of lakes. 
 
 Where a coast is very irregular, especially where there are pro- 
 jections of land into the sea, the waves attack projecting points of 
 
 w 
 
 s- 
 
 w 
 
 
 Sea 
 
LAKES AND SHORES 
 
 321 
 
 land more forcibly than they attack the reentrants, such as the 
 heads of bays. The projecting points are thus worn back more 
 than the heads of the bays. Where a coast is very irregular, there- 
 fore, wave erosion tends to reduce its irregularities, unless the pro- 
 jecting points are of rock which is much more resistant than that 
 of other parts of the coast. 
 
 FIG. 342. Diagram illustrating high sea cliff. It shows also a submerged 
 terrace, due partly to wave-cutting and partly to building. 
 
 We conclude, therefore, that wave erosion tends to develop 
 small irregularities of coast-line, but not great ones. Their extent 
 is dependent upon (1) the "fetch" of the waves, that is, the dis- 
 tance they have been traveling, (2) the strength of the winds, (3) 
 the depth of the water, (4) the exposure of the coast attacked, (5) 
 the kind and abundance of the tools (such as gravel, bowlders, etc.) 
 with which the waves work, and (6) the resistance of the rock against 
 which the waves beat, the resistance being determined partly by 
 hardness and partly by structure. 
 
 FIG. 343. Diagram showing a low sea cliff. 
 
 Without tools to work with, waves would be relatively 
 ineffective against hard rock which had no bedding or jointing 
 planes. Thus on the Outer Hebrides, barnacles are said to be as 
 abundant after a storm as before, where gravel and stones of 
 suitable size for the waves to move are absent. Rock affected by 
 cleavage planes, whether bedding or jointing, may be effectively 
 worn by waves, irrespective of the debris which they move. 
 
322 
 
 PHYSIOGRAPHY 
 
 FIG. 344. A high sea cliff, La Jolla, Cal. 
 
 FIG. 345. A high cliff with a beach, shore of Lake Michigan. 
 (U. S. Geol. Surv.) 
 
PLATE XXII 
 
 Tenrv&ssee' 
 Cave 
 
 Fi<L 1. PORTION OF THE CALIFORNIA COAST NEAR TAMALPAIS. 
 
 MAKLN- Co 
 JfRANClScJ*^ 
 
 L.H. #' Pt.Bonita 
 
 FIG. 1. A coast line developed chiefly by wave erosion. Scale 1 mile per inch. 
 (Tamalpais, Cal., Sheet, U. S. Geol. Surv.) 
 
 - ^^.-'A^^Qr *>(: _3 iU dfo. >/^V s 
 
 Bass PtX pig. 2. MASSACHUSETTS. ./ 
 
 FIG. 2. An island tied to the mainland by a 
 "beach." Scale 1 mile per inch. (Boston 
 Bay, Mass , Sheet, U. S. Geol. Surv.) 
 
LAKES AND SHORES 
 
 323 
 
 Irregularities developed by wave erosion are extremely numerous. 
 Here belong very many, if not most, of the small projections of 
 high land into the sea. Their outlines are often somewhat angular 
 (Fig. 1, PI. XXII). Here, too, belong the islands of some coasts, es- 
 pecially those of solid rock, many of which have been isolated from 
 
 FIG. 346. Steep cliff developed by waves; Allen Point, Grand Island, Lake 
 Champlam. (Perry.) 
 
 the mainland by wave erosion. Such islands are likely to be 
 destroyed in time by the same processes which gave them being. 
 
 The cutting of the waves affects the vertical as well as the 
 horizontal configuration of the shores. Where the sea is advanc- 
 ing upon the land, steep slopes, called sea cliffs (Figs. 342-346), 
 are developed. Sea cliffs may be high or low, according to the 
 elevation of the land into which the waves cut. Cliffs are of fre- 
 quent occurrence along seacoasts; and where they are absent. 
 
324 
 
 PHYSIOGRAPHY 
 
 the waves are not cutting, and the sea is not advancing on the 
 land, or at least not as a result of its own cutting. Slumping often 
 accompanies wave erosion. 
 
 tic. 347. Cross-section of a beach. (Gilbert.) 
 
 The sea cliff is often bordered by a wave-cut terrace a little 
 below the surface of the water (Fig. 342). The area of this terrace 
 often represents, in a rough way, the area which the sea has gained 
 from the land by wave-cutting. 
 
 FIG. 348. A lake beach (barrier) Griffins Bay, Lake Ontario. 
 
 Deposition by waves, shore currents, etc. Shore waters are 
 aggradational as well as degradational. The material cut from 
 the land by waves, or brought down by rivers, is shifted about by 
 the undertow and the shore currents, but it must ultimately come 
 
LAKES AND SHORES 
 
 325 
 
 to rest. While this material is being moved about by the shore 
 waters, it constitutes shore drift, whether brought in by rivers or 
 worn from the coast by waves. If shore drift is left at the shore- 
 line, it makes a beach (Figs. 348 and 349), which is sometimes de- 
 nned as the area of sand, gravel, etc., between high and low tides. 
 Deposits of gravel and sand continuous with those of the beach 
 are often made at greater depths, the material being carried out by 
 
 FIG. 349. A barrier beach, shutting in a marshy tract behind it, Lasells 
 Island, Penobscot Bay, A.e. (Bastin, U. S. Geol. JSurv.) 
 
 the undertow and by the shore currents which diverge from the 
 coast-line. It is sometimes carried out and deposited at the outer 
 edge of the wave-cut terrace (Fig. 342), and it is sometimes disposed 
 as a terrace along shore (Fig. 350) where there is no wave-cut 
 terrace. 
 
 Waves often build reefs or barriers a little out from the shore- 
 line. They are developed near the line of breakers, where the in- 
 
 FIG. 350. A wave-built terrace. (Gilbert, U. S. Geol. Surv.) 
 
 coming wave is no longer able to carry forward the bulk of the 
 debris which it is moving in toward the shore. The undertow 
 often contributes material to the reef. There are sometimes sev- 
 eral such reefs parallel to the coast and to one another. Bars, 
 reefs, etc., often hinder the movements of ocean vessels, as when 
 they close the entrances of harbors. A spit which does not 
 obstruct the entrance to a harbor, on the other hand, is sometimes 
 an advantage, since it breaks the force of the incoming waves in 
 
326 
 
 PHYSIOGRAPHY 
 
 storms, and so helps to form a harbor. In general, reefs discourage 
 navigation. 
 
 After the reef is developed, waves may build its crest above 
 the surface of the water, converting it into land (Fig. 351). 
 
 FIG. 351. Section of a barrier. (Gilbert, U. S. Geol. Surv.) 
 
 Such seems to have been the origin of many of the low, narrow- 
 belts of sandy land parallel to coasts, with marshes and lagoons- 
 behind them. This type of irregularity is illustrated by the coast 
 
 3ZL 
 
 Cape ffattei'a.s 
 
 Fro. 352. Map showing the early stages in the simplification of a shore- 
 line, and showing that at this stage the irregularities are increased. 
 
 of the United States at various points between New York and 
 Texas (Fig. 352). 
 
 Currents along the shore (littoral currents) shift sediment in 
 
LAKES'" AND SHORES 
 
 327 
 
 the direction of their motion; but where such a current reaches a 
 bay, it does not commonly follow the outline of the bay. It tends 
 rather to cross its debouchure in the direction in which it was 
 previously moving. Under such circumstances it tends to build 
 an embankment of gravel and sand across the bay. Such em- 
 bankments are spits. Currents do not build spits above the water, 
 but waves may accomplish this result by washing material from 
 their slopes up to their tops (Figs. 351-357). They may thus 
 become land, after which dunes often develop on them. When 
 
 FIG. 353. Map of the head of Lake Superior. (U. S. Geol. Surv.) 
 
 spits cress bays they become bars (Figs. 353 and 354). Spits 
 and bars are often hooked (Fig. 355), as the result of the shift- 
 ing of the currents while they are in process of building. 
 
 Spits and hooks often form harbors, and so have determined 
 the location of numerous settlements and towns. A great hook 
 makes Provincetown harbor, where the Pilgrims first landed, 
 while the harbor whose shores they finally chose for their settle- 
 ment is formed by a large spit. A hook-formed harbor upon an 
 otherwise regular coast determined the location of Erie, Pennsyl- 
 vania. 
 
 If the shore drift is deposited against the mainland, it may 
 make a flat extending out from the land into the water. A coast- 
 
328 
 
 PHYSIOGRAPHY 
 
 line developed by deposition is in contrast with one developed 
 by erosion, for the former has no sea cliff. 
 
 Land areas developed from reefs and spits often greatly in- 
 crease the irregularity of the coast-line temporarily (Fig. 352), 
 
 FIG. 354. Bar joining Empire and Sleeping Bear bluffs, Lake Michigan. 
 (Gilbert, U. S. Geol. Surv.) 
 
 but they really represent an initial stage in the development of 
 regularity, for after the reefs have become land, the lagoons behind 
 them are likely to be filled with sediment, organic matter, etc., 
 and converted into land (Fig. 356). The sediment which con- 
 
 FIG. 355. A recurved spit, Dutch Point, Grand Traverse Bay, 
 Lake Michigan. (U. S. Geol. Surv.) 
 
 tributes to this end is washed down from the land or blown in. 
 When the lagoon is filled, the shore-line is much more regular than 
 before, but the first effect of the making of the reef-land is to 
 make the coast more irregular. 
 
LAKES AND SHORES 
 
 329 
 
 The disposition of shore deposition to simplify coast-lines is 
 also shown in another way. Deposits are sometimes made be- 
 tween islands near the shore of the mainland, and the mainland 
 itself (PI. XXII and Fig. 357). Thus Nahant island, on the coast 
 ot Massachusetts, and the Rock of Gibraltar, on the coast of Spain, 
 
 FIG. 356. Sketch-map of a part of the New Jersey coast. The dotted 
 belt at the east is the barrier modified by the wind. The area marked 
 by diagonal lines is the mainland; the intervening tract is marsh-land. 
 The numbers show the depth of water in feet. Scale; J mch = l mile. 
 
 have been "tied" to the mainland by the deposits of waves and 
 shore currents. While this tying process gives rise to a notable 
 irregularity of the mainland, it simplifies the outline of the land 
 areas in the sense that it unites islands to mainland. 
 
 2. Rivers. Rivers erode and deposit at or near coasts. The 
 erosion of streams has little effect upon the coast-line, for a river 
 does not cut below sea-level more than the depth of its own water. 
 
330 
 
 PHYSIOGRAPHY 
 
 Working alone, therefore, rivers do not develop bays or other 
 similar bodies of water projecting into the land. 
 
 The deposition of sediment brought down to coasts by streams 
 is of more consequence in modifying the outline of the land. This 
 is especially the case where deltas are built into lakes or seas. At 
 the lower end of the Mississippi, for example, a great delta has 
 been built out into the Gulf (Fig. 209). The great irregularity 
 which the delta itselt constitutes has smaller irregularities about 
 its borders. Deltas in lakes often show the same general features 
 on a smaller scale. The forms of deltas have been noted (p. 200). 
 Delta-land is always low, unless affected by diastrophism, or by 
 lowering of the surface of the water in which it was built. 
 
 FIG. 357. Sheep Island, Penobscot Bay, Me., a land-tied island. 
 (Bastin, U. S. Geol. Surv.) 
 
 3. Winds. The chief effect of wind along the shores is to 
 blow about the dry sand. The sand is often piled up into con- 
 siderable dunes, as we have seen, but the shifting of the sand by 
 the wind does not commonly change the outline of the land area 
 to any great extent. The wind often piles up sand on low bars 
 and on low coasts, building them up much higher than they were 
 before, even though it does not change the position of the coast- 
 line. Plate V shows a coast where the land has been built up 
 notably by wind-driven sand. At Nag Head, N. C., the land is 
 said to have gained on the sea 350 feet in ten years as a result of 
 wind deposits. 
 
 4. Glaciers. Glaciers descend to the level of the sea in some 
 places, as in Greenland and Alaska. Where this is the case, they 
 usually move down to the sea through valleys. If the ice is thick, 
 the glaciers gouge out the valleys, sometimes to great depths 
 below the level of the sea. 
 
LAKES AND SHORES 
 
 331 
 
 When glaciers which have gouged out such valleys melt, the 
 lower ends of the valleys are filled with sea-water, making narrow 
 bays, or fiords. 
 
 This is the explanation, or a part of the explanation, of many 
 
 FIG. 358. Alaska fiords. (C. and G. Surv.) 
 
 of the fiords of Norway, Alaska (Fig. 358), Greenland, and Chile, 
 and some other coasts. 
 
 Glaciers which descend to the sea deposit their drift where they 
 end, but the drift, being of loose material, is usually soon washed 
 away by the waves, and rarely gives rise to enduring irregularities 
 of coast-line. Drift-made land in lakes would be less readily swept 
 Vway, because the waves are weaker. 
 
332 
 
 PHYSIOGRAPHY 
 
 S OUT HERN 
 ARGEN T I NA 
 
 FIG. 359. Fiords and other irregularities on the west coast of South America. 
 
LAKES AND SHORES 
 
 333 
 
 5. Shore ice is another agency which is working on the coast- 
 lines, but does not greatly modify their outlines. 
 
 EXTINCT LAKES 
 
 Many former lakes have become extinct. Extinct lakes are 
 recognized by various features. If a lake basin became extinct by 
 having its basin rilled, the former area of the lake is marked by a 
 flat (Fig. 360) covered with deposits such as are formed in lakes. 
 
 FIG. 360. A part of the flat of Lake Agassiz, Moorhead, Minn. (Goode.) 
 
 These deposits may be of gravel or sand along the shores, but the 
 materials deposited far from shore are fine. Such a flat is a lacus- 
 trine plain. A lacustrine plain is a minor type of plain, and may 
 He in mountains, on plateaus, or on plains of a larger type. 
 
 If a lake became extinct by the lowering of its outlet or by 
 evaporation, the old bed of the lake would be less flat, might even 
 depart much from flatness. 
 
 The former borders of an extinct lake are often marked by 
 various shore features, such as deltas, terraces, beaches, etc.; while 
 above the terraces, in places at least, old shore cliffs are often 
 found, especially if the lake was large. Conspicuous shore features 
 mark the former borders of Lake BonneviHe. Some of them are 
 shown in Fig. 361. The lower slope, marked by terraces developed 
 about the shores of the lake in relatively recent times, is in strik- 
 ing contrast with the upper slope, the topography of which was 
 developed by running water. The topography of the terraces is 
 young; that of the slopes above, much more advanced. This 
 relation between a slope of older topography above and a sur- 
 face of younger topography below, has been called a topographic 
 unconformity. In this case, the lower ends of the ravines and 
 
334 
 
 PHYSIOGRAPHY 
 
 valleys were filled and obliterated by the deposits along the shore 
 of the lake, the water of which stood at various levels at various 
 times. 
 
 Shore features, less conspicuous than those about Lake Bonne- 
 ville, but none the less distinctive, mark the borders of the extinct 
 
 FIG. 361. Shore of former Lake Bonneville, Wellsville, Utah. 
 (U. S. Geol. Surv.) 
 
 Lake Agassiz and many other extinct lakes. They also appear 
 about many- existing lakes well above their present shores, thus 
 showing their previous higher levels. 
 
 All shore features developed by lakes are likely to be destroyed 
 in time by the agents of degradation. The aridity of the Great 
 Basin has favored the preservation of the shore features of Lake 
 Bonneville and Lake Lahontan. 
 
LAKES AND SHORES 
 
 335 
 
 MAP EXERCISES 
 
 Maps Showing Lakes and Shores. 
 
 I. Study the following maps, group by group, as indicated by the letters. 
 in preparation for conference: 
 
 A. Browns Creek, Neb. 
 Great Bend, Kan. 
 
 B. St. Louis, Mo. 111. 
 Lancaster, Wis. la. 111. 
 Bodreau, La. 
 
 Bayou de Large, La. 
 
 C. Ft. McKinney. Wyo. 
 Greeley, Colo. 
 Granada, Colo. 
 
 D. Mt. Lyell, Cal. 
 Cloud Peak, Wyo. 
 Chief Mountain, Mont. 
 
 E. Paradox Lake, N. Y. 
 Berne, N. Y. 
 Webster, Mass. 
 
 F. Pingree, N. D. 
 Skanea teles, N. Y. 
 Penn Yan, N. Y. 
 Hammondsport, N. Y. 
 Cheian, Wash. 
 Methow, Wash. 
 Stehekin, Wash. 
 
 G. White Bear, Minn. 
 Minneapolis, Minn. 
 
 H. Falmouth, Mass. 
 
 Marthas Vineyard, Mass. 
 Nantucket, Mass. 
 Gay Head, Mass. 
 Provincetown, Mass. 
 
 I. Standingstone, Tenn. 
 Arredondo, Fla. 
 
 J. Crater Lake Special, Ore. 
 
 K. Boston Bay, Mass. 
 
 (Also maps of group H.) 
 Asbury Park, N. J. 
 Sandy Hook, N. J. 
 Atlantic City, N. J. 
 Southern California, Sheet 2. 
 Tamalpais, Cal. 
 San Mateo, Cal. 
 Erie, Pa. 
 Fairview, Pa. 
 Sodus Bay, N. Y. 
 Boothbay, Me. 
 Coast Survey Chart 103. 
 
 Note. Before taking up other details in the case of any map, note 
 its position in the country, its general topography, and the causes which 
 have developed its topography. Where maps are adjacent, it is well 
 to study them together; e.g., the New York maps under F, the Wash- 
 ington maps under F, some of the Massachusetts maps under H and K, 
 and two of the New Jersey maps under K. 
 
 II. 1. Each lake shown on the maps presents a series of problems, 
 among which are the following: 
 
336 PHYSIOGRAPHY 
 
 a. What is the origin (certain, probable, possible) of the basin? 
 6. What is the source of the water-supply? 
 
 c. Does the map indicate (certainly, probably, possibly) any 
 
 changes now in progress about the lake? 
 
 d. Is the lake likely to be destroyed soon? If so, what are 
 
 likely to be the chief factors in its destruction? 
 
 e. What inferences may be made from the map as to the 
 
 depth of individual lakes? What measure of certainty 
 or uncertainty attaches to the inference? 
 2.* Classify the lakes of each map under the headings indicated on pp. 
 
 303-311. 
 
 3.* Find as many types as possible of glacial lakes (see p. 311), speci- 
 fying examples. 
 
 4. In connection with each map, note whether there are areas (cer- 
 tain, probable, possible) which were once lake bottoms; i.e., 
 have lakes become extinct in the area represented? 
 5.* Are the waters in the ponds along the shore of Marthas Vineyard 
 
 probably fresh or salt? Give reasons. 
 6. In studying the ponds of the Standingstone Sheet, note carefully 
 
 the contour lines and the drainage about them. 
 
 III. The maps of group K are designed especially to illustrate shore 
 phenomena, most of them along the seashore. Interpret in the 
 light of pp. 320-329. 
 
 1. Indicate what parts of the coasts are being (or have recently 
 been) modified by (a) wave erosion and (6) shore deposition. 
 Give reasons for your conclusions. 
 2.* In general, how may (a) shore deposition and (6) wave erosion 
 
 be inferred from topographic maps of coasts? 
 3. From any good map or model of the United States, indicate 
 
 where (a) erosion, and (b) deposition prevails. 
 4.* What are the possible explanations, so far as the map shows, 
 of the marshes on the bay coast of the San Mateo region? 
 
 5. Are coastal features similar to those shown on the maps of 
 
 group K shown on the shores of lakes of preceding groups? 
 
 6. Make a careful study of Coast Survey Chart 103, noting all 
 
 the processes which may have played a part in the develop- 
 ment of the coast-line. 
 
 7.* Is there evidence on the Sodus Bay Sheet of change of rela- 
 tive level of land and lake? Give reasons. 
 
 8. Interpret the steep slope just south of the N. Y., C. & St. L. R. R. 
 
 9. What is the probable meaning of the low ridge extending east 
 
 from Fairview? 
 
 * Answer in writing. 
 
LAKES AND SHORES 337 
 
 REFERENCES 
 
 1. RUSSELL, Lakes of North America: Ginn & Co. 
 
 2. TAYLOR, Short History of the Great Lakes, in Studies in Indiana Geography. 
 
 3. CHAMBERLIN AND SALISBURY, Geology, Earth History, Vol. Ill, pp. 394- 
 403. See also index of same volume and of Volume I. 
 
 4. GILBERT, Topographic Features of Lake Shores, in 5th Ann. Rept. 
 U. S. Geol. Surv., and Mono. I, U. S. Geol. Surv. 
 
 5. DILLER, Crater Lake, etc.: Prof. Paper No. 3, U. S. Geol. Surv.; 
 Nat. Geog. Mag., Vol. VIII, pp. 33-48, and Am. Jour. Sci., Vol. Ill, 1897, 
 pp. 165-172. 
 
 6. RUSSELL, Geography of the Laurentian Basin: Bull. Am. Geog. Soc., 
 Vol. XXX, pp. 226-254, and Jour, of Geol., Vol. I, pp. 394-408. 
 
 7. TARR, For New York lakes, see Physical Geography of New York: 
 The Great Lakes: Bull. Am. Geog. Soc., Vol. XXXI, pp. 101-117, 217-235, 
 and 315-343; and Lake Cayuga, a Rock Basin: Bull. Geol. Soc. Am., Vol. V, 
 pp. 339-356. 
 
 8. UPHAM, Glacial Lakes in Canada: Bull. Geol. Soc. Am., Vol. II, pp. 243- 
 274. 
 
 9. HARRINGTON, Area and Drainage Basin of Lake Superior: Nat. Geog. 
 Mag., Vol. VIII, pp. 111-120. 
 
 10. BRIGHAM, Lakes: a Study for Teachers: Jour, of Sen. Geog., Vol. I, 
 pp. 65-72. 
 
 11. FENNEMAN, Lakes of Southeastern Wisconsin: Bull. 8, Wis. Geol. and 
 Nat. Hist. Surv. 
 
 12. KEMP, Physiography of Lake George: Annals N. Y. Acad. of Sci., 
 Vol. XIV, pp. 141-142, and Science, Vol. XIV, p. 774. 
 
 13. MURDOCH, Fall of Water Level in Great Salt Lake: Nat. Geog. Mag., 
 Vol. XIV, 1903, pp. 75-77. 
 
 14. MILL, Bathymetric Survey of the English Lakes: Geog. Jour., Vol. VI, 
 pp. 46-73 and 135-166. 
 
 15. MURRAY, Bathymetric Survey of the Fresh-water Lochs of Scotland: Scot. 
 Geog. Mag., Vol. XV11, p. 169; Vol. XIX, pp. 449 and 561; and Vol. XX, 
 pp. 1, 169, 235, 449, 589, and 628. , 
 
 EXTINCT LAKES, SHORE FEATURES, ETC. 
 
 16. RUSSELL, Present and Extinct Lakes of Nevada, in Physiography of 
 the United States: Am. Bk. Co. 
 
 17. SHALER, Sea-coast Swamps of the Eastern United States: 6th Ann. 
 Rept. U. S. Geol. Surv.; also Sea and Land: Scribner's. 
 
 18. GILBERT, Lake Bonneville: Mono. I, U. S. Geol. Surv. 
 
 19. RUSSELL, Lake Lahontan: Mono. XI, U. S. Geol. Surv.; and Mono 
 Lake: 8th Ann. Rept. U. S. Geol. Surv. 
 
 20. GEIKIE, J., Earth Sculpture, Chapter XV: Putnam. 
 
 21. GEIKIE, SIR A., Scenery of Scotland, Chapter III: Macmillan. 
 
CHAPTER VII 
 
 VULCANISM 
 
 A VOLCANO is a vent in the earth's crust out of which hot rock 
 issues. The hot rock may be liquid (called lava) and may flow 
 out; or it may be solid, when it is thrown out violently in pieces. 
 If the vent is in the form of a long crack or fissure, it is not com- 
 monly called a volcano. 
 
 The rock material which comes out of a volcano is generally 
 built up into mounds or cones (Fig. 362). They may be mere 
 
 FIG. 362. Fujiyama, a volcanic cone in Japan. 
 
 mounds or high hills, or even high mountains. The cones are 
 often called volcanoes, though they are really the results of volcanic 
 
 338 
 
VULCANISM 
 
 339 
 
 activity. The volcano from which lava flows makes a cone with 
 low slopes (Fig. 363). The volcano from which solid matter is 
 
 flavna Loa 
 
 Seata </ ffilet 
 
 FIG. 363. Profile of the cone of Mauna Loa. Vertical scale same as hori- 
 zontal. (U. S. Geol. Surv.) 
 
 thrown makes a cone with steeper slopes (Fig. 364). Many 
 volcanoes send out both liquid rock (lava) and solid rock. In 
 this case both may be issuing at about the same time, or lava 
 may flow out at one time and solid rock be thrown out at another. 
 Along with the hot rock, quantities of gases and vapors, some of 
 them poisonous, are discharged. So long as a volcano is active 
 
 FIG. 364. Typical cinder cone, Clayton Valley, Cal. (U. S. Geol. Surv.) 
 
 there is likely to be a hollow, called the crater (Figs. 365 and 366), 
 in the summit of its cone. From the crater an opening leads 
 down to the source of the lava, at an unknown depth. Craters 
 vary greatly in size. Some of them are a mile or more across, and 
 some but a small fraction of a mile. The sizes and shapes of the 
 openings leading down to the sources of the lava cannot be seen 
 while the volcano is active, but they doubtless vary much in size 
 and shape, and perhaps in length. 
 
 
340 
 
 Volcanoes exhibit two great types of eruption. These are 
 (1) the quiet type and (2) the explosive type. In the former the 
 liquid lava rises up into the crater, and either (a) flows over its 
 rim or (b) breaks through the cone and flows down its sides. In 
 
 FIG. 365. Panum crater, Cal.; Lake Mono and Paona Island in the dis- 
 tance. (U. S. Geol. Surv.) 
 
 the latter the material is blown out by explosions from within. 
 In this case the material may be either liquid or solid when it is 
 thrown out, but the liquid lava cools rapidly in the air and be- 
 comes solid quickly. Small masses of liquid lava blown out of a 
 volcanic vent are often solid when they fall, after even a few sec- 
 onds of flight through the air. 
 
 FIG. 366. Sketch of the crater of the cinder cone near Lassen Peak, Cal., 
 showing the peculiar feature of two rings. The funnel is 240 feet deep. 
 (U. S. Geol. Surv.) 
 
 Some volcanoes discharge quietly at one time and explosively 
 at another, and in some there is some measure of explosive violence 
 at all times, accompanied by some quiet discharge of lava. 
 
 From the follo"\fmg accounts of a few active volcanoes, many 
 of the features of volcanic action will be gathered. 
 
VULCANISM 341 
 
 Examples of Active Volcanoes 
 
 Stromboli. The cone of this volcano is an island 4 or 5 miles 
 in diameter, in the Mediterranean Sea, north of Sicily. The cone 
 is built up from tho bottom of the sea, and is about a mile high, 
 though but little more than half of it projects above the water. 
 About 1000 feet below its top there is an opening in the side of 
 the mountain, from which steam issues constantly. At a dis- 
 tance, the condensed water vapor looks like smoke. 
 
 It is sometimes possible to climb up to the opening or crater 
 and look in. The floor of the crater is then seen to be of black 
 rock composed of hardened lava. There are cracks in the floor, 
 and from some of them steam puffs out somewhat as from an 
 engine. In other cracks liquid lava may be seen to be boiling. 
 Bubbles form in it and burst, much as bubbles form and burst in a 
 pot of boiling mush. When they burst, fragments of the lava 
 of which the bubbles are composed are hurled hundreds of feet 
 into the air, and fall on the slopes of the cone, increasing its size. 
 
 At night the glowing lava in the cracks of the crater floor lights 
 up the clouds which hover over the mountain. For this reason 
 Stromboli is known as "the lighthouse of the Mediterranean." 
 
 The eruptions of Stromboli are occasionally so violent that the 
 roar of the escaping steam may be heard for miles, w r hile the ejected 
 material is hurled so high and so far that it is scattered not only 
 over the entire mountain, but into the surrounding sea. Stromboli 
 is an example of a volcano which is at the present time constantly 
 active. 
 
 Stromboli is one of many volcanoes which have existed in this 
 part of the Mediterranean Sea. Some of the others, such as Etna, 
 are still active, while others are dormant or extinct. 
 
 Vesuvius. Vesuvius is probably the best-known volcano. 
 Its cone is a mountain about 4000 feet high, on the shore of the 
 Bay of Naples, about 10 miles from the city of the same name. 
 The present cone of the volcano (Fig. 367) rises within the half- 
 destroyed rim of an older and much larger crater. 
 
 Previous to 79 A.D. Vesuvius was, so far as then known, only a 
 conical mountain in whose summit was a deep crater three miles 
 in diameter. The slopes and even the bottom of the crater were 
 covered with vegetation. In that year a most destructive ex- 
 
342 PHYSIOGRAPHY 
 
 plosion occurred, and blew away half the rim of the old crater. 
 Much of the rock blown out was broken into such small pieces 
 as to constitute dust (often called volcanic ash), and as it fell on 
 the surrounding country, it buried and destroyed not only plants, 
 but even cities. Pompeii, a city of some 20,000 inhabitants, was 
 thus buried, locally to a depth of 25 to 30 feet, and about 2000 of 
 its people were killed. During this eruption there were no streams 
 
 FIG. 367. Cinder cone forming the summit of Mt. Vesuvius. 
 
 of lava. Heavy rains accompanied or followed the eruption. 
 Falling on the volcanic dust, the rains gave rise to devastating 
 streams of hot mud. Herculaneum was overwhelmed by such a 
 stream, perhaps 60 feet deep at a maximum. The present cone of 
 Vesuvius has been built up inside the remnant of the rim of the 
 older cone since this eruption. 
 
 Since the outburst of 79 A.D., Vesuvius has had other violent 
 eruptions, separated by periods when it was quiet or when its 
 activity was mild. The eruption of 1631 was especially violent, 
 destroying 18,000 lives. The emission of steam and volcanic 
 dust was followed by outflows of lava, some of which reached the 
 sea. Other eruptions of importance occurred in 1737, 1794, 1822, 
 
VULCANISM 343 
 
 and 1872. For several months before the principal eruption of 
 1872 there had been mild eruptions, during which steam and fine 
 fragments of rock matter were ejected from the crater, and flows of 
 lava issued from cracks on the mountain-side. The activity grad- 
 ually increased in violence until April, when the eruption culminated. 
 Two huge fissures and several smaller ones opened on the flanks 
 of the cone, and from them great streams of lava flowed into the 
 neighboring valleys, overwhelming two villages. At the same 
 time, two large openings were made at the summit, from which 
 enormous quantities of steam, dust, and bomb-like masses of molten 
 rock were hurled 4000 feet or more into the air, with a noise which 
 could be heard for many miles. At night the cloud overhanging 
 the mountain was brightly illuminated by the glowing lava in the 
 crater. Earthquakes were felt throughout the entire region. 
 The discharges continued with great violence for four days. After 
 the eruption was over, two craters 750 feet deep, with nearly vertical 
 sides, were found at the summit. An enormous amount of loose 
 material had accumulated on the sides of the mountain, and the 
 lava which issued from the fissures lower down covered a large area. 
 
 When Vesuvius is but mildly active it is possible to climb to 
 the rim of its crater and look in. It is necessary to climb up on the 
 windward side, because of the noxious vapors which are blown 
 to leeward. Even on the windward side it is necessary to be 
 mindful of the course which is followed, for stifling and poisonous 
 gases are pouring out of numerous little vents. Fortunately the 
 poisonous gases have such a disagreeable odor that they are readily 
 detected. 
 
 The phenomena which may be seen and felt on the mountain 
 differ from time to time, but the conditions of a particular day 
 (in June, 1887) may be taken as fairly characteristic. Soon after 
 the ascent on that day began, rumbling noises were heard, accom- 
 panied by slight tremors or quakings. As the summit was ap- 
 proached, the noises grew louder, and the shaking of the moun- 
 tain more distinct, until, by the time the top was reached, both 
 noises and tremblings were nearly continuous. 
 
 From the rim of the crater it could be seen that there were 
 three places where the floor of the crater was not crusted over. 
 In these openings, the lava boiled and bubbled like thick liquid 
 in huge caldrons. About three times a minute there were ex- 
 plosions within these openings, which shook the whole top of the 
 
344 
 
 PHYSIOGRAPHY 
 
 mountain. At the same instant, hundreds of fragments of the 
 glowing lava were shot up into the air. After rising several 
 hundred feet, these fragments fell; but they were so scattered as 
 they fell, and they came down from such great heights, that it 
 was not difficult to avoid them. They were often glowing-hot as 
 they started upward, but they quickly cooled enough to stop glow- 
 ing, and when they reached the surface of the cone they were 
 
 FIG. 368. The Cauliflower cloud above Vesuvius, April 7, 1906. 
 (Jaggar, Nat. Geog. Mag.) 
 
 dark, slag-like pieces of rock, though not always thoroughly solid. 
 Some of the material ejected was in very small fragments, and 
 some of it in pieces weighing scores and hundreds of pounds. 
 Steam and many ill-smelling vapors were also constantly issuing 
 from the crater. The water vapor which issued was soon condensed 
 into clouds as it rose and cooled, so that clouds hung over the 
 mountain. 
 
 From the rim of the crater it was clear that the explosions 
 which blew out the lava were also the cause of the noises and the 
 quaking. At night the glowing lava of the uncrusted openings in the 
 bottom of the crater lighted up the clouds above, most brightly dur- 
 ing explosions, when hotter lava from greater depths was exposed. 
 
VULCANISM 345 
 
 Vesuvius was again disastrously active in the spring of 1906, 
 when quantities of dust and flows of lava were sent forth, causing 
 much destruction of property and some loss of life. 
 
 Professor Jaggar has described the conditions late in April as 
 follows: "The lava-fields of 1872 and 1898 were found buried under 
 5 or 6 inches of sand and dust, which formed a heavy mantle, but 
 not sufficient to wholly disguise the slaggy contortions beneath. 
 The whole cone of Vesuvius became cleared of clouds in the course 
 of the afternoon, and it was seen to be covered with straight sand- 
 
 v ; 
 
 FIG. 369. The new cone of Vesuvius, shrouded in snow-white ashes. 
 (Jaggar, Nat. Geog. Mag.) 
 
 slides of whitish-gray color, which occasionally slipped downward 
 as on the steeper slopes of a dune. Pure white steam boiled up 
 slowly from the crater. In one instance it burst out radially over 
 the edge of the crater, showing a ring on the border, a dome of 
 cumulus above and within, and a second still higher outer ring 
 made of an older rain-cloud which had been punctured and pushed 
 up bodily. The effect was like a hat on the mountain's crown. 
 At night the cone was clear and entirely without luminosity." 
 
 As seen from the top, the crater w r as so full of steam, etc., that 
 little could be seen; but occasionally "we could make out an in- 
 ward slope of 35 or more degrees, covered with hot sand and broken 
 rock fragments, terminated about 120 feet (vertically) below by 
 jutting ledges which appeared to be precipitous. Beyond was 
 
346 
 
 PHYSIOGRAPHY 
 
 steam and sulphurous heat and obscurity. The ledges fumed in 
 places. No noise could be heard above the howling of the wind. 
 The curvature of the crater edge was irregular with embayments, 
 and it showed much irregularity in height. We could not see 
 the opposite side of the caldron, but from the curvature it was 
 estimated that the crater could not be less than from one-fourth to 
 one-half mile in diameter unusually large for Vesuvius." 
 
 The history of the recent eruption is summed up by the same 
 author as follows: "In May, 1905, lava flowed from a split 
 
 FIG. 370. Vesuvius in 1906. (Hobbs.) 
 
 in the northwest side of the cone and continued in active motion 
 throughout the year. It ceased flowing at the time when the pres- 
 ent eruption opened a new vent on the south side of the cone. 
 On April 4, 1906, a splendid black ' cauliflower ' cloud rose from the 
 crater. On April 4, 5, 6, and 7 lava mouths opened along the south- 
 ern rift above mentioned, first 500 feet below the summit, then 1300 
 feet lower, and finally 600 feet low r er still, all in the same radial 
 line. The lowest mouth was more than half-way down the moun- 
 tain, and from this orifice came the destructive streams. It should 
 be borne in mind that these flows are not floods of lava which 
 cover the whole slope of the mountain, but relatively narrow, 
 snake-like trickles, none the less deadly when they push their way 
 through a closely built town. The molten rock crusted over and 
 cracked, making a tumble of porous bowlders at its front. 
 
VULCANISM 
 
 347 
 
 "At 8 P.M., April 7, a column of dust-laden steam shot up four 
 miles from the crater vertically. The cloud snapped with inces- 
 sant lightnings. New lava mouths opened, and the flows moved 
 forward, crushing and burning and swallowing parts of Boscotrecase, 
 the stream forking so as to spare some portions of the town. Mean- 
 time torrents of ashes fell on Ottajano, on the opposite side of the 
 volcano, and many roofs collapsed and lives were lost. At the 
 observatory Dr. Matteuci and his colleagues were obliged to re- 
 
 FIG. 371. The ruins of Ottajano. The roofs have fallen in under the load 
 
 of ashes. 
 
 treat, as the observatory was rocking violently and heavy stones 
 were falling. . . . 
 
 "Boscotrecase was ruined wholly by lava; Ottajano by falling 
 gravel. Boscotrecase is traversed in two places by the clinkery 
 lava stream, and in some cases houses were literally cut in two. 
 The stream of lava had forked about a spur of the mountain, leav- 
 ing the higher land with its vineyards untouched. The lower land 
 with its town was invaded. There is so little timber in the Italian 
 masonry construction that the uninvaded part of the town was not 
 burned at all. At Ottajano the roofs fell in under the weight of 
 sand and gravel. The roofs were largely flat or slightly sloping tiled 
 
348 
 
 PHYSIOGRAPHY 
 
 affairs. The ash and lapilli reached a depth of three feet on level 
 surfaces. The roofs carried the walls with them in many cases, 
 but there was no significant "earthquake. There was no fire, de- 
 structive lightning, nor strong wind. The persons who perished 
 were all found in the houses, where the sole cause of death was en- 
 tombment in the ruins." 
 
 Like Stromboli, Vesuvius is situated in a region where there 
 have been other volcanoes, some of which have been active within 
 historic times. 
 
 B 
 
 FIG. 372. Krakatoa after the eruption. A, as seen from the southwest, 
 and B, from the north. (Kept, of the Roy. Soc.) 
 
 Krakatoa. One of the most violent and destructive volcanic 
 explosions of which there is historical record was that of 1883, 
 in Krakatoa, a volcanic island in the Strait of Sunda, between 
 Sumatra and Java. 
 
 Previous to the great eruption, the island had been shaken by 
 earthquakes and minor explosions for some years. On the morn- 
 ing of the 27th of August there was a series of terrible explosions, 
 the sound of which was heard in southern Australia, 2200 miles 
 away. About two-thirds of the island was blown away (Fig. 372), 
 and the sea is now 1000 feet deep where the centre of the mountain 
 formerly stood. Enormous sea-waves were formed, which traveled 
 half-way around the earth. On the shores of the neighboring 
 islands the water rose 50 feet, causing great destruction. More 
 than 36,000 persons perished, mostly by drowning, and 295 villages 
 were wholly or partially destroyed. The sky over the island and 
 
VULCANISM 
 
 349 
 
 the bordering coasts became black as night from the clouds of dust. 
 It was estimated that steam and dust were shot up into the air 17 
 to 23 miles. The explosion produced great air-waves which traveled 
 three and more times around the earth. Its passage was recorded 
 by barometers in all parts of the world. The dust ejected during 
 this explosion has been noted already (p. 57). 
 
 Over a circle 10 to 12 miles from the centre of Krakatoa, the 
 sea bottom outside the crater was built up 10 to 12 feet. Along 
 
 a line to the west, the depth of the water was increased. 
 
 
 
 / \ 
 
 FIG. 373. A. Probable outline of the great crater ring of the Krakatoa 
 volcano after the ancient paroxysmal outbursts. The dotted line indi- 
 cates the mass which was blown away. 
 
 B. Probable outline of the Krakatoa volcano after the great crater indi- 
 cated by the dotted line had been filled up by growth of numerous small 
 cones within. 
 
 C. Form of Krakatoa in historical time, after the formation of the great 
 lateral cone of Rakata and the growth of other cones within the great crater. 
 
 D. Outline of the crater of Krakatoa as it is now. The dotted lines 
 indicate the parts blown away by the outburst of 1883 and the change in 
 form of the flanks by the fall of ejected matter. (Kept, of the Roy. Soc.) 
 
 The cause of this awful explosion was probably the same as 
 that of the milder eruptions of Stromboli, that is, the sudden es- 
 cape or explosion of superheated steam. 
 
 Something of the conjectured history of this volcano is shown 
 by Fig. 373, which is suggestive of the changes undergone by 
 volcanic cones. The explanation beneath the figure gives its 
 interpretation. 
 
350 
 
 PHYSIOGRAPHY 
 
 There are many other volcanoes, living and dead, in the vicinity 
 of Krakatoa. 
 
 Mont Pelee and Soufriere. The volcano of Mont Pelee is 
 situated on the island of Martinique (Fig. 376), one of the Lesser 
 
 Forsaken ( 
 
 Lang I. 
 
 FlO. 374. Krakatoa Island and surroundings before the eruption of 1883. 
 The numbers indicate the depth of the water in fathoms. 
 
 Antilles, at the eastern border of the Caribbean Sea. Its cone 
 descends by steep slopes to the sea on all sides but the south, where 
 it is bordered by a plain on which, prior to the eruption of 1902, 
 
 Forsak 
 
 1 | Lang I. 
 
 50 
 
 FIG. 375. Krakatoa Island and surroundings after the eruption of 1883. 
 The numbers indicate the depths of the water in fathoms. 
 
 stood the city of St. Pierre, with a population of about 26,000. 
 The crater of Pelee was half a mile in diameter, and its floor 2000 
 feet below the highest part of the crater rim. This rim was inter- 
 rupted at the southwest by a deep gash, through which a stream 
 
VULCANISM 
 
 351 
 
 FIQ. 376. Sketch-map of Martinique. (Nat. Geog. Mag.) 
 
 Pte.de Macouba 
 
 Cap St.ttartin 
 
 LEGEND : 
 
 Cnter 
 
 Hud Cnter 
 '' FumarolM 
 
 Annihilation Lint 
 
 Singe Lln 
 
 Ash Line 
 
 FIG. 377. Map of that part of Martinique devastated by the volcanic out 
 burst of 1902. (Hill. Nat. Geog. Mag.) 
 
352 
 
 PHYSIOGRAPHY 
 
 flowed. In the crater there was formerly a lake, but it is said to 
 have been dry for about half a century. 
 
 Previous to the eruption of 1902, Pelee had had two periods of 
 moderate activity within historic times, namely, in 1762 and in 
 1851. Neither was destructive to life. From 1851 to 1902 the 
 volcano slumbered. In the later part of April of the latter year 
 activity was renewed by (1) the discharge of steam, vapors, and 
 ashes, some of which were thrown 1300 feet above the top of the 
 mountain, and (2) by the opening of three vents in the basin of 
 the old crater. By April 25 sulphurous vapors had become so 
 
 FIG. 378. Mt. Pelee. (Am. Mus. Nat. Hist.) 
 
 abundant that horses dropped dead in the streets of St. Pierre, and 
 a little later the traffic of the streets was obstructed by the volcanic 
 dust or "ashes." On May 5 the mud which had accumulated in 
 the basin of the crater broke out and flowed down the valley, 
 overwhelming a factory and destroying a number of lives. Dur- 
 ing these early stages of activity there were numerous earthquakes, 
 and all cables from Martinique were broken. Detonations like 
 the report of artillery were heard even 300 miles away. 
 
 On May 8 the activity of the volcano reached its climax. On 
 that day a heavy black cloud swept down through the gash in the 
 crater rim over the plain to the southwest, and two minutes later 
 struck the city of St, Pierre, five miles distant. The city was at 
 once demolished. Buildings were thrown down, statues hurled 
 from their pedestals, and trees torn up. Explosions were heard in 
 
VULCANISM 
 
 353 
 
 FIG. 379. Successive stages of the dust-cloud of the eruption of Mt. Pele"e, 
 December 16, 1902. (La Croix.) 
 
354 
 
 PHYSIOGRAPHY 
 
 St. Pierre as the cloud reached it, and the city burst into flames, 
 fired either by the heat of the gases or the red-hot particles of 
 rock which the gases carried. A few moments later a deluge of 
 rain, mud, and stones fell, continuing the destruction. With very 
 few exceptions, the entire population, increased to some 30,000 by 
 refugees from the surrounding country, was wiped out of existence. 
 Study of the region after the eruption showed that the cloud 
 was probably composed of steam, sulphurous vapors, and dust. 
 
 FIG. 380. Outside of southern rim of crater of Pelee. The serrate edge 
 is due to landslides. (Hovey, Am. Mus. Nat. Hist.) 
 
 It is estimated to have had a temperature of 1400 to 1500 F. 
 (800 C.). The gases were heavier than air, and so swept along the 
 ground instead of rising. They may also have been kept down 
 by the clouds of steam and ashes thrown out just before the out- 
 burst of the destructive gases. Combustible gases seem not to 
 have been abundant, for the vegetation and thatched roofs in the 
 path of the blast were not burned, but only dried and withered. 
 The bodies of the victims were scorched, burned, or scalded. Ex- 
 cept in the axis of the blast, the clothing of the bodies was un- 
 burned, though the flesh beneath was burned and scalded. The 
 chief causes of death seem to have been suffocation by the noxious 
 vapors and gases and the great heat. Minor causes were blows 
 from stones thrown from the volcano, burns from hot stones, dust, 
 and steam, cremation in burning buildings, etc. 
 
VULCANISM 
 
 355 
 
 FIG. 381. Great rocks thrown out by the eruption of August 30, 1902. 
 (Hovey, Am. Mus. Nat. Hist.) 
 
 FIG. 382. St. Pierre after the eruption of Mt. Pelee, which is seen in the 
 distance. (Hovey, Am. Mus. Nat. Hist.) 
 
356 
 
 PHYSIOGRAPHY 
 
 Other eruptions occurred on May 20, 26, June 6, July 9, and 
 August 30. The first of these was similar in character and violence 
 to that of May 8, and destroyed such portions of the town as had 
 been spared by the first eruption. The blast of August 30 took a 
 path somewhat different from that of the earlier ones, and dev- 
 astated a number of villages in the vicinity of St. Pierre, adding 
 about 2000 to the list of human victims. Clouds of steam and 
 ashes were thrown to heights of 6 and 7 miles. 
 
 FIG. 383. Spine of Mt. Pelee. The spine rose about 1210 feet above the 
 crater rim. (Hovey, Am. Mus. Nat. Hist.) 
 
 The great crater of Mont Pelee is now occupied by a cone of 
 fragmental material and some lava. This cone now overtops the 
 crater rim, and terminates in a spire which rises hundreds of feet 
 above the shallow crater which occupies the apex of the cone, and 
 out of which it was thrust. Unlike the cone, the spire consists 
 of solid rock. It is believed to be the lava which filled the vent, 
 and which was pushed up by the expansive forces beneath. The 
 spire is reported to be rapidly crumbling. 
 
 An interesting case of sympathetic action was shown by a vol- 
 cano (Soufriere) on the island of St. Vincent (Fig. 385), about 90 
 
VULCAXISM 
 
 357 
 
 miles south of Martinique. After two days of premonitory symp- 
 toms the first eruption of the Soufriere occurred on May 7. The 
 
 Ekng So. / 
 700 m / 
 
 FIG. 384. Cross-section through the northern part of Mt. Pelee, showing 
 the growth of the spine. (Hovey, Am. Mus. Nat. Hist.) 
 
 FIG. 385. Sketch-map of the Island of St. Vincent, showing the zones of 
 devastation. On the black area the devastation of life was nearly 
 complete; in the checked area, slight. (Russell, Nat. Geog. Mag.) 
 
 eruption was similar to that of Mont Pelee, but as there was no con- 
 siderable city in the path of the steam-cloud, the loss of life was 
 
358 
 
 PHYSIOGRAPHY 
 
 much smaller, about 1350. The discharges from the vent were not 
 confined and directed by a valley so definitely as those of Mont 
 
 FIG. 386. The Soufriere, St. Vincent. (Hovey, Am. Mus. Nat. Hist.) 
 
 Pelee; hence they spread over a larger area, with less violence. 
 A later eruption, on May 18, preceded, by a short period, an 
 
 FIG. 387. Ash-filled gorge of the Rabaka, St. Vincent. 
 (Hovey, Am. Mus. Nat. Hist.) 
 
 outburst of Mont Pelee, and another, on September 3, followed a 
 great eruption of the sister volcano. 
 
 From both centres of activity the dust driven out was carried 
 long distances. On St. Vincent it formed beds 50 and 60 feet thick 
 
VULCANISM 
 
 359 
 
 in some places. There were no lava-flows in connection with any 
 of these eruptions. 
 
 FIG. 388. An eruption of steam from the ashes of the Walliban Valley 
 (Hovey, Am. Mus. Nat. Hist.) 
 
 FIG. 389. Ridge of Bunker Hill on the Richmond estate, St. Vincent. 
 Shows the devastation of trees and the accumulation of dust on the 
 crest of the ridge. (Hovey, Am. Mus. Nat. Hist.) 
 
 Earthquake tremors felt in China on May 8 are supposed to have 
 been connected with the violent eruption of that date. This is 
 
360 
 
 PHYSIOGRAPHY 
 
 FIG. 390. The Soufriere in eruption. Ruins of Walliban sugar-factory 
 in the foreground. (Photograph by Wilson.) 
 
 Fia. 391. A river of mud pouring from La Soufriere; the steam is rising 
 from hundreds of points in the hot stream. (Russell.) 
 
VULCANISM 
 
 361 
 
 the only case, with the exception of Krakatoa, in which tremors 
 are known to have been transmitted through the centre of the 
 earth to the opposite side. Earthquake shocks were felt in Vene- 
 zuela on August 30. 
 
 FIG. 392. Map of Hawaii. (U. S. Geol. Surv.) 
 
 Hawaiian volcanoes. The eruptions of the volcanoes thus far 
 described are more or less violent; but in the Hawaiian Islands 
 there are volcanoes whose eruptions are relatively quiet. Mauna 
 Loa is the largest of the four volcanic cones whose united mass 
 forms the island of Hawaii, which is 80 miles across. Mauna 
 Loa rises 14,000 feet above the sea. So far as known, almost the 
 
362 
 
 PHYSIOGRAPHY 
 
 whole island is made up of volcanic materials. The highest point 
 of the island is about 14,000 feet above the sea, but the island has 
 been built up from the sea bottom by the lava poured out from the 
 craters, and since the water about the island is about 16,000 feet 
 deep, the volcanic pile, the top of which is the island, is really 
 about 30,000 feet high. This is about the height of the highest 
 mountain above sea-level. 
 
 The crater of Mauna Loa (Fig. 392) is 3 miles long, 2 miles wide, 
 and about 1000 feet deep a very large crater. When the volcano 
 is not active, it is possible to descend into the crater and to walk 
 
 
 FIG. 393. View of crater of Kilauea. (U. S. Geol. Surv.) 
 
 about on its hard but hot floor. Cracks and other openings are, 
 however, generally present, and give evidence of the hot liquid 
 rock beneath. 
 
 Before an eruption the floor of the crater rises, and lakes of 
 lava appear in the enlarged openings in it. At intervals, foun- 
 tains of lava may rise from the lakes, sometimes to heights of sev- 
 eral hundred feet. Finally the eruption occurs; but the lava does 
 not usually flow over the rim of the crater. It generally comes 
 out through fissures which open on the side of the mountain, some- 
 times far from the top. Through them the liquid lava spouts, 
 sometimes for hundreds of feet, into the air, and then flows down 
 the sides of the mountain in streams. Such streams are some- 
 times half a mile in width, and flow for 50 miles. The lava streams 
 are somewhat like mountain glaciers in form. Their rate of ad- 
 vance is, however, much faster than that of glaciers, though much 
 slower than that of rivers. The lava flows faster at first, and more 
 slowly as it becomes cooler. Residents in the cities below go out at 
 intervals when the volcanoes are discharging, to see how the streams 
 
VULCANISM 
 
 363 
 
 of lava are coming on, and whether they are likely to descend 
 so far as to endanger life and property in the settled regions below. 
 As the lava streams reach natter ground, they spread out, and 
 the lava may collect in hollows, forming pools and lakes which 
 soon harden. The lava occasionally falls over cliffs (Fig. 395), 
 sometimes into the sea. 
 
 FIG. 394. The crater of Kilauea. (U. S. Geol. Surv.) 
 
 After it becomes hard the surface of a lava-flow may be nearly 
 smooth (Fig. 396), but it is often rough. It may be ropy (Fig. 397) 
 or clinkery (Fig. 398). The ropiness is due to movement of the 
 surface lava after it is partially hardened. The clinkery surface 
 is due to the breaking up of the hardened crust of the lava stream. 
 
 As the lava flows out, the lava lake in the crater at the summit 
 subsides, and great masses of the floor of the crater, formerly held 
 up by the lava below, sink. 
 
 During the eruptions of the Hawaiian volcanoes little steam 
 
364 
 
 PHYSIOGRAPHY 
 
 FIG. 395. Lava falling over cliffs, Kilauea. (H. M. S. Challenger Kept.) 
 
 FIG. 396. Relatively smooth lava surface near the Jordan craters, 
 Malheur Co., Ore. (U. S. Geol. Surv.) 
 
VULCANISM 
 
 365 
 
 is discharged, and there are no showers of dust or cinders, no loud 
 rumbling or explosive reports, and earthquakes are rare. The 
 
 FIG. 397. Ropy surface of lava, Mauna Loa, flow of 1881. (Calvin.) 
 
 Fro. 398. Clinkery lava, Cinder Buttes, Idaho. (U. S. Geol. Surv.) 
 
 eruption may continue for months at a time with so little dis- 
 turbance that only persons in the vicinity are aware of it. 
 
366 
 
 PHYSIOGRAPHY 
 
 Hawaii is one of a chnin of volcanic islands, 400 miles long. 
 Mauna Loa, therefore, like the other volcanoes studied, is one of a 
 considerable number in its region. 
 
 Common phenomena of an eruption. From the preceding 
 descriptions the essential features of eruptions may be gleaned. 
 In the explosive type of eruption, rumblings and earthquake shocks 
 
 FIG. 399. The volcano of Colima, Mex., in an active condition, March 24, 
 
 1903. (Arreola.) 
 
 due to explosions within the throat of the volcano often occur for 
 weeks or months previous to a violent outbreak. As the explo- 
 sions become violent, ashes, cinders, and bombs are shot forth and 
 fall upon the sides of the cone, while the summit of the mountain 
 is shaken. The clouds of condensed steam and dust rising from 
 the crater darken the sky, and torrents of rain, falling upon the 
 fine dust, form rivers of hot mud. Liquid lava may or may not 
 accompany the discharge of dust, cinders, etc. In the quiet type 
 of eruption, the lava rises in the crater and occasionally overflows 
 its rim; but more commonly a crack is opened in the side of the 
 
VULCAN ISM '367 
 
 cone by an earthquake shock, or by the pressure of the molten 
 rock within, and the lava issues below the top. 
 
 There is little or no burning in a volcano, for there is little or 
 nothing to burn. There is therefore no smoke. What appears as 
 smoke is mostly condensed water vapor (cloud), often blackened 
 by the dust. 
 
 The Products of Volcanoes 
 
 The materials which come out of volcanoes are partly solid, 
 partly liquid, and partly gaseous. The dust, the cinders, and the 
 larger pieces of rock are solid, the flowing lava is liquid, while the 
 number of vapors and gases which issue is large. 
 
 Lava. All the liquid rock which issues from a volcano is 
 lava. The term is also applied to the rock formed when the liquid 
 lava becomes solid on cooling. 
 
 Lava never flows so freely as water, and it is sometimes very 
 stiff or viscous. The distance to which it flows depends on its 
 amount, on the slope of the surface over which it flows, and on its 
 liquidity. The greater the amount of lava, the steeper the slope 
 on which it flows, and the more fluid it is, the farther it will flow. 
 
 As lava flows its upper surface cools and hardens. The sur- 
 face of a lava stream may thus become solid, while the interior is 
 still fluid. The fluid part may then break out at the side or end 
 of the hardened shell and flow away, leaving the hollow crust. 
 On further cooling the shell contracts and cracks, and sometimes 
 caves in. Sometimes the hardened surface is broken by the move- 
 ment of the fluid lava below, and the solid fragments, displaced 
 and upturned by the moving liquid, give the surface a jagged 
 appearance (Fig. 398). In 1872 the Modoc Indians of south- 
 eastern Oregon, from their nearly inaccessible retreat among the 
 lava-beds, waged a warfare which was for some time successful 
 against the United States troops. 
 
 Lava takes on various forms as it becomes solid. If it hardens 
 under little pressure, as at the surface, the gases and vapors which 
 it contains expand, and it is converted into a sort of rock froth. 
 If the lava solidifies quickly, without becoming frothy, it makes 
 volcanic glass or obsidian. If the lava cools slowly under pressure, 
 the substances of which it is composed crystallize into the form of 
 various minerals. The kinds and proportions of the minerals 
 depend upon the composition of the lava. 
 
368 PHYSIOGRAPHY 
 
 Cinders, ashes, etc. The fragmental materials which are blown 
 r out of a volcano are often nothing more than portions of lava which 
 solidified before ejection, or during their flight in the air. They 
 may be large or small. Masses of rock tons in weight are some- 
 times thrown out, and from such masses there are pieces of all 
 sizes down to minute dust particles. 
 
 The dust is often transported great distances from the volcano. 
 Being relatively light, it is thrown far into the air and, caught 
 by the winds, its particles are shifted incredible distances, as already 
 noted. While, therefore, the fluid lava and the larger fragmental 
 materials ejected from the volcano stay near the vent, the fine 
 materials are scattered broadcast. 
 
 Gases and vapors. The gases and vapors which issue from 
 volcanoes are of many kinds. Among the commoner ones are 
 those of water (H 2 O), carbon dioxide (CO 2 ), chlorine (Cl), hydro- 
 chloric acid (HC1), sulphur dioxide (SO 2 ), and hydrogen sulphide 
 (H 2 S) ; but with these more important ones there are many others. 
 Some of the gases are poisonous, and, as in the case of Pelee, their 
 temperature is sometimes so high as to be destructive to life. 
 
 Number, Distribution, etc. 
 
 Number. The number of volcanoes is not easily determined. 
 Various reasons make such determination difficult. In the first 
 place, it is often impossible to say whether a quiet volcano is dor- 
 mant or extinct. If the former, it should be counted; if the latter, 
 it should not. Again, the vent of a volcano often changes. In- 
 stead of discharging lava through a single crater, it may dis- 
 charge through several subordinate vents, more or less closely 
 associated with the main one. There may be differences of opinion 
 as to whether these several vents should be regarded as separate 
 volcanoes. For this and other reasons, the number of active vol- 
 canoes is not capable of definite statement. According to the 
 more common estimates there are between 300 and 400. Some- 
 thing like two-thirds of them are on islands, and the remainder on 
 the continents. There may be many in the sea which are not 
 known, for volcanoes in the deep sea might not be readily detected. 
 
 Distribution. The general distribution of active volcanoes 
 is shown in Fig. 400. Many of them are arranged in belts, and 
 within the belts they are sometimes in lines. The most marked 
 
VULCANISM 
 
 369 
 
 & 
 
 
 is 
 
 -s 
 
370 PHYSIOGRAPHY 
 
 belt nearly encircles the Pacific Ocean, as with a girdle of steaming 
 vents. This belt may be said to begin with the volcanic islands 
 south of South America, and includes the numerous vents in the 
 Andes and in the mountains of Central America and Mexico. The 
 belt widens in the western part of the United States, where the 
 volcanoes are extinct, but narrows again in Alaska and the Aleu- 
 tian Islands. On the west side of the Pacific, the volcanoes form 
 a well-marked belt with many active vents through Kamchatka, 
 Corea, Japan, the Philippine Islands, New Guinea, New Hebrides, 
 and New Zealand. A branch belt includes the volcanoes in the 
 islands of Java and Sumatra. The volcanoes of the West Indies 
 are sometimes considered as an eastern branch of the same belt. 
 Volcanoes are also numerous in the Mediterranean Sea, and there 
 are not a few which cannot be regarded as parts of any well- 
 defined belt. 
 
 Most volcanoes are in the sea or near it. Not a few of them 
 are in mountain regions, but it is by no means true that all moun- 
 tain regions have them. Not a few are on ridges or swells on the 
 sea bottom, or on ridges or swells which rise above the sea. Such, 
 for example, are the West Indian volcanoes. While the volcanoes 
 which are on the continents are on the whole near the shores, they 
 are not all near shores, nor do they occur along the borders of all 
 continents. There is an active volcano in Africa 700 miles from 
 the sea, and there are fresh cones of extinct volcanoes 500 to 800 
 miles from the sea in Arizona, Colorado, and Thibet. It cannot be 
 said, therefore, that nearness to "the sea or mountain ridges are con- 
 ditions necessary for volcanoes. 
 
 Many of the active volcanoes lie near the line where the con- 
 tinental plateau descends to the oceanic basins. Perhaps this is 
 the most significant feature of their distribution. 
 
 Volcanoes are, on the whole, not notably more abundant in 
 one latitude than in another. At any rate, they have a wide range 
 in latitude. 
 
 The data which are now in hand seem to point to the general 
 conclusion that volcanoes on land are commonly associated with 
 lands which have been recently warped. It is conceived that 
 these movements of the surface have some effect upon the pressures 
 and temperatures of the deeper zones beneath them, and that 
 these variations of pressure and temperature are among the con- 
 
VULCANISM 371 
 
 ditions necessary for the extrusion of lava from beneath the 
 surface. 
 
 Historical. Volcanoes have existed throughout the history 
 of the earth, so far as this history is now known, even back to the 
 earliest ages; but volcanic processes do not seem to have been 
 equally active at all times. There seem to have been periods of 
 great volcanic activity, alternating with much longer periods of 
 much less activity. There is no knowledge, however, that vul- 
 canism ever ceased altogether at any time. 
 
 While vulcanism seems to have been continuous, but more or 
 less periodic in its violence, the sites of volcanic activity have 
 shifted from time to time, and the areas where they now exist 
 are not the areas where they existed in former times. 
 
 What is now known of vulcanism seems to indicate that, gen- 
 erally speaking, a volcano has a beginning, runs a given course, 
 and dies. The vulcanism of a given region appears to have a 
 similar course. 
 
 It appears also that the phase of vulcanism sometimes changes 
 in a given region. In some volcanic regions fissure eruptions 
 came early in the course of the volcanic history. As activity de- 
 clined, fissure eruptions gave place to volcanoes, and the volcanoes 
 became less and less active, and finally extinct. 
 
 Even after vulcanism proper ceases, associated phenomena 
 are continued. Thus in the Yellowstone National Park there 
 are numerous geysers, hot springs, and other vents out of which 
 hot vapors issue. Such phenomena probably represent the last 
 phases of volcanic activity in the region. 
 
 IGNEOUS PHENOMENA NOT STRICTLY VOLCANIC 
 
 Fissure eruptions. Lava sometimes rises to the surface through 
 great fissures instead of through the relatively small vents of 
 volcanoes. From such fissures floods of lava spread over the 
 surrounding country sometimes for hundreds of miles. Such lava 
 floods once occurred in Oregon, Washington, and Idaho, where, by 
 successive flows, the pre-existing hills and valleys were buried, and 
 a vast plateau 200,000 square miles or more in extent was built up 
 (Fig. 401). Locally, the nearly level surface of the lava plateau 
 meets the mountains along its border, somewhat as the sea meets 
 the land, while islands of older rock rise above it. 
 
372 
 
 PHYSIOGRAPHY 
 
 In this lava plateau the Snake River (Fig. 26) has excavated 
 a great canyon 4000 feet deep in some places, and 15 miles wide. 
 The walls of the canyon show the structure of the plateau. They 
 show, among other things, the edges of the successive lava-flows, 
 sometimes separated by beds of sediment, with soils in which the 
 roots and trunks of trees are still preserved. These beds of sedi- 
 
 FIG. 401. Lava-flows of the northwestern part of the U. S. 
 
 ment, and these soils, show that long periods of time elapsed be- 
 tween successive lava-flows. At one point in the walls of the 
 canyon, a peak of older rock rising 2500 feet above the river is 
 buried by 1500 feet of lava. A rugged mountain region was here 
 
 FIG. 402. A. Ideal cross-section of a laccolith with accompanying sheet and 
 dikes. B. Ideal cross-section of a group of laccoliths. (Gilbert, U. S. 
 Geol. Surv.) 
 
 transformed into a plateau by the lava floods. A part of the 
 plateau has since been deeply dissected by streams, parts still 
 remain nearly plane, parts have been broken into blocks which 
 have been tilted into mountain ridges, while still other parts have 
 been arched up into great dome mountains. The Blue Mountains 
 
VULCAN ISM 
 
 373 
 
 of Oregon are the most conspicuous example of doming. Badger 
 Mountain of Washington is an elongate dome or anticline. 
 
 An older lava plateau of still greater size occurs in India. Ow- 
 ing to its greater age, its nearness to the sea, and the humid climate, 
 it is more dissected than the Oregon plateau. This lava has in 
 some areas weathered so as to form a soil of great fertility, to 
 which the Deccan owes its fame as a cotton-growing country. 
 Prominent hills of lava along the dissected edges of the flows have 
 frequently served as natural forts of great strength in the wars 
 of the country. Other dissected lava plateaus are found on the 
 north coast of Ireland and the west coast of Scotland. Some of the 
 islands off the coast of Scotland are remnants of an old lava plateau. 
 
 Fissure eruptions have occurred in Iceland within historic 
 times. In 1783 such flows took place from a fissure 20 miles or 
 so in length. The lava spread out in sheets on both sides of the 
 fissure, advancing in the valleys farther than on the uplands 
 between them. In this respect the lava-flows resemble the move- 
 ment of glacier ice. 
 
 While fissure eruptions of lava sometimes build up plateaus 
 or raise the level of the plains on which they spread, they do not 
 commonly give rise to mountains; but mountains are sometimes 
 developed from them, as they are dissected by stream erosion. 
 
 Intrusions of lava. Lava is sometimes intruded from below 
 into the crust of the lithosphere, without rising to the surface. 
 
 FIG. 403. Diagram of a bysmalith. 
 
 In such cases the surface strata may be arched up over the in- 
 trusion, making domes which sometimes reach the size of moun- 
 tains. Such mountains (Figs. 402, 404), of which the Henry 
 Mountains of Utah are examples, are called laccoliths. If the 
 roof of the intrusion is faulted up instead of being arched up, the 
 
374 
 
 PHYSIOGRAPHY 
 
 YHENRYMOUNTAINS 
 
 FIG. 404. Relief-map of the Henry Mountains. (Gilbert, U. S. Geol. Surv.) 
 
375 
 
 intrusion is called a bysmalith (Fig. 403). Intrusions of very 
 great size are batholiths. Lava is sometimes intruded between beds 
 
 FIG. 405. Diagrammatic representation of the relations of igneous rock 
 to stratified rock. The igneous rocks represented in black have been 
 forced up from beneath. 
 
 of stratified rock in sheets or sills (Fig. 405). Lava is also some- 
 times forced into cracks of the rock, solidifying there as dikes 
 (a, Fig. 405). 
 
 Causes of Vulcanism 
 
 The causes of vulcanism are somewhat outside the province of 
 physiography, but it may be stated that the lava of volcanoes does 
 not appear to come from a liquid interior, and the lavas from ad- 
 jacent vents do not appear to come from a common reservoir of 
 liquid rock. This is suggested by the fact that adjacent vents 
 frequently discharge different sorts of lava, and that the lava in 
 adjacent craters often stands at very different heights at the same 
 time. 
 
 The great pressure which exists in the interior of the earth 
 because of the weight of the overlying parts, insures a high tem- 
 perature to the interior. The heat thus developed is continually 
 working its way, in one way or another, to the outer portions of the 
 earth. It passes out by conduction everywhere, and locally, where 
 conditions favor, small amounts of rock may become liquid. This 
 liquid rock then works its way to the surface or toward it. Ac- 
 cording to this view, the extrusion of lava is to be looked upon as 
 one phase of the passage of interior heat to the surface. 
 
 In the explanation of volcanoes, two things are to be accounted 
 for: (1) the liquid lava and the heat necessary for its production, 
 and (2) the force which brings it to the surface. 
 
 Lava is to be regarded as a solution of mineral matter in min- 
 eral matter, rather than as melted rock. The solution, however, 
 takes place only at high temperatures. Various views have been 
 
376 PHYSIOGRAPHY 
 
 entertained as to the source of the heat necessary to cause minerals 
 to dissolve in one another. These views may be grouped into two 
 classes: (1) those according to which the heat is primary, that is, 
 that the interior of the earth has been hot always, or that it has 
 been hot since the earth attained its present size; and (2) those 
 according to which the heat which liquefies the rock is secondary, 
 and developed in rock (relatively near the surface) \vhich was once 
 cool. Some of the hypotheses of volcanic action based on these 
 views may be considered briefly. 
 
 1. It was formerly thought that the whole interior of the earth 
 might be liquid, and that the volcanic vents were connected with 
 this liquid interior. This view was based on certain familiar facts. 
 Deep mines and borings of all sorts show that the temperature 
 increases with increasing depth. The rate of increase varies 
 widely from 1 for 17 feet to 1 for more than 100 feet. The average 
 rate of increase is commonly stated as about 1 for every 50 to 60 
 feet; but if the estimate be based on the records of those deep 
 mines and other borings which seem to afford the most reliable 
 data, the rate is more nearly 1 for 100 feet, down to the greatest 
 depths yet penetrated. It is to be remembered, however, that 
 the deepest excavations are but little more than a mile in depth, 
 and that most excavations on which the generalizations are based 
 are much shallower. If the heat increases at the average rate 
 of 1 per 100 feet, a temperature of 3000 would be reached at a 
 depth of about 60 miles. Such a temperature would be enough 
 to liquefy rocks at the surface, but we are not to conclude that rocks 
 are liquid at this depth even if the temperature is 3000. At 
 this depth, the pressure due to the overlying rock is enormous. 
 Rock expands when it is liquefied, and the pressure at this depth 
 may be enough to prevent expansion, and so to prevent general 
 liquefaction. There are many reasons for believing that, though 
 the temperature of the interior of the earth is very high, the rock 
 is still solid. The fundamental element of the hypothesis, that all 
 volcanoes start from a common liquid center, is therefore believed 
 to be wrong. 
 
 2. It has been suggested that there is, at some depth beneath 
 the surface, a liquid layer below the solid crust and above a great 
 solid centre. This hypothesis does not seem to be well supported, 
 and does not seem to meet the objections to the hypothesis first 
 mentioned. 
 
VULCANISM 377 
 
 3. Another view has been that while the earth is virtually solid, 
 it is solid in spite of its internal temperature, and that if the pres- 
 sure were lessened at some point beneath the surface, the hot rock 
 would expand and become liquid. The pressure, it is conceived, 
 would be lessened where the outer part of the earth is folded up, 
 as in some mountains. This hypothesis has found much favor, 
 but it does not seem to account for some of the fundamental facts 
 connected with volcanoes, such as their distribution. 
 
 The hypotheses that the heat involved in volcanic action is 
 secondary, seek to explain the heat (1) by means of the crushing 
 of rock such as sometimes takes place when beds of rock are folded, 
 or (2) by chemical action between the elements of the rocks, or 
 between these elements and water which descends from the 
 surface. These hypotheses have little acceptance at the present 
 time. 
 
 No one of the preceding hypotheses, nor all combined, seem 
 to adequately explain vulcanism, and no hypothesis which seems 
 altogether satisfactory has been put into definite form. It seems 
 possible (1) that the local formation of liquid lava is a process 
 which is constantly but slowly going on in the deep interior, 
 perhaps where the rock material is more readily soluble than the 
 average; and (2) that the liquid rock is continually finding its 
 way to the surface, faster and in greater quantities at some times 
 than at others. The regions where the crust is least stable, that is, 
 where there is movement, are the regions most likely to afford the 
 lava a place of escape, for it is in such places that it is weakest. 
 
 The principal forces involved in the extrusion of lavas are ap- 
 parently two, (1) gravity and (2) the expansive and explosive 
 force of the vapors and gases contained in the lavas, especially 
 water vapor. 
 
 Lava beneath the surface would, if lighter than the solid rock 
 above, tend to find its way to the surface, or, more strictly, the 
 heavier rock above would tend to sink down, squeezing out the 
 lighter liquid rock beneath. This has probably been an important 
 factor perhaps the most important factor in the eruption of 
 some volcanoes and in some fissure eruptions. If at the same 
 time the region concerned is affected by lateral pressure, this 
 pressure might help to squeeze out the liquid lava. Pressure 
 from above or from the sides seems to be the principal factor 
 involved in the extrusion of lavas in quiet eruptions. Gases and 
 
378 PHYSIOGRAPHY 
 
 vapors in the lava tend to expand it, especially as pressure is 
 relieved, and so tend to diminish its specific gravity. 
 
 In the case of violent eruptions the gases and vapors, espe- 
 cially water vapor, appear to play a principal part. Even in these 
 cases, however, it is probable that gravity is the principal factor 
 in getting the lava up near to the surface, and that the vapors 
 and gases come into effective function only as the surface of the 
 lithosphere is approached. 
 
 The source of the vapors which issue from volcanoes is a matter 
 about which there is much difference of opinion. Among the 
 vapors which escape from volcanoes there are those which 
 might have been derived from sea-water. From this fact it was 
 inferred that sea-water had access to the sources of the lava. It 
 is now thought, however, that water probably does not descend 
 more than five or six miles beneath the surface of the lithosphere, 
 for below some such depth, pores and cracks, without which water 
 cannot descend, do not exist. It seems certain that the sources 
 of the lava are much deeper, and it therefore seems improbable 
 that descending water, either from the sea or from the land, reaches 
 the sources of vulcanism. 
 
 It seems probable that lava from depths far below all ground- 
 water is forced up to within a short distance of the surface before 
 coming into contact with water. In its passage through the outer 
 part of the crust, which contains water, the lava doubtless con- 
 verts water into steam; and the steam thus produced is possibly 
 an important factor in the rise of the lava through the outermost 
 portion of the earth's crust. But there is the best of reason for 
 believing that lava brings up vapors and gases, and among them 
 water vapor, from much greater depths. Such gases and vapors 
 must, it would seem, have been long within the earth. It is prob- 
 able, indeed, that some of them are now reaching the surface of 
 the earth for the first time. If this be true, they are to be looked 
 upon as original constituents of the earth. 
 
 Topographic Effects of Volcanic Action 
 
 Many volcanoes build up great cones, some of them mountain- 
 high, as already indicated. The first stages of growth have some- 
 times been observed. 
 
 Volcanic cones. In 1538 a small volcano appeared on the 
 
VULCANISM 379 
 
 north shore of the Bay of Naples, and built up a cone 440 feet high 
 and half a mile in diameter at its base in a few days. Its crater 
 was more than 400 feet deep. The development of the volcano 
 was preceded by earthquakes, which were felt in the same regions 
 for two years before the volcano was formed. 
 
 In 1770 the volcano Izalco in Central America broke out in 
 the midst of a plain which was then a cattle-ranch. Since that 
 time it has built up a symmetrical cone about 3000 feet high, with 
 steep slopes. In the earlier part of its history, lava-flows, attended 
 with streams of cinders, etc., were of frequent occurrence. For 
 many years no lava has flowed out, though the volcano has 
 remained active, discharging explosively. Earthquakes and rum- 
 blings preceded the original eruption. 
 
 In January, 1880, a volcano broke* out in Lake Ilopango, San 
 Salvador, Central America. The eruption continued more than 
 a month, heating the waters of the lake, killing the fish, and form- 
 ing a cone-shaped island rising 160 feet above the lake, which was 
 600 feet deep. A violent earthquake occurred in this region a 
 few months before the birth of the volcano. After the earth- 
 quake the water of the lake sank 35 feet. 
 
 Early in the last century a volcanic island (Graham Island) 
 arose in the Mediterranean, between Sicily and Africa, where the 
 water had been 800 feet deep. In 1831 a ship near the place felt 
 earthquake shocks. In July a sea-captain reported that he saw 
 a column of water 60 feet high and 800 yards in diameter rising 
 from the sea, and soon afterward a column of steam which rose 
 1800 feet. A few days later there was a small island 12 feet high 
 where the disturbance had been, and in its centre there was a 
 crater, from which eruptions were seen to be taking place. By 
 the end of the month the island was 50 to 90 feet high and f oi 
 a mile in circumference. On August 4 it was 200 feet high and 
 3 miles in circumference. Activity soon ceased, and early in 1832 
 the island had been destroyed by the waves. This volcano was 
 short-lived, as was the island which it built. 
 
 Volcanoes have recently built up islands off the coast of Alaska. 
 In 1795 such an island appeared about 40 miles west of Unalaska. 
 In 1872 this island was 850 feet above the sea, but had no crater. 
 In 1883 another island appeared close by, and was later joined to 
 the first. In 1884 it was 500 to 800 feet high. 
 
 Great mountains, as well as small ones, are often formed bv 
 
380 
 
 PHYSIOGRAPHY 
 
 FIG. 406. Mt. Shasta, a typical volcanic cone furrowed by erosion, but 
 retaining its general form. (U. S. Geol. Surv.) 
 
 FIG. 407. A part of the "crater" of Coon Butte, Ariz. The "butte" is 
 only the rim built up about the "crater" by the material blown out. 
 (R. T. Chamberlin.) 
 
VULCANISM 381 
 
 volcanoes. Thus Mt. Rainier in Washington (Fig. 408), Mt. 
 Hood in Oregon (Fig. 227), Mt. Shasta in California (Fig. 406), 
 and the San Francisco Mountain in Arizona (Fig. 411), as well 
 as numerous other high and well-known mountain peaks, were 
 built up by volcanoes. The volcanoes themselves have been 
 dead for long. Rainier, Hood, and Shasta are all so high that, 
 in spite of their origin, snow-fields and even glaciers are found on 
 them. 
 
 Many small islands, and some large ones, such as Iceland, are 
 due chiefly or wholly to the building up of volcanic cones which 
 have their foundations on the ocean bottom. The Aleutian Islands, 
 the Kurile Islands, and many of the islands of Australasia were 
 formed in the same way. Among the latter are the famous Spice 
 Islands (Moluccas) so important in connection with the early 
 history of America. 
 
 By the making of cones, volcanoes become an important factor 
 in shaping the surface of the lithosphere. The number of volcanic 
 cones which have assumed mountain size on land is large, and 
 the number in the sea still larger; but in spite of their great num- 
 ber, their aggregate area is relatively slight. The total area of 
 mountainous lands developed by volcanoes is but a fraction of the 
 area of mountain land developed in other ways. 
 
 Intrusions of lava may give rise to mounds, mountains, or even 
 plateau-like swells (laccoliths, bysmaliths, batholiths), as already 
 indicated. 
 
 In Arizona (near Canyon Diablo) there is a crater-like depression 
 with a distinct rim about it, composed of the material which was 
 blown out of the depression. The rim is high enough to be seen 
 from a great distance, and is known as Coon Butte (Fig. 407). There 
 is no lava about the butte, and it cannot therefore be called a 
 volcano. Apparently a great explosion beneath the surface blew 
 out a large quantity of rock from the crater-like depression, form- 
 ing the high rim. The formation of this depression may perhaps 
 be looked upon as a preparatory step for a volcano. If so, the 
 process was arrested before a volcano was developed. 1 
 
 1 The hypothesis was advanced long since by Gilbert that this so-called 
 butte, with its accompanying depression, was due to the fall of a large 
 meteorite. Mr. Gilbert abandoned this hypothesis, after somewhat full 
 investigation, but it has been revived recently, in modified form, by Fair- 
 child (Bull. G. S. A., Vol. XVIII, p. 493.) 
 
382 PHYSIOGRAPHY 
 
 Destruction of volcanic cones. 1. A volcanic cone may be 
 partially destroyed by a violent explosion, as in the cases of Kra- 
 katoa and Vesuvius, already cited. Enormous depressions, called 
 calderas, several miles in diameter and hundreds of feet deep, may 
 be developed in this way from previous cones and craters. 
 
 A volcanic cone may be undermined by the withdrawal of the 
 liquid lava in the core of the mountain, if it escapes at a lower 
 level. The entire summit of the mountain may then sink and be 
 engulfed, thus forming a caldera. Crater Lake, Oregon (p. 305), 
 occupies a caldera in the stump of a great volcanic cone (Figs. 
 333-336). There are several large calderas in the Azores, and the 
 floors of some of them are the sites of villages. 
 
 2. Volcanic cones are also destroyed by the less violent proc- 
 esses of weathering and erosion. The destruction of Graham 
 Island by the waves has already been cited. Wind and rain attack 
 volcanic cones as soon as they are formed, but their results are not 
 conspicuous until the volcano is extinct and the cone stops grow- 
 ing. Cones composed largely of cinders, etc., are worn away with 
 comparative ease, while those of lava resist longer. Glaciers fre- 
 quently aid in their degradation. Among the many extinct 
 volcanic peaks in the western part of the United States, it is pos- 
 sible to find illustrations of various stages in the process of destruc- 
 tion. Only those of relatively recent origin still show their un- 
 modified or but slightly modified forms All volcanic cones 
 except those of recent origin have lost their craters and the 
 symmetry of slope which they probably once possessed. 
 
 Examples of fresh cones. In Arizona, California (Fig. 364), 
 Idaho, Oregon, and elsewhere in the United States, there are 
 volcanic cones so recently formed that they have suffered little 
 destruction. Their conical forms have been little disfigured by 
 erosion, and their surface materials appear to have been but 
 little changed by weathering. Cones of similar freshness are found 
 in other lands, as in the Auvergne (France). 
 
 Examples of worn cones. Mt. Shasta (Fig. 406), in northern 
 California, is a volcanic cone which rises 2 miles above its base, 17 
 miles in diameter, to a height of 14,350 feet above the sea. It is 
 partly of lava and partly of fragmental material. Its upper 
 slopes are steep and furrowed with ravines. About 2000 feet 
 below the summit on the west side is a fresher and therefore 
 younger cone, known as Shastina, with a crater in its top. Re- 
 
VULCANISM 383 
 
 mains of more than 20 smaller cones also occur on the lower flanks. 
 Near the base are several lava-fields which, from the roughness of 
 their surfaces and the absence of soil, are known to have been 
 formed since the glaciers occupied the summit. The fact that the 
 Sacramento River has cut a narrow gorge 100 feet deep across 
 one of them proves, however, that the last eruption was many 
 years ago. Mt. Shasta is a good example of a volcano which has 
 suffered some erosion, but about which evidence of recent erup- 
 tions has not been destroyed. 
 
 The great changes which Shasta has undergone are made clear 
 by the fact that the once hot mountain is now the home of several 
 glaciers, which, moving over its slopes, are helping to waste it 
 away. 
 
 Mt. Rainier (Fig. 408) is another splendid mountain developed 
 by a former volcano. Various features of the mountain show 
 
 FIG. 408. Mt. Rainier, Wash. 
 
 that a second period of activity followed a long period of quiescence 
 in the history of this snow-capped mountain. Hot vapors still 
 issue from some small vents in the mountain, though the dis- 
 charge of rock material ceased long ago. The mountain is snow- 
 covered and the home of several glaciers. 
 
 Mt. Hood (Fig. 227), one of the peaks of the Cascade Range, 
 has been eroded more than Rainier. Only a part of the wall of its 
 summit crater remains, and its sides are deeply furrowed by ra- 
 vines, which are separated by sharp, jagged ridges, and precipices 
 hundreds of feet in height. Nevertheless, sulphurous fumes are 
 still escaping from openings in the rocks, even on the snow-clad 
 summit. 
 
384 PHYSIOGRAPHY 
 
 The Marysville Buttes. This circular cluster of hills (Figs. 
 409 and 410), 10 miles in diameter, rises 1700 to 2000 feet above 
 the level of the Sacramento River in California. The buttes are 
 composed of lava with an outer layer of fragmental material (or 
 tuff). The volcanic cone, which probably once rivaled Vesuvius, 
 
 FIG. 409 Marysville buttes in contour. (U. S. Geol. Surv.) 
 
 has been dissected into a group of hills with jagged and fantastic 
 outlines. So deeply have the roots of the mountain been laid bare 
 that the broken and distorted layers of sedimentary rock through 
 which the lava was erupted are exposed. 
 
 The San Francisco Mountain (Fig. 411) is another example 
 of a volcanic mountain partially destroyed by erosion. The 
 form of the old cone can be but imperfectly known. Numerous 
 minor volcanoes existed about the main ore after the latter be- 
 
VULCANISM 
 
 385 
 
 came extinct. It is said that the number of fresh volcanic cones 
 in this vicinity is more than 300. Many of them are so young that 
 
 they show little sign of erosion. 
 
 ..- "?#. . . ; " 
 
 Indirect Topographic Effects of Vulcanism 
 
 Volcanic necks. When a volcano becomes extinct, the throat, 
 or passage from the interior, may be filled with hardened lava. 
 This may be of much more resistant rock than the rest of the cone. 
 The cone may in time be worn away, but the plug, transformed 
 
 FIG. 410. Marysville volcanic cone. (Photograph of model by 
 Newsome.) 
 
 into a hill, may still mark the site of the former volcano. These 
 volcanic necks or plugs are sometimes conspicuous. East of the 
 Mt. Taylor plateau, in central Mexico, a number of them rise by 
 precipitous slopes 800 to 1500 feet above their surroundings. 
 Massive intrusions of lava have the same effect (Fig. 171). 
 
 Intrusions of lava may give rise to topographic features of 
 importance after erosion has affected the regions where they 
 occur, for the hardened lava or igneous rock is often harder than 
 its surroundings. Dikes often give rise to ridges (Fig. 412). Sills 
 also, if they have been tilted notably from a horizontal position, 
 give rise to ridges which may be so high as to be called moun- 
 tains. The Palisade Ridge of the Hudson (Fig. 413) and most of 
 the mountains of the Connecticut River Valley are illustrations. 
 A sheet of lava poured out on the surface and subsequently buried 
 
386 
 
 PHYSIOGRAPHY 
 
VULCANISM 
 
 387 
 
 by sediment may have the same effect on topography as a sill. 
 The Watchung Mountains of New Jersey are the projecting edges 
 
 FIG. 412. A dike isolated by erosion, Spanish Peaks region, Colo. 
 (U. S. Geol. Surv.) 
 
 FIG. 413. The Palisade Ridge. 
 
 of extrusive sheets of lava, once buried beneath sedimentary 
 rocks, then tilted, and later isolated by erosion. Sills and extrusive 
 
388 
 
 PHYSIOGRAPHY 
 
 sheets of lava may also give rise to buttes, mesas, rock terraces, 
 etc. indeed, to all the topographic forms which result from the 
 erosion of rock of unequal hardness (p. 163). 
 
 Columnar structure. Hardened lava sometimes assumes a 
 columnar structure (Figs. 414 and 415). This is probably the result 
 of the contraction incident to cooling. The surface of the homo- 
 geneous lava contracts about equally in all directions on cooling. 
 The contractile force may be thought of as centering about equi- 
 distant points. About a given point the least number of cracks 
 
 B 
 
 FIG. 414. A. Columnar structure in basalt, Giants Causeway, Ireland. 
 B. Columnar structure on a larger scale. 
 
 which will relieve the tension in all directions is three (A, Fig. 415). 
 If these radiate symmetrically from the point, the angle between 
 any two is 120, the angle of the hexagonal prism. Similar radi- 
 ating cracks from other centers complete the columns (B, Fig. 415). 
 A five-sided column would arise from the failure of the cracks to 
 develop about some one of the points (C, Fig. 415). 
 
 Mud Volcanoes 
 
 Mud volcanoes have some features in common with volcanoes 
 and some in common with geysers, while in others they depart 
 from both. Like volcanoes and geysers, they are eruptive, but, 
 as the name implies, their discharge is mud, instead of lava or 
 water. The general conditions which seem to be necessary for their 
 existence are (1) sufficient heat beneath the surface, presumably at 
 
VULCAXISM 
 
 389 
 
 a relatively slight depth, to produce steam, and (2) a surface layer 
 of earthy material, which when moist becomes mud. The steam 
 
 A B C 
 
 FIG. 415. Diagrams to illustrate the formation of columns in basalt: 
 A. The first stage i the development of a hexagonal column. B. The 
 completion of a hexagonal column. C. A pentagonal co.umn. 
 
 escaping through the mud forces some of it out, building up small 
 cones which simulate volcanic cones in form, though not in con- 
 stitution. They never reach great size. 
 
 FIG. 416. Mud cones Yellowstone National Park. (Fairbanks.) 
 
 Like geysers, mud volcanoes occur in regions of present or 
 relativelv recent vulcanism, for the most part. They are some- 
 
390 PHYSIOGRAPHY 
 
 times violently explosive, and sometimes not. Some of them 
 erupt at infrequent intervals, and some nearly continuously. 
 
 The "paint-pots" (Fig. 416) of the Yellowstone Park belong to 
 the same category, though from them there is little discharge, and 
 they do not build considerable cones. 
 
 When quantities of gas escape from beneath the surface through 
 wind, eruptions somewhat like those of mud volcanoes may take 
 place, even in the absence cf heat. 
 
 MAP EXERCISE 
 
 I. Study the following maps showing volcanic mountains, in prepara- 
 tion for the confrenece: 
 
 1. Marys ville, Cal. 
 
 2. Mt. Lyell, Cal. 
 
 3. Mt. Shasta, Cal. 
 
 4. San Francisco Mountain, Ariz. 
 
 5. Crater Lake, Ore. 
 
 Note, If models of any of these mountains are available, study them 
 in connection with the maps. Photographs of the mountains appear in 
 some of the works of the reference list, and these may well be studied 
 in connection with the maps. 
 
 For Crater Lake and its surroundings, see reference 10, and for 
 Mt. Shasta, reference 7, at end of chapter. 
 
 II. 1. Make a special study of the topography of the volcanic mountains 
 to determine how far their present configuration is the result 
 of the original cone-building, and how far the result of subse- 
 quent erosion. 
 2. Are craters in original or modified form shown on any of the maps? 
 
 REFERENCES 
 
 1. Standard text-books on Geology. 
 
 2. RUSSELL, Volcanoes 0} Xorth America: Macmillan. 
 
 3. JUDD, Volcanoes: Appleton. 
 
 4. BONNET, Volcanoes: G. P. Putnam's Sons. 
 
 5. Recent Eruptions in the West Indies. HEJLPRIN, Mt. PeUe and the 
 Tragedy of Martinique, Lippincott; HILL, Nat. Geog. Mag., Vol. XIII, pp. 
 223-267; RUSSELL, Nat. Geog. Mag., Vol. XIII, pp. 267-285, 415-435; 
 HOVEY, Nat. Geog. Mag., Vol. XIII, pp. 444-459, Am. Jour. Sci., Vol. XIV, 
 1902, pp. 319-358, and Bull. Am. Mu?. Nat. Hist., Vol. XVI, pp. 333-372. 
 
 6. Rept. of the Roy. Soc. on The Eruption of Krakatoa: Tiiibner & Co. 
 
 7. DILLER, Mt. Shasta, in Physiography of the United States: Am. Book Co. 
 
VULCAN ISM 
 
 391 
 
 8. BUTTON, Hawaiian Volcanoes: 4th Ann. Kept. U. S. Geol. Surv. ; 
 and Mt. Taylor and the Zuni Plateau: 6th Ann. Rept. U. S. Geol. Surv. 
 
 9. GILBERT, Geology of the Henry Mountains: U. S. Geol. Surv. 
 
 10. DILLER and PATTON, Geology and Petrography of Crater Lake Park, 
 Part I: Professional Paper 3, U. S. Geol. Surv. 
 
 FIG. 417. Mt. Wrangell, Alaska. (U. S. Geol. Surv.) 
 
CHAPTER VIII 
 CRUSTAL MOVEMENTS. DIASTROPHISM 
 
 . SECULAR CHANGES 
 
 IN many places the coastal lands appear to have recently 
 risen from the sea, while in others coastal tracts appear to have 
 been recently submerged. The most obvious evidence of the 
 apparent rise is found in the beaches and other shore features 
 now well above sea-level, and the most obvious evidence of 
 sinking is found in the drowned lower ends of rivers (p. 173). 
 These relative changes of level of the land are best seen along the 
 sea-shore, because the sea-level is the place from which land ele- 
 vations are measured. Movements of the outer portions of the 
 solid part of the earth are also in progress, or have recently taken 
 place, far from coasts, but they are not so readily detected, and 
 are therefore less well known. Some of them have been referred 
 to incidentally in other connections (p. 174). The changes of level 
 are in general so slow that no motion is seen from day to day, or 
 even from year to year. All that is seen is the result of changes 
 which have been going on slowly for centuries. 
 
 Movements of the earth's crust (outer parts of the lithosphere) 
 were first inferred from various phenomena along the shores. 
 They were subsequently demonstrated by careful measurements 
 which have shown not only the fact of movement, but in some 
 cases its rate. 
 
 It is to be especially noted that beaches or other shore marks 
 above the present level of the sea do not necessarily mean that the 
 land has risen. They might mean depression of the sea-level 
 instead, but in either case they mean increased emergence of the 
 land. Similarly, the lower ends of valleys would be drowned by 
 the rise of the sea just as effectively as by the sinking of the land; 
 but in either case there has been a depression of the land relative 
 
 392 
 
CRUSTAL MOVEMENTS. DIASTROPHISM 393 
 
 to sea-level. In some situations and relations it may be possible 
 to say whether it is the land or the sea surface which has changed 
 its position; but in general it is best to think of the changes as rela- 
 tive. 
 
 Evidences of (Relative) Elevation of Land 
 
 1. Human structures. In certain regions which have been long 
 inhabited, structures which were once at sea-level are now above it. 
 Thus on the island of Crete, in the Mediterranean Sea, the remains 
 of old docks are in some places as much as 27 feet above the water. 
 This is the more extraordinary since other parts of the same 
 island have sunk so as to submerge human structures, the ruins 
 of which are still visible beneath the water. 
 
 2. Rocks. Several rocks in the Baltic Sea which within his- 
 toric time were at sea-level, or so little below it as to be dangerous 
 to navigators, are now well above the water. One is said to have 
 risen about 60 feet in 50 years. From an inscription supposed to 
 be about 500 years old in Lake Maelar (near the Baltic), the land 
 at that point is inferred to have risen about 13 feet in the 50 years 
 preceding 1730. 
 
 3. Measurements. Changes of level were recognized long ago 
 in Scandinavia, and they excited so much interest that marks 
 were made on the shores at different places and kept under obser- 
 vation for a period of years in order to determine the rate of change 
 of level between land and water. These observations showed 
 that the coast was gradually becoming higher relative to the 
 Baltic. 
 
 In recent times it has been found that the larger part of the 
 coast of Scandinavia is rising relative to sea-level, but that it is 
 rising at unequal rates, and that the southern extremity is sink- 
 ing. In some localities the rate of rise has been determined to be 
 about 2 \ feet per century. 
 
 4. Organic remains ; fossils. Another line of evidence pointing 
 to the rise of coasts is found in the shells, tests, etc., of marine animals 
 found above sea-level. Thus barnacle shells are sometimes found 
 above the surface of the water, attached to the rocks where they 
 grew. There is no escape from the conclusion that the sea-level 
 has sunk, or the land risen, to the extent of their height above sea- 
 level. Beds of marine shells accumulated beneath the sea are 
 also sometimes found above sea-level. Such shells are conclusive 
 
394 PHYSIOGRAPHY 
 
 of the relative rise of the land, in case they are known to have been 
 deposited by the sea-water. The evidence from unattached shells 
 must, however, be carefully scrutinized, since birds and other 
 animals frequently carry marine shells inland and up to con- 
 siderable heights. 
 
 Beds containing sea-shells which were certainly deposited be- 
 neath the sea are now found above the water at various points, as 
 along the coast of Sweden, and at some points in North Greenland, 
 where they occur up to heights of 100 to 200 feet. The shells 
 here are so fresh that in some cases they are still covered with the 
 epidermis. The sand in which they are imbedded is frozen during 
 a large part or all of the year, and the low temperature undoubtedly 
 keeps the organic matter from decay. Darwin long ago found 
 shells, etc., along the west coast of South America up to eleva- 
 tions of 1300 feet above sea-level. On the coast of Peru a reef 
 of corals of modern species is said to be found at an elevation of 
 nearly 3000 feet. On the coast of the New Hebrides, coral reefs 
 occur up to 2000 feet, and on the coast of Cuba up to heights of 
 1000 feet or more. 
 
 5. Raised beaches, etc. Raised beaches and terraces along the 
 shore are also evidences of change of level. Such features are 
 found along many parts of the northern coast of western Europe 
 and eastern North America, about the West Indies, on the Cali- 
 fornia coast, and in many other places. Along certain coasts, for 
 example those of California and Scotland, towns are situated on 
 these elevated terraces, and w r agon-roads and railroads follow them 
 for considerable distances. 
 
 One of the significant facts concerning the elevated beaches and 
 other shore features is that they are no longer horizontal. The 
 island of Crete, already cited, affords an illustration, and the coast 
 of Scandinavia another. 
 
 6. Sea cliffs. Sea cliffs (Fig. 418) developed by wave-cutting 
 are sometimes found above the elevated shore terraces. 
 
 7. Sea caves. Waves sometimes develop sea caves at water 
 level. Caves developed in this way are now sometimes found at 
 levels considerably above the sea. Such caves occur on the coast 
 of Scotland up to levels of 100 feet. 
 
 All these phenomena are evidences that the land has risen 
 relative to the sea, in many places, in recent times. 
 
PLATE XXIII 
 
 A section of the California coast, showing lands, near the coast, which have 
 recently emerged. Scale 1 mile per inch (Oceanside, Cal., Sheet, U. 8. 
 Geol. biirv.) 
 
PLATE XXIV 
 
 Cushetunk and Round Mountains, New Jersey ; examples of isolated moun- 
 tains left by the removal of less resistant surroundings. Scale 1 mile 
 per inch. (High Bridge Sheet, U. S. Geol. Surv.) 
 
CRUSTAL MOVEMENTS. DIASTROPHISM 395 
 
 Evidences of Relative Depression 
 
 Evidences of the sinking of land are, in the nature of the case, 
 less readily seen, because they are for the most part beneath the 
 water. 
 
 1. Human structures. It has already been noted that at the 
 east end of the island of Crete ancient buildings are submerged. 
 
 FIG. 418. Wave-cut terraces. West side of Ojai Valley, Cal. (Arnold.) 
 
 Certain portions of the coast of Greenland have likewise been sink- 
 ing recently, for various human structures on low coasts have 
 sunk, and still stand beneath the water. The southern end of 
 
 FIG. 419. Wave-cut terraces. Bottle and glass. St. Vincent. 
 (Hovey, Am. Mus. Nat. Hist.) 
 
 Scandinavia, as already noted, has been sinking recently, while 
 the rest of the peninsula appears to have been rising. "At Malmo 
 one of the present streets is over-flooded by the waters of the Baltic 
 when the wind is high, and excavations made some years ago dis- 
 closed an ancient street at a depth of eight feet below the present 
 one." 
 
396 
 
 PHYSIOGRAPHY 
 
 2. Submerged forests. Along some coasts there are evidences of 
 submerged forests. This is the case, for example, at some points 
 north of Liverpool, England (Fig. 420). Here, when the tide is 
 
 FIG. 420. Stumps laid bare on the beach at low tide. Lease we, Cheshire. 
 
 Eng. (Ward.) 
 
 out, numerous stumps may be seen standing on the beach where the 
 trees once grew. Since trees of the varieties represented by these 
 stumps do not grow in salt water, there is no alternative but to 
 conclude that the land where they grew, once dry, has since sunk 
 
 FIG. 421. Masses of peat and tide marsh sod, broken up by storm waves 
 and washed ashore near Sea Isle City, N. J. (Knapp, N. J. Geol. Surv.) 
 
 below the level of high water. On the coast of New Jersey 
 stumps have been found seven feet below sea-level at low tide. 
 Old marsh-lands are submerged beneath the sea along some 
 shores. From them the strong waves of storms sometimes 
 wrench large masses of peat and toss them up on the shore. 
 
CRUSTAL MOVEMENTS. DIASTROPHISM 
 
 397 
 
 This was the case at Cape May in New Jersey in a violent storm 
 but a few years ago (Fig. 421). 
 
 3. Submerged valleys. Some river valleys on land are con- 
 tinuous with valleys in the shallow sea bottom far out beyond the 
 coast-line (Fig. 422). Such submerged valleys indicate that the 
 surface in which they are cut was land when they were excavated, 
 and that subsequent sinking has submerged them. The numerous 
 bays along the eastern coast of the United States, especially between 
 New York and the Carolinas, indicate recent sinking of the land, 
 
 FIG. 422. The submerged valley which is believed to be the continuation 
 of the Hudson Valjey. The position of the valley is indicated by the 
 contours. Depths in fathoms. (Data from C. and G. Survey.) 
 
 enough to carry the lower ends of the former valleys below sea- 
 level, thus converting them into bays. Submerged valleys of this 
 sort occur in many parts of the earth, and show that submergence 
 of coastal lands is now taking place, or has recently taken place, 
 along many coasts. 
 
 4. An Italian temple. One of the most striking cases of change 
 of level appears to involve both upward and downward move- 
 ment. On the shore of Italy, near Naples, are the ruins of an old 
 temple. From inscriptions it is known to have been above water 
 as late as 235 A.D. In 1749 several columns of the temple were 
 
398 PHYSIOGRAPHY 
 
 found still standing. Their bases were buried to a depth of 12 feet 
 in sediment deposited by the sea. For 9 feet above the sediment, 
 the columns were filled with holes bored by marine animals. It 
 is inferred, therefore, that between the years 235 and 1749, the land 
 on which the temple stood sank until the water stood 21 feet above 
 the bases of the columns, and then rose again so that its floor was 
 above sea-level. 
 
 7s it the Land or the Sea which Changes its Level f 
 
 It is clear that if the outside of the lithosphere, commonly called 
 the crust of the earth, is subject to warping, some parts rising and 
 some sinking, the observed phenomena of coasts could be readily 
 explained; and, in most discussions of the changes of relative level 
 of land and sea, it is commonly assumed that the land, rather than 
 the sea, is the element which changes. The validity of this assump- 
 tion has, however, been questioned, and with reason. The alter- 
 natives are several, namely: (1) Is it the sea rather than the land 
 which changes its level? Or (2) do both land and sea change their 
 levels? In the latter case (a) does each rise and fall, or (6) does 
 one rise and fall while the other rises or falls only? Certain gen- 
 eral considerations will make it clear that some of these alternatives 
 are not tenable. 
 
 Other conditions remaining constant, the sea-level would be de- 
 pressed everywhere if depressed at one point, because all oceans 
 are connected. In this case, all coasts would appear to rise. Simi- 
 larly, if the sea-level rose at one point, it would seem that it should 
 rise everywhere, in which case all coasts should appear to be sink- 
 ing at the same time. Since some coasts seem to be rising while 
 others seem to be sinking, it is clear that the changes of level of 
 the sea surface, taken alone, do not explain the observed phe- 
 nomena. It does not follow, however, that such changes may not 
 be one of the elements involved in the explanation of the observed 
 phenomena. 
 
 Without discussing all the other alternatives separately, the 
 principles involved in the problem may be readily understood. 
 
 Let us suppose the sea-level to be lowered at all points, as it 
 would be by the sinking of the bottom of any one of the ocean 
 basins, and let us suppose further that the continents sink at the 
 same time. Under this general conception various cases arise. 
 
CRUSTAL MOVEMENTS. DIASTROPHISM 
 
 399 
 
 (1) If the lands are lowered as much as the sea-level, and lowered 
 equally everywhere, the relation of coasts to sea-level would not 
 be changed. (2) If the land were lowered more than the sea, and 
 lowered equally everywhere, the old coast-lines would all be sub- 
 merged. (3) If the land were lowered less than the sea, and 
 lowered equally everywhere, all coasts would appear to have risen. 
 (4) If different parts of the coasts sank unequally, those parts 
 which sank less than the sea would appear to have risen, those 
 which sank as much as the sea would not have changed their 
 position relative to sea-level, and those which sank more than the 
 sea would appear to have sunk. 
 
 All these relations are illustrated by Figs. 423 to 425. The 
 sea-level, which in Fig. 423 is at AB, is represented in Fig. 424 as 
 
 FIG. 423. Diagram showing a coastal tract rising above sea-level. 
 
 having sunk from A' to B f . In Fig. 425 the sea-level of Fig. 423 is 
 represented as having sunk, but the land of the old coast-line has 
 sunk more than the sea from A" to C, as much as the sea at C, and 
 
 FIG. 424. Diagram showing the same coast as Fig. 423, after the sea-level 
 has been lowered uniformly. The land appears to have risen. 
 
 less than the sea from C to B". From A" to C the coast seems to 
 have sunk, from C to B" it seems to have risen. 
 
 It appears, therefore, that all apparent risings and sinkings 
 along coasts might be explained by unequal sinking of land while 
 the sea-level is being lowered; but it does not follow that this is 
 necessarily the true explanation. Existing phenomena about 
 coasts may be explained on the supposition that coastal lands 
 rise locally and sink locally, as well as on the supposition that they 
 sink only. They may be equally well explained on the supposi- 
 tion that the sea-level sometimes rises and sometimes sinks, at the 
 same time that the coasts are being warped, in some places up and 
 in some places down. On the whole, it seems probable that the 
 
400 
 
 PHYSIOGRAPHY 
 
 sea-level does change, sometimes rising and sometimes sinking, 
 and that coastal lands are warped upward in some places and 
 downward in others, and that observed phenomena involve all 
 these movements. 
 
 Current theories of the origin and history of the earth all pro- 
 ceed on the assumption that the earth is a cooling and therefore a 
 
 
 
 FIG. 425.- Diagram showing the same area as the preceding. The sea has 
 sunk as much as in Fig. 424, but the land at the left has also sunk, and 
 has sunk more than the sea-level has. At the left, therefore, the land 
 seems to have sunk and at the right it seems to have risen, while at one 
 point, C, it appears to have neither risen nor sunk. 
 
 contracting body. If this be true, it is clear that depressions rather 
 than elevations of the surface must be the rule, and that such 
 elevations of the surface of the lithosphere as take place are in- 
 cidental to the general lowering of surface which results from 
 contraction. 
 
 We may now inquire into the causes which make, or which may 
 make, the surface of the sea rise and fall, and also into the causes 
 which make, or may make, the surface of the lithosphere rise and 
 fall locally. 
 
 Why the Sea-level Changes 
 
 Sedimentation. Rivers carry a large amount of sediment 
 from the land to the sea each year (p. 154). This material, de- 
 posited on the sea bottom, tends to fill the ocean basins. The 
 result must be rise of the sea-level. The detritus worn from the 
 shores by waves, blown from the land by winds, and brought to 
 the sea by glaciers, is likewise deposited beneath the water and 
 produces the same result. The material extracted from the sea- 
 water by plants and animals, and deposited on its bottom when 
 they die, likewise helps to raise the surface of the sea, for the space 
 occupied by shells, etc., exceeds the reduction of volume suffered 
 by the water when the material of the shells was extracted from 
 it. Rise of the surface of the sea due to sedimentation would be 
 universal. 
 
CRUSTAL MOVEMENTS. DIASTROPHISM 401 
 
 The rise of the sea due to sedimentation is extremely slow, too 
 slow to be obvious from year to year, or perhaps even in a lifetime. 
 But if existing lands were base-leveled, the resulting rise of sea- 
 level would be hundreds of feet, enough to submerge a very con- 
 siderable part of the existing land, and a much larger part of the 
 land as it would be after base-leveling. There is reason to be- 
 lieve that great areas have been nearly base-leveled in the past. 
 Degradation of the land and aggradation of the sea bottom may, 
 therefore, have been very important factors in the repeated sub- 
 mergence of great areas of land in past ages. 
 
 It is presumed that evaporation from the sea is balanced by 
 rainfall and by the inflow of rivers. If this is the case, evapora- 
 tion and precipitation do not affect its level. But if great quanti- 
 ties of water evaporated from the sea were to be retained on the 
 land in the form of ice, as it was in the glacial period, the result 
 would be a lowering of the sea-level. The melting of the ice, on 
 the other hand, would cause the sea-level to rise. The ice of the 
 glacial period was of such quantity that its withdrawal, on the one 
 hand, and its return, on the other, would have influenced the 
 surface of the sea sensibly. 
 
 Submarine volcanic extrusions. Extrusions of lava beneath 
 the sea likewise cause its surface to rise, as would also laccolithic 
 and other intrusions beneath its bottom. 
 
 Diastrophism. While sedimentation and vulcanism have 
 surely caused changes in the level of the sea, it is not believed that 
 they have been the only causes which have produced such changes; 
 neither is it believed that, in the long course of time, they have 
 been the causes of the most profound changes. 
 
 The progressive cooling of the earth has resulted in its pro- 
 gressive shrinking since the time when it attained its growth and 
 its maximum heat. This shrinking must have resulted in the 
 deformation of its outer part, for the same reason that the skin 
 of an apple wrinkles when its juice evaporates. As the cooling 
 and contraction are constant, it appears that slow warpings of the 
 crust may also be constant; but it appears also that the rigidity 
 of the earth may be such that its outer parts are able to withstand 
 for a time the strain set up by contraction. As the strain accumu- 
 lates, it ultimately overcomes the resistance, and the outer part 
 of the earth yields. If the yielding results in the sinking of the 
 ocean basin, the surface of the water is drawn down, and the sur- 
 
402 PHYSIOGRAPHY 
 
 rounding lands seem to rise, unless they sink as much as the surface 
 of the sea does at the same time. The lowering of the sea surface, 
 because of the sinking of the sea bottom, is probably the most 
 fundamental single cause of the apparent rise of land. 
 
 The periodic emergences of the continents, alternating with 
 periodic submergences in the course of geological history, are 
 perhaps to be thus explained. Periodic submergences, on the 
 other hand, might be explained by the sinking of the continental 
 segments of the earth, or by such sinking combined with the proc- 
 esses already referred to which cause the rise of the sea. 
 
 Why the Land Changes Level 
 
 The reasons assigned for changes of level of the lithosphere 
 beneath the sea have equal force when applied to the land. It 
 is probable that the continents sink more in the course of ages than 
 they rise, but that they sink less than the ocean basins. A local rise 
 of the surface may result from a more general sinking. A de- 
 pression of ocean basins, for example, may crowd up the conti- 
 nental area between, and the same principle may be applicable to 
 smaller areas. Again, in volcanic regions, the intrusion of lava 
 may cause the surface to rise, as over a laccolith, and the bringing 
 of the hot-rock material near to the surface heats the surface rocks 
 and perhaps expands them enough to cause sensible rise of the 
 surface. 
 
 We are to conclude, therefore, that the apparent rise of the 
 land along coasts is probably due in part to the sinking of the sea, 
 in part to the lesser sinking of the coastal lands, as compared with 
 the sea, and in part to the actual rise of the land itself. 
 
 Changes of Level in the Interiors of Continents 
 
 General facts. Changes of level are perhaps as common in 
 the interiors of continents as along coasts, though less easily de- 
 tected. There are raised beaches about many lakes, as about the 
 Great Lakes and Great Salt Lake (Fig. 361). Raised beaches 
 about lakes result from the lowering of the level of the lakes, either 
 by the cutting down of their outlets or by evaporation. They 
 do not, therefore, indicate a rise of the land. In both cases cited, 
 however, the old shore-lines should remain horizontal. But about, 
 many lakes the old shore-lines are not level, as they must have 
 
CRUSTAL MOVEMENTS. DIASTROPHISM 403 
 
 been when formed. Some parts of one of the old shore-lines about 
 Lake Bonne ville (p. 314) are 300 feet higher than other parts 
 formed at the same time. An old shore-line about the east end of 
 Lake Ontario is more than 400 feet above the lake, while the same 
 shore-line traced westward passes beneath the water at the west 
 end of the lake. Similar phenomena are found about the shores of 
 all the Great Lakes, though the departure from horizontality is 
 not everywhere so great. Such deformed shore-lines show that the 
 surface about the lake basins has warped since the old shore-lines 
 were formed. 
 
 The former shore-lines of many smaller lakes and of some ex- 
 tinct lakes are also well known, and they tell the same story. 
 This is notably the case with the shore-lines of the extinct Lake 
 Agassiz (p. 282). 
 
 Changes of level are still in progress. The accurate observations 
 and measurements of recent years have shown that the area of the 
 Great Lakes is being tilted upward, relatively, to the northeast, and 
 downward, relatively, to the southwest. The rate has been shown 
 to be less than six inches per hundred miles per century. 
 
 Extent. So wide-spread are the evidences of changes of level 
 that it may be said that, within regions so situated as to furnish 
 evidence, more of the earth's surface has been sinking or rising in 
 recent times, than has been standing still. 
 
 This general statement seems to point to great instability of 
 the earth's crust, but it should be supplemented by the statement 
 that these changes go on, as a rule, with extreme slowness and, 
 in general, probably without violence. The amount of move- 
 ment is, perhaps, to be reckoned in small fractions of an inch per 
 year, more commonly than in larger terms. At times and places, 
 however, it is probable the movements have been more rapid, but 
 even in these cases it is not to be supposed that the movements were 
 always violent. 
 
 The instability of the earth's exterior is believed to indicate that 
 it is not in perfect adjustment to the interior, and that the con- 
 tinued lack of adjustment is the result of the continued shrinking of 
 the earth as a whole. 
 
 Ancient changes of level. Old shore-lines and all features 
 connected with ocean shores are destroyed in time by erosion; but 
 the evidence of movements which took place so long ago that no 
 traces of shore-lines remain, is firm. Thus layers of rock which 
 
404 PHYSIOGRAPHY 
 
 were once deposited as sediment (sand, mud, etc.) beneath the 
 sea are now found over great areas, far above sea-level. Most of 
 the solid rock beneath the Mississippi basin, for example, was de- 
 posited as sediment beneath the sea. The land has emerged, 
 perhaps because the sea-level has been drawn down by the sink- 
 ing of the sea basin. In the Appalachian Mountains, rocks simi- 
 larly formed are found up to altitudes of a few thousand feet. In 
 the Rocky Mountains they occur as high as 10,000 feet or even 
 more. In the Andes Mountains they are found in limited areas 
 up to 16,000 feet or more, and in the Himalaya Mountains at 
 still greater heights. In these extreme cases, at least, it seems 
 probable that there has been an actual rise of the crust, that is, 
 that the outside has been bent up, or thrust up locally, so as to 
 be farther from the centre of the earth than when in its former 
 position. 
 
 Future changes of level. Not only have changes of level 
 between land and sea been taking place for untold ages, but they 
 are likely to continue. The wear of the land and the transfer of 
 sediment to the sea tends to raise the sea-level, as already noted 
 (p. 400). This tends to increase the area of the sea and to corre- 
 spondingly diminish the area of the land. In the past there seem 
 to have been occasional sinkings of the ocean basins, increasing 
 their capacity and drawing down the level of the sea, thus caus- 
 ing the continents, as a whole, to appear to rise, and such changes 
 are likely to occur in the future, so far as can now be seen. On 
 the average, the lowering of the sea-level, due to the subsidence 
 of its basins, has probably been greater than the rise of the sea- 
 level, due to sedimentation from the land. The result is that as 
 the continents have been brought low by wind and water and 
 ice, they have been renewed, periodically, by the sinking of the 
 sea. 
 
 In this general sequence of events appears to lie the explana- 
 tion of the fact that though rain and river and ice erosion tend to 
 bring the land to base-level, and though wave erosion tends to 
 reduce it even below sea-level, the land is not destroyed, and is not 
 even completely reduced to base-level. 
 
 As the great ocean-basin segments of the earth settle down, 
 it seems possible that the smaller continental segments between 
 may sometimes be wedged up, and perhaps warped and deformed 
 at the same time. In this process may lie the explanation of many 
 
CRUSTAL MOVEMENTS. DIASTROPHISM 
 
 405 
 
 mountains, plateaus, and plains, the physiographic features of the 
 second order. 
 
 Crustal Deformation 
 
 The foregoing discussion of changes of level has implied a 
 measure of deformation of the outside of the solid part of the earth. 
 
 FIG. 426. Open anticlinal fold, near Hancock, Md. (U. S. Geol. Surv.) 
 
 This deformation sometimes takes the form (1) of gentle warping, 
 sometimes (2) of folding, and sometimes shows itself (3) in faulting. 
 
 Warping and folding. The warping may be gentle, resulting in 
 slight arching or tilting of the beds of rock, or it may be so great 
 that the arches grade into folds (Figs. 426 and 427). Most rock 
 strata of the land are at least gently deformed, and folding is com- 
 mon in many mountain regions, and in some plains which were 
 once high, but which have been brought low by erosion. 
 
 Warping and folding give rise to great topographic features, 
 but in most mountains of folded rock, the present topography is 
 the result of erosion rather than of the original folding. The 
 structure of the rock resulting from the folding has, in many 
 cases, determined or greatly influenced the topography which has 
 resulted from erosion. 
 
406 
 
 PHYSIOGRAPHY 
 
 Faulting. At many times and in many places portions of 
 the earth's surface have sunk or risen along a plane of fracture, 
 as shown by Figs. 429 and 430. Such movements are faults. Fig. 
 428 represents a normal or gravity fault, and Fig. 429 a reversed 
 or thrust fault. The former implies tension when it was made, 
 and the adjustment takes place under the control of gravity. 
 The latter implies lateral thrust, and the adjustment takes place 
 under the control of such pressure. 
 
 Faults of both types are common, but reversed faults are 
 
 FIG. 427. Closed anticlinal fold, near Levis Station, Quebec. 
 (U. S. Geol. Surv.) 
 
 common only in regions where the rock strata have been folded. 
 Fig. 429 suggests the relation between such a fault and a fold. 
 A fold which is not faulted sometimes passes into a reversed or 
 thrust fault. Normal faults may also grade into folds, especially 
 monodinal folds (Fig. 430). Faulted blocks of the earth's crust are 
 sometimes tilted. They may be of such size and so displaced as 
 to give rise to mountains, basins, etc. (Figs. 331 and 433). Nor- 
 mal faulting has taken place on a great scale in the plateau region 
 of the West, between the Rocky Mountains and the Sierras, and 
 many of the more striking topographic features of that region, 
 
CRUSTAL MOVEMENTS. DIASTROPHISM 
 
 407 
 
 including numerous mountain ranges and lines of cliffs (Fig. 433), 
 are the result of such movements. Cliffs or mountain slopes due 
 to faulting are called fault-scarps. The steep slopes at the borders 
 of the continental segments are probably fault-scarps in some 
 
 FIG. 428. Diagram of a 
 normal fault. 
 
 FIG. 429. Diagram of a reversed or 
 thrust fault. (U. S. Geol. Surv.) 
 
 cases (p. 13). Great faults are probably not developed by a single 
 movement, but by repeated displacements along the same plane. 
 
 Though faults are common phenomena, only those of recent 
 times now show themselves in the topography of the surface. 
 
 FIG. 430. Fault passing into a 
 monoclinal fold. 
 
 FIG. 431. A branching fault. (Powell, 
 U. S. Geol. Surv.) 
 
 Fault-scarps of earlier ages have been obliterated by erosion, 
 though the faults still show themselves in the structure of the rock. 
 Faults have determined the positions of valleys in many cases, 
 so that present topographic features are often closely associated 
 with them, even though the fault-scarps have vanished. 
 
408 
 
 PHYSIOGRAPHY 
 
 Earthquakes 
 
 Definition. Earthquakes are tremors or quakings of the 
 earth's surface, due to causes which cannot be referred to human 
 agencies. The passing of a railway-train causes the surface near 
 the track to vibrate, and this vibration is often enough to be felt 
 in adjacent buildings. In this case, the cause is readily under- 
 stood, and the shaking is not called an earthquake; but an equal 
 
 FIGS. 432^434. Diagrams to illustrate the history of a fault-scarp. Fig. 432 
 shows an unfaulted block with a line of cliffs due to the superior hard- 
 ness of one formation. Fig. 433 shows the same faulted, and with a 
 pronounced fault-scarp. Fig. 434 shows the fault-scarp partly worn 
 down. (Huntington and Goldthwait.) 
 
 amount of quaking, due to causes which were not known, would 
 be called an earthquake, especially if felt over a considerable area. 
 
 Strength and destructiveness. Earthquakes vary much in 
 strength. Some are so gentle that they can barely be felt; others 
 are so violent that buildings are overthrown, crevasses opened in 
 the surface of the land, and masses of rock loosened from cliffs 
 and precipitated into the valleys below. Earthquakes some- 
 times disturb the waters of the seas, causing destructive sea waves. 
 
 Besides quakings which are sensible, there are many earth 
 
CRUSTAL MOVEMENTS. DIASTROPHISM 409 
 
 FIG. 435. Monument 
 disturbed by earth- 
 quake. (Falb.) 
 
 FIG. 436. A chapel in Kasina injured in 
 an earthquake of November 9, 1880. 
 (Wahner.) 
 
 FIG. 437. Horizontal and vertical displacement during an earthquake. 
 Bengal-Assam earthquake of July 12, 1897. (From Button's Earth- 
 quakes, by permission of G. P. Putnam's Sons.) 
 
410 
 
 PHYSIOGRAPHY 
 
 tremors so slight as not to be felt. They are known only by means 
 of delicate instruments set up for the purpose of recording all 
 
 FIG. 438. Craterlets observed after the Calabrian earthquake of 1783. 
 
 (Sieberg.) 
 
 vibrations of the surface. The - number of such tremors too 
 slight to be generally noticed is much greater than the number of 
 earthquakes strong enough to be felt. 
 
 Although violent earthquakes are sometimes very destructive 
 
 FIG. 439. Great sea-wave on the coast of Ceylon. (Sieberg.) 
 
 to buildings and to life, the amount of movement of the surface is 
 usually so slight as to be measured in millimetres (a millimetre is 
 
CRUSTAL MOVEMENTS. DIASTROPHISM 
 
 411 
 
 about -2*5 of an inch) rather than in inches or feet. Bodies on the 
 surface often move much more than the solid crust. The relation 
 
 Phase of Vibrations 
 
 Scale of Minutes 
 
 FIG. 440. Seismograph of earthquake in Punjab, India, April 4, 1905, show- 
 ing the actual amount of movement. (De Montessus de Ballore.) 
 
 is illustrated by the fact that a blow on the floor may cause a ball 
 which rests upon it to bound up inches or even feet, though the 
 floor itself moves but a small fraction of an inch. 
 
 While earthquakes are among the most disastrous and appal 
 
 FIG. 441. The bending of railway track in India, earthquake of 1897. 
 
 (Oldham.) 
 
 ling of natural phenomena, so far as human affairs are concerned, 
 those of historic times, at least, have left few important marks on 
 
412 
 
 PHYSIOGRAPHY 
 
 the surface of the earth. Their destructiveness to human life 
 comes largely from the fall of buildings and from the "great sea 
 waves " caused by the earthquakes. Destruction of life results 
 from the advance of these waves upon a low coast which is densely 
 populated. In the Lisbon earthquake of 1755 a wave 60 feet high 
 swept upon the shore and destroyed some 60,000 human lives. 
 
 FIG. 442. Fault in Japan, 1891. (Koto.) 
 
 Vessels in harbors have been swept in by waves and left high and 
 dry above the water-level. 
 
 Examples. Some of the principal features of earthquakes 
 may be brought out by the study of a few striking examples. 
 
 1. On October 28, 1891, an earthquake on the main island 
 of Japan opened a fissure traceable for over 40 miles. The ground 
 on one side of this fissure sank 2 to 20 feet (a fault) below that 
 on the other. At the same time, the east wall of the fissure was 
 pushed horizontally about 13 feet northward. In some places the 
 cracking of the rock "showed itself at the surface as a cracked 
 ridge, like the track of a mole just below the surface." Several 
 tracts of land became lakes, one on the depressed side of the 
 
CRUSTAL MOVEMENTS. DIASTROPHISM 
 
 413 
 
 fissure, the others in hollows formed by the shocks. Fig. 443 
 shows the distribution of earthquakes in Japan from 1883 to 1902. 
 2. On the evening of August 31, 1886, the city of Charleston, 
 South Carolina, was disturbed by an earthquake which was 
 felt over a large part of the United States. Strange noises and 
 
 FIG 443. Distribution of earthquakes in Japan, 1885-1892.^ (Davison. 
 From Button's Earthquakes, by permission of G. P. Putnam s Sons.) 
 
 slight tremblings of the earth had been noted for several days 
 previous to the destructive quaking, but they excited no great 
 alarm. About ten o'clock in the evening of the fateful day, a 
 low rumbling sound was heard, which rapidly deepened into an 
 awful roar. The slight trembling of the ground increased until 
 it became destructively violent. The motion then subsided 
 
414 
 
 PHYSIOGRAPHY 
 
 slightly, but increased again in intensity and then died away. The 
 violent disturbance lasted 70 seconds. A second shock, almost 
 as severe as the first, occurred eight minutes later, and six or seven 
 more or less severe ones were felt before morning, and slight trem- 
 ors occurred at intervals until the following April. During the 
 shocks, buildings swayed, chimneys were thrown down, walls 
 
 105 100 95 90 85 80 75 70 J 65 s 
 
 FIG. 444. Isoseismal curves (that is, curves connecting points having the 
 same amount of disturbance) of the Charleston earthquake. (Button 
 U. S. Geol. Surv.) 
 
 were cracked, houses moved from their foundations, railroad-tracks 
 displaced sidewise and the rails bent, and trees disturbed in the 
 ground. Numerous fissures were formed in the earth, and from 
 some of them streams of water, mud, and sand were forced out. 
 Hardly a large building in the entire city but was damaged, and 
 27 persons were killed, chiefly by falling masonry. The people 
 
CRUSTAL MOVEMENTS. DIASTROPHISM 
 
 415 
 
 fled in terror from their homes, and for several days and nights a 
 large part of the population camped in the public parks. 
 
 Outside the vicinity of Charleston the earthquake shock was 
 less violent, but the quaking was felt over an area of between 
 
 Tli* two epicentral tract* are indicated by arbitrar t 
 iwseigmal curvet; tint heavy line being tfa index. 
 
 SCALE OF MILES . 
 
 012 i 6 8 10 
 
 FIG. 445. Epicentral tracts (i.e., tracts over the centers of disturbance) of the 
 Charleston earthquake, with isoseismal lines. (Button, U. S. Geol. burv.J 
 
 2,000,000 and 3,000,000 square miles. It was felt earliest near 
 Charleston, and later at increasing distances from the city (Fig. 
 444). There were two centers of disturbance (Fig. 445), and 
 the earthquake wave spread at the rate of about 150 miles per 
 minute. 
 
416 PHYSIOGRAPHY 
 
 There was no volcano near Charleston, and this earthquake 
 appears to have been altogether independent of any volcanic 
 eruption. 
 
 3. In 1822 and again in 1835 the coast of Chile was shaken by 
 earthquakes for 1200 miles. In both years the shocks continued 
 for several months. When they were over, it was found that 
 the coast-lands had been elevated 2 to 4 feet. 1 In 1835 a volcano 
 
 FIG. 446. Wreck of the Charleston earthquake. (U. S. Geol. Surv.) 
 
 broke out beneath the sea (San Fernandez Island)- at the time of 
 the earthquake shocks on the coast, and many of the Andean vol- 
 canoes were active. 
 
 The same region was again shaken disastrously in August, 1906, 
 causing great destruction of life and property in some of the 
 principal cities cf the country. 
 
 4. In 1819 a part of the delta of the Indus River experienced a 
 series of shocks lasting four days. During the disturbance an 
 area some 2000 square miles in extent subsided so as to be covered 
 by the sea, while a neighboring belt, 50- miles long and 16 miles 
 wide, rose about 10 feet. 
 
 5. The earthquake of Kangra, in the same country, April, 1905, 
 affected an area of 1,625,000 square miles, and killed about 20,000 
 people. In this case, the vibrations spread from two centers, and 
 
 1 This statement has been disputed, and the records which bear on the 
 point seem to be less perfect than could be desired. 
 
CRUST AL MOVEMENTS. DIASTROPHISM 417 
 
 traveled at the rate of about two miles per second. The same 
 region had been shaken in 1897. 
 
 FIG. 447. A large craterlet formed during the Charleston earthquake. Hun- 
 dreds of them were formed near Summerville, S. C. (U. S. Geol. Surv.) 
 
 FIG 448. Sand-cones and craterlets observed after the Achaischen earth- 
 quake of 1861. (Schmidt.) 
 
 6. A series of earthquake shocks, lasting from 1811 to 1813, af- 
 fected the Mississippi Valley just below the mouth of the Ohio. 
 
418 PHYSIOGRAPHY 
 
 Many fissures were formed in the deposits of the flood plain of 
 the Mississippi, and some of them remained open for years. Parts 
 of the flood plain sank, and the sunken portions gave rise to marshes 
 and lakes, some of which still remain. The trunks of the drowned 
 trees are in some cases yet standing above the water of these lakes 
 and marshes. The Orleans, the first steamboat built west of the 
 
 FIG. 449. Map showing the intensity of earthquakes in Italy. 
 (Baratta and Gerland.) 
 
 Appalachians, narrowly escaped destruction at New Madrid during 
 this earthquake. 
 
 7. At about the same- time, 1812, nearly 10,000 persons were 
 killed in a violent earthquake which destroyed Caracas, Venezuela. 
 
 8. Earthquakes have been most destructive in southern Italy. 
 Some 20,000 lives were lost here in 1688; 43,000 in 1693, and 32,000 
 in 1783; in all about 100,000 in a single century. 
 
CRUSTAL MOVEMENTS. DIASTROPHISM 
 
 419 
 
 9. On April 18, 1906, there was a destructive earthquake on the 
 coast of California, in and about San Francisco. Many buildings 
 were injured by the earthquake and some practically destroyed, 
 both in San Francisco and elsewhere. Fire broke out at various 
 points in San Francisco after the shocks, and as the quaking had 
 
 Fro. 450. Chart of epicentra and outer limits of sensibility of the earth 
 quakes of the eastern Mediterranean, from 1846 to 1870. (After J. 
 Schmidt. From Button's Earthquakes, by permission of G. P. Putnam's 
 Sons.) 
 
 cut off the city's supply of water, none was available for fighting 
 the flames, and a large part of the city was burned. This earth- 
 quake, the most disastrous in North America in historic times, 
 was caused by a horizontal fault of eight to twenty or more feet, 
 which was promptly traced some 300 miles. Figs. 451 to 459 show 
 some of the phenomena of this seismic disturbance. 
 
420 
 
 Earthquakes starting beneath the sea. Earthquakes some- 
 times seem to start beneath the sea, and to spread thence to the land. 
 The record of the accompanying changes beneath the sea is rarely 
 distinct, but in a few cases some facts are known about them. 
 
 FIG. 451. Map showing the position of the San Francisco earthquake fault. 
 The line north of Point Arena is quite uncertain. (Gilbert.) 
 
 This is especially the case with reference to some of the earthquakes 
 which have occurred about the coast of Greece, for, in a number of 
 cases, cables have been broken, and soundings taken when they 
 were repaired gave some indication of what had happened. In 
 
CRUSTAL MOVEMENTS. DIASTROPHISM 
 
 421 
 
 FIG. 452 A. An "earth flow," or landslide, which occurred during the Cali- 
 fornia earthquake, several miles west of the fault line east of Half Moon 
 Bay. (Photo, by Dudley.) 
 
 FIG. 452 B. Characteristic surface appearance of the fault line, south end 
 of Tomales Bay, Cal. (Photo, by Newsom.) 
 
PHYSIOGRAPHY 
 
 one case, the soundings from the bow and the stern of the vessel 
 which repaired the cable show differences of more than 1500 feet 
 
 FIG. 453. Deformed railway, Seventh and Mission Streets, 
 San Francisco. (Lindgren, U. S. Geol. Surv.) 
 
 FIG. 454. A fissure on East Street, San Francisco, near the water front. 
 "Made " ground. (Lindgren, U. S. Geol. Surv.) 
 
 in the depth of the water, where the bottom had been nearly level 
 when the cable was laid. 
 
CRUSTAL MOVEMENTS. DIASTROPHISM 
 
 423 
 
 FIG. 455. The fault line two miles southeast of Portola, Cal. 
 some vertical displacement at this point. 
 
 There was 
 
 FIG. 456. A broken (now mended) and offset fence. That in the foreground 
 was formerly in line with the single length directly behind the man. 
 
424 
 
 PHYSIOGRAPHY 
 
 The earthquake wave. An earthquake usually spreads over the 
 surface, somewhat as a water wave spreads from a center. Hence 
 we have come to speak of the earthquake wave. The actual move- 
 
 FlG. 457. A water-pipe buckled out of the ground by the earthquake. 
 Alpine road, Portola Valley, Cal. (Photo, by Dudley.) 
 
 Fio. 458. County Bridge, Pajaro River, Chittenden Cal 
 (Photo, by Dudley.) 
 
 ment is a wave, but the wave is unlike that of water, though there 
 are some points of similarity. The center of an earthquake wave 
 may be a line, a belt, or a point, and in many cases it is not readily 
 located. 
 
CRUSTAL MOVEMENTS. DIASTROPHISM 
 
 425 
 
 Points distant from the center feel the earthquake shock later 
 than those nearer to it. Lines drawn upon the surface connecting 
 points where a given earthquake wave is felt at the same time are 
 coseismic lines (Fig. 462). These lines are rarely circles, but they 
 are usually closed curves, and are often irregular. Their irregu- 
 larity appears to be due to the fact that the wave motion spreads 
 faster in some directions than in others. This is another way of 
 saying that the rocks in some parts of the earth's crust transmit 
 the motion faster than others. If a circular metal plate were set 
 
 FIG. 459. Tree uprooted by the earthquake. 
 (Photo, by Dudley.) 
 
 Searsville road, Cal. 
 
 vibrating at its center, the vibrations would spread from the 
 center in all directions at about the same rate, and would reach the 
 border of the plate at about the same time. But if the plate were 
 made of sectors of different material, one sector being of steel, 
 another of hard wood, and another of cork, the vibrations started 
 at the center would pass outward through these sectors with dif- 
 ferent velocities, and would extend to different distances in a 
 given time. The more porous and open the medium, the less the 
 distance which the vibrations would travel before being com- 
 pletely deadened. ... The principle illustrated here has some applica- 
 tion in earthquakes.- 
 
426 
 
 PHYSIOGRAPHY 
 
 FIG. 460 A. Track of electric railway, be- FIG. 460 B. A street railway on loose 
 tween South San Francisco and San ground. Union Street, near Pierce 
 
 Bruno Point. (Photo, by Moran.) Street. (Photo, by Moran.) 
 
 E 
 
 FIG. 461. Scenes on the Campus of the Leland Stanford University, after the earth- 
 quake of April, 1906. 
 
 A. The Agassiz statue. (Branner.) C. The University Chapel. (Branner.) 
 
 B. The great arch. (Branner.) D= The Library. (Branner.) 
 
CRUST AL MOVEMENTS. D1ASTROPH1SM 
 
 427 
 
 In general, the earthquake waves diminish in violence with 
 increasing distance from the centers of disturbance (Fig. 463). 
 
 Frequency. Earthquakes are of very common occurrence, 
 though fortunately those which are violent enough to be destruc- 
 
 FIG. 462. Coseismic lines for each minute, Herzogenrath (Germany), earth- 
 quake of October 22, 1873. (Lasaulx.) 
 
 tive are rare. From 1889 to 1899 an average of 36 per year were 
 recorded in California alone, but most of them were so slight as to 
 cause no destruction. In Japan, earthquakes have been recorded 
 at the rate of several per day for many years, but this includes 
 
 FIG. 463. Diagram illustrating the dimensions and intensity of vibration 
 with increasing distance from the epicentrum. (Belax.) 
 
 many very trivial quakings, and only a few of sufficient violence 
 to be destructive. 
 
 The Isthmus of Panama and its surroundings have been under 
 careful observation with reference to earthquakes for a few years, 
 
428 
 
 PHYSIOGRAPHY 
 
 because the frequency and violence of earthquakes had a bearing 
 on the site which was to be selected for the canal which was to 
 join the Atlantic and Pacific. In 40 months, between January, 
 1901, and April, 1904, 169 earthquakes were recorded at San Jose, 
 near the eastern end of the proposed Nicarauguan route. Of 
 
 FIG. 464. Map showing in black the principal earthquake regions of the Old 
 World. (Montessus de Ballore.) 
 
 these, 43 were mere tremors, 91 slight shocks, and 35 strong shocks. 
 During the same period, 6 tremors and 4 slight shocks were 
 recorded at Panama. 
 
 The slight tremors would probably not have been known but 
 for the observatories, many of which have been established in 
 recent times, where all earth-tremors, however slight, are recorded 
 by delicate instruments (seismographs) devised for this purpose. 
 
CRUSTAL MOVEMENTS. DIASTROPHISM 
 
 429 
 
 In view of the records at the stations in various parts of the 
 civilized world where such apparatus has been set up, it has been 
 said that some part of the earth's surface is probably shaking all the 
 time. 
 
 Distribution. Earthquakes are perhaps most common in 
 volcanic regions, though not confined to them. It can hardly be 
 
 FIG. 465. Map showing the principal earthquake regions of the New World. 
 
 (Montessus de Ballore.) 
 
 said that all such earthquakes are caused by volcanoes, since many 
 of them do not occur at the time of volcanic eruptions. It is 
 perhaps better to regard earthquakes and volcanoes as the result 
 of a common cause, rather than to regard one of them as the gen- 
 eral cause of the other. 
 
 Many great earthquakes have occurred near the edges of the 
 continental platforms. Mountain regions in general seem to be more 
 
430 PHYSIOGRAPHY 
 
 subject to earthquakes than plains, though earthquakes originat- 
 ing in mountain regions sometimes spread to plains. Earthquakes, 
 on the other hand, do not always start in mountain regions. As in 
 the case of the Charleston earthquake, they sometimes originate 
 beneath plains. 
 
 Causes of earthquakes. Earthquakes are probably due to 
 various causes. Small ones are perhaps sometimes due to the 
 falling in of the roofs of underground caves. If the roof of Mam- 
 moth Cave, for example, were to fall in, the disturbance would 
 cause an earthquake of small extent. Earthquakes accompany 
 violent volcanic eruptions, and in these cases the explosions 
 which cause the eruption are doubtless also the cause of the earth- 
 quakes. Great landslides and avalanches are the causes of some 
 minor earthquakes, and it is probable that slumping on the slopes 
 of deltas and on the outer faces of the continental shelves pro- 
 duces similar results. 
 
 Many great earthquakes appear to be connected with other 
 forms of crustal movement. As already noted, fissures are some- 
 times opened in the surface of the land during an earthquake. 
 This is best seen where there is little or no soil, and where the solid 
 rock lies close to the surface. There is a great crack of this sort 
 in Arizona (Fig. 467), and similar fissures have been formed in 
 New Zealand, Japan, and elsewhere during earthquakes. It is not 
 always clear whether the fissure should be looked on as the cause 
 or the result of the earthquake. In some cases it is found that one 
 side of such a fissure is higher than the other after the earthquake, 
 indicating that the rock on one side was raised or that on the other 
 sunk, or both in other words, that the strata have been faulted. 
 Faulting is probably the principal cause of earthquakes, for the 
 slipping of one great body of rock past another would cause 
 vibrations which would spread far from the center of disturbance. 
 
 There is sometimes horizontal as well as vertical displacement 
 along the cracks, as already noted, and the horizontal thrust or 
 fault is sometimes the principal one, as in the recent earthquake 
 of California. Horizontal displacement shows itself in the distor- 
 tion or breaking of lines which were straight or continuous before 
 the faulting. Thus fences or rows of trees which were straight 
 before an earthquake may be bent or broken and offset at the 
 fissure. The force which causes this displacement is the real cause 
 of the earthquake. 
 
CRUSTAL MOVEMENTS. DIASTROPHISM 
 
 431 
 
 Again, great thicknesses of rock strata are sometimes found 
 folded and crumpled. The process of mountain folding has never 
 been seen, and it is probably much too slow to be seen from day 
 to day or from year to year. But there can be no doubt that 
 
 FIG. 466. Faulting accompanying the Sinjan earthquake of 1898. 
 (Faidiga.) 
 
 beds now folded so as to stand on edge were once horizontal or 
 nearly so. No series of horizontal beds can be folded, as many 
 beds have been, without more or less slipping of layer on layer. 
 The amount of slipping at any one time may be slight, but it 
 
 FIG. 467. Fissure produced by earthquake. Arizona. 
 
 must be real. This, too, is probably a cause of earthquakes, and of 
 earth-tremors which are not sensible. 
 
 It is probable that most earthquakes are to be looked upon 
 as but one expression of the wide-spread movements to which 
 
432 PHYSIOGRAPHY 
 
 the crust of the earth is subject, movements which are due pri- 
 marily to the continued adjustment of the outside of the earth to 
 a shrinking interior. In general, these movements are too slow 
 to produce sensible vibrations; but locally and periodically they 
 are sufficient to cause distinct quakings. 
 
 Surface changes caused by earthquakes. The changes in 
 the surface of the land made by earthquakes are numerous if not 
 important. In addition to the cracks and fissures, and the risings 
 and sinkings of surface which have been noted, drainage is often 
 disturbed. This is partly because of the cracks and fissures which 
 are opened, and partly for other reasons. If an open fissure is 
 developed athwart the course of a stream, the stream will plunge 
 into it. Springs are often disturbed, old ones ceasing to flow and 
 new ones appearing. This is probably because the earthquake 
 movement has ruptured the rock beneath the surface, and so 
 changed the course of the ground-water circulation. Temporary 
 spouting springs are sometimes formed, water being forced up 
 violently through them (p. 414). Earthquakes sometimes cause 
 landslides, and if the material from a mountain-side slides down, 
 it may dam the valley below so as to disturb its drainage (p. 313). 
 
 From fissures and from lesser vents noxious gases sometimes 
 issue. In some cases, too, loose sand is thrust up into cracks from 
 beneath during earthquakes. 
 
 Earthquake waves have a singularly destructive effect upon 
 aquatic life. It has been recorded in many cases that the animals 
 of rivers, bays, and even of the ocean are killed in extraordinary 
 numbers during an earthquake. 
 
 MAP EXERCISES 
 
 Study the following maps in preparation for the conference: 
 
 A. Choptank, Md. 
 Tolchester, Md. 
 Boothbay, Me. 
 Coos Bay, Ore. 
 Oceanside, Cal. 
 Honey Lake, Cal. 
 Erie, Pa. 
 Fairview, Pa. 
 
 B. U. S. Coast and Geodetic Survey charts, 10, 124, 125, 188, 210, 
 
 5100, 5500. 
 
CRUSTAL MOVEMENTS. DIASTROPHISM 433 
 
 C. Mt. Trumbull, Ariz. 
 
 Echo Cliffs, Ariz. 
 
 Diamond Creek, Ariz. 
 
 Coast Survey Chart, 5100. 
 
 It is to be borne in mind that the relative change of water and 
 land levels along the shores of lakes may be due to the lowering of 
 the lake by the cutting down of its outlet, and not to diastrophism 
 at all. The topographic features which give evidence of the lowering 
 of a lake-level are, however, the same in kind as those which arise from 
 the uplift of the shore-land or from the sinking of the sea-level. 
 
 D. Study the maps of group A for evidences of change of relative 
 level of land and water. Answer in writing the questions marked *. 
 In the case of each map and chart, 
 
 1. Is submergence, emergence, or warping (part up and part down) 
 of the land in recent times indicated by the map? Reasons. 
 
 2.* Specify three well-defined cases of coast-land (or shore-land) 
 emergence suggested by the topographic maps, with reasons therefor. 
 
 3.* Specify three cases of coast-land (or shore-land) submergence 
 suggested by the topographic maps, with reasons therefor. 
 
 4. What factors besides diastrophism have probably been operative 
 in shaping the coast in the Boothbay region? In the Alaskan region? 
 Distinguish, if possible, between the features due to diastrophism 
 and those due to other causes. 
 
 5.* Make an interpretation of the topography of the coastal part 
 of the Oceanside Sheet, indicating the degree of confidence with which 
 your conclusions are held. 
 
 E. Study the charts of group B for evidences of submergence and 
 emergence of coastal lands. Evidences of submergence are to be found 
 largely in the configuration of the submerged surface. Evidences of 
 emergence are much the same as on the contour maps. Note the 
 unit (feet, fathoms) in which soundings are given, in the case of each 
 chart. 
 
 1. Note cases of well-defined apparent submergence of former lands, 
 as indicated by the configuration of the bottom (see p. 397). 
 
 2. Note cases of well-defined apparent emergence shown on these 
 charts (see p. 394). 
 
 3. What are the possible explanations of the larger features of the 
 coast-line shown on Chart 5500? 
 
 F. In group C, the Hurricane Ledge of the Mt. Trumbull Sheet, the 
 Echo Cliffs of the Sheet of the same name, the Grand Wash Cliffs of 
 the Diamond Creek Sheet, and the steep cliff southwest of Honey Lake, 
 on the Honey Lake Sheet, are fault-scarps now somewhat eroded. This 
 could, however, not be certainly known from the topographic map. 
 
434 PHYSIOGRAPHY 
 
 The steep cliffs on the northeastern coast of San Clemen te Island, Chart 
 5100, also represent an old fault-scarp. 
 
 REFERENCES 
 
 1. Standard text-books on Geology, under Changes of Level, Secular 
 Changes of Level, Crustal Movements, Earthquakes, etc. 
 
 2. MILNE, Earthquakes: Appleton, and Movements of the Earth's Crust: 
 Geog. Jour., Vol. VII, 1896, pp. 229-250. 
 
 3. DUTTON, Earthquakes: Putnam. 
 
 4. HOBBS, Earthquakes: Appleton. 
 
 5. BUTTON, The Charleston Earthquake. 9th Ann. Rept. U. S. Geol. Surv. 
 
 6. GILBERT, A Theory of Earthquakes of the Great Basin, with a Practical 
 Application: Am. Jour. Sci., Vol. XXVII, 1884, pp. 49-53. 
 
 7. GILBERT: Pop. Sci. Mo., Vol. LXIX, 1906, p. 97. 
 
 8. JORDAN: Pop. Sci. Mo., Vol. LXIX, 1906, p. 289. 
 
 9. SHALER: Chapter in Aspects of the Earth (Scribners). 
 
 10. McGfiE, The Gulf of Mexico as a Measure of Isostasy. Am. Jour. Sci., 
 Vol. XLIV, 1892, pp. 177-192. 
 
 11. SHEPARD, New Madrid Earthquake: Jour. Geol., Vol. XIII, pp. 45-56. 
 
 12. LE CONTE, Earth Crust Movements and their Causes: Science, Vol. V, 
 1897, pp. 321-330". 
 
 Many of the references under Volcanoes (p. 390) also touch upon earth- 
 quakes. 
 
CHAPTER IX 
 ORIGIN AND HISTORY OF PHYSIOGRAPHIC FEATURES 
 
 WE may now review the principal physiographic types in the 
 light of the knowledge afforded by the preceding study of physio- 
 graphic processes. 
 
 Plains 
 
 Plains, considered as one of the three great divisions of land 
 surfaces, have arisen in various ways, as already noted. In many 
 cases their materials show that they were once below sea-level. 
 From this position they were (1) bowed up, (2) faulted up, (3) or 
 built up so as to emerge from the water; or (4) the sea-level may 
 have been drawn down so as to leave them dry. Plains have arisen 
 also (5) by the degradation of plateaus or mountains. In many 
 cases, two or more of these processes have operated jointly in the 
 development of plains. 
 
 After plains have come into existence they are modified by 
 gradation, generally by stream erosion and often by glacier erosion; 
 by diastrophism, which may deform them; or by vulcanism, 
 which may build them up by lava-flows or diversify them by 
 the development of volcanic cones. 
 
 All existing plains of great extent have been modified in some 
 or all these ways. Thus the Coastal Plain of the Eastern United 
 States, developed by aggradation and diastrophism, has been much 
 changed by erosion, and perhaps somewhat by warping, since its 
 origin. The great Interior plain of the United States has been 
 much modified by rain and river erosion, and at the North by 
 glaciation. These processes have led to the development of many 
 minor features, as already indicated. Valleys have been made 
 by running water, and ridges and hills left in the process. Mounds, 
 hills, and ridges, with associated kettle-like, saucer-like, trough- 
 
 435 
 
436 
 
 PHYSIOGRAPHY 
 
 like, and irregular depressions, have been made by the continental 
 glaciers. Many of these depressions have become the sites ot lakes, 
 
 fiu. 468. A semi-arid plain, with sinks and one water-hole, in the western 
 part of the United States. (U. S. Geol. Surv.) 
 
 ponds, and marshes. Lands thus modified are sometimes called 
 glacial plains. In the bottoms of the larger valleys, river plains 
 
 FIG. 469. Arid plain, western United States. A mesa or plateau at the 
 right. (U. S. Geol. Surv.) 
 
 have been developed, and about many lakes whose basins have 
 been partly filled, or whose levels have been drawn down by the 
 
ORIGIN AND HISTORY OF PHYSIOGRAPHIC FEATURES 437 
 
 lowering of their outlets, lacustrine plains have been developed. 
 Similar flats occupy the sites of many former lakes which have 
 become extinct. 
 
 In addition to the changes produced 
 by gradation, the Interior Plain has 
 probably been somewhat changed by 
 unobtrusive expressions of diastro- 
 phism. 
 
 Various plains in the West have 
 been modified by the ejection of volcanic 
 matter, as well as by gradation and 
 diastrophism, and most plains have 
 been affected to some extent by the 
 wind. 
 
 Plateaus 
 
 Plateaus may originate through the 
 operation of some of the processes 
 which give rise to plains (Chap. 1). Suffi- 
 cient up-warping or up-wedging of the 
 sort which gives rise to plains would 
 give rise to plateaus. So also would 
 sufficient up-building, especially per- 
 haps by lava-flows. It may be doubted 
 whether the sea-level was ever lowered 
 enough at one time to convert coastal 
 plains into plateaus, and plateaus are 
 not made by the degradation of higher 
 lands. 
 
 After they come into existence, ^ jji II) ^ |y\ 
 plateaus are modified by all the proc- 
 esses which modify plains. All exist- 
 ing plateaus have felt the effect of some 
 of these processes, and most of them of 
 several. 
 
 MOUNTAINS 
 
 Some study has already been made 
 of mountains as topographic features, but some points concerning 
 them could not well be considered until the processes of vulcanism 
 
438 
 
 PHYSIOGRAPHY 
 
 and diastrophism had been outlined. We have now to see what 
 various forms mountains assume, how they are grouped, what 
 their structure is, and what purposes they serve in the economy of 
 nature. 
 
 Mountains have been defined as masses of land high enough 
 to be very conspicuous in their surroundings, but without a great 
 expanse of surface at the top. It is to be understood, however, 
 that between large-topped mountains and small plateaus there 
 are all gradations. 
 
 Those who have never seen mountains, but who have seen hills 
 and ridges, may perhaps best get their conceptions of mountains 
 
 FIG. 471. Dome-shaped mountain in the Uinta Mountains. (Church.) 
 
 by thinking of them as hills and ridges which, in their surround- 
 ings, appear to be very high. They may be but a few hundred 
 feet above their environs, or they may be many thousand feet, and, 
 as in the cases of hills and low ridges, their slopes may be steep or 
 gentle. 
 
 A single mountain may be but a big hill (Fig. 471 and PI. XXIV). 
 But, as already stated, there are all gradations between a big hill 
 and a little mountain, and whether an isolated elevation is called 
 a hill or a mountain depends on its surroundings, or on the judg- 
 ment of those who named it. A single mountain may stand in the 
 same relation to a mountain region that a single hill does to a hilly 
 region. 
 
 A mountain may be a high ridge rather than a high hill (PI. 
 XXV). A mountain ridge is often called a mountain range. A 
 
PLATE XXV 
 
 Dunning Mountain, Pennsylvania; a good example of a mountain ridge due to 
 the superior hardness (resistance) of a tilted layer of rock, the outcrop 
 of which was left as a ridge after the less resistant surroundings were 
 worn away. Scale 1 mile per inch. (Everett Sheet, U. S. Geol. Surv.) 
 
PLATE XXVI 
 
 An area southwest of Denver, showing a mountain ridge dissected by erosion. 
 The outcropping hard layer appears in the form of a series of short 
 ridges, or " hog-backs." (Compare PI. XXV.) Scale 2-miles per inch. 
 (Denver, Colo , Sheet, U. S. Geol. Surv.) 
 
ORIGIN AND HISTORY OF PHYSIOGRAPHIC FEATURES 439 
 
 mountain range may have a nesrly even crest (PI. XXV), or its crest 
 may be a series of high points partially separated from one another 
 by depressions (PI. XXVI). A mountain range or ridge may have 
 
 FIG. 472. An asymmetrical mountain ridge. 
 
 its opposite slopes much alike, or they may be very unequal (Fig. 
 472). 
 
 A mountain group is made up of several or many mountain 
 peaks, or of short mountain ridges. The Catskill Mountains (Fig. 
 473) and the Black Hills may serve as examples. 
 
 A mountain chain or system is an elongate mountain group, 
 made up of many single mountains or of mountain ranges, or of 
 both. The individual ridges commonly have a pronounced trend 
 in a common direction. The Appalachian Mountain system is an 
 example (Fig. 23, p. 29). A few mountain systems, like the Alps, 
 are not conspicuously longer in one direction than in another. 
 The more conspicuous elevations of a system have separate names, 
 and are individually mountains. 
 
 Distribution of mountains. In some of the continents the 
 more important mountains are toward the borders of the land 
 rather than in the interiors. It is to be noted, however, that 
 even in some of the continents where this is true, the moun- 
 tains are not all near the coast. In the western part of North 
 America, for example, some of the highest ranges are nearly 1000 
 miles from the Pacific, while portions of the eastern mountains 
 are some 400 miles from the Atlantic. In the narrow part of the 
 continent at the south, nearly all the land is mountainous. 
 
 In South America the high mountains (the Andes) are con- 
 fined to a belt rarely exceeding 300 miles in width near the coast, 
 while some of the lower mountains to the east are farther from 
 the sea. 
 
 In Africa, the highest mountains are near the southeastern 
 border of the continent. Mountains also occur on the northwest 
 border and at some other points; but, on the whole, it can hardly 
 
440 PHYSIOGRAPHY 
 
 be said that mountainous borders are especially characteristic of 
 this continent. In Australia, also, the more important moun- 
 tains are near the coast, though most of the coasts are not mountain- 
 ous. 
 
 The mountains of Europe and Asia, taken as a whole, can 
 
 FIG. 473. Photograph of a model of the Catskill Mountains. (Ho well.) 
 
 hardly be said to be near the oceans, though some of them have 
 such positions. 
 
 Heights. Mountains range in height from large hills or ridges 
 but a few hundred feet high, to elevations of nearly 30,000 feet. 
 The highest mountains in the United States, outside of Alaska, 
 
ORIGIN AND HISTORY OF PHYSIOGRAPHIC FEATURES 441 
 
 are found in the Sierra Nevada range of California, where the 
 highest peak (Mt. Whitney) reaches an elevation of nearly 15,000 
 feet. The highest mountains of the Rockies are but little lower, 
 many peaks exceeding 14,000 feet in elevation. In Colorado alone 
 there are about 40 peaks reaching an elevation of between 14,000 
 
 FIG. 474 A mountain valley. The narrow part of the canyon shown here 
 is to be the site of a dam 310 feet high, for a reservoir for irrigation pur- 
 poses. (U. S. Geol. Surv.) 
 
 and 14,500 feet. Mt. Rainier in Washington also reaches an 
 elevation of a little more than 14,000 feet. The highest moun- 
 tain in Alaska, Mt. McKinley, has an altitude of 20,300 feet. 
 
 The highest points in the Andes Mountains attain an elevation 
 of abcut 23,000 feet, and many peaks rise above 20,000 feet. 
 
 The highest peaks of the highest mountains of Europe, the Alps, 
 
442 PHYSIOGRAPHY 
 
 attain an elevation of nearly 16,000 feet, and the highest peaks in 
 the Caucasus are but little less. In the Himalayas, the loftiest 
 mountains of Asia and of the earth, the highest peak, Mt. Everest, 
 is nearly 30,000 feet above sea-level. 
 
 The mountains of Africa and Australia are, for the most part, 
 much lower. A few volcanic peaks in the former attain an eleva- 
 
 FIG. 475. King's River Valley, Cal. (U. S. Geol. Surv.) 
 
 tion of nearly 20,000 feet, while the greatest elevation of Australia 
 falls short of 8000 feet. 
 
 Oceanic mountains. Mountains exist in the ocean basins as 
 well as on the continental platforms. Many oceanic mountains are 
 partly or wholly beneath the water, but the crests of some of them 
 are not. 
 
 If the height of a mountain be reckoned in terms of elevation 
 above its base, rather than in terms of elevation above sea-level, 
 some of the volcanic cones of the ocean would rank among the 
 highest mountains of the earth. Thus Mauna (Mount) Kea (Fig. 
 
ORIGIN AND HISTORY OF PHYSIOGRAPHIC FEATURES 443 
 
 392), on the island of Hawaii, rises nearly 14,000 feet above 
 the sea. Measured from the ocean floor, from which the island 
 rises, its height is more than 30,000 feet. It is, in one sense, 
 nearly or quite the highest mountain of the earth, though not the 
 highest above sea-level. Parts of the Antillean mountain system 
 (including the West Indies and the mountains of Central America, 
 etc.) also rise from a depth of 16,000 to 18,000 feet below sea-level 
 to a maximum height of more than 10,000 feet above. They are 
 therefore among the greatest mountains of the earth, if their 
 elevation be reckoned from their real base. 
 
 Changes taking place in mountains. Most of the processes 
 of degradation already studied are in operation in mountain 
 regions, but their relative importance is not the same as in lower 
 lands. The differences are due partly to the steepness of the 
 mountain slopes, and partly to the differences of climate incident 
 to altitude. 
 
 Because of their steep slopes, erosion by mechanical processes 
 is more rapid in mountains than on plains. Streams in mountains 
 are, as a rule, rapid, at least in the early stages of an erosion cycle, 
 and make deep valleys. Chiefly for this reason, mountains are the 
 roughest portions of the earth's surface. Rapid erosion means 
 that weathered rock is promptly removed. The accumulation of 
 mantle rock is therefore less in mountains than where erosion 
 is less rapid, and bare rock is, accordingly, more common. 
 
 The temperature decreases on the average about 1 Fahr. for 
 every 300 feet of rise. If a mountain is 3000 feet higher than its 
 surroundings, the temperature at the top is therefore some 10 
 colder than at the bottom. Because of their low temperature, 
 high mountains have little vegetation. The absence of vegetation 
 allows running water and wind to remove weathered rock readily. 
 When a mountain is so cold as not to allow the growth of vegetation, 
 the absence of the plants, together with the steepness of slopes 
 characteristic of mountains, leaves the bare rock freely exposed 
 to all the processes of weathering. Daily changes of rock tempera- 
 ture are great in high altitudes, especially on sunny days, and 
 rock breaking, due to this cause, is most effective. The steep 
 slopes allow the rock-masses broken off in this w r ay to fall or to 
 be carried down readily (Fig. 476), thus exposing fresh surfaces 
 of rock to the same changes. 
 
 In general, there is more precipitation (rain and snow) in moun- 
 
PHYSIOGRAPHY 
 
 tains than on plains, and more of it falls as snow. The snow accu- 
 mulates through a considerable part of the year, to be melted at a 
 later time. When it melts, the water runs off and has much the 
 effect of concentrated rainfall. If it accumulates in sufficient 
 quantity, it will give rise to glaciers, which, except in very high 
 latitudes, do not occur outside of mountain, regions. On the 
 
 FIG. 476. Quartzite Peak, Wasatch Mountains, with quantities of talus at 
 its base. (Chamberlin.) 
 
 whole, therefore, erosion is more rapid in mountains than else- 
 where. 
 
 The deposition of sediment, on the other hand, is relatively 
 less in mountains than on plains, because of the steep slopes and 
 the swift streams. Much of the debris which falls or is carried 
 down steep slopes is however temporarily lodged at their bases 
 (p. 182). 
 
 Winds are often strong in mountain regions, though they pro- 
 
ORIGIN AND HISTORY OF PHYSIOGRAPHIC FEATURES 445 
 
 duce relatively little direct effect on the land, (1) because it is less 
 commonly dry, and (2) because there is little material fine enough 
 to be blown. 
 
 The winds in the mountains have a notable effect on the char- 
 
 FIG. 477. A mountain tree. Near Granite, Colo. (Capps.) 
 
 acter of the trees (Fig. 477), especially where they are scattered, 
 or near the upper limit of their growth (the timber line). 
 
 Origin of Mountains 
 
 Volcanic mountains. Mountains originate in very different 
 ways. Figs. 362 and 364 show isolated mountains of volcanic origin. 
 Single mountains having this structure are among the high moun- 
 tains of the earth. Besides Shasta, Rainier, and others already 
 mentioned (pp. 378-84), the Spanish Peaks of Colorado (13,620 
 feet) and Mt. Wrangell of Alaska (17,500 feet) belong to this class. 
 So, also, do Orizaba (18,200 feet) and Popocatepetl (17,523 feet) 
 of Mexico; Tajamulco (18,317 feet) and others of Central America; 
 Aconcagua (22,860 feet), Chimborazo (21,498 feet), and numerous 
 others in the Andes; Elbruz (18,470 feet), Demavend (18,000 
 feet), Great Ararat (nearly 17,000 feet), Fuji-yama (12,365 feet), 
 and others in Asia; and Kilimanjaro (19,780 feet) and Kenia 
 (18,000 feet) in Africa. The highest mountains of Africa and 
 South America are volcanic. 
 
 The origin of many other mountains is also clearly suggested 
 by their structure. 
 
 Mountains produced by erosion. One type of mountain struc- 
 ture is represented by Figs. 478 and 479. Mountains of this sort 
 
446 
 
 PHYSIOGRAPHY 
 
 may occur singly, but they are often in groups. They are clearly 
 the result of erosion, their surroundings having been worn away. 
 
 FIG. 478. Mountains of horizontal strata, Timpanagas Mountain, Utah. 
 
 (Church.) 
 
 Mountains of this sort are developed from plateaus in the course 
 of their degradation. The Catskill Mountains are an illustration. 
 Other illustrations occur in the arid regions of the West, where 
 the isolated masses of rock are often called buttes (Fig. 173). 
 
 FIG. 479. Mountains shaped by the erosion of horizontal beds of stratified 
 rock. Castle Group, Colo. (Holmes, Hayden Surv.) 
 
 Other illustrations of mountains, the outlines of which were pro- 
 duced by erosion, are shown in Figs. 479 and 480. 
 
ORIGIN AND HISTORY OF PHYSIOGRAPHIC FEATURES 447 
 
 Intrusion and uplift. Fig. 37 (p. 39) represents another common 
 type of mountain structure. Mountains of this type may con- 
 stitute single mountains or groups of mountains. They may be 
 large or small. In most cases the bedded rocks which lie on the 
 
 FIG. 480. Mountains shaped by erosion, where the rock is massive. Elk 
 Mountains, Colo. (Holmes, Hayden Surv.) 
 
 sides once extended over the crests, and have been cut away by 
 erosion. The Black Hills and the Adirondacks are examples of 
 large groups of mountains of this type. Numerous small ones 
 occur about the Black Hills and at many other points in the 
 West. The Henry Mountains of Utah (Figs. 402 and 404), made 
 classical by the exhaustive study of Gilbert, are a well-known 
 illustration of this general type. 
 
 Mountains of this type of structure may be linear, making a 
 mountain range or system rather than a mountain group. The 
 Sierra Nevada Mountains of California are an example. 
 
 Mountains produced by folding. Fig. 481 represents still 
 another type of mountain structure illustrated by the Jura Moun- 
 tains, while Figs. 482 and 483 represent variations of the type. 
 
 FIG. 481. Section of the western Jura Mountains. 
 
 The Jura Mountains as they now are, are the result of the upswell- 
 ing (without renewed folding) of a worn-down mountain system 
 produced originally by folding. In other words, the folded struc- 
 ture of the former mountains has been lifted up bodily in recent 
 times. The details of the present mountains are the result of 
 erosion on this upwarped structure. 
 
448 
 
 PHYSIOGRAPHY 
 
 It is to be noted that the present topography of mountains 
 whose component strata are folded was not always produced 
 by folding. It was indeed rarely produced in this way. The 
 
 folding doubtless gave rise 
 to ridges, sometimes to ridges 
 of great height, as in the case 
 of the Appalachians. The 
 
 FIG. 482. Section across the Shcrtenkopf, mountains thus produced 
 Bavarian Alps. (Frass.) 
 
 were then brought low by 
 
 erosion. Later, the planed-down region of folded rocks was bowed 
 up, relative to its surroundings, but bowed up as a unit, without 
 further folding. The present mountain crests are the outcrops 
 of the harder layers, isolated by erosion subsequent to this later 
 
 rfiiS*^- ^jlSr^-^iiillsii^iiS^ -^w^T^S^Sir"^ 
 
 FIG. 483. Appalachian structure. (Rogers.) 
 
 uplift (Figs. 484 and 485). It is now known that many other 
 mountains of folded structure have had a similar history. 
 
 Most mountains produced by folding have been extensively 
 
 FIG. 484. Diagram suggesting the type of structure possessed by the simple 
 folding of strata. The diagram shows the folded surface worn down. 
 
 modified by erosion, as the accompanying figures suggest, but 
 there are occasional exceptions, as shown by Fig. 486. 
 
 Mountains produced by faulting. Figs. 487 and 488 represent 
 another type of mountain structure. Such mountains are some- 
 times called block mountains, because great blocks of the earth's 
 crust, bordered by distinct planes of fracture, have been tilted so 
 that one edge at least is well above its surroundings. In many 
 such cases, it is probable that the surroundings have sunk rather 
 than the mountains themselves that have been elevated. Many of 
 
ORIGIN AND HISTORY OF PHYSIOGRAPHIC FEATURES 449 
 
 the mountains between the Rockies and the Sierras belong to this 
 type, as elsewhere noted. This type of mountains is sometimes 
 known as the Basin Range type. 
 
 FIG. 485. The same as Fig. 484, after an upwarp and subsequent erosion. 
 No further folding is shown, and erosion has isolated the hard layers as 
 mountain ridges. This represents, in a general way, the present condi- 
 tion of the Appalachians. 
 
 These illustrations show, that mountains have various struc- 
 tures; they also suggest how mountains originated, though they 
 do not in all cases indicate the causes which brought them into 
 existence. 
 
 Summary. It will be seen from the foregoing that mountains 
 are developed (1) by the degradation of their surroundings (Fig. 
 
 CLEMAN MOUNTAIN 
 
 UMNUM RIDGE 
 
 i. ::-..."... .i :- .-^-^ " ' '^- -^ 
 
 FIG. 486. Open low mountain folds, not greatly modified by erosion. Cle- 
 man Mountain and Umnum Ridge, Washington. (U. S. Geol. Surv.) 
 
 471); (2) by the subsidence of their surroundings, either by down- 
 warping or down-wedging (Fig. 487); (3) by elevation, either by 
 up-warping (folding) (Fig. 486) or up-wedging; (4) by up-swelling, 
 due to intrusions of igneous rock (Fig. 403) ; and (5) by up-building, 
 
 FIG. 487. Ranges of the Great Basin. Length of section, 120 miles. 
 
 (Gilbert.) 
 
 as in the case of volcanic cones. Mountains which originate by 
 diastrophism or vulcanism are subject to erosion, and most existing 
 mountains of volcanic or diastrophic origin have been so largely 
 
450 
 
 PHYSIOGRAPHY 
 
 modified by erosion, that the details of their present surfaces 
 are the result of degradation. 
 
 FIG. 489. The divide between the head- 
 waters of the Lake Fork of San Miguel 
 River and Cascade Creek, southwestern 
 Colorado. The lowest point in the di- 
 vide, in -the center of the photograph, is 
 12,700 to 12,800 feet above the sea. 
 (Hole.) 
 
 Jx 
 
 "V 
 
 FIG. 490. Cascade Pass, Washington. 
 The trees have an Alpine aspect. 
 (U. S. Geol. Surv.) 
 
 Effects of Mountains on Mankind 
 
 Climctic effects. Directly and indirectly, mountains play an 
 important part in the affairs of men. In the first place, they affect 
 
ORIGIN AND HISTORY OF PHYSIOGRAPHIC FEATURES 451 
 
 climate greatly. The winds blowing over them are cooled, and 
 as they are cooled a part of their moisture often condenses and 
 falls. Thus mountains, and especially the windward sides of 
 mountains, are generally the sites of heavy rainfall, and there- 
 fore become the sources of important streams. On the other 
 hand, plains and plateaus on the leeward side of mountains often 
 have light rainfall, because the air, after passing over the moun- 
 tains where it has left much of its moisture, is drying rather than 
 rain-giving. This is the reason why the tracts east of the Sierra 
 and Rocky mountains are arid or semi-arid (Fig. 492), and therefore 
 sparsely settled. The state of Nevada, with an area of more than 
 
 FIG. 491. A view in the Sierras from University Peak. Ko-ip Crest, 
 Sierra Nevada Mountains. 
 
 100,000 square miles, had a smaller population in 1900 than the 
 city of Peoria, Illinois. 
 
 Though the mountains make the country to leeward arid, 
 they sometimes furnish water which may be utilized in irrigating 
 these lands, for as the water which falls in the mountains flows 
 out from them, it may be diverted from its natural courses and 
 carried out by ditches to the fields. 
 
 The work of storing water in preparation for irrigation has 
 been well started in the western part of the United States (p. 193), 
 and still more extensive work in this direction is already planned. 
 But a small part of the arid lands of the West will ever be irrigated, 
 however, for the amount of water available is too small to supply 
 more than a fraction of all the land which needs water. 
 
 The effect of mountains on the temperature, winds, cloudiness, 
 
452 
 
 PHYSIOGRAPHY 
 
 etc., of their surroundings is considerable, though perhaps less 
 important than their effect on precipitation. 
 
 Mountains are barriers to transportation. It is true that 
 railroads now cross mountain systems, but the cost of building 
 and operating them after they are built is much greater than 
 on the plains. A railroad map of the United States shows that 
 there are few railroads in the eastern or western mountains, as 
 compared with the number in the interior. Mountains are how- 
 ever much less effective barriers to mankind now than in earlier 
 
 FIG. 492. Rainfall of the United States. (U. S. Weather Bureau.) 
 
 times, before railway engineering had reached its present develop- 
 ment. 
 
 Mountains are effective barriers to animals and plants. 
 Most of the animals lower than man do not possess, and are unable 
 to devise, means of crossing mountains, and to many of them high 
 mountains are effective barriers. The climate of high altitudes 
 is often such as to prevent the migration of plants from one side 
 to the other, except by human help. 
 
 Mountains often contain ores of various metals. The fact 
 that mining is the most distinctive industry of many mountains 
 has been referred to elsewhere. It may be added here that most 
 of the gold and silver and much of the copper of the United States 
 come from the western mountains. From the same sources also 
 
ORIGIN AND HISTORY OF PHYSIOGRAPHIC FEATURES 453 
 
454 
 
 PHYSIOGRAPHY 
 
 IOO 35 9O 85- 60 T5 TO 
 
 GOLD AND SILVER 
 
 FIG. 494. Map showing the areal distribution of gold and silver in the United 
 States. The relative importance of the productions of different areas is 
 not indicated. Circles = gold, crosses = silver, and the two combined = 
 gold and silver. (After Ransome.) 
 
 FIG. 494a. Map showing the distribution of copper ores in the United States. 
 The sizes of the dots indicate approximately the relative amounts pro- 
 duced. (U. S. Geol. Surv.) 
 
ORIGIN AND HISTORY OF PHYSIOGRAPHIC FEATURES 455 
 
 125 iaO IIS IIO 105 100 95 90 65 6O 75 . 7O 65 
 
 FIG. 495. Map showing the distribution of iron ores in the United States. 
 In 1904 Pennsylvania produced about 350,000 tons; Colorado, about 
 400,000 tons; Virginia and New Jersey, about 500.000 tons each; New 
 York, about 789,000 tons; the southern Appalachian region, chiefly 
 Alabama, about 4,459,000 tons; while the Lake Superior region pro- 
 duced 20,198,000 tons. The production of the Lake Superior region had 
 increased by 1906 to more than 33,000,000 tons. (Based on an unpub- 
 lished map prepared by W. O. Hotchkiss, with C. K. Leith.) 
 
 FiG. 496. Map showing the distribution of coal in the United States. 
 (U. S. Geol. Surv.) 
 
456 
 
 PHYSIOGRAPHY 
 
 comes much, but not all, of the lead and zinc. Iron and coal, 
 on the other hand, the two most important products of mining, 
 are not won chiefly from mountain regions, though some iron 
 and much coal is mined both in the eastern and western moun- 
 tains. 
 
 >LEAO AND ZINC 
 
 FIG. 497. Map showing the distribution of lead and zinc ores in the United 
 States. Circles = lead, crosses = zinc, and circles and crosses combined 
 = lead and zinc. (After Ransome.) 
 
 Agriculture in mountains. Mountain valleys are often fertile, 
 and many of them are under cultivation. Colorado, a moun- 
 tainous state, produces more mineral wealth than any other state 
 in the West; but the value of the products of the soil is greater 
 than that of the products of its mines. A considerable portion of 
 the cultivated land is in the mountains. 
 
 Scenic effects. Quite apart from economic considerations, 
 mountains have a value not to be estimated in dollars and cents, 
 in the scenery which they afford. The man who has not seen 
 mountains, and who has not lived with them long enough to really 
 make their acquaintance, has missed one of the good things of life. 
 The Adirondacks, the Catskills, and the other mountains within 
 a few hours' ride of such great cities as New York, Philadelphia, and 
 
ORIGIN AND HISTORY OF PHYSIOGRAPHIC FEATURES 457 
 
 Boston, are of inestimable value. The mountains of the West are 
 far grander. They are, unfortunately, frequented by fewer people, 
 because much farther from great centers of population. 
 
 The Outlines of the Continents 
 
 Any good map showing the outlines of the continents makes 
 it clear that some coast-lines are regular while others are most 
 irregular. The coasts of the northern part of North America and 
 Eurasia are strikingly irregular, and in strong contrast with the 
 outlines of South America, Africa, and Australia. The west 
 coast of the southern part of South America is, however, very 
 irregular. 
 
 Even the more regular coast-lines present contrasts, for some 
 of them are nearly straight, while others are notably curved; and, 
 where the continental outlines present large irregularities, certain 
 portions of the coast, considered by themselves, are regular, and 
 these may be straight or curved. Illustrations of such regularity 
 are found on the west coast of India and the southeast coast of 
 Arabia, though India and Arabia themselves constitute coastal 
 irregularities of great size. 
 
 The irregular coasts present much greater variety. Broadly 
 speaking, it may be said that there are two great types of irregu- 
 larity, namely (1) projections of the water into the land, and (2) 
 projections of the land into the water. The former are bays, gulfs, 
 etc., and the latter are peninsulas, capes, etc. 
 
 Coastal irregularities may be further classified in various ways, 
 and each classification brings out certain significant features. 
 They may be classified on the basis of (1) size, (2) position with 
 reference to the general trend of the coast, (3) relief, (4) origin, 
 etc. 
 
 Size. The projections of water into the land, like the pro- 
 jections of land into the water, may be either large or small. The 
 Gulf of Mexico, Hudson Bay, the Bay of Bengal, and the Baltic Sea 
 are examples of large projections of water into the land, while Dela- 
 ware, Chesapeake, Narragansett, and San Francisco bays, and Puget 
 Sound are examples of smaller projections of water into the land. 
 The little bays on the sides of Chesapeake Bay (Fig. 177) are 
 examples of still smaller indentations of the same sort. On the 
 coasts of Alaska (Fig. 498), Norway, Chile (Fig. 499), and some 
 
458 
 
 PHYSIOGRAPHY 
 
 FIG. 498. Map of part of the coast of southern Alaska, showing islands, 
 which were once a part of the mainland, isolated by glacial and wave 
 erosion, and by sinking. 
 
ORIGIN AND HISTORY OF PHYSIOGRAPHIC FEATURES 459 
 
 other places, there are numerous narrow but deep bays or 
 fiords (p. 248). 
 
 Examples of large projections of land into the sea are afforded 
 by the north coast of Australia, the east coast of Africa, the south 
 
 SKETCH MAP Or 
 
 S O UT HERN 
 
 ARGEN TI NA 
 
 FIG. 499. Islands on the west coast of South America which were once a 
 part of the continent. They have been isolated by erosion of glaciers 
 and waves and by subsidence. 
 
 coast of Asia, and both the south and west coasts of Europe. 
 North America, too, furnishes illustrations of this sort, especially 
 in Alaska and Labrador, and, on a somewhat smaller scale, in 
 
460 PHYSIOGRAPHY 
 
 Florida, Yucatan, and Lower California. Small projections of 
 land into the sea abound on most coasts. Cape Cod, Cape May, 
 and Cape Mendocino are examples. 
 
 The projections of water into the land and of land into the 
 water are often closely related. If two bodies of water not far 
 from each other project into the land, they leave between them a 
 projection of land into the water. An illustration is afforded by 
 the Bay of Bengal and the Arabian Sea with India between. India 
 may therefore be looked upon as a projection of land into the 
 sea, or as an area of land left by the projection of two great arms 
 of the sea into the land. Similarly, if two areas of land not far apart 
 project into the sea, they enclose a body of water. The land east 
 and west of the Gulf of Carpentaria on the north coast of Australia 
 is an example. 
 
 Position. Arms of the sea may project into the land, or 
 areas of land into the sea, so as to be somewhat nearly at right 
 angles to the trend of the coast; or they may have positions essen- 
 tially parallel to the general trend of the coast. Florida, India, 
 Hudson Bay, and the Gulf of Mexico are examples of the former, 
 and the Gulf and Peninsula of Lower California and many small 
 irregularites along the Atlantic coast of the United States (Fig. 
 500) are examples of the latter. It will be seen that both large 
 and small irregularities may occupy either position. They also 
 occupy positions intermediate between these two. 
 ; ; Relief. Some coasts are high and some low, and the differences 
 are so great that the lands present strong contrasts when seen 
 from the sea. The irregularities of coast-lines as seen on maps 
 affect both the coasts which are low and those which are high. 
 Some of the great peninsulas which project out into the water, 
 such as Scandinavia, the Iberian Peninsula, India, Lower Cali- 
 fornia, and much of Alaska, are high, and others, such as Florida 
 and Yucatan, are low. Small projections of land into the sea 
 present the same contrasts. Some of them, as those on the coasts 
 of Maine, Alaska, and Chile, are high, while others, such as those 
 along the eastern coasts of the United States south of New York, 
 are low. 
 
 The water in bays, gulfs, fiords, etc., is sometimes deep and 
 sometimes shallow, and its depth is measurably independent of 
 area. It is deep, for example, in the Gulf of Mexico, the Gulf of 
 California, the Mediterranean Sea, the Arabian Sea, in many 
 
ORIGIN AND HISTORY OF PHYSIOGRAPHIC FEATURES 461 
 
 fiords, etc., but shallow in the Baltic Sea, Hudson Bay, and the 
 Gulf of Carpentaria. 
 
 Distribution of various types of irregularities. From the 
 maps showing the outlines of the continents, it appears that great 
 irregularities are distributed with less inequality than the small 
 ones. While the northern continents have both large and small 
 
 FIG. 500. Portion of the coast of Texas, showing the tendency of shore 
 deposition to simplify the coast line. The deposits (narrow necks of 
 land parallel to the coast) shut in bays. (Coast and Geodetic Surv.) 
 
 irregularities in greater numbers than the southern continents, 
 the contrast between the small irregularities of the northern 
 and southern continents is greater than that between the large 
 ones. 
 
 The great irregularities are not notably greater in the northern 
 parts of the northern continents than in their southern parts. So far 
 as this class of irregularities is concerned, the coast-lines of southern 
 Asia and Europe are as irregular as those of other parts of these 
 continents. The small irregularities of northern Europe, and 
 especially of northwestern Europe, are, however, more conspicuous 
 
462 PHYSIOGRAPHY 
 
 than those of southern Europe. The same holds, in a general way, 
 for North America. While this continent has great and small 
 irregularities both at the north and south, small irregularities 
 are more numerous in high latitudes than in low. 
 
 Again, the small irregularities in the southern part of North 
 America are more commonly low, while those in the northern part 
 often have greater vertical range. The former are often parallel 
 to the trend of the coast, while the latter are more commonly at 
 right angles to it. 
 
 The irregularities of coasts stand in some relation to the width 
 of the continental shelf. Large irregularities of outline are, in 
 general, more common where the continental shelf is wide than 
 where it is narrow. High shores are, on the whole, more irregular 
 in outline than low ones, though to this general rule there are many 
 exceptions. 
 
 The islands along many of the coasts of continents are really 
 to be looked upon as parts of the coastal irregularities, for, as we 
 shall see later, many islands along coasts were once part of the 
 mainland. Here belong many of the islands off the coast of Alaska 
 (Fig. 498), Chile (Fig. 499), Scandinavia, etc. 
 
 All these numerous and varied irregularities call for explana- 
 tion, and our studies of processes now in /Operation have furnished 
 the data necessary for understanding why some coast-lines are 
 regular and others irregular, why some coasts are high and others 
 low, why the slopes of some coasts are steep and those of others 
 gentle. They have also given us a basis for some conception of 
 the origin of great and small projections of land into the sea, and 
 of great and small projections of the sea into the land. 
 
 Agents of gradation. In preceding chapters w r e have seen 
 the results produced by agents of gradation on the horizontal con- 
 figuration of coasts. We have seen (p. 320) that waves tend to 
 develop indentations of water where the rock is weak, leaving 
 projections of land where the rock is resistant, and that irregu- 
 larities thus developed are relatively small. The capes, etc., thus 
 formed will be low or high, depending on the relief of the land from 
 which they were developed. The reentrants of water developed 
 by wave erosion are always shallow-. 
 
 We have also seen that deposition along shores develops irregu- 
 larities, especially by the formation of strips of land roughly parallel 
 to the trend of the coast, across the debouchures of bays, etc., and 
 
ORIGIN AND HISTORY OF PHYSIOGRAPHIC FEATURES 463 
 
 that the irregularities thus developed are a step in the direction 
 of final simplification of the shore-line (Fig. 500). The lands de- 
 veloped by shore deposition are always low, as left by the waves, 
 and the lagoons shut in behind them are shallow. 
 
 We have also seen (p. 248) that glaciers descending to the sea 
 may gouge out deep valleys, which become fiords when the ice 
 melts. Glacial erosion may otherwise modify the coast-line, both 
 by erosion and deposition. Glaciation, once much more exten- 
 sive than now, affords the explanation, or at least a part of the 
 explanation, of the many fiords of high latitudes. Subsidence may 
 also be a factor in the development of fiords. 
 
 Rivers make coast-lines irregular by building deltas at their 
 debouchures, but through their erosive work they do little to 
 make coast-lines irregular horizontally. On the other hand, they 
 make high coast-lands irregular vertically, by developing valleys 
 in them. 
 
 Winds have little effect on the horizontal configuration of 
 coasts, but by piling up dunes they affect the relief of coast-lands 
 to some extent. 
 
 This brief review makes it clear that agents of gradation are 
 competent to produce many irregularities of coast, especially those 
 of small size. 
 
 Diastrophism. If the bottom of the ocean were somewhat 
 depressed, increasing the capacity of the basin, the water would be 
 drawn down about the borders of the continents and all the coast- 
 lines would be shifted seaward. On such a coast as that of the 
 eastern part of the United States, the border of the continent would 
 become notably more regular than now, because the topography of 
 the continental shelf, now submerged, is nearly plane. Some coasts 
 which are comparatively regular owe their regularity to recent 
 emergence. 
 
 If, on the other hand, the borders of continents were depressed, 
 the coast-lines would in general become somewhat more irregular 
 than now, for the depression of the land would allow the sea-water 
 to extend up the valleys, developing bays where there are none 
 now, and extending those which now exist (p. 174). 
 
 Some indented coasts, like that of the United States between 
 New York and the Carolinas, owe their numerous bays to recent 
 subsidence. Where the river valleys were normal to the coast, as 
 is most commonly the case, roughly speaking, the bays are normal 
 
464 PHYSIOGRAPHY 
 
 to the coast. If the valleys drowned in the making of the bays 
 were not normal to the coast, the bays would not be. 
 
 Again, the sufficient up-warp of a submerged tract along the 
 coasts of continents would develop peninsulas. These penin- 
 sulas might be normal to the coast or roughly parallel to it, or at 
 any angle between. A corresponding down-warp would develop 
 a bay or gulf, and many large bays and gulfs have probably arisen 
 in this way. The elevated or depressed area might be faulted 
 instead of warped, with similar results so far as the horizontal 
 configuration of the coast is concerned. 
 
 Vulcanism. Volcanoes affect coast-lines locally, but their 
 influence is relatively slight as compared with that of gradation 
 and diastrophism. Volcanoes make islands near coasts more com- 
 monly than they produce modifications of the coasts of mainlands. 
 Igneous rocks are often more resistant than sedimentary rocks, 
 and so affect the forms of coast lines developed by erosion. 
 
 Application 
 
 By the application of the above principles to coast-lines, the chief 
 features of many of them may be readily understood. Where 
 there are numerous bays along the coast-line projecting into the 
 land at right angles, roughly speaking, to its general trend, it may 
 be inferred with some confidence either that the region has re- 
 cently sunk, drowning the lower ends of the rivers, or that it has 
 been glaciated recently, converting the lower ends of the valleys 
 into fiords, or both. If the area concerned is in low latitude, the 
 chances are in favor of the first interpretation; if in high latitude, 
 and especially if the altitude be high, glaciation is a probable or 
 partial cause of the indentations. 
 
 Chesapeake Bay and the numerous bays tributary to it, Dela- 
 ware Bay and others of the same sort on the eastern coast of the 
 United States, point clearly to recent submergence of the land. 
 Farther north, the indentations of the coast of Maine find their 
 explanation partly in subsidence perhaps, but largely in glacier 
 erosion, for the ice of the continental glacier passed out to sea 
 over this coast. The fiords of such coasts as that of Alaska, 
 Chile, Scandinavia and Scotland are largely the result of glacial 
 erosion, though subsidence may have deepened and extended the 
 indentations of water. 
 
ORIGIN AND HISTORY OF PHYSIOGRAPHIC FEATURES 465 
 
 Where there are long, narrow belts of low land, roughly parallel 
 to the general trend of the coast, deposition by waves and shore 
 currents is to be inferred. Illustrations are afforded by many 
 parts of the coast between New York and Texas. 
 
 Where there are great irregularities of outline, such as the 
 Gulf of Mexico, the Gulf of California, the Adriatic Sea, the Bay 
 of Bengal, the Arabian Sea, the Iberian Peninsula, Italy, India, 
 Kamchatka, the peninsulas of Lower California, Yucatan, Florida, 
 etc., diastrophism has probably been the chief factor concerned. 
 
 The irregularities of coast produced by diastrophism are not 
 all of great size. Puget Sound, though large, is much smaller 
 than most of the irregularities mentioned, and is believed to have 
 had its origin in a down-warp. 
 
 Where coasts are high, diastrophism or wave-cutting, or both, 
 are suggested. Steep slopes, even where not high, give the same 
 suggestions, while low coastal lands without cliffs are character- 
 istic of areas of shore deposition. 
 
 It is to be borne in mind that coast-lines are not permanent, 
 and that the coast-lines of to-day are not precisely the same as 
 the coast-lines of yesterday, and those of to-morrow will not be 
 precisely the same as those of to-day, for gradation and diastrophism 
 are constantly changing them, and vulcanism occasionally. 
 
 Historical bearing. The character of the coast-lines has had 
 an important influence upon the development of many countries. 
 The irregular coasts of northwestern Europe and the northeastern 
 part of the United States abound in harbors, and favor the de- 
 velopment of ocean commerce. On the other hand, a smooth, 
 regular coast has always rendered difficult, and sometimes com- 
 pletely discouraged, sea trade. The southeastern states, eastern 
 Mexico, Africa, and India have all experienced, in varying degrees, 
 the disadvantages of such a coast. The greatest motive in Russian 
 expansion has been the possession of ice-free harbors. 
 
 Wherever a people has occupied an indented coast, with off- 
 lying islands and an infertile hinterland, it has early turned to 
 the sea for a living, and has developed daring seamen; and this 
 regardless of race or inherent abilities. Examples are the North- 
 men, the Indians of southern Alaska, the blacks of northwestern 
 Madagascar, and the Malays of the Tenasserim coast. On the 
 other hand, a harborless coast has invariably prevented the de- 
 velopment of the sea-going habit. 
 
466 PHYSIOGRAPHY 
 
 ISLANDS 
 
 As already indicated, many islands are really shore features, 
 being developed by the same agents and processes which develop 
 the horizontal configuration of coasts. 
 
 Like other natural features, islands may be classified in various 
 ways, and each classification brings out certain significant facts. 
 Thus, on the basis of size, they are large and small; on the basis 
 of height, they are high and low; on the basis of position, they are 
 continental and oceanic; and on the basis of fertility, they are 
 fertile and barren. Between the extremes of each of the above 
 groups there are all gradations. Other comparable bases of group- 
 ing may be suggested. The most significant classification, from the 
 physiographic point of view, is based on origin. Islands arise 
 through the processes of diastrophism, vulcanism, and gradation, 
 and, if the action of organisms be excluded from gradation, by 
 organic action. 
 
 1. By diastrophism. The rise of any portion of sea bottom 
 enough to cause it to emerge from the w r ater, gives rise to an island, 
 if the new land is not connected with a continent. Similarly, the 
 subsidence of the sea might cause the emergence of elevated por- 
 tions of sea bottom, giving rise to islands. Cuba and the other 
 large islands of the West Indies belong to this general class. 
 
 The rise of the sea-level might transform the elevations of a 
 coastal plain into islands, by submerging the surrounding land. 
 The same result might be brought about by the sinking of coastal 
 lands of strong relief. Great Britain was thus separated from the 
 mainland. Had it remained connected with the continent, the 
 course of European history would probably have been very dif- 
 ferent. 
 
 2. By vulcanism. Many submarine volcanoes have built up 
 their cones so that their tops emerge. Far from coasts, islands of 
 this sort are more common than any others. Volcanic islands 
 are, however, not confined to the deep sea. 
 
 3. By gradation, (a) By erosion. Islands arise both by aggra- 
 dational and degradational processes, and both aggradation and 
 degradation are effected by different agents. 
 
 Waves often so erode a coast as to isolate small areas of resistant 
 rock, converting them into islands (Fig. 501). 
 
ORIGIN AND HISTORY OF PHYSIOGRAPHIC FEATURES 467 
 
 Glaciers descending from the land to the sea may, by erosion, 
 isolate coastal promontories, converting them into islands. It is 
 probable that some of the islands on glaciated coasts arose in this 
 
 way. 
 
 Islands are sometimes formed in rivers by the erosion of the 
 stream (Fig. 502). The jointing of the rock seems often to afford 
 the conditions for the development of such islands. The erosive 
 action of the James River transformed the Jamestown peninsula 
 into an island toward the end of the 17th century, a thing the 
 
 FIG. 501. Finn Rock and Cape Blanco, Oregon. (U. S. Geol. Surv.) 
 
 colonists had planned to do for purposes of defence. River islands 
 are also sometimes developed through the meandering of the 
 streams (Fig. 197). 
 
 (b) By deposition. Islands arise by the deposition of sediment 
 along sea and lake shores and in rivers. Such islands are usually 
 low and sandy, and always near other land. The processes which 
 give rise to them have been indicated (p. 324). They are often 
 affected by dunes. Glacier deposits also give rise to islands, as in 
 Boston harbor. Islands which have cores of solid rock are often 
 enlarged by various processes of deposition. 
 
 4. By combinations of diastrophism, gradation, and vulcanism. 
 Many existing islands owe their origin and form to the com- 
 bination of two or more of the above agents. River or glacier 
 erosion often develops an uneven topography along shore, and a 
 
468 PHYSIOGRAPHY 
 
 slight subsidence of the coast, or a rise of the sea-level there, gives 
 rise to islands, because the land has been properly prepared in ad- 
 vance. It is to such a combination of degradation and diastroph- 
 ism that many of the islands of glaciated coasts, such as those of 
 Maine, Alaska, Norway, etc., are due. 
 
 Other combinations, too, of the several agents operative on 
 coasts may give rise to islands. Thus, an island which was pri 
 marily volcanic may be enlarged in area by the deposition of sedi- 
 
 FIG. 502. Lone Rock. An island in the Wisconsin River, isolated as an 
 island by the notable widening of a series of joints in the sandstone. 
 (Meyers.) 
 
 ment about it, the sediment being brought down from the higher 
 parts of the island. Iceland is an example. 
 
 Whatever their origin, most existing islands have been more 
 or less notably modified by erosion. 
 
 Island coasts are subject to all the changes which affect the 
 coasts of continents. Islands are subject to destruction by the 
 waves, on the one hand, and they may cease to be islands by 
 being attached to continents. Such connection may be brought 
 about by diastrophism or gradation (PI. XXII). Thus, a former 
 island may be tied to the mainland by deposition. After being 
 joined to the mainland the former island becomes a part of a 
 striking irregularity of the coast. 
 
 5. By organic processes. There are in some parts of the world 
 numerous islands composed of coral. The little animals (polyps) 
 
ORIGIN AND HISTORY OF PHYSIOGRAPHIC FEATURES 469 
 
 which secrete the coral live (1) where the water is 120 feet or less 
 in depth; (2) where the temperature never falls below about 68 
 F.; (3) where the water has the saltness of normal sea-water; (4) 
 where the water is nearly free from sediment; and (5) where it 
 is subject to some movement by the wind. In such situations 
 they thrive, and sometimes make reefs and sometimes islands. 
 
 Polyps are not free-moving animals, except in the early part of 
 their lives, before they begin coral-making. Through the larger 
 
 FIG. 503. Diagram of a fringing 
 reef. 
 
 FIG. 504. Diagram of a barrier 
 reef. 
 
 part of their lives they are attached to the bottom. They flourish 
 about many islands of volcanic origin and along some continental 
 coasts, as along the east coast of Australia. They also flourish 
 in some places far from islands or continents, if there is shallow 
 water of the right temperature. 
 
 Figs. 503 and 504 show coral reefs. Those which are far 
 enough from the land to leave a somewhat wide and deep lagoon 
 
 FIG. 505. Diagram suggesting the development of a barrier reef and an 
 atoll, successively, from a fringing reef by sinking. 1. Fringing reef, 
 formed in shallow water; 2, barrier reef, developed from fringing reef 
 after subsidence; 3, the atoll which succeeds the barrier reef. 
 
 inside are barrier reefs; those close to the land are fringing reefs. 
 li seems probable that fringing reefs sometimes become barrier 
 reefs by the sinking of the island or coast where they occur, as 
 illustrated by Fig. 505. The sinking should not proceed faster 
 than the polyps build up the reef. Barrier reefs, the bottoms of 
 which are in deep water, were formerly thought to prove sub- 
 sidence; but this conclusion is questioned. A reef in shallow water 
 may come to have a long outer slope, with its bottom in water far 
 below 120 feet, if coral be broken off from the upper part of the 
 
470 
 
 PHYSIOGRAPHY 
 
 reef and caused to descend the slope into deeper water. This 
 process is illustrated by Fig. 506. It is probable that barrier reefs 
 have been developed in both these ways. Coral reefs are usually 
 
 FIG. 506. Diagram suggesting the origin of a barrier reef without subsidence. 
 The reef starts in shallow water near shore. Material broken from it 
 falls down, making a sort ol talus slope, the lower part of the shaded 
 portion, and the polyps build out on this slope, but always remain in 
 shallow water (1, 2, 3, 4). The outer edge of the reef thus comes to be 
 in deep water. 
 
 interrupted where fresh water descends from the land, so that a 
 reef rarely surrounds an island, and is rarely continuous for great 
 stretches along any coast. 
 
 It is manifest that the barrier reef about a small island may 
 
 FIG. 507. An atoll. (From Dana's Corals and Coral Islands by permission 
 of Dodd, Mead & Co.) 
 
 become an island or atoll by subsidence. This is illustrated by 
 
 FIG. 508. Coral island developed from a submerged volcano (or other rock). 
 
 Fig. 505. Coral islands might also arise by the development of 
 reefs on volcanic cones which did not rise into islands (Fig. 508). 
 
ORIGIN AND HISTORY OF PHYSIOGRAPHIC FEATURES 471 
 
 The polyps do not build the reef or the atoll above water; but 
 when they have built up to water-level the waves may build it 
 higher, as they convert sand reefs into land. Once land appears, 
 the wind may make it higher by piling up coral sand. The growth 
 
 FIG. 509. Coral growing. 
 
 of vegetation may help along the building, both by its own growth 
 and by helping the lodgment of wind-blown sediment. 
 
 Coral islands and reefs would always remain low if it were not 
 for diastrophism. There are indeed no very high coral islands, 
 but there are coral reefs 2000 or 3000 feet above sea-level. Either 
 the land where such reefs occur has risen greatly, or the sea-level 
 has been depressed. 
 
 TOPOGRAPHIC MAP STUDIES. 
 
 The following maps illustrate types of plains, plateaus, and mountains. 
 
 See also the list of maps on pp. 54 and 479. 
 
 The folios (see foot-note, p. 79) of the areas shown on the maps 
 marked * are published, and are always serviceable in the interpreta- 
 tion of the topography. 
 
472 PHYSIOGRAPHY 
 
 A. PLAINS 
 
 1. Coastal 
 
 Deal Island, Md. Asbury Park, N. J. 
 
 Dennisville, N. J. Sandy Hook, N. J. 
 
 Cape May, N. J. * Coos Bay, Ore. 
 
 Atlantic City, N. J. * Norfolk, Va. 
 Great Egg Harbor, N. J. 
 
 2. Interior 
 a. Flat: 
 
 Maxwell, Cal. Chocowinity, N. C. 
 
 * Casselton, N. D. (Casselton-Fargo Bowling Green, 0. 
 folio). 
 
 6. Uneven and well-drained' 
 Petersburg, Ind. Sullivan, Mo. 
 
 * Ditney, Ind. Lancaster, Wis. la. IJ1. 
 Tuscumbia, Mo. 
 
 c. Uneven and ill-drained, glaciated: 
 Crystal Falls, Mich. Geneva, Wis. 
 
 White Bear, Minn. Madison, Wis. 
 
 Oswego, N. Y. (special map). Briggsville, Wis. 
 
 B. PLATEAUS 
 
 Chino, Ariz. * Canyon, Wyo. (Yellowstone Na- 
 
 Burnsville, W. Va. tional Park folio). 
 
 * Charleston, W. Va. 
 
 C. MOUNTAINS 
 
 * Marysville, Cal. Kaaterskill, N. Y. 
 San Mateo, Cal. Mt. Marcy, N. Y. 
 Mt. Lyell, Cal. Saluda, N. C. S. C. 
 Shasta, Cal. * Mt. Mitchell, N. C. Tenn. 
 
 * Tellunde, Colo. Millersburg, Pa. 
 Leadville, Colo. Lykens, Pa. 
 
 * Pikes Peak, Colo. * Greenville, Tenn. N. C. 
 Huerfano Park, Colo. * Morristown, Tenn. 
 
 * Holyoke, Mass. Conn. * Maynardville, Tenn. 
 Greylock, Mass. Vt. Henry Mountains, Utah. 
 Saypo, Mont. * Tmtic, Utah. 
 Hamilton, Mont. Ida. Tooele Valley, Utah. 
 Sumpter, Ore. * Ellensburg, Wash. 
 
 Ishawoot, Wyo. 
 
 The types of plains are indicated in the above classification, and 
 their distinctive features should be observed. The plateaus show 
 various degrees of dissection. 
 
ORIGIN AND HISTORY OF PHYSIOGRAPHIC FEATURES 473 
 
 The mountain maps should be studied with a view to distinguishing 
 topographic types (pp 435-437). An attempt should also be made to 
 group them according to their origin (p. 445). The results of this 
 latter classification should be tested by the Structural Sheets of the folios 
 so far as possible. 
 
 MAPS FOR REVIEW 
 
 I. List of Maps. 
 
 1. Fire Island, N. Y. 7. Glacier Peak, Wash. 
 
 2. Marsh Pass, Ariz. 8. Frostburg, Md. 
 
 3. Princess Anne, Md. Va. 9. Stoughton, Wis. 
 
 4. Abilene, Tex. 10. Marseilles, 111. 
 
 5. Hahnville, La. 11. Mt. Taylor, N. M. 
 
 6. Gay Head, Mass. 12. Savanna, la. 111. 
 
 These maps touch most of the topics studied in preceding pages. 
 
 II. Questions to be answered in writing. In the case of each map 
 
 1. State whether the area is plain, plateau, or mountain, or if 
 more than one of these great types appears, the location of each. 
 
 2. Name the several agencies which have shaped the surface, and 
 indicate their relative importance. 
 
 3. State age of the topography in terms of erosion. If different 
 parts of the area are in different stages, bring out this point. 
 
 4. How many cycles of erosion are shown? The evidence for the 
 conclusion stated. Is it conclusive? 
 
 5. Is there any indication as to the position of the strata under- 
 lying the region? 
 
 6. What indications are there of inequalities of hardness of rock? 
 
 7. State what inferences may be made (certain, probable, possible) 
 concerning the climate, and the evidence on which they are based. 
 
 8. Note any important features not brought out by the preceding 
 questions. 
 
 REFERENCES 
 
 1. Standard text-books on Geology. 
 
 2. Folios of the U. S. Geol. Surv., of areas in mountain regions. 
 
 3. WILLIS, Mechanics of Appalachian Structure: 13th Ann. Rept. U. S. 
 Geol. Surv., Pt. II, pp. 217-283. 
 
 4. LE CONTE, On the Structure and Origin of Mountains, etc.: Am. Jour. 
 Sci., Vol. XXXVIII, 1889, pp. 257-263; Theories of the Origin of Mountain 
 Ranges: Jour. Geol., Vol. I, pp. 543-573. 
 
 5. DANA, On the Origin of Mountains: Am. Jour. Sci., Vol. V, 1873, pp. 
 347, 423, and 474, and Vol. VI, pp. 6, 104, 161, 304, and 381. 
 
474 PHYSIOGRAPHY 
 
 6. POWELL, Types of Orographic Structure: Am. Jour. Sci., Vol. XII t 
 1876, pp. 414-428. 
 
 7. TARR, Mountains of New York, in Physical Geography of New York 
 State. 
 
 8. READE (T. MELLARD), The Origin of Mountain Ranges. 
 
 9. GULLIVER, Shoreline Topography: Am. Acad. Arts and Sci., Vol. 
 XXXIV. 
 
 10. DARWIN, The Structure and Distribution of Coral Islands: Appleton. 
 
 11. DANA, Corals and Coral Islands: Dodd, Mead & Co. 
 
 12. AGASSIZ, Various Papers in Bull. Mus. Comp. Zool., Harvard. 
 
 13. HEILPRIN, The Bermuda Islands: Appleton. 
 
 14. Various Papers on coral islands in Nature, Vols. 22, 35, 37, 39, 
 40, 41, 42, 51, and 55. 
 
CHAPTER X 
 TERRESTRIAL MAGNETISM 
 
 THE earth is a great magnet and, like the small magnet with 
 which we are familiar, has two poles. One of these poles is called 
 the North Magnetic Pole and the other the South Magnetic Pole. 
 Generally speaking, one end of the magnetic or compass needle 
 points toward one of these poles, and the other toward the other. 
 If we were to follow the directions pointed by the compass needle, 
 we would be led to the North Magnetic Pole in the one case, and to 
 the South Magnetic Pole in the other. The lines connecting these 
 poles are magnetic meridians (Fig. 510). 
 
 The North Magnetic Pole is in latitude a little above 70, and 
 in longitude about 97 W., as nearly as known. The South Mag- 
 netic Pole is in latitude about 72, and in longitude about 152 E. 
 These positions have been calculated from the directions in which 
 the compass needles point in various positions in the vicinity of the 
 magnetic poles. The South Magnetic Pole has never been reached, 
 and Captain Amundsen reports that "the North Magnetic Pole 
 has no immediate situation," which probably means that it is not 
 a fixed point. 
 
 It will be seen from the foregoing that the magnetic poles are 
 far from the geographic poles, and that they are not exactly oppo- 
 site each other. It is believed, too, that they are not quite constant 
 in position, though they are not known to wander widely. The 
 North Magnetic Pole has been thought to have shifted its position 
 some 50 or 60 miles in as many years, following 1830, though this 
 determination does not appear to be conclusive. 
 
 Since the north end of the magnetic needle points to the North 
 Magnetic Pole, it follows that the compass does not indicate true 
 north and south in many places. At points northward from the 
 North Magnetic Pole, the "north" end of the needle points 
 in a southerly direction. At points to the south it points to 
 
 475 
 
476 
 
 PHYSIOGRAPHY 
 
TERRESTRIAL MAGNETISM 
 
 477 
 
 the northward, at points east, westward, and at points west, 
 eastward. The departure of the needle from the true north and 
 south is magnetic declination. Lines connecting places of equal 
 declination are isogonic lines. A line connecting places of no 
 declination is an agonic line. 
 
 Fig. 511 shows an agonic line in the United States passing from 
 Lake Superior to South Carolina. Along this line the magnetic 
 needle points due north and south. All places east of this line 
 
 FIG. 511. Isogonic lines for the United States, 1902. The heavy line is an 
 agonic line, or line of no declination. (U. S. Coast and Geodetic Surv.) 
 
 have west declination, and all places west of this line have east 
 declination. In general, declination increases with increasing dis- 
 tance from the agonic line. In Maine, for example, the declination 
 is more than 20 W. at a maximum, and in Washington more than 
 20 E. (Fig. 511). At Chicago the declination is about 3 E.; 
 at New York nearly 10 W.; at Denver about 13 E.; and at San 
 Francisco about 16 E. It will be seen that it is important to 
 know the magnetic declination of a region, if the compass is to be 
 used there for determining directions. 
 
 The declinations shown in Figs. 511 and 512 are interfered 
 with locally by certain rock formations, especially magnetic iron 
 ore. In the vicinity of such ore, especially if it be in large bodies, 
 the needle may depart widely from the declination indicated by 
 these lines. 
 
478 
 
 PHYSIOGRAPHY 
 
 
 1C 
 
 
 
 s 
 1 
 
 I 
 
 
TERRESTRIAL MAGNETISM 479 
 
 Since the magnetic poles shift slowly, the declination at any place 
 also shifts in harmony. It is not certain, however, that all varia- 
 tions in magnetic declination are due to the shifting of the mag- 
 netic pole. The declination at Chicago has shifted more than 2 
 since 1820. 
 
 Dip. The magnetic needle does not usually take a horizontal 
 position. At the magnetic poles it should be vertical, and the 
 north end would be down at the North Magnetic Pole. Its position 
 would be reversed at the South Magnetic Pole. Half-way between 
 the magnetic poles, that is, at the magnetic equator, the needle 
 should be horizontal. A compass which is constructed so as to 
 show the "dip or magnetic inclination is a dip compass. 
 
 Intensity. Magnetic intensity varies greatly from place to 
 place, and slightly from time to time in the same place. 
 
 The causes and the conditions of change of terrestrial magnetism 
 are not well understood. 
 
 SUPPLEMENTARY LIST OF MAPS 
 I. Topographic Maps, U. S. Geological Survey 
 
 The following maps will afford opportunity for more extended map 
 study. The use of as many of them as time permits will be profitable. 
 These, together with those already mentioned in preceding pages, make 
 a fairly adequate equipment so far as topographic maps are concerned. 
 The maps mentioned in this volume should be supplemented by those 
 of the home region, and the maps of the home region should be used 
 in the field, as much as possible. In no other way will the maps be so 
 well understood. For the areas marked *, folios have been published 
 (see foot-note, p. 79), and they are helpful in the study of topography. 
 See especially the structure-section sheets. 
 
 CHAPTER I 
 
 Batesville, Ark. Minneapolis, Minn. 
 
 Tamalpais, Cal. Hamilton, Mont.-Ida. 
 
 Tipton, la. Dennisville, N. J. 
 
 * Cottonwood Falls, Kan. Everett, Pa. 
 Montross, Md.-Va. * Gaines, Pa.-N. Y. 
 Frostburg, Md.-W. Va.-Pa. 
 
 CHAPTER II 
 
 Lakin, Kan. Browns Creek, Neb. 
 
 Oceanside, Md.-Del. 
 
 CHAPTER III 
 
 * London, Ky. * Kingston, Tenn. 
 
480 PHYSIOGRAPHY 
 
 CHAPTER IV 
 
 A. River erosion, especially river valleys. 
 
 Morrilton, Ark. Oberlin, 0. 
 
 Dunlap, 111. * Wartburg, Tenn. 
 
 New Harmony, Ind.-Ill. Anson, Tex. 
 
 Medicine Lodge, Kan. Abajo, Utah-Colo. 
 
 Tell City, Ky.-Ind. * Monterey, Va.-W. Va. 
 
 Palmyra, Mo. St. Croix Dalles, Wis.-Minn. 
 
 Oak Orchard, N. Y. * Gallatin, Wyo. (Yellowstone Na- 
 
 * Mt. Mitchell, N. C.-Tenn. tional Park folio). 
 
 Parmelee, N. C. * Shoshone, Wyo. (Yellowstone Na- 
 
 * Fargo, N. Dak. (Casselton- tional Park folio). 
 
 Fargo folio). 
 
 B. Topographic effects of unequal hardness after notable erosion. 
 
 Denver, Colo. High Bridge, N. J. 
 
 * Rome, Ga.-Ala. Passaic, N. J. 
 
 * Holyoke, Mass.-Conn. Hollidaysburg, Pa. 
 Saypo, Mont. * Uvalde, Tex. 
 
 C. Piracy and adjustment. 
 
 * Stevenson, Ala.-Ga.-Tenn. * Franklin, W. Va.-Va. 
 
 * Piedmont, Md.-W. Va. * Lake, Wyo. (Yellowstone Na- 
 
 * Chattanooga, Tenn. tional Park folio). 
 
 * Ringgold, Tenn.-Ga. 
 
 D. Attuviation. 
 
 Morrilton, Ark. * Three Forks, Mont. 
 
 Sierraville, Cal. Lexington, Neb. 
 
 Hartford, Conn. Paxton, Neb. 
 
 Camas Prairie, Ida. Silver Peak, Nev.-Cal. 
 
 Mountain Home, Ida. Cohoes, N. Y. 
 
 Waukon, la.-Wis. Watkins, N. Y. 
 
 East Delta, La. Williamston, N. C. 
 
 Hahnville, La. Portland, Ore.-Wash. 
 Independence, Mo. 
 
 See also River Charts (IV, below). 
 
 E. Cycles of erosion. 
 
 Echo Cliffs, Ariz. Everett, Pa. 
 
 San Francisco Mt., Ariz. Harrisburg, Pa. 
 
 Tusayan, Ariz. Huntingdon, Pa. 
 
 Batesville, Ark. Delaware Water Gap, Pa.-N. J. 
 
 Marshall, Ark. Pala Pinto, Tex. 
 
 Mountain View, Ark. Wausau, Wis. 
 
 * Ditney, Ind. Winchester, W. Va.-Va. 
 Watrous, N. M. 
 
 CHAPTER V 
 
 * Colfax, Cal. Deer Isle, Me. 
 
 Durant, la. Chief Mountain, Mont. 
 
 Clinton, la.-Ill. Hamburg, N. J. 
 
 Boothbay, Me. Plainfield, N. J.-N. Y. 
 
SUPPLEMENTARY MAPS 481 
 
 CHAPTER V Continued 
 
 Greenwood Lake, N. Y.-N. J. Weedsport, N. Y. 
 
 Elmira, N. Y.-Pa. Tower, N. Dak. 
 
 Little Falls, N. Y. Pingree, N. Dak. 
 
 Hammondsport, N. Y. Masontown, Pa. (Masontown- 
 
 Harlem, N. Y.-N. J. Uniontown folio). 
 
 Niagara Falls, N. Y. Methow, Wash. 
 
 Tonawanda, N. Y. Chelan, Wash. 
 
 Skaneateles, N. Y. Snoqualmie, Wash. 
 
 Penn Yan, N. Y. Delavan, Wis. 
 
 Tully, N. Y. Briggsville, Wis. 
 
 Rosendale, N. Y. Baraboo, Wis. 
 
 Rochester, N. Y. Denzer, Wis. 
 
 Syracuse, N. Y. The Dells, Wis. 
 
 CHAPTER VI 
 Group K. 
 
 Cayucos, Cal. Northport, N. Y. 
 
 Haywards, Cal. Babylon, N. Y. 
 
 Hueneme, Cal. Fire Island, N. Y. 
 
 San Francisco, Cal. Hamlin, N. Y. 
 
 Oceanside, Cal. Euclid, 0. 
 
 Biddeford, Me. * Port Orford, Ore. 
 
 Tolchester, Md. Erie, Pa 
 
 Muskegat, Mass. 
 
 CHAPTER VII 
 
 San Francisco Mt., Ariz. Crater Lake (special), Ore. 
 
 * Lassen Peak, Cal. Terlingua (special) Tex 
 
 Mt. Lyell, Cal. Abajo, Utah-Colo 
 
 Mt. Taylor, N. M. Henry Mts., Utah. 
 
 II. Coast and Geodetic Survey Charts l 
 
 Charts bearing the numbers 8, 10, 19, 21, 103, 105, 109, 110, 120, 
 122, 123, 124-126, 131-136, 146, 156, 157, 161, 167-169, 177, 184, 204, 
 1000, 1001, 1002, 1007, 5100, 5106, 5143, 5200, 5500, 5581, 6300, 6450, 
 6460, 8100, 8300, 9302, S, and T. 
 
 III. Lake Survey Charts l 
 
 The general charts of Lakes Superior, Michigan, Huron, St. Clair, 
 Erie, and Ontario. Charts of most parts of the shores of these lakes, 
 on a much larger scale, are also published. 
 
 IV. River Charts 
 
 Charts 9, 13, 14, 18, 19, 20, and 27, and Index charts I, II, and III 
 of the Mississippi River, issued by the Mississippi River Commission. 3 
 Students interested in any special portion of the Mississippi River will 
 do well to get the charts for those regions. Similar charts are published 
 for certain other large rivers, such as the Missouri, the Tennessee, etc. 
 
 1 See foot-note, p. 203. 
 
 1 Issued by the War Department, Washington, D. C. 
 8 These maps may be purchased of the Mississippi River Commission, 
 St. Louis, Mo. 
 
PART II 
 
 CHAPTER XI 
 EARTH RELATIONS 
 
 Form. The form of the earth is very much like that of a 
 sphere, but, since it is not exactly a sphere, it is generally said 
 to be a spheroid. The form has been determined in various ways: 
 (1) Ships have sailed quite around it. This proves that it is 
 everywhere bounded by curved surfaces, though it does not prove 
 that it is a sphere or even a spheroid, for, if it had the shape of 
 an egg, it would be possible to sail around it. (2) It has been 
 found that when vessels go to sea their lower parts disappear 
 first. When a vessel has gone four miles, the lower five feet of 
 its hull is out of sight to an observer on the shore, if his eye is 
 five feet above the level of the sea. Similarly, when a vessel 
 approaches land, its highest parts are seen first by observers on 
 the land, while to observers on the vessel the high lands are seen 
 first and the low ones later. From the vessel the spires and 
 chimneys of houses appear before the roofs, and the roofs before 
 the lower parts. These phenomena show only that the earth 
 has a curved surface; but it is found that in whatever direction 
 vessels sail, and from whatever port they start, objects on land 
 disappear at about the same rate. This means that the curva- 
 ture is nearly the same in all directions. A body whose curva- 
 ture is the same in all directions is a sphere, and a body whose 
 curvature is nearly the same in all directions is nearly a sphere. 
 This is the condition of the earth. (3) Again, the earth some- 
 times gets directly between the sun and the moon. It then casts 
 a shadow on the moon, and this shadow always appears to be 
 circular, though its edges are not very clearly defined. (4) The 
 direction of the plumb-line (the perpendicular to a horizontal surface 
 on the earth) changes from point to point on the earth's surface, 
 and it changes by an angle which is almost exactly proportional 
 
 482 
 
EARTH RELATIONS 
 
 483 
 
 to the distance between the points, wherever they are. If the 
 change of direction were exactly proportional to the distance 
 between two points, wherever taken, the earth would be a sphere 
 (Fig. 513). Since it is only approximately true, the earth is only 
 approximately spherical. 
 
 This point may be put in another way. The stars are very far 
 from the earth. As one travels along the earth's surface, the 
 
 Fio. 513. The circle represents the earth's circumference. The extensions 
 of the radii represent the directions of the plumb-lines at various posi- 
 tions. The distance from a to b is the same as that from b to c and 
 c to d, and the change in the direction of the plumb-line, that is, the 
 angle between aa' and 66', is essentially the same as that between 66' and 
 cc', cc' and dd', etc. This is true for all parts of the earth. 
 
 apparent directions of the stars change, and the angle of change 
 is almost exactly proportional to the distance traveled, wherever 
 the starting-point, and whatever the direction of travel. 
 
 The significance of this change in the position of the stars 
 appears to have been correctly interpreted, in general terms at 
 least, by certain Greek students (e.g., Thales of Miletus) as early 
 as 640 B.C. The same idea appears to have been entertained 
 at various subsequent times by individual students. Columbus 
 recognized it in the statement: "I have always read that the 
 
484 PHYSIOGRAPHY 
 
 world, comprising the land and the water, is spherical, as testified 
 by the investigations of Ptolemy and others, who have proved 
 it by the eclipses of the moon and other observations made from 
 east to west, as well as by the elevation of the pole [pole star] from 
 north to south." 1 
 
 In these and other ways 2 it is known that the form of the 
 earth does not depart greatly from that of a sphere. 
 
 Size. The circumference of the earth is nearly 25,000 miles, 
 and its diameter nearly 8000 miles. Since the earth is not a 
 perfect sphere, its various diameters and circumferences are not 
 exactly equal. Its longest diameter is 7926.5 miles, and its 
 shortest nearly 27 miles less (7899.7 miles). The shortest cir- 
 cumference is about 42 miles shorter than the longest. 
 
 The surface area of the earth is nearly 197,000,000 square miles, 
 and its volume, exclusive of the atmosphere, about 260,000,000,000 
 cubic miles. The earth is between five and six times as heavy as 
 an equal volume of water would be. 
 
 Motions 
 
 The earth has two principal motions. These are (1) rotation, 
 and (2) revolution around the sun. The earth rotates on its 
 shortest diameter, which is called its axis. The ends of the axis 
 of rotation are the poles; and the circumference midway between 
 the poles is the equator. The equator is the longest circumfer- 
 ence of the earth. The lines that pass from pole to pole on the 
 earth's surface are meridians. All meridians converge at each 
 pole. Meridians are parallel with one another at the equator, 
 but nowhere else. 
 
 Rotation. The rotation of the earth may be demonstrated by 
 simple experiments. 1. If a body be dropped from a high tower, 
 it does not fall so as to reach a point immediately beneath that 
 from which it fell. Instead, it always falls a little to the east of 
 the point from which it started. This is explained as follows: 
 If the earth rotates, any point must move faster than any other 
 point which is nearer its center, for the same reason that a point 
 on the rim of a wheel moves faster than a point between the rim 
 and the hub. If the earth be rotating, the top of a tower must 
 
 1 Hakluyt Soc. Pub., History of Columbus's Third Voyage, Vol. II., p. 129. 
 'See Moulton's Introduction to Astronomy, pp. 114-124. 
 
EARTH RELATIONS 
 
 485 
 
 be moving forward faster than the bottom. In this case, the 
 falling body, starting from the top of the tower, has a forward 
 velocity greater than that possessed by the base of the tower. 
 Under these circumstances, the falling body must gain on the 
 base of the tower in the direction of rotation; that is, if the earth 
 rotates to the east, the falling body would be farther to the east, 
 relative to the tower, when it reached the ground than when 
 it started. If the earth rotated to the west, the body would 
 fall the other way. Since the body always falls to the east, and 
 since nothing but the rotation of the earth to the east seems to 
 explain this fact, it is taken to be a proof that the earth rotates 
 
 FIG. 514. The leaning tower of Pisa, 
 where some of Galileo's famous 
 experiments on falling bodies were 
 performed. 
 
 FIG. 515. Figure to illustrate the 
 effect of rotation on a falling 
 body as explained in text. 
 
 in that direction. The actual deviation in our latitude is about 
 one inch for 500 feet of fall. 
 
 Fig. 515 illustrates the principle involved in the falling body. 
 Let AB = the earth's radius, and ra a point on a tower (height 
 greatly exaggerated) above the earth's surface. Suppose the 
 mass m is dropped from the top. If the earth were not rotat- 
 ing, it would fall in the direction of the plumb-line, and would 
 strike the surface at B. Suppose, however, the earth is rotating 
 at such a rate that BA turns to B'A while m is falling to the 
 surface. If it were not for the attraction of the earth, m would 
 go in a straight line to m'. Gravitative attraction is at right 
 angles to this line mm', and, though it does not change the 
 amount of motion of m in this direction, it impresses upon it a 
 
486 
 
 PHYSIOGRAPHY 
 
 new motion toward the earth. The result is that it describes the 
 curved line mR, and strikes the earth at R, a little beyond the 
 foot of the perpendicular m'B'. 1 
 
 2. Another experiment, which shows the same thing, may be 
 performed with a pendulum (known as Foucaidt's pendulum). 
 If a pendulum attached to a ceiling is set swinging parallel to 
 a given line on the earth's surface, as, for example, parallel to a 
 line on the floor, it will be found a little later to be swinging in 
 a plane which is not parallel to the original line. The pendulum 
 changes its direction, with reference to the line along which it 
 
 FIG. 516. Diagram to illustrate the fact that the direction of the swing of 
 the pendulum changes more rapidly in high latitudes than in low lati- 
 tudes. A pendulum set swinging with the central meridian of the dia- 
 gram, in different latitudes, will depart from the meridians, as shown at 
 the right, in six hours. There is no departure at the equator, much in 
 middle latitudes, and still more in high latitudes. 
 
 was started, more rapidly near the poles and less rapidly near 
 the equator. If it could be set swinging along a meridian so that 
 one end of the swing barely reached the pole, it would be found 
 that the pendulum was swinging at right angles to that meridian 
 after the earth had turned a quarter of the way around. This is 
 illustrated by Fig. 516. If the pendulum were set swinging half- 
 way between one of the poles and the equator, it would have 
 departed from the plane in which it was started much less when 
 the earth had turned a quarter of the way around. This is also 
 
 1 Moulton's Introduction to Astronomy, pp. 148-149. 
 
EARTH RELATIONS 487 
 
 illustrated by Fig. 516. If the pendulum were set swinging at 
 the equator so that half the swing was on either side of it, the 
 swing would remain parallel with its original position (Fig. 516). 
 
 The departure of the pendulum, except at the equator, from 
 the plane of the meridian in which it was set swinging is often 
 said to mean that the meridian in the plane of which it first swung 
 has changed its position, and in its new position it is not parallel 
 to the position in which the pendulum started to swing. The 
 pendulum itself continues to swing in its original plane, but this 
 plane is no longer parallel to the meridian in its changed position. 
 According to this statement, the pendulum seems to have changed 
 its direction, because we determine direction by meridians, and 
 the successive positions of a meridian on a spherical rotating 
 body do not remain parallel with one another, except at the 
 equator of the rotating body. 
 
 This change in the direction of the pendulum, which is uni- 
 versal except at the equator, is always in the same direction in the 
 northern hemisphere, and always in the same direction in the 
 southern hemisphere, and proves that the earth rotates. The 
 change in the direction of the pendulum does not take place at 
 the equator, so the explanation runs, because the meridians there 
 are all parallel with one another and the successive positions of a 
 given meridian therefore are all parallel. According to this state- 
 ment of the case, the apparent change in the direction of swing of 
 the pendulum takes place less rapidly midway between the equator 
 and the poles than near the poles (Fig. 516), because the meridians 
 are more nearly parallel with one another in the former position 
 than in the latter. 
 
 If this were the full explanation of the matter, the swing of the 
 pendulum should always be parallel to its original position at 
 the end of 24 hours, whether the pendulum was near the equator 
 or near the pole. This is not the case, and the above statement 
 is therefore not an adequate explanation of the phenomenon. 
 Though the dependence of the rate of variation of the direction 
 of the pendulum's swing on latitude cannot be given here, it is 
 well understood. 
 
 The form of the earth is consistent with its rotation, but can 
 hardly be said to prove it. Any body which is not perfectly 
 rigid (and no body is) would be somewhat flattened at its poles, 
 and somewhat bulged at its equator, by rotating. This is the 
 
488 PHYSIOGRAPHY 
 
 condition of the earth, for the diameter between the poles is the 
 shortest diameter, and the diameters in the plane of the equator 
 are the longest. The amount of flattening which would result 
 from rotation depends on (1) the rate of rotation, and (2) the rigid- 
 ity of the body. The faster the rotation and the less rigid the 
 body, the greater the polar flattening. There are other ways of 
 proving that the earth rotates, but they need not be cited here. 
 
 The rate at which a point on the surface of the earth moves, as 
 a result of rotation, varies greatly. Points on the equator move 
 fastest, because they have farthest to go in the time of one com- 
 plete rotation. At the equator, where the circumference is nearly 
 25,000 miles, a point moves nearly 25,000 miles a day, as a re- 
 sult of rotation. Half-way between the equator and either pole, 
 a point moves about 17,600 miles per day, while at the poles the 
 rate of motion resulting from rotation is zero. 
 
 Effect of rotation. The most obvious effect of rotation is 
 the alternation of day and night, for one side of the earth and then 
 the other is turned toward the sun during each rotation. But it is 
 to be noted that the alternation of day and night does not of 
 itself prove rotation. Day and night might be brought about 
 equally well by the revolution of the sun around the earth each 
 day. The period of rotation, 24 hours, determines the length of a 
 day (day and night). 
 
 Revolution. The second principal motion of the earth is its 
 revolution about the sun. No simple experiment can be cited to 
 prove this motion; but the fact of revolution may be illustrated 
 in various ways. 
 
 If the positions of individual stars be observed for long periods 
 of time, they appear to describe small circuits each year. Some 
 of the circuits are nearly circular and some are nearly straight lines. 
 Some of them are larger and some smaller. This annual change in 
 the apparent position of the stars is their annual parallax. Either 
 the stars make this annual circuit, and all of them in the same 
 length of time, or the earth makes a yearly circuit in space, which 
 causes the apparent annual movement of the stars. The fact that 
 these apparent circuits of the stars are all made in the same length 
 of time makes it more probable that they are due to the motion 
 of the earth, than that they are due to the individual motions of 
 the stars themselves. The varying sizes of the apparent annual 
 paths of the stars is accounted for by the fact that some of them 
 
EARTH RELATIONS 
 
 489 
 
 are nearer to the earth than others, and the nearer they are, the 
 larger the annual circuits they appear to describe. The varying 
 shapes of the annual paths would be accounted for by the direc- 
 tions of the stars, some being in a polar direction from the ob- 
 server and some in an equatorial direction. 
 
 Various other physical and astronomical phenomena, which need 
 not be cited here, also demonstrate that the earth makes an annual 
 circuit around the sun. 
 
 The length of time which the earth requires to make its revolu- 
 tion about the sun determines the length of the year. It is a little 
 more than 365 days. 
 
 The path of the earth around the sun is its orbit. The orbit of 
 the earth is not a circle, but an ellipse (Fig. 517), and the sun is in 
 one of the foci, more than 1,500,000 miles from the center of the 
 ellipse. When the earth is nearest the sun, the distance between 
 
 FIG. 517. The orbit 
 of the earth is an 
 ellipse, with the sun 
 in one of the foci. 
 Eccentricity of the 
 orbit exaggerated 
 10 times; diameter 
 of the sun exagger- 
 ated nearly 10 
 times; diameter of 
 the earth exagger- 
 ated about 50 times. 
 
 the earth and sun is more than 3,000,000 miles less than when they 
 are farthest apart. It so happens that the earth is nearest (about 
 91,500,000 miles) the sun in the early winter (early in January) 
 of the northern hemisphere, and farthest (about 94,500,000 miles) 
 from it in early summer (early in July). The perihelion (nearest 
 the sun) and aphelion (farthest from the sun) dates are subject to 
 slow periodic change. The perihelion date in 4000 B.C. was Sep- 
 tember 21. It will be March 21 in 6590 A.D. 
 
 The motion of the earth through space during its revolution 
 about the sun is at the rate of about 600,000,000 miles a year. 
 This means that the earth travels about 1,600,000 miles daily, or 
 about 66,666 miles hourly. 
 
 The earth's axis is inclined toward the plane of its orbit about 
 23^ (Fig. 518). This position of the axis, together with the 
 
 Aphelion 
 
490 PHYSIOGRAPHY 
 
 motions of the earth, have much to do with the distribution of the 
 heat and light received from the sun, and so with the changes in 
 the length of day (daylight) and night (darkness), and with the 
 succession of the seasons. But, before attempting to see how 
 
 FIG. 518. Diagram to show the effect of the inclination of the earth's axis 
 upon the distribution of light, heat, etc., on the earth. The line-shading 
 represents the plane of the earth's orbit. Half the earth is above this 
 plane, but the plane does not cut the earth symmetrically with refer- 
 ence to the parallels. In the position E t more than half the northern 
 hemisphere is being heated and lighted. In position E, less than half 
 of the same hemisphere is heated and lighted. In positions E* and Z? 4 
 the half of the northern ani of the southern hemispheres is being lighted 
 and heated. 
 
 these changes are brought about, we must become familiar with 
 certain terms which are to be used in the discussion of these changes. 
 
 Latitude, Longitude, and Time 
 
 Latitude. The equator has been defined as the circle about 
 the earth midway between the poles. Circles parallel to the 
 equator are parallels. The number of parallels which might be 
 drawn is infinite, though but a few are represented on maps. On 
 maps of small scale parallels are drawn every 5 or 10. On 
 maps of large scale they are drawn for every 1 or 2, or sometimes. 
 even for fractions of a degree. The length of parallels varies greatly, 
 those near the equator being longer, and those near the poles 
 shorter. 
 
 The planes of all parallels are perpendicular to the earth's 
 axis, but no circle perpendicular to the axis, except the equator, 
 is a great circle, for no other passes through the ends of a diameter 
 of the earth. This is shown in Fig. 519, which represents the 
 earth in two positions. In the left-hand part, the half of each 
 parallel and meridian represented is shown. In the right-hand 
 part, the relation of parallels to the North Pole is shown. The 
 
EARTH RELATIONS 491 
 
 distance between the equator and either pole is a quadrant (i.e., a 
 quarter of a circle) and is divisible into 90 parts (90) called 
 degrees. The degrees are numbered from the equator to the poles. 
 Each degree is divided into 60 parts (600 called minutes, and the 
 minutes, like the degrees, are numbered from the equator toward the 
 poles. Each minute is divided into 60 parts (60") called seconds, 
 and the seconds are numbered in the same direction as the larger 
 divisions. Distance north or south of the equator may there- 
 fore be indicated exactly by means of parallels. This distance 
 is called latitude, the latitude of the equator being 0. 
 
 In reality, geographic latitude, as distinct from astronomic latitude and 
 geodetic latitude, is the angle between the plane of the equator, and the 
 perpendicular to the standard spheroid at the place of observation. The 
 angle is measured by the arc at the surface, and the length of the arc is 
 commonly called the latitude. 
 
 If the latitude of a place is 40 40' 40" N., its distance and its 
 direction from the equator are accurately known; but, since the 
 parallel of 40 40' 40" runs quite around the earth, it is clear that 
 the statement of the latitude of a place indicates only what paral- 
 lel it is on, but not its position on that parallel. 
 
 Longitude. Position on a parallel is indicated by means of 
 meridians (p. 484). The number of possible meridians is infinite, 
 but, as in the case of parallels, only a few are commonly indicated 
 on maps. One meridian, that passing through Greenwich, Eng- 
 land, was long ago arbitrarily chosen as the meridian from which 
 distances east and west are to be reckoned. This meridian is the 
 meridian of zero degrees (0) . Distance east or west of this meridian 
 is known as longitude. Places east of long. are in east longitude, 
 and those west of it are in west longitude. East and west longitude 
 respectively are regarded as extending 180 from the meridian 
 0; that is, half-way around the earth. The degrees of longitude 
 are divided into minutes and seconds, the same as the degrees of 
 latitude. 
 
 The position of a place on the earth's surface may be absolutely 
 fixed by means of meridians and parallels. If a place is in longi- 
 tude 30 E., its distance east of the meridian is known. If, at 
 the same time, it is in latitude 30 N., it must be where the parallel 
 of 30 N. crosses the meridian of 30 E. This gives its position 
 on the earth's surface exactly. 
 
492 
 
 PHYSIOGRAPHY L 
 
 Every meridian reaches each pole. It might seem therefore 
 that each pole has all longitude. But longitude is distance east 
 or west of the meridian 0, and at the poles there is neither east 
 nor west. At the north pole the only direction is south, and at 
 the south pole the only direction is north. The poles therefore 
 cannot be said to have longitude, since they are not east or west 
 of the meridian of 0. 
 
 Longitude and time. There is a definite relation between 
 longitude and time. Since the earth turns through 360 in 24 
 
 FIG. 519. Parallels and meridians. 
 
 hours, it turns 15 in one hour, or 15' of longitude in one minute 
 of time. The sun therefore rises one hour earlier at a place in 
 longitude than at a place in the same latitude in longitude 
 15 W., and one hour later than at a place in the same latitude 
 in longitude 15 E. Similarly, noon comes an hour earlier in longi- 
 tude than in longitude 15 W. and an hour later than in longi- 
 tude 15 E. All places on a given meridian have noon and mid- 
 right at the same time, and such places are said to have the same 
 time; but places on different meridians have different times. 
 Thus, when places on the meridian of Chicago have noon, it is 
 afternoon on meridians farther east, and before r.oon on meridians 
 farther west. If the longitude of two places is known therefore, 
 their difference of time may be readily calculated. Fig. 520 repre- 
 sents three cities on or near the parallel of 40, and about 15 apart 
 in longitude. On June 21 the sun would be 73^ above the horizon 
 at noon in latitude 40. Fig. 520 may be taken to represent noon 
 at Philadelphia. At this hour the sun is not so high above the 
 horizon at St. Louis, which is 15 farther west, and it is still lower 
 at Denver, which is 15 farther west than St. Louis. After the 
 earth has turned 15, the sun will be 73 above the horizon at 
 St. Louis, while it will have become lower at Philadelphia and 
 
EARTH RELATIONS 493 
 
 higher at Denver. When the' earth has turned 15 more (i.e., an 
 hour later) the sun will be 33$ above the horizon at Denver. It 
 will then be noon at Denver; ah hour past noon at St. Louis, and 
 two hours past noon at Philadelphia. 
 
 Though all places on a given meridian have noon and midnight 
 at the same time, they do not al ways have sunrise and sunset at 
 the same hour, for reasons which will appear later. 
 
 The variations of time with changes of longitude become appar- 
 ent when long journeys are made either east or west. Thus a 
 
 FIG. 520. Diagram to illustrate the change in the altitude of the sun from 
 hour to hour, in places in the same latitude. The diagram represents 
 noon at Philadelphia at the time of the summer solstice. At this time 
 the sun is there but a few degrees from the zenith, as represented by the 
 dotted line. At St. Louis, in about the same latitude, but farther west, 
 the sun is much farther from the zenith at the same hour; but when 
 the noon hour arrives at St. Louis the sun will be as near the zenith 
 there as it is at Philadelphia in the diagram. At Denver, which is still 
 farther west than St. Louis, the sun is farther from the zenith than at 
 St. Louis at the noon hour of Philadelphia.. When it is noon at St. 
 Louis the sun will be as far from the zenith at Denver as it is in the 
 diagram at St. Louis. At this hour the sun will be about equally distant 
 from the zenith at Denver and at Philadelphia, but at Philadelphia it 
 will be west of south and at Denver east of south. When it is noon at 
 Denver the sun will be as near the zenith there as it is in the diagram 
 at Philadelphia, and the position of the sun in St. Louis will be as far 
 from the zenith as it is in the diagram, but the sun will be west of 
 south instead of east of south. 
 
 watch which has the correct local time in New York has not the 
 correct local time when it is carried to Chicago. To avoid the 
 difficulties of timekeeping growing out of travel, railroads have 
 adopted a system of standard time. Under this system the country 
 is divided into north-south belts, about 15 wide, and all places in 
 each belt use the time which is correct for the central meridian of 
 that belt. The railway time in adjacent belts differs by one hour. 
 By this system, the clocks and watches do not show correct local 
 time anywhere except on the central meridians of each belt. 
 Fig. 521 shows the standard-time zones. 
 
494 PHYSIOGRAPHY 
 
 Lengths of degrees. The length of a degree of longitude, as 
 measured on the surface of the earth, is the ^-$ part of a parallel. 
 Since the parallels are very much shorter near the poles than 
 near the equator, the length of a degree of longitude varies with the 
 latitude. At the poles, where the length of the parallel becomes 
 zero, the length of a degree of longitude also becomes zero. At 
 the equator the length of a degree of longitude is 69.652 miles; 
 in latitude 30, 59.955 miles; and in latitude 60, 34.914 miles. 
 
 Degrees of latitude are measured along meridians. They also 
 vary in length. The length of a degree of latitude has been meas- 
 ured in several places. In India it is about 68| miles, while in 
 Sweden, the most northerly point where it has been measured, 
 it is 69} miles. At the poles, it is calculated that it must be about 
 69-^ miles. In the United States, the average length is about 69 
 miles. The lengths of degrees of latitude and longitude in cer- 
 tain selected latitudes, are shown in the following table: 
 
 Latitude. Longitude. 
 
 In latitude 0, 1 = 68.704 miles. 69. 1 72 miles. 
 3 o^ 1 o ==68 881 59.956 " 
 
 " 45, 1 = 69.054 " 48.995 " 
 
 " " 60, 1 = 69. 230 " 34.674 " 
 
 " " 90, 1 = 69.407 " 0.000 " 
 
 All measurements which have been made show that the length 
 of a degree of latitude increases as the poles are approached. In 
 other words, the nearer the pole the longer the degree of latitude, 
 or, more strictly, the longer the arc which subtends a degree. This 
 means that the earth is flattened at the poles. 
 
 That this is the meaning of the variation in the length of the 
 degree is shown by Figs. 522 and 523. In the study of Fig. 522 
 it is to be remembered that a degree is -j^ of the angular distance 
 about a point, and, measured on a circumference, it is the 5 ^- F part 
 of the circumference described about the point from which the 
 angle is measured. Since the degree is longer in high latitudes 
 than in low, it means that the arc on which it is measured is the 
 arc of a larger circle than that on which the degree in low latitudes 
 is measured. The -g^ of a larger circumference is longer than the 
 y^ff part of a smaller circumference. Thus, the distance on the 
 circumference between and 18 is much less than that between 
 
EARTH RELATIONS 
 
 495 
 
 72 and 90 (= 18). In other words, the center of the circum- 
 ference on which a high-latitude degree is measured, is not the 
 same as the center from which a low-latitude degree is measured 
 (Fig. 522). 
 
 Fig. 523 shows the same thing in another way. The oblate 
 curve S represents a meridional section of the earth, with the flat- 
 tening greatly exaggerated. Circle C coincides with S at the equa- 
 
496 
 
 PHYSIOGRAPHY 
 
 tor E, while the circle M coincides with it at one pole, P. A degree 
 of arc on the curve S near P is about as long as a degree on M, 
 
 M 
 
 FIG. 522. 
 
 FIG. 523. 
 
 FIG. 522. Figure to illustrate the fact that the longer degrees of latitude 
 toward the poles means polar flattening. The curve is the half of a 
 spheroid, more oblate than the earth is. The radiating lines are repre- 
 sented as 18 apart; that is, the distance from 0to 1S is 18/360 of the 
 circle of which this arc is a part. Similarly the distance from 18 to 
 36 is 18/360 of the circumference of which this curve is an arc, and so 
 on. The curve between 72 and 90 is much longer than the curve 
 between and 18. 
 
 FIG. 523. The curve S represents a meridian section of the earth (the flat- 
 tening being greatly exaggerated). The circle C coincides with S near 
 the equator E, and the larger circle M coincides with it near the pole. 
 A degree of arc on S near P is of about the same length as one on M , 
 while one on S near E is of about the same length as one on C. Since 
 the circle M is larger than the circle C, a degree on S near P is longer 
 than one near E. 
 
 while a degree of arc on S near E is about as long as a degree on 
 C. A degree of arc on M is clearly much longer than a degree on C. 
 
 The actual measurement of the length of a degree of latitude is a diffi- 
 cult matter, but the principle on which it is measured is easily understood. 
 At any given point in the northern hemisphere the north star is a certain 
 number of degrees above the horizon. When the observer, starting from 
 a given point, has gone directly northward until the star appears one degree 
 higher above the horizon at the corresponding hour, he has gone one degree 
 (Fig. 524). In practice, the measurement is complicated, because the surface 
 of the land is always somewhat uneven, and allowance must be made for 
 every irregularity. A line measured along the uneven land surface would 
 be too long. Again, the degree is to be measured at sea-level. The land is 
 above sea-level, and therefore the measurement on the land surface must 
 
EARTH RELATIONS 
 
 497 
 
 be corrected, not only for all unevennesses, but for its elevation above sea- 
 level. 
 
 Inclination of axis and its effects. The sun's rays illuminate 
 one-half of the earth all the time. The border of the illuminated 
 
 FIG. 524. Diagram to illustrate the way in which a degree of latitude is 
 measured. When the observer has traveled so far along the surface 
 that the position of the pole-star has changed 1, the distance between 
 the two stations, A and B, is a degree. 
 
 half is called the circle of illumination (Fig. 525). All places within 
 the circle of illumination have day, while all places outside it 
 have night. If the axis about which the earth rotates were per- 
 pendicular to the plane in which the earth revolves about the 
 sun, the circle of illumination would always pass through the 
 poles. Under these conditions the half of each parallel would be 
 
 FIG. 525. Diagram to illustrate the fact that half of the earth is illuminated 
 by the sun at any one time. The line between the illuminated half and 
 the half which is not illuminated, is the circle of illumination. 
 
 illuminated all the time. If the half of each parallel was con- 
 stantly illuminated, the days and nights on each parallel would be 
 equal, for it takes just as long for a place at A (Fig. 525) to move to 
 B (half of a day) as for it to move from B to A' (half of a night). 
 
498 PHYSIOGRAPHY 
 
 If, then, the axis of the earth were perpendicular to the plane of 
 its orbit, days and nights would always be equal everywhere. 
 
 Since days and nights are not equal at all seasons on most 
 parts of the earth, it follows that the axis on which the earth rotates 
 is not perpendicular to the plane of its orbit. 
 
 Again, if the earth rotated on an axis perpendicular to the 
 plane of its orbit, the sun's rays would always fall on a given place 
 at the same angle at the same hour of the day. Thus at A, Fig. 
 
 525, the sun's rays would fall vertically at noon; while at the 
 same hour (noon at A) they would fall at a lesser angle at C', 
 but the angles of the rays at A, C, and B would always be 
 the same at the same hour of the day, in whatever part of its 
 orbit the earth found itself. The same relations would hold for 
 points on all parallels. Now, the sun's rays do not fall at the same 
 angle at the same place at the same hour at all times of the year. 
 In middle northern latitudes, for example, the sun is much higher 
 above the horizon at noon in summer than in winter. This varia- 
 tion of the angle at which the sun's rays strike the earth at a 
 given time and place, as well as the unequal lengths of days and 
 nights in most places, is the result of the inclination of the axis 
 on which the earth rotates as it revolves around the sun. The 
 position of the axis is essentially constant throughout the year, and 
 though its changes are more considerable in long periods of time 
 they may be disregarded when short periods are concerned. 
 
 The effect of the inclination of the axis is illustrated by Fig. 
 
 526, which represents the earth in four positions in its orbit. In 
 the position marked March 21, the half of each parallel is illu- 
 minated. At this time, therefore, days and nights are equal every- 
 where. In the position marked June 21, more than half of all the 
 parallels of the northern hemisphere are illuminated, and there 
 the days are more than 12 hours long and the nights correspond- 
 ingly shorter. In the southern hemisphere the nights are longer 
 than the days. In the third position, September 22, the days and 
 nights are again equal everywhere, for the circle of illumination 
 bisects every parallel. In the fourth position, December 22, more 
 than half of each parallel in the southern hemisphere is within the 
 circle of illumination, and there the days are longer than the nights, 
 while in the northern hemisphere the nights are longer than the 
 days. Twice during the year, therefore, on March 21 and Septem- 
 ber 22, the days and nights are equal everywhere. These times 
 
EARTH RELATIONS 499 
 
 are known as the equinoxes. The equinox in March is the vernal 
 equinox, and that in September is the autumnal equinox. 
 
 When the earth is in the relation to the sun shown in the posi- 
 tion marked June 21, Fig. 526, the days are longest in the north- 
 ern hemisphere, and the rays of the sun fall perpendicularly on the 
 surface of the earth farther north (in lat. 23 27^') than at any other 
 time. This is the summer solstice. The winter solstice occurs six 
 months later, when the sun's rays strike the earth vertically 23^ 
 (nearly) south of the equator, and when the days of the southern 
 hemisphere are longest and those of the northern shortest. The 
 
 SEPT. 22 
 
 FIG. 526. Diagram showing the position of the earth and of its illumination 
 at the solstices and equinoxes. 
 
 distribution of light and the relative lengths of day and night in 
 various latitudes are further shown for the solstitial dates by 
 Figs. 527 and 528. 
 
 These figures also show that the days and nights are always equal 
 at the equator, since the equator is always bisected by the circle of 
 illumination (Figs. 527, 528, and 536). Days and nights are not 
 always equal in any other latitude, unless at the poles, where there 
 is one day of six months and one night of six months, each year. 
 
 Apparent motion of the sun. The effect of the inclination of 
 the axis of the earth is to make the sun appear to move north and 
 south once during each revolution of the earth about the sun. 
 The effect on the earth is illustrated by Fig. 529. That is, the revo- 
 lution of the earth about the sun, while it rotates on an axis in- 
 
500 
 
 PHYSIOGRAPHY 
 
 clined toward the plane of its orbit, makes the sun appear to move 
 from a place where his rays are vertical 23 (nearly) north of the 
 equator (direction S, Fig. 529), to a place where they are vertical 
 
 FIG. 527. Diagram to illustrate the effect of inclination of the earth's axis 
 on the length of day and night. In the figure, more than half of every 
 parallel of the northern hemisphere is illuminated. The days are there- 
 fore more than twelve hours long, and the nights less, since the half of 
 each parallel is the measure of 180 of longitude, and ISO of longitude 
 corresponds to twelve hours of time. Similarly less than half of every 
 parallel of the southern hemisphere is illuminated, and the nights are 
 therefore more than twelve hours long. 
 
 FIG. 528. The relation of the earth to the sun's rays at a time six months 
 later than that represented in Fig. 527. The conditions of day and 
 night in the hemispheres are reversed. 
 
 23 (nearly) south of the equator (direction W), and back again, 
 in one year. 1 The result, so far as the earth is concerned, is as 
 
 1 The inclination of the earth's axis is not quite constant. Its present 
 inclination (1908) is 23 27' 4.5". Three thousand years ago its inclination 
 was about 23. The extreme variation possible is 2 37'. 
 
EARTH RELATIONS 501 
 
 if the sun moved from S, which corresponds to the time of the sum- 
 mer solstice, to A, which corresponds to the time of the autumn 
 equinox, to W, which corresponds to the time of the winter sol- 
 stice, then back again to Sp, which corresponds to the spring 
 equinox, and to S, while the earth is making one circuit about the 
 sun. 
 
 When the sun is vertical in latitudes north of Sp, the days are 
 longer than the nights in the northern hemisphere, and the sun's 
 rays strike the surface in the northern hemisphere less obliquely 
 than they do in the southern hemisphere. When the sun is in the 
 position Sp, days and nights are equal everywhere, and when the 
 
 Sp&A 
 
 FIG. 529. The inclination of the earth's axis, as it revolves about the sun, 
 makes the sun appear to travel north and south. The sun is vertical at 
 the equatoi on the 21st ol March (Sp.), then appears to move northward 
 until it is vertical 23 north ol the equator (S), then appears to move 
 southward until it is vertical again at the equator (A), then south until 
 it is vertical 23$ south of the equator (W.), and then north again until 
 it is vertical at the equator. These chances are accomplished in the 
 course of one year as a result of the revolution. 
 
 sun is vertical south of Sp, days are longer than nights in the 
 southern hemisphere, and the sun's rays are more nearly vertical 
 than in the northern hemisphere. 
 
 The northernmost parallel where the sun's rays are ever verti- 
 cal is called the tropic of Cancer. The corresponding southernmost 
 parallel is the tropic of Capricorn. The tropics are nearly 23 
 (23 27') from the equator, because the axis of the earth is in- 
 clined by that amount toward the plane of its orbit. The sun is verti- 
 cal at the tropic of Cancer at the time of the summer solstice, and 
 at the tropic of Capricorn at the time of the winter solstice. The 
 parallels just touched by the circle of illumination at the time of the 
 
502 PHYSIOGRAPHY 
 
 solstices are the polar circles. They are as far from the poles as 
 the tropics are from the equator. They are, therefore, in latitude 
 about 66 (66 33'). The one in north latitude is the Arctic circle, 
 and the one in south latitude the Antarctic circle. 
 
 The effects of inclination of the earth's axis on the length of days and 
 nights may well be emphasized by comparing the lengths of days and 
 nights as they now exist, in our own region, with those which would exist 
 if the axis of the earth were inclined 45 toward the plane of its orbit 
 instead of 23J. It is also instructive to study the conditions which would 
 exist with reference to day and night (1) if the earth did not rotate during 
 its revolution around the sun, and (2) if it rotated once in the period of 
 its revolution. In the latter case, the results depend on the direction of 
 rotation. 
 
 Latitude and sun altitude. The solution of certain problems 
 in the determination of latitude and sun altitude will help to a 
 clearer understanding of the changes in the relations of sun and 
 earth due to the movements of the latter. 
 
 At the time of equinox, the sun is directly overhead at the 
 equator at noon. One degree north of the equator, i.e., in latitude 
 1 N., the sun will appear 1 from the zenith (i.e., the point directly 
 overhead) at noon, or 89 above the horizon. This is the same as 
 saying that the altitude of the sun is 89. Five degrees north of 
 the equator (lat. 5 N.) the sun will appear 5 from the zenith at 
 noon, and his altitude (above the horizon) is 85. 
 
 If, therefore, the altitude of the sun at a given place at noon 
 at the time of an equinox is known, the latitude may be determined. 
 Thus if the altitude of the sun is 30 at the time of an equinox, 
 the observer must be 60 from the place where it is vertical, that 
 is, in latitude 60 N. or 60 S. Similarly if the latitude is known, 
 the altitude of the sun at noon at the time of equinox may be 
 determined. Thus in latitude 40 the sun must be 50 above the 
 horizon at noon, for latitude 40 is 40 from the place where the 
 sun is vertical. 
 
 Any other dates besides the equinoxes may be used if the lati- 
 tude where the sun is vertical is known. Thus at the time of the 
 summer solstice, when the sun is vertical in latitude 23i N., it is 
 23 from the zenith, or has a noon altitude of 66 J, at the equator. 
 It has the same altitude in latitude 47 N., for this place, like the 
 equator, is 23J from the place where the sun is vertical. 
 
 Reversing the problem, we may determine the latitude of a 
 
EARTH RELATIONS 503 
 
 place if we know the noonday altitude of the sun there, and the 
 latitude where the noonday sun is then in the zenith. Suppose, 
 for example, the altitude of the sun is 40 at noon at the time 
 of the June solstice, what is the latitude of the place? Since 
 the altitude of the sun is 40, the place must be 50 from the place 
 where the sun is vertical, that is, 50 from latitude 23J N. This 
 is 73i N. or 26^ S. 
 
 Problems. 
 
 Note. In the solution of these problems the student will find it 
 helpful, in many cases, to make diagrams representing the conditions 
 of the problem. 
 
 1. What is the altitude of the sun at noon at the time of an equinox, 
 
 (1) In latitude 50 N.? 
 
 (2) In latitude 50 S.? 
 
 (3) In latitude 75? 
 
 2. What is the altitude of the sun at noon at the time of the summer 
 solstice, 
 
 (1) In latitude 30 N.? 
 
 (2) In latitude 30 S.? 
 
 (3) In the latitude of New York? 
 
 (4) In the latitude of Vancouver? 
 
 (5) In latitude 75 N.? 
 
 (6) In latitude 66i S.? 
 
 (7) At the north pole? 
 
 3. Formulate a rule for finding the altitude of the sun (a) at the 
 time of an equinox, and (6) at the time of a solstice, the latitude of the 
 place being given. 
 
 4. In what latitude or latitudes is the sun 30 above the horizon at 
 noon at the time of an equinox? 
 
 5. In what latitude or latitudes is the sun 75 above the horizon at 
 noon at the time of an equinox? 
 
 6. In what latitude or latitudes is the sun 40 above the horizon at 
 noon at the time of the June solstice? 
 
 7. In what latitude or latitudes is the sun 80 above the horizon at 
 noon at the time of the December solstice? 
 
 8. What is the latitude of the place or places where the sun is 10 
 above the horizon at noon at the time of the June solstice? 
 
 9. Formulate a rule for finding the latitude of a place from the noon 
 altitude of the sun. 
 
 10. In what direction and at what altitude would the sun appear (a) 
 at midnight, and (6) at noon, to an observer in latitude 75 N. at the 
 time of the summer solstice? See 2 (5) above. 
 
504 
 
 11. To an observer at the equator, in what direction would the sun 
 appear to rise on June 21? What would be the noon altitude of the sun 
 at the equator on the same day? a , 
 
 12. What would be the noon altitude of the sun at Chicago on June 
 21? On December 21? 
 
 13. In what latitude is the altitude of the sun the same, at noon, at 
 the time of a,n equinox and on June 21? 
 
 The Solar System 
 
 The solar system includes the sun and all the bodies which 
 revolve about it. There are eight planets, of which the earth is one. 
 Named in the order of their distance from the sun, commencing 
 with the nearest, the planets are: Mercury, Venus, Earth, Mars, 
 Jupiter, Saturn, Uranus, and Neptune. Most of the planets 
 have satellites corresponding to our moon. The following table 
 shows some of the more important facts about the planets: 
 
 
 
 
 
 
 g.sJ 
 
 
 
 
 
 
 
 
 II 
 
 7 
 
 q 
 
 7 
 
 is* 
 
 IB 
 
 "3 
 
 3.2 
 
 sl 
 
 
 - oa 
 
 a>:2 
 
 
 >> iJ 
 
 * o 
 
 1 ' a; 
 
 '^ *^ ^. 
 
 - 
 
 
 88 
 
 |1 
 
 CQ 
 
 ** "rf 
 
 ES 
 
 2jM 
 
 15 
 
 ^3 
 
 
 Is 
 
 
 
 
 N 
 
 S o-^J 
 
 |.s 
 
 .- i, 
 
 B os 
 
 02 
 
 
 Q 
 
 > 
 
 an 
 
 Q 
 
 S 
 
 n 
 
 
 
 
 
 
 
 i 
 
 
 
 
 
 
 Mercury. . . . 
 
 2,765 
 
 0.05 
 
 
 3.70 
 
 36.0 
 
 0.24 
 
 7 0' 
 
 
 
 9,64 /, 000 
 
 
 
 
 1 
 
 
 
 
 
 
 V ATlllfl 
 
 7,826 
 
 0.89 
 
 
 4.89 
 
 67.2 
 
 0.62 
 
 3 24 
 
 
 
 
 405,000 
 
 F.rth 
 
 f 7,926.5* 
 { 7,899.7f 
 
 1.00 
 
 1 
 
 5.53 
 
 92.9 
 
 1.00 
 
 
 
 1 
 
 
 332,000 
 
 Mcirs 
 
 / 4,352* 
 I 4,312f 
 
 0.14 
 
 1 
 
 3.95 
 
 141.5 
 
 1.88 
 
 1 51 
 
 2 
 
 
 3,020,000 
 
 Jupiter 
 
 1 90,190* 
 \ 84,570t 
 
 1264.00 
 
 1 
 
 1.33 
 
 483.3 
 
 11.86 
 
 1 19 
 
 7 
 
 
 1,047 
 
 Saturn 
 
 1 76,470* 
 1 69,780f 
 
 759.00 
 
 1 
 
 0.72 
 
 886.0 
 
 29.46 
 
 2 30 
 
 10 
 
 3,502 
 
 
 
 
 1 
 
 
 
 
 
 
 T T Y>n iiii 
 
 34,900 
 
 63.40 
 
 
 1 .22 
 
 1781.9 
 
 84. C2 
 
 46 
 
 4 
 
 
 22,760 
 
 Neptune. . . . 
 
 32,900 
 
 82.30 
 
 1 
 
 1.11 
 
 2791 . 6 
 
 164.78 
 
 1 47 
 
 1 
 
 19,500 
 
 Equatorial. 
 
 t Polar. 
 
EARTH RELATIONS 505 
 
 Besides the planets and their satellites, the solar system in- 
 cludes numerous (more than 400) asteroids, bodies much smaller 
 than the planets, intermediate in position between Mars and 
 Jupiter, and those comets which revolve about the sun. These 
 bodies have little influence on the earth, and nothing further need 
 be said of them in this place. 
 
 REFERENCES 
 
 Text-books on Astronomy. Among the available recent ones are Moul- 
 ton's (Macmillan) and Comstock's (Appleton). Todd's (Am. Book Co.) 
 and Young's (Ginn & Co.) are less recent, but serviceable; also Johnson's 
 Mathematical Geography (Am. Book Co.). 
 
PART III 
 THE ATMOSPHERE 
 
 CHAPTER XII 
 GENERAL CONCEPTION OF THE ATMOSPHERE 
 
 Substantiality. When the atmosphere is still, we are hardly 
 conscious of its existence. We walk through it without realizing 
 that we are forcing our way through a real substance. Compared 
 with land or even with water, it seems most unsubstantial. But 
 when the air is in motion, that is, when the wind blows, we are con- 
 scious that it is very real and substantial, for the force of the 
 wind may be so great that it is difficult to stand or walk against it. 
 Trees and buildings are occasionally blown down by it, and quanti- 
 ties of dust and sand are picked up and sometimes carried up to 
 great heights. These familiar phenomena show that the air is a 
 real substance, and that, when it moves rapidly, even strong 
 objects give way before it. 
 
 A strong wind is not equally strong at every instant; it comes 
 in gusts. When a strong gust of wind strikes a high building, 
 the air is reflected from the wall, somewhat as a ball is when thrown 
 against it. If a strong gust of wind is followed the next instant by 
 a weak wind or a lull, the air rebounding from the wall may have 
 great force in a direction opposed to that of the main wind. These 
 reflected winds occasionally blow people down, for they blow in 
 the direction opposite to that against which the body is braced. 
 In cities where there are high buildings, the streets are sometimes 
 protected from the direct winds; but the currents of air, whether 
 direct or reflected, are often concentrated at the street level, where 
 they sometimes have force enough to overturn cabs. 
 
 The substantiality of the air may be shown in still another way. 
 If the air be pumped out of a cylinder whose top is covered by a 
 thin piece of rubber, the rubber covet is pressed down into the 
 
 506 
 
GENERAL CONCEPTION OF THE ATMOSPHERE 507 
 
 cylinder, and may even be broken. The force which presses it 
 down is the weight of the air above. If the cylinder be of weak 
 material, such as thin glass, while the cover is strong, the pressure 
 of the air outside may burst the cylinder when the air inside is 
 pumped out. The cylinder does not break when full of air, because 
 the pressure on the inside balances that on the outside. If the 
 cylinder be of tin instead of glass, the pressure of the air on the 
 outside may bend in its walls when the air is pumped out. These 
 experiments, as well as the phenomena of the wind, show that 
 the air is something real and that it has weight. 
 
 The amount of pressure which the air exerts, that is, its weight, 
 may be determined. At sea-level, it is found to be nearly 15 
 (14.7) pounds to the square inch. 
 
 Relation to the rest of the earth. The atmosphere is com- 
 monly called an envelope of the earth. More properly it is an 
 envelope of the rest of the earth, for it is itself as much a part of the 
 earth as the rocks ^are. It goes with the rest of the earth through 
 space, and it is essential to the life of the earth and to most of the 
 processes which are in operation on the earth's surface. It is the 
 medium through which moisture is distributed, and it has much 
 to do with the temperature of the earth, for without an atmosphere 
 the earth would be very much colder than now. Without the 
 air, therefore, the earth would be a very different body. Some 
 conception of its functions may be gained by trying to conceive 
 what the earth would be without it. This point may be recalled 
 from time to time, as our study proceeds. 
 
 The atmosphere is in reality a little more than an envelope of 
 the rest of the earth, for it penetrates the soil and rocks as far down 
 as there are holes and cracks, and its constituents are dissolved in 
 the waters of the sea, in all the waters on the land, and in all the 
 waters beneath its surface. 
 
 Density and altitude. Many of the laws which govern the 
 distribution of gaseous matter are known. From these laws it is 
 known that the air, which is but a mixture of gases, must be most 
 dense below and less dense above. This is the same as saying that 
 there is more air in a cubic foot of space at sea-level than in a 
 cubic foot of space at higher levels. To this general rule there are 
 exceptions, locally and temporarily, but they need not be con- 
 sidered here. Similarly there is more air in a cubic foot of space 
 1000 feet above sea-level than in the same space 2000 feet above 
 
508 PHYSIOGRAPHY 
 
 sea-level, and so on. This means that the particles of which 
 gases are composed are nearer together at low altitudes than at 
 high altitudes. The reason is readily understood. 
 
 If a cubic foot of air were pressed from all sides, it could be 
 squeezed into a smaller space, and the more the pressure, the 
 smaller the space into which it could be compressed. Now at the 
 bottom of the atmosphere the air is pressed down by all the air 
 above. At the height of 1000 feet above the sea, the air is pressed 
 down by all the air above that level, and so on. Hence the 
 lowest air is under most pressure, and is therefore (with certain 
 exceptions) densest. 
 
 It is largely because the air gets rarer with increase of altitude 
 that mountain-climbing is difficult. As the climber gets higher 
 and higher, it becomes more and more difficult to breathe. He 
 may take in the same number of cubic inches of air each time he 
 inhales, but each cubic inch contains less air the higher he goes. 
 Finally the air gets so rare as to stop further ascent. It should be 
 noted, however, that it is not simply the decreasing amount of air 
 taken into the lungs which makes it difficult to ascend to great 
 heights. The cold, the snow and ice, and often the very steep 
 slopes of high altitudes, are all obstacles, and the body is not 
 adjusted to the lessened pressures of the higher altitudes. 
 
 Height. How high above the sea and land does the air ex- 
 tend? No positive answer can be given to this question, though 
 something is known about it. 
 
 1. The greatest altitude reached by any mountain-climber is 
 between four and five miles. At this altitude there was air enough 
 to make breathing possible to one exercising actively. A con- 
 siderable quantity of air, therefore, exists at a height of more 
 than four miles. Its upper limit must be considerably higher. 
 
 2. Men have ascended to heights of over six miles in balloons, 
 and in one case (Coxwell and Glaisher, 1802) to a height of more 
 than seven miles. In some cases the occupants of the balloons 
 have become unconscious at an elevation of about 29,000 feet, 
 but in one case (Dr. Berson, Berlin, 1894) this difficulty was over- 
 come by carrying oxygen for breathing. Balloons without human 
 occupants have risen eighteen miles. At the upper limit of their 
 ascent, the air was still dense enough so that the amount displaced 
 by the balloon was at least equal to the weight of the balloon. 
 
GENERAL CONCEPTION OF THE ATMOSPHERE 503 
 
 3. From the phenomena of twilight, due to the refraction of the 
 light as it passes through the atmosphere, it may be demonstrated 
 that the air extends up to a height of forty-five miles. 
 
 4. On almost any clear night "shooting stars" may be seen. 
 These shooting stars, or meteors, are small solid bodies which come 
 into the earth's atmosphere from outside space. When they 
 enter the atmosphere, they are very cold, for the temperature 
 of space, outside the earth's atmosphere, is believed to be about 
 459 F. As they approach the earth, they are traveling (fall- 
 ing) very fast, say 12 to 45 miles per second. In passing through 
 the atmosphere, their movement is resisted by the air. The re- 
 sult is friction, and the friction with the air heats them. When 
 they get hot enough to glow (red-hot), they may be seen. Now 
 the height at which they begin to glow has been estimated in 
 some cases, and is found to be, at a maximum, more than 100 miles 
 above sea-level. This shows that the atmosphere is much more 
 than 100 miles high, for the meteors must have come through the rare, 
 cold upper air a very considerable distance before becoming red-hot. 
 
 5. The aurora or "northern lights" sometimes seen in high 
 latitudes is believed to be an electric phenomenon in very rare 
 air. The height of the aurora is sometimes more than 100 miles. 
 The southern ends of the streamers have even been calculated to 
 be as much as 400 miles high. This shows that the air is dense 
 enough to show electric phenomena at that height. It is believed, 
 however, that the density of the air 100 miles above sea-level is 
 not much more than one billionth of its density at sea-level. 
 
 6. We know the weight of the atmosphere. We know also 
 the rate at which a gas, or a mixture of gases, like the air gets 
 lighter with increase of altitude. The law is that the density is 
 proportional to the pressure.. If we go up till half the air is below 
 us, the air at that height should be half as dense as it was at the 
 bottom. If we rise again until half of the upper half of the air is 
 below us, the air at that level is half as dense as it was at the first 
 station. On this principle it would appear that there should be 
 no upper limit; the air should simply get rarer and rarer without 
 having a definite upper surface. 
 
 Though the above law holds in all places where experiments can 
 be carried on, there is some reason to believe that it may cease to 
 hold when the air becomes very rare. 
 
 7. All gases have a tendency to fly away from the earth, but 
 
510 PHYSIOGRAPHY 
 
 are Tield by gravity. Gravity gets weaker and weaker with in- 
 creasing distance from the earth's center, and, at a sufficiently 
 great distance from the center, the earth would not be able to hold 
 any of the gases of its atmosphere. That distance would be less 
 for lighter gases, and greater for heavy ones. It is calculated that 
 none of the gases of the atmosphere could be held by the earth at 
 a distance greater than 620,000 milesfrom the center of the earth. 
 
 From the above considerations it appears to be certain that 
 the air extends much more than 100 miles above the rest of the 
 earth, but how much more is unknown. Whatever its height, one- 
 half the atmosphere (by weight) lies below a plane about 3.6 miles 
 above sea-level, three-fourths of it below a plane 6.8 miles above 
 the same level, and seven-eighths of it below a plane 10.2 miles 
 up. The highest mountain is about 6 miles high, so that nearly 
 three-fourths of the atmosphere lies below the level of its top. 
 
 Volume. Since the height of the air is not known, its volume 
 cannot be determined. If it extends up but 200 miles, its volume 
 is about one-sixth that of the rest of the earth ; if it extends up 500 
 miles, its volume is nearly one-half that of the rest of the earth. 
 
 Mass. Great as is the volume of the atmosphere, its mass 
 (measured by its weight) is far less than that of the solid part of 
 the earth, or even than that of the water. It has been estimated 
 at about ^TT that of the water, and about i^org- that of the solid 
 part of the earth. Its weight is about equal to that of a layer of 
 water completely covering the earth to a depth of 33 feet. Sir 
 John Herschel estimated the weight of the air resting on the other 
 parts of the earth at 11 trillion pounds. 
 
 History. It is probable that the atmosphere has undergone 
 changes in mass and volume in the course of its history. It was 
 formerly assumed that the atmosphere is being gradually diminished, 
 and that it would in time disappear, as the moon's atmosphere 
 was assumed to have disappeared. But this assumption does not 
 appear to be well founded. It is more probable that the moon 
 never had an atmosphere, than that it has lost one it once had. 
 Furthermore, the atmosphere is now gaining various gases from 
 volcanic and other vents (p. 368), and probably has always done 
 so. It is probably acquiring gases from space also. Though 
 contributions from this source are inconsiderable now, they may 
 not always have been so. The atmosphere is losing as well as 
 gaining. Some gases, especially light ones like hydrogen, prob- 
 
GENERAL CONCEPTION OF THE ATMOSPHERE 511 
 
 ably escape the attractive control of the earth and pass off into 
 space. Other constituents of the air, like oxygen (p. 72) and 
 carbonic acid, are withdrawn from the air and locked up for long 
 periods at least, if not permanently, in the rocks. The rates both 
 of supply and loss fluctuate. When loss exceeds supply, the mass 
 of the atmosphere must decrease; when supply exceeds loss, the 
 mass must increase. So far as can be judged from present phe- 
 nomena, slight fluctuations of mass must have taken place. As 
 will be seen in the next chapter, the fluctuations in composition 
 may have been more significant than the fluctuations of mass and 
 volume. 
 
 REFERENCES 
 
 The following references apply to Chapters XII to XIX. Numbers 
 1 to 5 are the most serviceable for the general student; Number 13 is help- 
 iul for the topic indicated by the title, and 14 is constantly valuable. 
 
 1. DAVIS, Elementary Meteorology: Ginn & Co. 
 
 2. WALDO, Elementary Meteorology: Am. Bk. Co.; Modern Meteorology. 
 Scribners. 
 
 3. BARTHOLOMEW, Physical Atlas; Meteorology: Constable, London. 
 
 4. WARD, Practical Exercises in Meteorology: Ginn & Co. Also Bull. 
 Am. Geog. Soc., Vol. XXXVII, 1905, p. 385, and Vol. XXXVIII, 1906, p. 401. 
 
 5. HANN, Handbook of Climatology: The Macmillan Co. 
 
 6. GREELT, American Weather: Dodd, Mead & Co. 
 
 7. FERREL, Popular Treatise on the Winds: Wiley & Sons. 
 
 8. RUSSELL, Meteorology: Macmillan. 
 
 9. Illustrated Cloud Forms: U. S. Hydrographic Office, Washington. 
 
 10. DAVIS, The Temperature Zones: Jour. Sch. Geog., Vol. I, pp. 139-143. 
 
 11. HAYDEN, The Great Storm off the Atlantic Coast of the United States, 
 March 11-14, 1888: Nat. Geog. Mag., Vol. I, pp. 40-58. 
 
 12. GREELY, Rainfall Types of the United States: Nat. Geog. Mag., Vol. V, 
 pp. 45-58. 
 
 13. MOORE, Storms and Weather Forecasts: Nat. Geog. Mag., Vol. VIII, 
 pp. 65-82, and Vol. XVI, pp. 255-305. 
 
 14. The Monthly Weather Review: The Weather Bureau, Washington. 
 
CHAPTER XIII 
 CONSTITUTION OF THE ATMOSPHERE 
 
 Principal constituents. The atmosphere is remarkably con- 
 stant in composition, and is made up chiefly of two gases, namely, 
 nitrogen, which makes up nearly 78 per cent, of dry air, and oxygen, 
 which makes about 21 per cent. 
 
 Minor constituents. Beside these two principal constituents, 
 the proportions of which do not vary much, there are several minor 
 constituents, of which argon, about one per cent, of the whole, is 
 most abundant. Argon was not separated from nitrogen until 
 recent years. Another minor constituent of dry air, carbon dioxide 
 or carbonic-acid gas, is of great importance. It makes up about 
 TBT?nr by volume of the whole atmosphere, and its amount is 
 nearly constant from year to year. 
 
 There is also a considerable amount of water vapor, that is, 
 water in particles so small as to be invisible in the air. The total 
 amount in the atmosphere at any one time varies within rela- 
 tively narrow limits; but the amount varies greatly from place 
 to place at the same time, and from time to time in the same place. 
 Since this is the case, and since it is separated frequently from the 
 atmosphere, in the form of rain, snow, etc., it is often regarded as 
 something in the air, rather than as a part of the air. The weight 
 of the total amount in the air at one time has been variously 
 estimated at from 3-^3- (1 per cent.) to ? ^ ( per cent.) that of dry 
 air. The smaller figure is probably nearer the truth than the 
 larger. The water-vapor pressure at the bottom of the atmosphere 
 is not a measure of the amount of water vapor above. It was 
 this assumption which gave the larger of the figures cited above. 
 The water vapor may make as much as 3 per cent, of the air (by 
 volume) in moist tropical regions. 
 
 512 
 
CONSTITUTION OF THE ATMOSPHERE 513 
 
 Impurities. The air always contains some other gases which 
 are commonly looked upon as impurities, though they are not neces- 
 sarily injurious to life or to natural processes in general. Gases 
 arise from combustion and decay of organic matter, from various 
 chemical processes used in manufacturing, from volcanic and other 
 vents in the earth's crusts, etc. Their aggregate amount is small, 
 but locally, as about some vents, they are so abundant as to be 
 injurious to life. This is the case in Death Valley in the Yellow- 
 stone Park, where animals straying into certain parts of the valley 
 are often overcome and killed. 
 
 The air always contains numerous solid particles, which may, 
 collectively, be called dust. Though the dust in the air serves 
 important functions, it is to be looked upon as an impurity rather 
 than as a constituent. 
 
 Relations of constituents to one another. The various gas- 
 eous constituents of the air are mixed with one another, and each 
 of them retains its own characteristics. The oxygen behaves essen- 
 tially as if no nitrogen were present, and the nitrogen as if no 
 oxygen were present. That the several constituents of the air 
 are merely mixed, and not chemically united, may be shown in 
 various ways. One of them is as follows: When air is liquefied 
 and allowed to stand, its constituents evaporate independently. 
 Nitrogen and carbon dioxide evaporate faster than oxygen, so that 
 as the liquid air stands, the proportion of oxygen increases. Again, 
 heat is given off whenever a chemical compound is formed. When 
 nitrogen and oxygen are mixed, no heat is developed. 
 
 > The Functions of the Atmospheric Elements 
 
 . 'i 
 
 The various constituents of the air play various roles in the 
 economy of the earth. 
 
 Nitrogen is inactive. Though it is inhaled with the oxygen 
 in breathing, it does not appear to be of direct use to animals. 
 Both animals and plants need nitrogen, but they cannot use the 
 nitrogen of the air directly. Before they can make use of it, 
 it must be combined with something else, making what are 
 known as nitrogenous compounds. From some of these compounds 
 animals and plants derive the nitrogen they need. Plant decay 
 sets some nitrogen free, but the aggregate effect of plant life and 
 plant decay on the amount of nitrogen in the air is not known. 
 
514 PHYSIOGRAPHY 
 
 Since nitrogen makes up the larger part of the atmosphere, air- 
 pressure and wind-strength are due chiefly to it. 
 
 Oxygen from the air is being consumed constantly by all 
 animals. Air-breathing animals take it directly from the air, and 
 water-breathing animals -take it from the water in which it is dis- 
 solved. Oxygen is consumed by plants also, especially by green 
 plants in the dark. Oxygen is consumed wherever combustion 
 is going on, for combustion is primarily the union of oxygen 
 with other substances, especially carbon. When the oxygen 
 enters into combination, it loses its distinctive character- 
 istics. Whenever organic matter decays, oxygen is also con- 
 sumed, for the decay of such matter is but slow combustion. In 
 spite of the constant and rapid consumption of atmospheric oxygen 
 by animals and in all combustion, its amount does not appear to 
 decrease. We must therefore infer that oxygen is supplied to 
 the air about as fast as it is consumed. The sources of supply 
 are several. Plants break up the CO 2 into its elements, C and O, 
 and set some of the oxygen free. This is perhaps the greatest 
 source of supply of free oxygen. It is to be noted that oxygen 
 received by the air in this way is not (or may not be) new to the 
 air. Much of it at least is only returned to the air, after having 
 been temporarily withdrawn. Oxygen also reaches the atmosphere 
 from volcanic vents, by changes (deoxidation) which take place 
 in certain kinds of rocks, and perhaps from other sources. 
 
 The carbonic-acid gas of the atmosphere, though a very minor 
 constituent so far as quantity is concerned, is most important. 
 We have already seen that it is being constantly produced by the 
 burning of coal, wood, peat, gas, etc., and by the decay of all or- 
 ganic matter. It is also added to the air by all animal respiration, 
 and it is poured into the air from volcanic vents, often in great 
 quantity. It is probable that "shooting stars" sometimes con- 
 tain carbon, for the corresponding bodies (meteorites), which are 
 so large that they are not reduced to dust in the atmosphere, but 
 reach the earth as masses of rock or metal, sometimes contain 
 carbon (sometimes in the form of diamonds). Any carbon which 
 meteors contain must be burned to CO 2 in the upper air. It is 
 probable that there are still other minor sources of this gas. 
 
 Carbonic-acid gas is supplied to the atmosphere very rapidly 
 from these various sources. For example, about 75 per cent, of 
 common bituminous coal is carbon. When burned, a ton of such 
 
CONSTITUTION OF THE ATMOSPHERE 515 
 
 coal would make about 2| tons of carbonic-acid gas, all of which 
 goes into the atmosphere. A ton of hard coal, which contains a 
 higher proportion of carbon, would produce still more carbonic-acid 
 gas. If we knew the number of tons of coal burned daily, we could 
 calculate the amount of C0 2 poured into the atmosphere daily 
 as a result of its combustion. Nearly a billion l tons of coal are 
 mined each year, and if each ton of coal makes 2| tons of car- 
 bonic-acid gas, it will be seen that the atmosphere would be 
 supplied with C0 2 at the rate of more than 1\ billion tons a 
 year from this source alone. This figure takes no account 
 of other fuels, such as wood, peat, natural gas, oil, etc. Neither 
 does it take account of the slow combustion (decay) of vegetable 
 matter, nor of the CO 2 produced by respiration. When these and 
 all other sources of carbonic-acid gas are considered, it seems safe 
 to say that carbonic-acid gas is being supplied to the atmosphere 
 at the rate of several billions of tons per year; yet the amount of 
 CO-2 in the air does not increase enough to be noted from year 
 to year, or even from generation to generation. It must there- 
 fore be taken out of the atmosphere about as rapidly as it 
 enters. 
 
 The losses of carbonic-acid gas from the air come chiefly (1) 
 through green plants, of which it is the chief food, and (2) through 
 combination with mineral matter; for the CO 2 of the air is con- 
 stantly uniting with mineral matter in the solid part of the earth. 
 It will be seen therefore that some of the C0 2 is making a con- 
 tinuous round of change. It is taken out of the air by plants, and 
 its constituents, or some of them, become a part of the woody 
 tissue of the plant. In this process of transformation some of the 
 oxygen is set free in the air. The carbon of the plant is then burned, 
 either in a fire or by decay, and the carbonic-acid gas produced 
 passes back into the air to be used by plants again. Much car- 
 bonic-acid gas goes through this round each year, for much vege- 
 tation grown during one warm season is burned or partially de- 
 cayed before the next. It will be readily seen, too, that some of 
 this gas might go through a cycle of change involving its return 
 to the atmosphere more than once in a season. 
 
 The various sources of supply of CO., are not always equal at 
 the same place, and are not equal at different places. Thus, the 
 
 . 'See Mineral Resources of the Drilled States, an annual publication 
 ot the U. S. Geol. Survey. 
 
516 PHYSIOGRAPHY 
 
 amount produced by combustion is much greater in winter than 
 in summer, while the amount produced by the decay of plant and 
 animal matter is much greater in summer than in winter. It is 
 to be remembered, however, that the warm season in one hemi- 
 sphere corresponds to the cold season in the other; but since there 
 are fewer people in the southern hemisphere than in the northern, 
 there is less burning of coal there, and since there is much less 
 land in the southern hemisphere, there is less decay of land vege- 
 tation. Volcanoes are more active at some times than others, and 
 probably give forth most CO 2 when most active. The amount 
 produced by animal respiration is probably nearly the same 
 throughout the year. 
 
 The rate at which CO 2 is taken from the air also varies. Since 
 plants use it during the growing season only, the plants of middle 
 and high latitudes draw on the supply in the atmosphere chiefly 
 during the summer. Though summer alternates in the hemi- 
 spheres, there is much more plant life on land in the northern hemi- 
 sphere than in the southern, and, so far as land plants are concerned, 
 CO 2 must be consumed more rapidly in the northern summer than 
 in the northern winter. Carbonic-acid gas also enters into com- 
 bination with mineral matter more readily when it is warm than 
 when it is cold, so that there must be some variation from season 
 to season in the amount taken out in this way. 
 
 At first thought it would seem that carbonic-acid gas should 
 greatly increase in one hemisphere during the winter season, and 
 diminish in the same hemisphere during the summer; but this is 
 not the case. The reason is twofold. In the first place, the 
 winds distribute the carbonic-acid gas. In the second place, even 
 without winds, the carbonic-acid gas tends to diffuse equally 
 through the atmosphere. It is, for example, produced in great 
 quantities in a large city in winter, for the thousands of tons of 
 coal consumed daily in such a city produce enormous quantities 
 of carbon dioxide. But instead of accumulating in great quanti- 
 ties over the city, it diffuses through the atmosphere, so that, 
 even without winds, there would be no great excess over the region 
 where it is produced. A slight excess in such situations is often 
 noticed, for diffusion does not bring about equality of distribu- 
 tion instantaneously. 
 
 At present the supply and loss of carbon dioxide so nearly 
 balance that no change in the amount of C0 2 in the air is noted; 
 
CONSTITUTION OF THE ATMOSPHERE 517 
 
 but it seems quite possible that in the course of long periods of 
 time the supply may have exceeded the loss, or vice versa. While 
 therefore the amount of C0 2 remains nearly constant from year 
 to year, there is no warrant for the inference that it has remained 
 so from age to age. 
 
 Though a very minor constituent of the atmosphere, carbonic- 
 acid gas has an important function other than in supplying food 
 to plants. It has the power of retaining some of the heat radiated 
 from the solid part of the earth into space. It therefore serves as 
 a blanket to hold in the heat of the earth, and thin (tenuous) as 
 the blanket now is, it is more effective, in this respect, than the 
 denser blanket of oxygen and nitrogen. If it were thicker, it would 
 be still more effective, making the earth warmer. So important 
 is its function in this respect, that, if the amount of this gas were 
 doubled, the temperature of the earth, and especially of high lati- 
 tudes, would be notably increased. It has been estimated that 
 if its amount were doubled or trebled, magnolias might grow 
 again in Greenland, as they once did. On the other hand, it has 
 been estimated that if the amount of carbonic-acid gas in the 
 atmosphere were decreased one-half, the climate would be so much 
 colder than now, that much of the land in the northern part of our 
 continent would be covered by a sheet of snow and ice, some- 
 what as it was in the glacial period (p. 271). While these con- 
 clusions have been called into question, so far as the amount of 
 change of temperature for a given increase or decrease of carbonic- 
 acid gas is concerned, there seems to be no doubt that an increase 
 of carbonic-acid gas in the atmosphere would ameliorate climate, 
 while a decrease would make it more rigorous. 
 
 It has been noted that the water vapor in the atmosphere is a 
 variable quantity. It is constantly entering the atmosphere, and it 
 is constantly being condensed and precipitated in the form of rain, 
 snow, dew, frost, etc., to be again evaporated, condensed, and pre- 
 cipitated. Like much of the C0 2 , it is making continuous rounds. 
 The amount which the atmosphere may contain at any time is de- 
 pendent on temperature; but various other factors, especially the 
 available local supply, help to determine the amount which is 
 actually held. The importance of water vapor in the general econ- 
 omy of the earth will be referred to in later chapters, but it may be 
 stated here that, like the carbonic-acid gas, it serves as a blanket 
 to keep the earth warm. Furthermore, it is to be remembered 
 
518 PHYSIOGRAPHY 
 
 that the water vapor of the air, constantly renewed, is the source 
 of all the rain and snow, the work of which has been described 
 on preceding pages. 
 
 The dust in the atmosphere includes all solid particles held in 
 it. We do not ordinarily see them, though clouds of dust are some- 
 times visible on windy days. The settling of dust from the air 
 on all objects in doors or out is sufficient evidence of its universal 
 presence (p. 55). It may be readily seen in indoor air if the 
 room be darkened and the light allowed to enter only through a 
 narrow crack or small hole. Even air which appears clear may 
 in this way be seen to contain innumerable particles of solid matter. 
 The amount of dust is sometimes very great, as over cities and in 
 dry and windy regions. During the fogs of February, 1891, it was 
 estimated 'that the amount of dust deposited on glass roofs in and 
 near London was six tons per square mile. The variety of matter 
 in the dust was also great, carbon (soot) being most abundant. 
 
 Some years since a method was devised for counting the dust 
 particles in a given volume of air. The result showed that in the 
 air of great cities there are hundreds of thousands of dust particles 
 in each cubic centimetre (a centimetre is less than fo of an inch) 
 of air; and that even in the pure air of the country, far from towns 
 and factories, there are hundreds of motes per cubic centimetre. 
 It has been estimated that "every puff of smoke from a cigarette 
 contains about 4000 million separate granules of dust." 
 
 The amount of dust in the air is greater over the land than 
 over the sea, and in the lower atmosphere than in the upper. 
 
 The dust particles consist of inorganic materials, such as (1) tiny 
 particles of mineral matter blown up from dry roads and fields, (2) 
 particles of smoke from chimneys, (3) frequently of tiny bits of rock 
 matter blown out of volcanoes, and (4) meteoric dust, or the dust 
 which comes to the .earth from outside space such as the dust to 
 which "shooting stars " are reduced in the atmosphere; and organic 
 particles. Among the organic dust particles are bacteria of various 
 sorts, and the spores of many plants. The dust that is thrown 
 into the air when a dry puffball is broken may serve as an illus- 
 tration of the spores of plants which are often abundant in the air. 
 The fact that plant spores are nearly universal in the air is shown 
 by the promptness with which a moist piece of bread or cake, or a 
 moist piece of leather, gets mouldy, especially in a dark, warm place. 
 The moulds are plants, and the spores from which they grow were 
 
CONSTITUTION OF THE ATMOSPHERE 519 
 
 floating in the air, until they found a lodging-place suitable for their 
 growth. In the blossoming season also, the winds get much pollen 
 dust from flowers. The scattering of pollen by the wind serves 
 an important purpose in the plant world. 
 
 Some diseases are spread by means of germs in the air, though 
 fortunately most of the germs in the air are not injurious. 
 
 The number of bacteria found in a cubic metre of air at Mont- 
 souris (France) Observatory was 345, while in the same amount 
 of air in the heart of Paris the number was 4790. These figures 
 give some idea of the relative purity of country and city air. 
 
 The dust particles in the atmosphere play an important role 
 in various other ways. They "scatter" the light of the sun, so 
 as to illuminate the whole atmosphere. Without the dust in the 
 air, all shady places would be in darkness. The sun would prob- 
 ably appear in dazzling brilliance, shining from a black sky in 
 which the stars would be visible even in the daytime. The blue 
 color of the sky, and the sunset and sunrise tints, are determined 
 or affected by the dust in the atmosphere. Dust particles also serve 
 as nuclei about which water vapor condenses. It was formerly 
 held that they were necessary for the condensation of water vapor 
 in the atmosphere, but this appears not to be the case. 
 
CHAPTER XIV 
 TEMPERATURE OF THE AIR 
 
 THE temperature of the air varies from season to season, from 
 day to day, and even from one part of a day to another. Because 
 of the importance of temperature in all human affairs, it is con- 
 venient to have some method of measuring and recording it. 
 
 The thermometer. The temperature is measured by means 
 of the thermometer. The principle of the thermometer is readily 
 understood. It consists of a glass tube of uniform diameter, except 
 for a bulb at one end. The bulb and the lower part of the tube 
 are filled with some liquid, generally mercury. The mercury is 
 then heated to its boiling temperature, so as to expel all air. 
 When the tube is full of boiling mercury from which all air has 
 been driven by the heat, it is sealed. 
 
 The mercury contracts on cooling, so that it but partly fills the 
 tube. Above it is a vacuum. When the temperature rises, the 
 mercury in the tube expands and rises. When the temperature 
 falls, the mercury contracts and sinks. The amount of rise or 
 fall of the mercury in the tube indicates the amount of the change 
 of temperature. 
 
 That the temperature may be read directly from the ther- 
 mometer, it is necessary to have a scale marked on the tube. Two 
 scales are in common use the Fahrenheit and the Centigrade. 
 The scales are made as follows: The thermometer tube is placed 
 in boiling water, or in steam just over boiling water, at sea-level 
 (760 mm. pressure), and allowed to remain there until the tube and 
 its contents acquire the temperature of the water. The point to 
 which the mercury rises in the tube under these circumstances 
 is then marked 212, if the Fahrenheit scale is to be used. The tube 
 of mercury is then put into moist pounded ice or snow, where it 
 
 520 
 
TEMPERATURE OF THE AIR 521 
 
 remains until the level of the mercury in the tube becomes stationary, 
 and the level at which the mercury then stands is marked 32. 
 The space between the 212 mark and the 32 mark is then divided 
 into 180 equal parts, each being called a degree (1 Fahr.) The 
 marks on the tube may be made for each degree, for every two 
 degrees, or for every five degrees, according to the delicacy which 
 is required of the thermometer. 
 
 The space below the freezing temperature is divided similarly 
 into degrees, each degree below 32 having the same length on the 
 tube as each degree above. The of this scale is 32 below the 
 freezing-point. The scale is carried still lower on the tube, and 
 the temperature below is called "below zero." Thus 20 be- 
 low zero means 52 below the freezing-point, and is written 20 
 Fahr. or -20 F. 
 
 The Centigrade scale is much simpler and better, though un- 
 fortunately not in common use in English-speaking countries. The 
 height of the mercury at the freezing temperature under normal 
 atmospheric pressure is marked 0, and the boiling temperature 
 100. The space between is divided into 100 parts, each of which 
 is a degree (1 .) The degrees below zero have the same length 
 on the scale as the degrees above. It will be seen that 1 C. is 
 equal to lf Fahr. If this relation of degrees is remembered, de- 
 grees Fahrenheit may be readily changed to degrees Centigrade, 
 or vice versa. 
 
 The Heating of the Atmosphere 
 
 Sources of heat. The atmosphere receives heat from several 
 sources, but that received from the sun so far exceeds that from 
 all other sources that the others may almost be neglected. 
 
 That much heat is received from the sun is shown by the fact 
 that the temperature generally rises when the sun rises and sinks 
 when the sun goes down, and by the further fact that the tempera- 
 ture is generally warmer on a sunny day than on a cloudy one. 
 It is true there are occasional exceptions to these general rules, 
 for now and then a night is warmer than a day, and a cloudy day 
 is sometimes warmer than a sunny one of the same season. But 
 these exceptions do not interfere with the truth of the general 
 statement. 
 
 The source of atmospheric heat which is second in importance 
 is the interior of the earth; but the heat from this source is not 
 
522 PHYSIOGRAPHY 
 
 enough to affect the temperature of the atmosphere sensibly. This 
 is indicated, in a general way, by the fact that in polar regions, 
 during the long night, the temperature is very low, and all the heat 
 received from the interior of the earth has no apparent effect on 
 the snow and ice. 
 
 Sun heating : insolation. The temperature of space is supposed 
 to be about 273 C. ( 459 F.). The more genial temperature 
 which we enjoy results chiefly from the heat received from the sun; 
 yet the earth receives less than 5-<rffWo~o<jotf f ^ ne nea ^ given off by 
 that luminary. The amount received each year, if equally dis- 
 tributed, is enough to melt a layer of ice about 141 feet thick over 
 the entire earth, or to evaporate a layer of water 18 feet deep. 
 
 Each hemisphere receives the same amount of heat from the 
 sun each year (Fig. 518), but. because of the inclination of the 
 earth's axis, the heat is very differently distributed in different 
 latitudes. Other things being equal, the earth gets most heat 
 per unit area where the sun shines the greatest number of hours per 
 day. In summer, the days are longest in the highest latitudes. 
 So far as length of day is concerned, therefore, the highest lati- 
 tudes, namely the poles, should get more heat than any other part 
 of the earth in summer. 
 
 Again, other things being equal, the earth (land or water sur- 
 face) gets most heat per unit area where the sun's rays fall most 
 
 FIG. 530. Diagram to illustrate the unequal heating power of the sun at 
 different altitudes. When its rays are vertical they are concentrated on 
 less space on the surface of the earth, and at the same time pass through 
 less atmosphere, than when they strike the surface of the earth obliquely. 
 
 nearly vertically, both because they are there most concen- 
 trated, and because they there pass through a lesser thickness 
 of the air, which absorbs some of their heat. This is shown by 
 Fig. 530. A given bundle of rays, 1, falling vertically on the sur- 
 face, is distributed over a given space, while an equal bundle of 
 rays, 2 or 3, falling obliquely on the surface, is distributed over a 
 much greater area, and therefore heats each part less. Again, 
 
TEMPERATURE OF THE AIR 523 
 
 the oblique rays, 2 and 3, pass through a greater thickness of at- 
 mosphere, and more of their heat has been absorbed before they 
 reach the surface of the solid part of the earth. The angle at which 
 the sun's rays reach the earth varies from place to place. It also 
 varies at the same place from time to time, because the earth's 
 axis of rotation is inclined to the plane of its orbit as the earth 
 revolves about the sun. This is illustrated by Fig. 529, which has 
 already been explained. 
 
 Primary distribution of heat. Remembering that it is the 
 rotation of the earth on an inclined axis while it revolves about the 
 sun (Fig. 526) which makes the sun appear to move north and 
 south during the year (Fig. 529), we may study the effects of this 
 apparent motion of the sun on the distribution of heat received by 
 insolation. From Fig. 528 we see that when the sun's rays are 
 perpendicular to the surface of the earth 23i south of the equator, 
 they are most oblique at all points in the northern hemisphere, 
 and least oblique at all points in the southern hemisphere. At this 
 time, therefore, the southern hemisphere is receiving more heat 
 than the northern, because of the direction of the sun's rays. At 
 the same time, the days are longer in the southern hemisphere 
 than in the northern, and this is a second reason why the southern 
 hemisphere is receiving more heat than the northern at this time. 
 After the time (winter solstice, December 22) when the sun's 
 rays are vertical at 23^ S., they become perpendicular to the sur- 
 face at points farther and farther north, and on March 21 they 
 are vertical at the equator (Fig. 529). Days and nights are then 
 equal everywhere, and the sun's rays are equally oblique in corre- 
 sponding latitudes north or south of the equator. Any latitude 
 in one hemisphere is then receiving the same amount of heat as 
 the corresponding latitude in the other hemisphere. 
 
 After March 21, the sun appears to continue its journey 
 northward until, on June 21, its rays are vertical at the tropic 
 of Cancer, 23^ N. (Fig. 529), when the days of the northern hemi- 
 sphere attain their greatest length, and the nights of the same 
 hemisphere become shortest in all latitudes where there is alterna- 
 tion of day and night (Fig. 527). At the same time, the rays of the 
 sun are less oblique in the northern hemisphere, as a whole, than 
 at any other time. In the southern hemisphere the conditions are 
 reversed. At this time, therefore, the northern hemisphere is being 
 heated by the sun faster than at any other time of the year, while the 
 
524 
 
 PHYSIOGRAPHY 
 
 southern hemisphere is receiving less heat than at any other 
 time. 
 
 From June 21 to December 22 the sun appears to move 
 so that its rays become vertical farther and farther south, and the 
 preceding sequence of events is reversed. 
 
 The latitudes where the sun's rays fall vertically range from 
 the tropic of Cancer to the tropic of Capricorn. For the whole 
 year, however, the sun's rays are, on the average, least oblique in 
 the lowest latitudes. This is why the low latitudes are, on the 
 whole, warmer than the high latitudes. 
 
 The actual amount of sun heat received in different latitudes 
 is determined by the length of day (hours of sunshine) and the 
 direction of the sun's rays ; but it is to be noted that the latitudes 
 which have the longest days never have the vertical rays of the sun. 
 Calculations based on these two factors have been made, showing 
 the proportion of heat received in different latitudes during the 
 whole year and during different seasons. For the year, the 
 equator receives more heat than any other part of the earth. If 
 the average amount of heat received there each day be taken as 
 1, the amount of heat received in a year is 365.2. The proportionate 
 amount received in various other latitudes is shown in tne following 
 table: 
 
 Latitude 
 
 
 
 10 
 
 20 
 
 30 
 
 40 
 
 50 
 
 60 
 
 70 
 
 80 
 
 90 
 
 Thermal days, 
 
 
 
 
 
 
 
 
 
 
 
 or relative 
 
 
 
 
 
 
 
 
 
 
 
 amount of 
 
 
 
 
 
 
 
 
 
 
 
 heat yearly 
 
 365.2 
 
 360.2 
 
 345.2 
 
 321.0 
 
 288.5 
 
 249.7 
 
 207.8 
 
 173.0 
 
 156.6 
 
 151.6 
 
 From this table it is seen that latitude 40 receives about three- 
 fourths as much heat as the equator, and latitude 70 a little less 
 than one-half as much. 
 
 During the half of the year when the sun's rays are vertical north 
 of the equator, most heat is received in latitude 25 N. During this 
 half of the year the sun's rays are most nearly vertical, on the 
 average, in latitude 11| (half-way between the equator and lati- 
 tude 23); but the days are longer farther north. During the 
 three months centering about June 21, the zone of greatest heat 
 is iri latitude 41 N. The sun's rays are here less nearly vertical 
 than in latitudes about the tropic of Cancer, but the days are much 
 
TEMPERATURE OF THE AIR 
 
 525 
 
 longer. Between May 31 and July 16 l the north pole receives 
 more heat than any other part of the earth, the continuous day 
 offsetting the great obliquity of the sun's rays at this time. At 
 the time of the summer solstice, the area immediately about the 
 north pole receives 20% more heat than an equal area at the 
 
 FIG. 531. Diagram showing receipt of heat in different latitudes of the 
 northern hemisphere for four dates between the vernal equinox and the 
 summer solstice. The latitudes are indicated at the top of the figure, 
 and the relative amounts of heat at the right. (After Wiener.) 
 
 equator ever receives, and 36% more than the equatorial region 
 receives at that time. Fig. 531 shows the amount of heat re- 
 ceived from the sun in various latitudes of the northern hemi- 
 sphere from the time of the vernal equinox to the time of the sum- 
 mer solstice. 
 
 The temperature of one place is not necessarily higher than 
 that of another, because it receives more heat. No amount of heat, 
 for example, would make the temperature of Greenland warm 
 until after the snow and ice was melted. All the heat received tend- 
 ing to raise the temperature above 32 F. would be expended in 
 melting and evaporating the snow, without raising its tempera- 
 ture above 32 F. (0 C.). The region about the north pole does 
 not get very warm, even when it receives more heat than the 
 equator, because much of the heat is expended in melting ice and 
 in warming up ice-cold water, which heats very slowly and runs 
 away as soon as the heating is well begun. 
 
 What the sun does for the earth in the matter of heat is shown 
 
 1 Hann gives these dates May 10 to August 3, a period of 56 days. 
 
526 
 
 PHYSIOGRAPHY 
 
 by the following table, which gives the estimated temperatures 
 (Centigrade) which would exist on the earth in different latitudes 
 if there were no atmosphere. The figures in the upper part of the 
 table are for the warmest and coldest months. 
 
 Equator 
 
 10 
 
 20 
 
 30 
 
 40 
 
 50 
 
 60 
 
 70 
 
 80 
 
 Pole. 
 
 67 
 
 67 
 
 70 
 
 74 
 
 75 
 
 75 
 
 73 
 
 76 
 
 80 
 
 82 
 
 56 
 
 50 
 
 36 
 
 16 
 
 -10 
 
 -45 
 
 -103 
 
 -273 
 
 -273 
 
 -273 
 
 ANNUAL MEANS 
 
 62 
 
 61 
 
 57 
 
 50 
 
 39 
 
 24 
 
 1 
 
 -43 
 
 -81 
 
 -105 
 
 HEATING AND COOLING 
 
 There are three ways in which the air receives, loses, and trans- 
 fers heat. These are radiation, conduction, and convection. 
 
 1. Radiation. When the sun shines, the surface which its rays 
 strike is warmed by the absorption of heat radiated from the sun. 
 An object in front of a fire is warmed by heat radiated from it. 
 A body need not be glowing hot, like the sun, to radiate heat. A 
 hot stove would continue to radiate heat if all the fire were taken 
 out. The body which radiates heat is itself cooled. The hot 
 stove from which the fire has been taken presently ceases to radiate 
 heat. The land warmed by radiation from the sun during the 
 day is cooled by the radiation of its heat during the night. The 
 rate at which a given body loses heat by radiation depends upon 
 the difference of temperature between it and its surroundings. 
 Thus a hot stove will cool much more quickly in a cold room than 
 in a warm one. 
 
 2. Conduction. If one end of a bar of iron, such as an iron 
 poker, be put in the fire, the other end soon becomes hot. The 
 heat passes along the iron rod from one end to the other. 
 This means that the molecular motion, known as heat or heat- 
 energy, set up in one end of the rod by the fire, is passed along from 
 particle to particle to the other end. This method of transmitting 
 heat is conduction. Any cold body in contact with a hot body is 
 warmed by conduction. Thus, the bottom of the air is warmed 
 
TEMPERATURE OF THE AIR 527 
 
 by contact with the land, that is, by conduction, wherever the 
 temperature of the land is higher than that of the air. 
 
 3. Convection. When a kettle of water is placed on a hot stove, 
 the water in the bottom is heated by conduction, that is, by con- 
 tact with the hot kettle. The heating of water causes it to expand, 
 and when the water in the bottom of the kettle expands, it becomes 
 lighter than the water above it. The heavier water above then 
 sinks and pushes the lighter water below up to the top. This sort 
 of movement is convection. Other illustrations of convection are 
 afforded by stoves, fireplaces, etc. A thin sheet of light paper 
 may be momentarily sustained in the air over a hot stove, or even 
 carried up by the rising air of the convection current. Again, 
 as the air in a chimney is heated, it expands and becomes less 
 dense than the air about it. The cooler, denser air about the base 
 
 \ 
 
 It 
 
 FIG. 532. Diagram to illustrate convection in a vessel of water heated at 
 one point at the bottom. 
 
 of the chimney or stove crowds in below the expanded air in the 
 chimney, and crowds it up out of the chimney. Since the air com- 
 ing into the chimney is continually being expanded, the up-draught 
 continues as long as there is fire. Every draught from a chimney 
 is therefore an example of convection. 
 
 It will be seen that in convection the molecules of the gas or 
 liquid change their position relative to one another, while in con- 
 duction in a solid, they do not. 
 
 Convection is of so much importance in connection with the 
 temperature of the air and of water that the process may be 
 analyzed a little more fully. Suppose a vessel of water (Fig. 532) 
 heated at the central point of its bottom. (1) The water heated at 
 a expands, lifting the overlying column of water, producing a very 
 low dome on the surface at 1. (2) Under the influence of gravity, 
 
528 
 
 PHYSIOGRAPHY 
 
 water flows off the dome. There are now unequal pressures at the 
 bottom of the dish. It is greater at c than at a, because there are 
 more molecules above c than above a. (3) Because of the excess 
 of pressure at c, water moves from c to a, displacing (lifting) the 
 warmer water at that point, and producing the upward move- 
 ment indicated in the center of the figure. (4) The centerward 
 movement from c causes the water above c to sink and to occupy 
 the abandoned space, while the lifting of the water over a, by the 
 inflow from c, renews the dome, and the lateral motion from 
 center to side at the surface. 
 
 When the surface of the land is warmed by the heat radiated 
 from the sun, it heats the air above, partly by conduction, but 
 chiefly by radiation. Some parts of the surface are heated more 
 than others. The heated air expands and rises. The beginning 
 of the rise is due to expansion (Fig. 533). If the air in a given 
 region were expanded as shown in Fig. 533, the air at the top of 
 the expanded column would run over (spread), much as water 
 would under similar conditions. After this has taken place, the 
 
 FIG. 533. The initial rise of air, as a result' of heating, is due to the 
 expansion of the part heated. 
 
 amount of air at the base of the column h will be less than the 
 amount at the same level outside the heated are"a, and air from 
 outside the heated column will flow in to balance the deficiency. 
 This inflow will push up the column of warmed and expanded air, 
 and further overflow above will cause more inflow at the bottom. 
 If the heated area continues to be heated, a permanent convec- 
 tion current will be established in the heated area (Fig. 534). 
 
 It is not necessary that the expanding air actually raise the 
 upper surface of the air sensibly, as shown in Fig. 534, to establish 
 a convection current. As it expands upward it compresses the 
 air above the lower heated part (Fig. 535). Where this compres- 
 sion takes place, the air is denser than that at the same level about 
 it, and flows sideways to balance the discrepancy. This is what 
 
TEMPERATURE OF THE AIR 
 
 529 
 
 actually takes place in the air. It will be seen that convection 
 gives rise to horizontal, as well as to vertical, air movements, 
 and that the horizontal movements take place at various levels. 
 
 ^^, 
 
 FIG. 534. The permanent heating of the air over a given region gives rise to 
 permanent convection currents. 
 
 How the sun heats the atmosphere. The atmosphere is 
 heated by the sun in two principal ways: (1) It is warmed by 
 heat radiated from the sun as the rays of the sun come through it 
 and (2) the land and water below the atmosphere are warmed by 
 absorbing the heat radiated from the sun, and then radiate much 
 
 Condensed Air 
 f f 
 
 FIG. 535. Flow of air from above a heated area would take place even if 
 the surface of the air were not raised. 
 
 of the heat they have received. Much of the heat thus radiated 
 is absorbed by the air, which is thus warmed. 
 
 The amount of heat absorbed by the air from the direct rays 
 of the sun is different in different latitudes, and depends chiefly 
 on the distance the rays travel in the atmosphere; that is, on the 
 verticality of the sun's rays (Fig. 530). The amount for different 
 altitudes of the sun is shown in the following table: l 
 
 Altitude of the sun 
 
 
 
 5 
 
 10 
 
 20 
 
 30 
 
 50 
 
 70 
 
 90 
 
 Thickness of the atmos- 
 phere in units 
 
 35 5 
 
 10 2 
 
 5 56 
 
 2 90 
 
 1 99 
 
 1 31 
 
 1 06 
 
 1 00 
 
 Proportion of solar radia- 
 tion reaching the bottom 
 of the atmosphere 
 
 00 
 
 05 
 
 20 
 
 43 
 
 56 
 
 69 
 
 74 
 
 75 
 
 
 
 
 
 
 
 
 
 
 1 Copied from Waldo's Elementary Meteorology, p. 28. 
 
530 PHYSIOGRAPHY 
 
 The heat radiated into the air from land and water is more 
 readily absorbed by the air than the luminous heat radiated by the 
 sun, so that the atmosphere is heated more by radiation from 
 below than by direct insolation. The lowest air is heated most 
 by both earth radiation and insolation, because it is densest, and, 
 being warmed, it gives rise to convection currents which warm 
 the air above. Since convection currents involve horizontal as 
 well as vertical movements, regions which are heated much give 
 of their heat to regions which are heated less. 
 
 Whenever the land and water are .warmer than the air which 
 rests on them, they also warm it by conduction, and convection 
 results. Warmer air also radiates heat to cooler air. 
 
 The heating of land and water. Land and water are heated 
 unequally by the sun, the former being heated four or five times 
 as fast as the latter by insolation. The reasons are several: 
 
 1 . The absorption of a given amount of heat by a given amount 
 of soil or rock raises the temperature of the soil or rock more 
 (about four times as much) than that of the same amount of water; 
 that is, the specific heat of water is higher than that of the land. 
 
 2. Water is a good reflector, while the land is not, and the 
 latter therefore absorbs a larger proportion of the heat of the sun's 
 rays. 
 
 3. The land cools more readily than water. 
 
 4. Convection currents or movements are established in water 
 as soon as its surface is heated locally. This prevents excessive 
 heating at any one point. The land, on the other hand, being 
 solid, is without movements of convection. 
 
 5. There is more evaporation from a water surface than from 
 a land surface, other conditions being the same, and evaporation 
 cools the surface from which it takes place. 
 
 G. Soil and rock are essentially impenetrable to light and heat 
 rays, while water is not. The heat of insolation is, therefore, dis- 
 tributed, at first hand, through a greater thickness of water than 
 of land. Being confined essentially to the surface of the latter, 
 the temperature of the surface is made higher. 
 
 7. Rock is a poor conductor of heat, but water is even poorer. 
 
 Secondary distribution of heat. From what has preceded, it is 
 clear that after the heat is received by the earth from the sun, it 
 is to some extent re-distributed. This re-distribution is accom- 
 plished in part in ways just noted. It is also effected by air move- 
 
TEMPERATURE OF THE AIR 531 
 
 merits (other than convection currents) and water movements 
 (especially ocean currents) both of which are of great importance 
 in the secondary distribution of heat. It has been estimated that 
 without them the average temperature of the equator would be 
 about 131 F., instead of about 80 F. as now, and that of the 
 poles about 108 F. ; instead of as now. 
 
 THE SEASONS 
 
 We are now prepared to understand the seasons, and the 
 reasons for their differences so far as temperature is concerned. 
 In most latitudes the seasons are usually said to be four spring 
 summer, autumn, and winter; but they are not sharply separated 
 from one another, each grading into the one which follows. 
 
 The exact limits of the four seasons are arbitrarily defined. 
 In the United States, March, April, and May are commonly called 
 the spring months ; June, July, and August, the summer months ; 
 September, October, and November, the autumn months; and 
 December, January, and February the winter months. Some- 
 times, however, spring is defined as the time between the vernal 
 equinox and the summer solstice. On this basis, summer is the 
 time between the summer solstice and the autumnal equinox, 
 autumn the time between the autumnal equinox and the winter 
 solstice, and winter the time between the winter solstice and the 
 vernal equinox. In the southern hemisphere spring comes in 
 September, October, and November; summer in December, Janu- 
 ary, and February; and so on. The vernal equinox of the north- 
 ern hemisphere is the autumnal equinox of the southern, and the 
 summer solstice of the northern is the winter solstice of the southern. 
 
 The first of these subdivisions is based primarily on tempera- 
 ture. The summer is made up of the three warmest months, so 
 far as intermediate (temperate) latitudes are concerned, and the 
 winter of the three coldest. The second has an astronomical basis. 
 The limits of the seasons are defined in still other ways in some 
 countries even in middle latitudes, and in some cases the lengths 
 of the seasons, according to popular use of the terms, are not equal. 
 
 In middle latitudes the distinction between the seasons is 
 primarily one of temperature; but in some parts of the earth 
 the distinction between the seasons is based partly, or even largely, 
 on elements other than temperature. Thus, in some regions the 
 
532 PHYSIOGRAPHY 
 
 wet and dry seasons are more distinct than the warm and cold 
 ones. This is true, for example, in some low latitudes where the 
 temperature is always high. In the polar regions, on the other 
 hand, while the temperature of the cold seasons is very much 
 lower than that of the warm ones, there is also a striking difference 
 in the matter of light. At the poles the w r arm season is the light 
 season, and the cold season is the dark one. 
 
 Differences between summer and winter. N Aside from the 
 higher temperature of summer in our latitude (middle latitude of 
 the northern hemisphere), there are certain other obvious dif- 
 ferences between summer and winter. (1) In summer the days 
 are more than 12 hours long, and the nights less. (2) The sun is 
 much higher above the horizon at noon in summer than at the cor- 
 responding hour in winter. This is the same as saying that the sun's 
 rays are less oblique, at any given hour of the day, in summer than 
 in winter (Fig. 526). (3) A third difference between summer and 
 winter in our latitude is the direction in which the sun rises and 
 sets. In summer the sun rises to the north of east, and sets to 
 the north of west. At the equinoxes it rises in the east and sets 
 in the west. In the winter it rises to the south of east and sets 
 to the south of west. (4) The amount of moisture in the air often 
 varies with the season ; but in some regions it is the warm season 
 which is wet, while in others it is the cool season. (5) In some 
 regions the winds change their direction and force with the change 
 of seasons, as will be noted later. The first and second of these 
 differences are the most important, so far as concerns the seasons 
 of middle latitudes. 
 
 The differences between summer and winter, other than the 
 differences of temperature, are dependent primarily upon the 
 differences in temperature. 
 
 Why we have summer when we do. Since the earth receives 
 most of its surface heat from the sun, it follows that the period of 
 the year when the days are long and the nights short must be 
 warmer than the period when the days are short and the nights 
 long; for long days and short nights mean long periods of heat- 
 ing and short periods of cooling daily, while short days and long 
 nights mean short periods of heating and long periods of cooling 
 daily. Not only this, but the sun's rays are more nearly vertical 
 when the days are long, as shown by Fig. 536, and so have greater 
 heating power. It follows, therefore, that during the summer the 
 
TEMPERATURE OF THE AIR 533 
 
 surface is not only heated more hours a day than during the winter, 
 but the heat per hour is greater while the sun shines. These are 
 the immediate reasons why summer is warmer than winter. 
 
 The reasons why the days are longer at one time of the year 
 than another have already been given (p. 498). 
 
 Change of seasons. The change of seasons may be understood 
 from a study of Figs. 526 and 536. We have already seen (1) that 
 the sun's rays are vertical at the equator at the equinoxes, and 
 that the days and nights are then equal everywhere; (2) that the 
 northern hemisphere is being heated most by the sun at the time 
 of the summer solstice, and least at the time of the winter solstice; 
 (3) that the days are longer than the nights in the northern hemi- 
 sphere (except where there is continuous day) from March 21 
 to September 22; (4) that the sun's rays are less oblique in either 
 hemisphere during the half of the year when the days are longer 
 than the nights; and (5) that the relative lengths of day and night, 
 and the angle of the sun's rays, are reversed in each hemisphere 
 every half-year. 
 
 Since the northern hemisphere is being heated most at the 
 time of the summer solstice and least at the time of the winter sol- 
 stice, it would seem, at first thought, that these dates, respectively, 
 should be the middle points of the hot and cold seasons. This is 
 not the case. It therefore follows that the temperature of any 
 given latitude is not altogether dependent on the amount of heat 
 it is receiving from the sun (p. 524). Again, since the value of 
 insolation in corresponding latitudes in the two hemispheres is 
 equal at the equinoxes, it would seem, at first thought, that 
 corresponding latitudes in the two hemispheres should have the 
 same temperature at these times; but this, again, is not the case. 
 In our own latitude, for example, March 21 is much colder than 
 September 22. There is some discrepancy, too, between the tem- 
 peratures of the northern and southern hemispheres on corre- 
 sponding dates in corresponding latitudes, because of the greater 
 preponderance of water in the latter. 
 
 The reason why a place in our latitude is warmer at the time 
 of the autumnal than at the time of the vernal equinox is because 
 the warmth of the summer just passed has not all been lost. At 
 this time, ^theref ore, the northern hemisphere has a temperature 
 higher than that which it would have if it depended solely on 
 daily insolation. On the other hand, the temperature at the time 
 
534 PHYSIOGRAPHY 
 
 of the vernal equinox is lower than that which would seem appro- 
 priate from the insolation then taking place, because the cold of 
 the winter just passed has not been altogether overcome. The 
 cold of the spring is rather more enduring than the heat of the 
 autumn, for it is in some sense "stored up" in the snow, the ice, 
 and the frozen ground. 
 
 Similarly, our summer solstice is not the hottest part of the 
 year in the northern hemisphere, or the coldest in the southern, 
 for the summer's heat has not altogether overcome the effect of 
 the preceding winter in the northern hemisphere, or the effect of 
 the preceding summer in the southern hemisphere. The time of 
 maximum heat therefore lags behind the season of maximum 
 heating. Similarly, the time of maximum cold does not come 
 till after the season of minimum heating. In middle latitudes 
 the lag is about a month, but it is greater over the ocean than over 
 the lands, because the latter are heated and cooled the more readily. 
 
 Seasons in other latitudes. Attention to the subdivisions of 
 the year in latitudes other than our own will help to an under- 
 standing of the fundamental principles involved. At the equator, 
 for example, the sun's rays are vertical twice each year, that is, 
 at the time of the equinoxes. Twice a year, too, the sun's rays 
 are vertical 23^ from the equator, once to the north and once to 
 the south. The equator, therefore, has two seasons, occurring at 
 the time of our spring and autumn, which are somewhat warmer 
 than two other seasons occurring at the time of our summer and 
 winter. The variations in temperature are much less than in our 
 own latitude, for the length of day and night never varies, and the 
 angle of the sun's rays varies but 23^, while with us, in middle 
 temperate latitudes, it varies 47. At the equator, therefore, there 
 is a fourfold division of the year, but the divisions do not corre- 
 spond very closely with those of middle latitudes. 
 
 In high latitudes the conditions are still different. The suc- 
 cession of seasons in latitude 75 N. may be taken to illustrate 
 the conditions in latitudes above the polar circles generally. When 
 the sun's rays are vertical 15 south of the equator, the sun would 
 appear on the horizon at noon in latitude 75 N. (Fig. 536), for this 
 latitude is 90 from the place where the sun's rays are vertical. 
 When they are vertical farther south than 15 S., points on the par- 
 allel of 75 N. will not see the sun. When the sun's rays are vertical 
 in latitude 15 N., or in any latitude farther north, no point on the 
 
TEMPERATURE OF THE AIR 
 
 535 
 
 B 
 
 FIG. 536. Diagram to illustrate seasons in latitude 75. When the sun's 
 rays are vertical at C, the circle of illumination is represented by the 
 line 90-90. The half of each parallel of 75 is then illuminated, and 
 days and nights on that parallel are therefore equal. The same is true 
 of all other latitudes. When the sun's rays are vertical at B, in latitude 
 15 N., the circle of illumination is represented by b-b, the whole of the 
 parallel of 75 N. is illuminated, and daylight is continuous throughout 
 the twenty-four hours. No part of the parallel of 75 S. is illuminated 
 at this time, and on that parallel darkness is continuous. When the sun 
 is vertical at A, in latitude 23 J N., the circle of illumination is repre- 
 sented by a-a. While the sun appears to move from position B to 
 position A and back again to B, the parallel of 75 N. is continuously 
 illuminated, while the parallel of 75 S. at the same time is continuously 
 in darkness. When the sun appears to move from the position where its 
 rays are vertical at B to the position where its rays are vertical at D, a 
 part of each parallel of 75 is illuminated, and during this time, therefore, 
 there is light and darkness in the course of the twenty-four hours. When 
 the sun's rays are vertical between B and C, more than half of the paral- 
 lel of 75 N. is illuminated, and less than half of the parallel of 75 S. 
 When the sun is vertical at C the half of each parallel of 75 (and of all 
 other parallels) is illuminated, and days and nights are equal. While 
 the sun appears to be passing from C to D less than half of the parallel 
 of 75 N. is illuminated, and more than half of the parallel of 75 S. 
 During this time, therefore, nights are longer than days in latitude 75 
 N., and days are longer than nights in latitude 75 S. When the sun is 
 in a position where its rays are vertical at D, the circle of illumination is 
 d-d. At this time all of the parallel of 75 N. is in darkness, and all of 
 the parallel of 75 S. is in light. This condition continues while the sun 
 appears to move on from the position where its rays are vertical at D 
 to the position where its rays are vertical at E, and back again. 
 
536 PHYSIOGRAPHY 
 
 parallel of 75 N. will be in darkness during any part of the twenty- 
 four-hour day (Fig. 536). When the sun's rays are vertical in 
 any latitude between 15 S. and 15 N., a part of the parallel of 
 75 N. will be illuminated, and all points on that parallel will have 
 alternating light and darkness in the course of the twenty-four- 
 hour day. (See also explanation below Fig. 536.) 
 
 Here, too, there is a natural fourfold division of the year: one 
 (summer) when daylight is continuous, one (winter) when dark- 
 ness is continuous, one (spring) when there is alternating day and 
 night with the days lengthening, and one (autumn) when there 
 is alternating day and night with the nights lengthening. In 
 other words, summer, according to this subdivision of the year, is 
 the time during which the sun appears to move from 15 N. to 
 23 N. (B to A, Fig. 536) and back again to 15 N. Autumn is 
 the time during which the sun appears to pass from a position 
 where its rays are vertical 15 N. to a position where its rays are 
 vertical 15 S. (B to D). Winter is the time during which the sun 
 appears to pass from 15 S. to 23J S. (D to E) and back again to 15 
 S., and spring the time when the sun is passing from 15 S. to 15 
 N. (D to ). 
 
 It will be noted that the lengths of the several seasons denned 
 in this way are not the same. In latitude 75 the summer would 
 be as long as the winter, and the spring as long as the autumn; 
 but the spring and autumn would be nearly twice as long as the 
 summer and winter, for during each of the former the sun moves 
 through 30, and during each of the latter but 17. Not only this, 
 but the lengths of the several seasons would vary with the latitude. 
 In latitude 85 the summer and winter would be longer than in 
 latitude 75, and the springs and autumns correspondingly shorter. 
 
 There is a prevalent idea that in polar regions there is a day 
 of six months and a night of six months each year; but it will be 
 seen from the above, as well as from what has been stated before, 
 that this notion is not correct. There is a six-months day and a 
 six-months night at the poles only. 
 
 Effect of varying distance of the sun. Since the orbit of 
 the earth is an ellipse, the distance of the earth from the sun varies 
 in the course of a year. On this account, the amount of heat 
 which the earth receives daily varies a little, being somewhat 
 greater when the earth is nearer the sun, somewhat less when it 
 is farther from it. But the variations in the amount of heat re- 
 
TEMPERATURE OF THE AIR 537 
 
 ceived by the earth, because of its varying distance from the sun, 
 are of relatively little importance in comparison with effects which 
 result from the inclination of the axis. At the present time, the 
 northern hemisphere has its summers when the earth is farthest 
 from the sun (aphelion), and its winters when it is nearest (peri- 
 helion). The southern hemisphere, on the other hand, has its 
 summers when the earth is nearest the sun and its winters when 
 it is farthest from it. This condition of things is reversed every 
 10,500 years. At the present time, the southern hemisphere 
 receives more heat from the sun in a day at the time of the winter 
 solstice in the northern hemisphere, than the northern hemisphere 
 does at the time of its summer solstice. The difference is consid- 
 erable. 
 
 Effect of altitude on temperature. High altitudes are colder 
 than low ones, and the average rate of decrease of temperature is 
 about 1 F. for 330 feet (1 C. for 594 feet) of rise, for altitudes 
 where observations are common. It varies, however, from time to 
 time and from place to place, being especially influenced by the 
 temperature of the surface beneath it. The rate of decrease of 
 temperature for the first 100 feet or so of rise at the bottom of the 
 atmosphere is much more rapid where the land or water is warm. 
 
 The average decrease of temperature with increase of altitude 
 is about 800 times as rapid as its decrease with increase of latitude. 
 In other words, one mile of ascent in the air means about the 
 same decrease of temperature as a poleward movement of 800 
 miles. 
 
 When air rises it expands, because there is less weight of air 
 above it tending to compress it. As a gas expands it is cooled, 
 and as it is compressed it is warmed. Dry air should be cooled 
 about 1 F. for every 183 feet it rises (1 C. for 329 feet). .Moist 
 air cools much less rapidly with expansion, for reasons which will 
 appear later (p. 572). Conversely, air is warmed as it descends 
 and becomes denser. The presence of moisture makes much less 
 difference in the case of descending air, which is warmed at about 
 the same rate as dry air is cooled during its ascent. 
 
 High altitudes are colder than low, primarily because the air 
 is thinner; but, in the case of isolated elevations, also because of the 
 more complete exposure. 
 
 Since the air is thinner, it (a) absorbs less heat from the direct 
 rays of the sun, chiefly because there is less carbonic-acid gas, less 
 
538 PHYSIOGRAPHY 
 
 water vapor, and less dust; and (6) being thinner, it is less effective 
 in retaining the heat radiated from the earth below. 
 
 In sunny days in summer the sunny sides of bare mountain 
 surfaces, when free from snow, get very warm. If the air re- 
 mained in contact with the warm rock surface for long periods of 
 time, it would be notably warmed; but since it is, as a rule, moved 
 on quickly, especially about isolated elevations of notable height, 
 it is not greatly heated before it passes on, and the new air by 
 which it is replaced is much colder than air which has been resting 
 directly on the land. 
 
 On the other hand, there are likely to be many cloudy days 
 in the mountains, and the clouds shelter the rocks from the sun. 
 This tends to reduce the average temperature of the mountain, 
 as compared with that of low land. 
 
 Again, where mountains are sufficiently high and not too steep 
 to retain snow throughout the year, their surfaces are never warmed 
 
 FIG. 537. Diagram to show that the sun's rays may fall less obliquely on a 
 mountain slope than on the plain adjacent. Under these circu.nstances 
 they have greater heating power, so far as the surface ot the land is con- 
 cerned, on the mountain than on the plain. 
 
 above a temperature of 33 F., the melting-point of snow (p. 525). 
 The temperature has been observed in balloons up to elevations 
 of about 30,000 feet, where it was found to be -54 F. This is 
 doubtless much colder than it would be at the top of a mountain 
 30,000 feet high, and very much colder than it would be on a pla- 
 teau at that elevation. The temperature has been recorded by 
 self-registering thermometers in balloons set up to altitudes of 
 ten miles, where the temperature was 104. 
 
 It is to be noted that land surfaces at high altitudes may be 
 heated quite as effectively by the sun as land surfaces at low 
 altitudes. That this is the case is shown by familiar experiences 
 in high mountain regions, where the surface of the rock may be 
 very warm though the air is cool. A mountain surface such as 
 that shown in Fig. 537 may receive the sun's rays much more 
 
pq 
 
TEMPERATURE OF THE AIR 539 
 
 perpendicularly than a flat surface. The rock is correspondingly 
 heated while the sun shines; but as the sun goes down, the heated 
 rock surface cools readily, and may, during the night, become 
 much cooler than the surface of the lower land. 
 
 It is to be noted that only the equatorward sides (the southern 
 sides in the northern hemisphere and the northern sides in the 
 southern hemisphere) of mountains receive the sun's rays more 
 perpendicularly than a flat surface. The poleward slopes of moun- 
 tains (outside tropical latitudes and sometimes within them) receive 
 the sun's rays much more obliquely than flat surfaces, and they 
 receive them fewer hours per day. This serves to reduce the average 
 temperature of mountain regions. 
 
 Representation of Temperature on Maps 
 
 It is desirable to have some method of representing not only 
 the general distribution of temperature over the earth, but various 
 other facts concerning temperature and its variations. Maps 
 showing such phenomena are thermal maps or charts. The princi- 
 ple of thermal charts is simple. 
 
 Isotherms. A line may be drawn on the surface of the earth 
 connecting points having the same temperature. Such a line is 
 an isotherm. An isotherm connecting places having the same 
 average temperature for the year is an annual isotherm.. An 
 isotherm connecting places which have the same summer or the 
 same winter temperature is a seasonal isotherm. Similarly there 
 may be monthly isotherms, daily isotherms, etc. A map show- 
 ing the distribution of isotherms for a year, a season, a month, or 
 a day, is an isothermal map or chart. 
 
 The line of highest temperature about the earth is the thermal 
 equator. This line is not straight, and in general it lies a little 
 north of the geographic equator. 
 
 Isothermal charts. Fig. 538 shows the annual isotherms. 
 It shows an isotherm of 80 enclosing a considerable area in the 
 tropical region extending from the Americas eastward to north- 
 ern Australia. This isotherm shows that all points enclosed by 
 it have an average temperature of more than 80. There are two 
 isotherms of 70, one north of the equator and one south of it. 
 All points between the isotherm of 70 and the isotherm of 80 
 have an average annual temperature of more than 70 and less 
 than 80. In the Pacific, all points between the two 70 isotherms 
 
540 PHYSIOGRAPHY 
 
 have a temperature of more than 70 and less than 80. The 
 map also shows two isotherms of 50, one in the northern hemi- 
 sphere and one in the southern. All points between the isotherms 
 of 50 and 70 have an average temperature between these limits. 
 The warmer portion of these zones in either hemisphere is the 
 portion near the higher isotherm, that is, nearer the equator. 
 
 The chart expresses the general fact that the temperatures are 
 higher in the equatorial regions and lower toward the poles, and 
 this shows that there is a relationship between isotherms and lati- 
 tude. The reason for this relationship has already been explained. 
 
 Fig. 539 shows the isotherms for the month of January. As 
 compared with the preceding map, this shows that the zone of 
 highest temperature, and all isotherms, have been shifted to the 
 south. The fact that the sun is shining vertically some distance 
 south of the equator at this season, seems to be a sufficient reason 
 for the change. This conclusion may be tested by referring to the 
 isothermal chart for July (Fig. 540), for if the conclusion be right, 
 the thermal equator and all isotherms should there be found farther 
 north than in Fig. 538 or Fig. 539. Fig. 540 shows this to be the 
 case. 
 
 Fig. 539 shows that the thermal equator is mostly south of the 
 geographic equator in January, and Fig. 540 shows that the thermal 
 equator is wholly north of the geographic equator in July. In the 
 former case it is in latitude 20 S. (nearly) in South Africa, and in 
 the latter in latitude 40 N. (nearly) in southwestern Asia. In both 
 charts it is farther from the equator on land than on sea. In Africa, 
 the thermal equator is fully 40 farther north in July than in Janu- 
 ary, and in the Americas the shifting is still greater. 
 
 A comparison of Figs. 539 and 540 shows that the range of tem- 
 perature between January and July is greater in high latitudes than 
 in low. Thus in the southern part of Hudson Bay it is 80; at 
 Montreal about 50; in Florida less than 30; and at the equator 
 in South America, less than 10. The same charts show that the 
 range is greater in the interiors of continents than on coasts or over 
 the sea in the same latitude. 
 
 The general distribution of atmospheric temperature in latitude 
 is shown in Fig. 541. 
 
 What determines the positions and courses of isotherms? 
 1. A relationship between isotherms and parallels is suggested further 
 by the fact that the isotherms have a general east-west direction. 
 
TEMPERATURE OF THE AIR 
 
 541 
 
 Some of them are notably irregular, but none of them runs north and 
 south, or anywhere nearly north and south, for any considerable 
 distance. Some of them have a nearly straight east-west course, 
 and, in all, the east-west direction is the general one. Since, how- 
 ever, the isotherms do not follow the parallels exactly, it Is clear 
 that latitude is not the only factor which determines their position. 
 
 Northern Hemisphere Southern Hemisphere 
 
 ^ 
 
 Mean Temperature f< 
 
 .- Globe 
 
 I: 
 
 Fia. 541. Figure showing distribution of atmospheric temperature in lati- 
 tude for the year, for January, and for July; also the mean temperature 
 of the year for the globe. The figures at the left are Fahrenheit , those at 
 the right Centigrade. The numbers at the top represent degrees of 
 latitude. 
 
 Some other cause or causes besides the length of day and the angle 
 of the sun's rays must therefore influence temperature, and so the 
 position of the isotherms. 
 
 2. From Figs. 538, 539, and 540 it is seen that the isotherms are 
 straightest where there is least land, and most crooked where there 
 is much land. This suggests that the land and water have some- 
 thing to do with their positions. Following this idea, it is to be 
 
542 PHYSIOGRAPHY 
 
 noted that, on the January chart, there is an area in South Africa, 
 and another in north Australia, surrounded by the isotherm of 90. 
 Both of these areas are on land, and there is no corresponding area 
 over the sea. It is to be noted also that the areas where the tem- 
 perature is above 80 are wider on the land, and in the vicinity of 
 land, than on the open sea; and furthermore, that in the widest 
 ocean there is no area where the January temperature reaches an 
 average of 80. All these facts tend to confirm the conclusion that 
 the sea and the land influence the position of the isotherms. 
 
 Following this idea still further, it is seen that the isotherms of 
 this map (Fig. 539) frequently bend somewhat abruptly in passing 
 from water to land, or vice versa. Thus the isotherm of 40 in the 
 northern hemisphere turns abruptly to the south when it reaches 
 North America, and again on the coast of Europe. In the southern 
 hemisphere, the isotherms of 80 and 70 make abrupt turns at the 
 west coast of Africa and on or near the west coast of South America. 
 This tends to confirm the conclusion that the relation of land and 
 water has something to do with the position of isotherms. It will 
 be seen later that ocean currents have something to do with the 
 peculiar courses of the isotherms here referred to. 
 
 So far as this chart (Fig. 539) is concerned, it will be seen that 
 the isotherms south of the equator bend poleward on the land in 
 passing from west to east, while those north of the equator bend 
 equatorward. 
 
 The land and the sea are affected differently by the sun's rays 
 (p. 530). The land is heated more readily than the sea in the 
 summer, and therefore becomes warmer. The land also gives up 
 its heat much more readily than the sea, and becomes cooler in 
 winter. The fact that an isotherm, for example the January iso- 
 therm of 40 in the northern hemisphere, bends equatorward in 
 crossing the northern continents, shows that the land is cooler than 
 the water in the same latitude, for the isotherm, in crossing the con- 
 tinent, bends toward the equator to find the same temperature 
 which it had on the water. In the southern hemisphere, on the 
 other hand, where it is summer, the corresponding isotherm, on 
 reaching the land, bends toward the pole in order to find a temper- 
 ature like that of the sea. 
 
 All these phenomena clearly indicate that the position of the 
 land and the sea has something to do with causing the isotherms 
 to depart from the parallels. 
 
TEMPERATURE OF THE AIR 543 
 
 If the preceding inferences are correct, the July isotherms 
 should be in contrast with the January isotherms. The former 
 should bend poleward on the continents in the northern hemisphere, 
 and equatorward in the southern. In Fig. 540, which shows the 
 July isotherms, it is seen that every isotherm crossing North Amer- 
 ica bends poleward on the land, while those crossing the southern 
 continents bend equatorward. The reason is that this is the 
 season when the lands of the northern hemisphere are warmer than 
 the seas of the same latitude, and when the lands of the southern 
 hemisphere are cooler than the seas about them. 
 
 It will be noted that the irregularities of the isotherms of the 
 northern hemisphere in July are much greater than those of the 
 southern hemisphere in January. This is probably because there 
 is much more land in the northern hemisphere than in the southern, 
 and the larger land areas have a greater effect on the isotherms than 
 the smaller ones. 
 
 These facts seem to confirm the inference that land and water 
 influence the position of the isotherms; but does the distribu- 
 tion of land and water account for all the irregularities of the 
 isotherms? 
 
 If the unequal heating of land and sea were the only factor 
 concerned in deflecting the isotherms from the parallels, the bends 
 of the isotherms should be as pronounced on the east sides of the 
 continents as on the west. This is not the case, as shown by Figs. 
 538 and 539. Again, the January isotherm of 50 near the west 
 coast of North America bends chiefly on the land, not at the coast. 
 On the eastern side of the continent the bend of the isotherm of 30 
 is chiefly on the sea, not at the coast. Other isotherms have similar 
 courses. We infer, therefore, that though land and water have much 
 to do with the irregularity of the isotherms, other factors also are 
 involved. 
 
 3. The peculiarities just cited may be explained in part by the 
 winds. The prevailing winds in the middle latitudes of North 
 America are from the west, and the westerly winds tend to carry the 
 temperature of the sea (warmer in winter) over onto the land on 
 the western side of the continent, and the temperature of the land 
 (cooler in winter) over onto the sea on its eastern side. This 
 appears to afford a partial explanation of the bends of the iso- 
 therms of 30 and 50 in the northern hemisphere in January; but 
 it does not afford an explanation of the remarkable northward 
 
544 PHYSIOGRAPHY 
 
 loop of the isotherm of 30 over the eastern side of the North 
 Atlantic, nor of the lesser one over the corresponding part of the 
 Pacific. 
 
 Other illustrations of the effects of winds are furnished by the 
 west coast of the United States. Thus in July (Fig. 540) the land 
 is warmer than the sea, and the cooler temperature of the latter is 
 carried over to the former. The winds therefore make it clear 
 why the bends in the isotherms here are on the land, rather than 
 at the coast or on the sea. 
 
 On the whole, the influence of the winds on the position of the 
 isotherms is less clear from these charts than the influence of land 
 and sea. This is partly because the winds are inconstant, and 
 their effects at one time tend to counteract their effects at another, 
 and the maps show only averages. 
 
 4. The great bend in the isotherm of 40 in the North Atlantic in 
 January is not explained by the relations of land and sea, or by 
 winds. It is due to a warm current of ocean water flowing north- 
 eastward, in the direction of the pronounced loop of the isotherm. 
 The same isotherm is held off the eastern coast of North America 
 by a cold current which flows southward along the east side of the 
 continent. Ocean currents are, therefore, a fourth cause of the 
 irregularities of isotherms. 
 
 The amount of heat carried northward by the ocean currents 
 of the Atlantic and Pacific is very great. Croll 1 has estimated 
 that conveyed from the tropics by the Gulf Stream to be equal 
 to two-fifths of that received by the Arctic regions from the sun. 
 It has been estimated that the temperature of England is raised 
 10 F., that of Norway 16, and that of Spitzbergen 19 by the 
 warm poleward movement of waters in the North Atlantic. These 
 figures have been called into question and are very likely too 
 high; but there can hardly be a reasonable doubt that the north- 
 ward movement of relatively warm water helps to ameliorate the 
 temperature of northwestern Europe, especially in winter. The 
 tempering influence of the poleward drift 2 of warm water is indi- 
 rect. The air over the water is warmed and made moist, and 
 
 1 Climate and Time, p. 27. 
 
 2 The term "Gulf Stream" is of doubtful propriety as applied to the 
 poleward movement of water in the high latitudes of the North Atlantic. 
 The "current" is very indefinite north of the latitude of Newfoundland. 
 
TEMPERATURE OF THE AIR 545 
 
 it is this warmed and moistened air, carried over to the land, which 
 raises the temperature of northwestern Europe. 
 
 It should be noted that the milder climate of northwestern 
 Europe, as compared with northeastern North America, is not due 
 wholly to the poleward drift of warm waters. Even if there were 
 no Gulf Stream, the climate of northwestern Europe would be 
 much more temperate than that of the corresponding latitudes of 
 North America, because the ocean, whence the winds of winter 
 blow to that part of Europe, is warmer than the land whence the 
 winter winds blow to the corresponding latitudes on the west side 
 of the Atlantic. Similarly the heat of summer is less extreme 
 in northwestern Europe than in northeastern North America. 
 
 5. Other minor causes of irregularities in isotherms are found in 
 topographic relations, in the character of the surface, the amount of 
 moisture, etc. A basin region shut in by mountains gets hotter 
 in summer than a region not so surrounded, partly because the air 
 is warmed by heat reflected and radiated in from all sides, as well 
 as by heat reflected and radiated from the bottom, and partly 
 because the enclosing mountains prevent free circulation of the 
 air. There is less evaporation from a dry surface than from a 
 moist one, and since evaporation cools the surface notably, a dry 
 surface will be warmer than a moist one, if other conditions are 
 the same. The color of the soil, the presence or absence of vegeta- 
 tion, etc., also affect the absorption and radiation of heat. 
 
 Topographic relations have much to do with the high tempera- 
 ture (90 and above) in the southwestern part of the United States 
 in July. The dryness of the soil and of the air above it also tends 
 to raise the temperature. Aridity also helps to make the tempera- 
 ture high in the high-temperature area (90 and above) in northern 
 Africa (July) and Australia (January). 
 
 Altitude has a pronounced effect on temperature, as already 
 pointed out; but a study of Figs. 538 to 540 seems to show no 
 relation between isothermal lines and surface relief. The reason 
 is that on isothermal charts all isothermal lines are represented 
 as at sea-level. This is done by making allowance for altitude at 
 the average rate of 1 F. for about 330 feet. Thus, if the tempera- 
 ture of a place at an altitude of 3300 feet is 60, it is put down on 
 the chart as 70 (60 +10). If the place were 6600 feet above 
 sea-level, 20 F. would be added to the temperature recorded by 
 the thermometer. Isothermal charts, therefore, are intended to 
 
546 PHYSIOGRAPHY 
 
 show the temperature as it would be if the land 'were at sea- 
 level. 
 
 Thermal charts may be made to show many other features. A chart 
 may be made to show the departure of the temperature of each place, from 
 the temperature normal to its latitude. Such departure is abnormal tem- 
 perature. Lines connecting places having the same abnormal temperature 
 are is-abnormal or is-anomalous lines. They may be made for the year, 
 for any season, or for any month (Figs. 542 and 543). Charts may be made 
 showing the lines of equal annual range of temperature (Fig. 544), and 
 they may also be made to show the average maximum temperatures (Fig. 
 545) and the average minimum temperatures (Fig. 548). The former are 
 obtained by averaging the highest temperatures of successive years, and 
 the latter by averaging the lowest temperatures of successive years. The 
 absolute maximum and minimum for any place would be the highest and 
 lowest temperatures, respectively, ever recorded for that place. Fig. 547 
 shows the absolute maximum temperature to be more than 120 in the 
 Sahara, and but little less in New South Wales and the southwestern part 
 of the United States. The lowest temperature recorded is in northeastern 
 Asia. 
 
 Isothermal surfaces. A surface might be drawn connecting 
 all points having the same temperature. The annual isothermal 
 surface of 30, for example, would be at sea-level, where the iso- 
 therms of 30 appear in Fig. 538. One of these isotherms is north 
 of the equator and one south of it. Equatorward from these lines, 
 in either hemisphere, the isothermal surface would rise above sea- 
 level. The temperature at sea-level in the northern part of South 
 America is about 80. Its temperature is therefore about 50 above 
 that of the isothermal surface of 30. That surface is here about 
 50 times 330 feet, or 16,500 feet, above sea-level. Where the iso- 
 therm of 50 (Fig. 538) crosses North America, the temperature at 
 sea-level is 20 too high. To find the temperature of 30 in this 
 latitude we must rise high enough into the air to get a reduction of 
 20; that is, 20 times 330 feet, or 6600 feet. 
 
 North of the isotherm of 30 in the northern hemisphere (Fig. 
 538) the temperature at sea-level is less than 30. To find a tem- 
 perature of 30 in this latitude, therefore, we must go beneath sea- 
 level. 
 
 The rate of increase of temperature below sea-level is not the 
 same as above it, and it is not the same for the land as for the sea. 
 Beneath the land the rate of increase is about 1 F. for 60 to 75 
 feet. (The observed rates of increase beneath the surface vary 
 
TEMPERATURE OF THE AIR 
 
 547 
 
 I 
 
 Q 
 
 I 
 
 i 
 
 25 
 
548 
 
 PHYSIOGRAPHY 
 
 .1 
 
 CO 
 
 
 
 . 
 
 S :J 
 
 g 
 1 
 
TEMPERATURE OF THE AIR 
 
 549 
 
550 
 
 PHYSIOGRAPHY 
 
 J 
 
 
 
 O 
 
 pq 
 
 c 
 
 J 
 
 C5 
 I 
 
 1 
 
TEMPERATURE OF THE AIR 
 
 551 
 
 pg 
 
 a, 
 -S 
 
 1 
 
 I 
 
552 
 
 PHYSIOGRAPHY 
 
 w 
 
 I 
 
TEMPERATURE OF THE AIR 
 
 553 
 
 from about 1 F. for 17 feet, to 1 for more than 100 feet.) To find 
 the isothermal surface of 30 F. where the isotherm of 20 F. crosses 
 the continent, we should have to go down far enough to gain 10 F. 
 This would be 600 feet if the rate of increase is 1 for 60 feet, or 
 1000 feet if the rate be 1 for 100 feet. 
 
 The proper conception of isothermal surfaces will be of impor- 
 tance when we come to consider the circulation of the atmosphere. 
 
 FIG. 548. Isothermal chart cf the United States for the year. 
 (U. S. Weather Bureau.) 
 
 Fig. 554 shows sections of the isothermal surfaces, along the 
 meridian of 100 in January and July. These sections are based 
 on the data of Figs. 539 and 540. 
 
 It will be seen from the above that isothermal lines are the lines 
 where the corresponding isothermal surfaces touch the level of the 
 earth which corresponds to sea-level. 
 
 Daily Range of Temperature 
 
 It has been found by experience that the average daily tem- 
 perature of a place may be found by averaging the temperatures 
 of 7 A.M., 2 P.M., and 10 P.M. It is found that the daily range is 
 less above the bottom of the atmosphere than at the bottom, since 
 the lower air is heated much by contact with the land during sunny 
 
554 
 
 PHYSIOGRAPHY 
 
 130' lir 110' 105' 100- 95- 90* 85" 80' 70' 65" 
 
 FIG. 549. Isothermal chart of the United States for January. 
 tU. S. Weather Bureau.) 
 
 125' 120- 115' 110' 105' 100" 95* 90* 65' 8O' K' TV 
 
 10- 105" 100' 95' 90- 95' 80' 75' 70" 
 
 FIG. 550. Isothermal chart of the United States for April. 
 (U. S. Weather Bureau.) 
 
TEMPERATURE OF THE AIR 
 
 555 
 
 FIG. 551. Isothermal chart of the United States for July. 
 (U. S. Weather Bureau.) 
 
 125 ' 120' US' 110 105' 100* 95' 90 65 80 75" 70' 
 
 FIG. 552. Isothermal chart of the United States for October. 
 (U. S. Weather Bureau.) 
 
556 
 
 PHYSIOGRAPHY 
 
TEMPERATURE OF THE AIR 
 
 557 
 
 days, and also because the denser air below is heated more than the 
 rarer air above by direct insolation. It is also found that the daily 
 range is greater when the air is dry than when it is moist, for in 
 the former case less of the heat radiated from the land is absorbed 
 by the air. The daily range is greater far from the sea than in 
 proximity to it, for the sea gets neither so warm nor so cold as the 
 land. Other things being equal, the daily range is greatest when 
 days and nights are nearly equal. 
 
 The daily range of temperature is often as much as 40 or 50 F. 
 in dry, interior regions, and in the Sahara it is sometimes 70. 
 
 The temperature of the day is highest somewhat after noon, for 
 somewhat the same reason that the greatest heat of summer follows 
 the time of greatest heating. To understand this point it should 
 
 ISOTHERMAlg,, p 
 
 FIG. 554. Curves of the isothermal surface of 30 F. A. Along the meridian 
 of 100 W., corresponding to Fig. 539. B. Same, corresponding to Fig. 
 540. The numbers below the horizontal line indicate latitude. 
 
 be understood that the land surface and the air just above it are 
 losing heat by radiation all the time, and receiving heat by insola- 
 tion only when the sun shines. 
 
 Let us suppose the day to be twelve hours long. When the sun 
 rises, the daily heating begins, and the rate of heating increases as 
 the sun climbs above the horizon. The land and the air receive 
 most heat when the sun is highest, that is at noon. After noon the 
 heating becomes less and ends at sunset. Fig. 555, A,shows the curve 
 of insolation. 
 
 When cooling by radiation exceeds heating, the temperature 
 falls. This is the case, as a rule, at night, for radiation goes on 
 after insolation ceases. The temperature becomes lowest about 
 sunrise, when radiation without insolation has been going on longest 
 
558 
 
 PHYSIOGRAPHY 
 
 The land and the lower air continue to radiate heat after sunrise. 
 but both are then heated, and heated as a rule faster than they are 
 cooled by radiation, for the temperature rises. As the temperature 
 rises, radiation increases (Fig. 555, B); but it does not commonly 
 keep pace with insolation, for the temperature continues to rise till 
 some time after noon. The fact that the temperature then begins 
 
 12 246 
 
 10 12 2 4 6 8 10 12 
 
 122468101224681012 
 
 Midnight 
 
 12 2 4 6 
 
 C 
 
 Fio. 555. 
 
 A. Curve of insolation. B. Curve of radiation. 
 
 C. Curves of insolation and radiation combined. The maximum tempera- 
 ture of the day occurs at the higher crossing, the minimum temperature 
 of the day just to the left of the lower crossing of these two lines. 
 
 to fall shows that radiation then exceeds insolation. Fig. 555, C 
 shows the curve of insolation in its relation to the curve of radiation. 
 The average daily range of temperature by months for six 
 places is shown in Fig. 556. San Diego, Phrenix, Shreveport, and 
 Charleston are in about the same latitude (about 33). All are in 
 the zone of westerly winds. San Diego, on the Pacific coast, has 
 an average daily range of about 14 F. Phosnix, which is inland, 
 much higher, and in a dry region, has an average daily range of 
 about 33. Shreveport, which is inland, but low, and in a region 
 
TEMPERATURE OF THE AIR 
 
 559 
 
 of abundant moisture, has an average daily range of about 17. 
 Charleston, on the eastern coast, has a daily range of about 14. 
 Tampa and San Antonio are farther south, and both are affected 
 somewhat by the trades. Tampa is on the coast, while San Antonio 
 
 40 
 35 
 30 
 
 25 
 "0 
 
 Jan. Feb. March April May June July Aug. Sept. Oct. Nov. Dec. Average 
 
 
 
 
 
 , >^ 
 
 o, 
 
 S., 
 
 
 
 
 
 
 
 * 
 
 ^ 
 
 ^ 
 
 
 
 
 
 '^"' 
 
 Fho 
 
 enix, Ai 
 
 iz. 
 
 
 33.1 
 
 
 
 
 "^ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 " 
 
 ESs 
 
 
 
 ?2?i.-- 
 
 ntor.io. 
 
 :~-ri 
 
 Texas 
 
 =--' 
 
 
 
 20.8 
 
 13 
 10 
 6 
 
 
 ^. 
 
 ** 
 
 ^ 
 
 
 1 . 
 
 ^^~. 
 
 . Tan 
 
 Pa, Fla[ 
 
 , 
 
 ~"-l~- 
 
 19.1 
 
 
 
 
 17.3 
 13.6 
 
 
 
 - 
 
 "^ 
 
 ^3 
 
 -^ 
 
 jewofl. 
 
 .S.C. 
 
 --^-* 
 
 ^ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 FIG. 556. Curve showing the average daily range of temperature for cer- 
 tain type stations, for each month. The figures at the right show the 
 average daily range for each station. The great range at Phrenix is the 
 result of high altitude and aridity. (U. S. Weather Bureau.) 
 
 is inland. The former has an average daily range of about 19, 
 while the latter has a range of about 21. 
 
 The Seasonal Range of Temperature 
 
 The seasonal range of temperature is affected by various condi- 
 tions, such as (1) latitude, (2) position with reference to land and 
 sea, (3) prevailing winds, (4) presence of snow during the warmer 
 season, and (5) humidity. 
 
 1. In general the seasonal range of temperature increases with 
 the latitude (compare Figs. 539 and 540), because the range of in- 
 solation increases with the latitude. This range is greatest at the 
 poles, where there is six months of insolation and six months free 
 from it. The great range of seasonal temperature to which the 
 poles would be entitled by their latitude is greatly modified by (2) 
 and (4) of the preceding paragraph. 
 
 2. Islands have a lesser range of temperature than continental 
 lands in the same latitude, and coasts have a lesser range than 
 
560 
 
 PHYSIOGRAPHY 
 
 interiors, because the range of sea temperature is less than the range 
 of land temperature (Fig. 557). A striking and rather extreme illus- 
 tration of the difference between the range of temperature on an 
 island and inland is afforded by Thorshavn (Faroe Islands, Lat. 
 62 N.), where the annual range is 7.9 C., and Yakutsk, in the 
 same latitude, where the range is 61.6 C. 
 
 3. A coast to which the prevailing winds blow from the ocean 
 has a less range of temperature than a coast to which the prevailing 
 winds blow from the land. Thus the range of temperature is less 
 on the Pacific coast of the United States than on the Atlantic in the 
 
 Fio. 557. Diagram illustrating the difference between continental and 
 oceanic temperatures, the former indicated by the full line and the latter 
 by the dotted line. The letters stand tor the months. The numbers are 
 the degrees above and below the average annual temperature of the place. 
 (After Hann.) 
 
 same latitude, the winds being chiefly from the west in both cases. 
 Hann has shown that in Europe, between the latitudes of 47 and 
 52, the temperature changes from west to east are as follows: 
 With every 10 of longitude there is a decrease of 3.1 C. in winter, 
 an increase of 0.7 C. in summer, and a decrease of 1.3 C. in the 
 mean annual temperature. 
 
 4. The presence of snow during the warm season, as in high 
 latitudes and high mountains, prevents a high temperature in sum- 
 mer, even though insolation be strong (p. 525). 
 
 The annual range of temperature is of much importance on vari- 
 ous human affairs. It has some effect on vegetation, and so on all 
 industries connected with the soil. The range of temperature, or 
 more exactly the temperature of winter, has some effect on trans- 
 portation, especially transportation by means of water. Naviga- 
 tion ceases, for example, on the Great Lakes, because ice forms 
 about their borders in winter. The lower limit of temperature also 
 
TEMPERATURE OF THE AIR 561 
 
 affects some phases of mining. Placer mining, for example, is sus- 
 pended in winter in high latitudes and high altitudes, not only 
 because the gravel and sand are frozen, but also because the 
 water needed in the mining is frozen. Other effects of great 
 seasonal range will be readily suggested. 
 
 Effect of Atmospheric Temperature on Atmospheric Movement 
 
 When air is heated it expands and becomes lighter, volume 
 for volume. If we think of the air over a given area as shut in 
 from its surroundings on all sides, but not shut in above, it would 
 expand upward when heated. The result would be that its surface 
 would rise above that of its surroundings. If its surface became 
 higher than that of its surroundings, the upper part of the air 
 would spread (run over) sideways, much as water would under 
 the same circumstances. If some of the air at the top of a heated 
 column runs over, the pressure of the air at the bottom of the 
 heated column is less than that at the bottom of the surrounding 
 air, and, if the air of the surrounding area were not shut off, it would 
 move in from the area of greater pressure (where the air is denser) 
 to the area of less pressure (where the air is lighter). The result 
 would be a horizontal movement at the bottom of the atmosphere 
 (Fig. 534) ; that is, a wind. Unequal heating of the air is, therefore, 
 a cause of air movements, and since the air is being unequally heated 
 constantly, it follows that unequal heating is a constant cause of 
 atmospheric movement. Some of the movements are horizontal 
 and some vertical; some are in the lower part of the air and some 
 in the upper. 
 
 The unequal heating of the air is the immediate cause of certain 
 familiar winds and breezes. 
 
 1. Land- and sea-breezes. During a sunny summer day the 
 land near a lake or sea becomes warmer than the water. The result 
 is that the air over the land becomes sensibly warmer than that 
 over the water on a hot day. The expanded lower air over the 
 land crowds the air above it, and so increases the pressure above 
 the bottom of the air. The result is that the pressure above the 
 bottom of the air over the land is greater than that at the same 
 level over the sea. There is, as a result, (1) a movement of air 
 from the land to the water somewhat above the bottom of the 
 air. This movement diminishes the pressure at the bottom of 
 
562 PHYSIOGRAPHY 
 
 the air on land, and increases the pressure at the bottom of the 
 air over the sea. This gives rise to (2) a breeze from sea to land 
 at the bottom of the atmosphere. This is the sea (or lake) breeze. 
 At night, the air over the land cools and contracts below, and 
 pressure above becomes less than at the same level over the sea. 
 Above the bottom of the atmosphere, therefore, air flows in (3) 
 from sea to land. This increases the pressure at the bottom 
 of the atmosphere on land, and decreases that at the bottom 
 over the sea. A breeze therefore sets seaward (4) from the land 
 to the sea at the bottom of the atmosphere. This is the land- 
 breeze. 
 
 The sea-breeze is best developed in middle and low latitudes 
 during the hot part of the day in summer. When it has the same 
 direction as the prevailing wind, it occasionally develops such 
 strength that business is interrupted and people forced to seek 
 shelter. This is sometimes the case at Valparaiso. Along cer- 
 tain coasts fishermen put to sea in the early morning with the 
 land-breeze, and return at night with the sea-breeze. Land- 
 and sea-breezes will be referred to again in connection with atmos- 
 spheric circulation. 
 
 2. Monsoons. Some lands near the sea become so much heated 
 in summer that the sea (from-sea) wind continues during the hot 
 season, not merely through the hot part of the day, while the 
 land (from-land) wind holds sway during the winter. This is the 
 case, for example, in India. Such winds, which change their direc- 
 tions with the seasons, are monsoon winds. 
 
 The monsoon winds of the Indian Ocean exerted a great 
 influence on the early trade of India. European sailing-vessels 
 formerly timed their outward voyages so as to take advantage of 
 the southwest monsoon, and their return voyages so as to take 
 advantage of the northeast monsoon. 
 
 3. Mountain and valley breezes. At night the air of the moun- 
 tain tops becomes cold, because of its prompt radiation of heat. 
 It thus becomes denser than the air below, and descends, giving the 
 mountain (that is,from-mountain) breeze. The downward flow (or 
 blow) of the air is, however, not confined to mountain valleys, but 
 affects the slopes of the mountains generally. In the morning, 
 especially on sunny days, the air next the land becomes heated, 
 and most at the lower levels. It therefore expands upward, and 
 expands more over the lowland than over the mountains. The 
 
TEMPERATURE OF THE AIR 563 
 
 result is that air moves toward the mountains at this time of day, 
 and strikes them at such an angle as to be deflected upward. This 
 is the valley (really toward-mountairi) breeze. There is also an up- 
 ward movement of the ah* from the mountain slope. This air 
 tends to go straight up, but is crowded over against the mountain 
 by the mountain-ward movement, and so strengthens the moun- 
 tain breeze. At the top of the mountain, horizontal winds take 
 away the air which tends to accumulate there as a result of the 
 valley breeze. 
 
 Mountain and valley, and land and sea breezes, and monsoon 
 winds, are not the only ones due to differences of atmospheric 
 temperature, but they afford the simplest illustrations of air move- 
 ments due to this cause. 
 
 Vertical movements and temperature. It has been stated 
 already that when air rises it expands, and that as it expands it 
 becomes cooler; and, conversely, that when air descends it becomes 
 denser and warmer. These changes of temperature have an im- 
 portant bearing on the condensation and precipitation of atmos- 
 pheric moisture, and will be considered in connection with that 
 topic. 
 
CHAPTER XV 
 THE MOISTURE OF THE AIR 
 
 THE atmosphere always contains some water vapor, even in 
 the driest deserts. We can neither see nor smell water vapor, nor 
 can we recognize it as such by feeling, though air with much water 
 vapor has a different feeling from air with little. 
 
 The presence of moisture in the air, under ordinary conditions, 
 is proved by various familiar phenomena. If a pitcher of ice 
 water is allowed to stand in a warm room, drops of water often 
 appear on the outside of it. This water could have come only 
 from the air. Again, if a vessel of warm air be closed and 
 its temperature lowered sufficiently, the inside of the vessel will 
 become coated with droplets of water. The amount of reduction 
 of temperature necessary to bring about this result is great or slight, 
 according as the amount of water vapor in the air is small or large, 
 or, more exactly, it depends upon the amount which is in the air 
 compared with that which the space occupied by the air is capable 
 of holding. Water vapor sometimes condenses into water in the 
 air, and then becomes visible as clouds or fog. 
 
 Water vapor may be looked upon as an atmosphere by itself, 
 for it is distributed very much as it would be if there were no other 
 atmosphere. Water vapor is five-eighths as dense as dry air; that 
 is, a cubic foot of water vapor would weigh five-eighths as much 
 as a cubic foot of dry air under the same conditions of temperature 
 and pressure. The water vapor of the air displaces some of the 
 oxygen, nitrogen, etc., and so makes the air lighter. 
 
 Function of atmospheric moisture. The function of the mois- 
 ture in the atmosphere is a most important one. Without it no 
 life could exist. In addition to furnishing the rain, the snow, and 
 all the water upon which land life depends, it serves a most im- 
 portant function in connection with temperature, as already in- 
 
 564 
 
THE MOISTURE OF THE AIR 565 
 
 dicated. It appears to be the most important constituent of the 
 air in the absorption both of the heat radiated from the sun and 
 of that radiated from the earth. It increases the average tem- 
 perature at the bottom of the atmosphere and it reduces the 
 extremes of heat and cold which would exist if the air were 
 altogether dry. 
 
 Sources of water vapor: evaporation. It is a familiar fact 
 that a moist surface exposed to the air soon becomes dry, and that 
 water left standing in an open dish will presently disappear. Any 
 fluid, such as ink, which contains much water, also dries up if 
 left uncovered. These familiar experiences illustrate what is 
 taking place all the time wherever moist surfaces are exposed to 
 the air. They are constantly losing water to the atmosphere. 
 We conclude therefore that the water vapor of the air is being derived 
 constantly from all exposed moist surfaces. The conversion of 
 liquid water into water vapor is evaporation. It consists of the 
 passage of molecules from the surface of a liquid or a solid into 
 the vaporous condition. The molecules of a liquid, for example, 
 are in active movement. If they move with sufficient velocity 
 when near the surface of the liquid, they may pass out of the range 
 of the attraction of the other molecules of the liquid, in which case 
 they become vapor. 
 
 Evaporation also takes place from land surfaces which seem 
 dry, for even here the rock, subsoil, etc., beneath the surface is 
 more or less moist, and moisture is continually passing from be- 
 neath up into the atmosphere. Evaporation also takes place from 
 snow and ice, even though the temperature is far below that of 
 melting. This is shown by the fact that snow and ice slowly dis- 
 appear in a temperature below 32 F. A wet cloth, put into a 
 very low temperature, say F., freezes stiff; but if it remains in 
 the same temperature, it presently becomes dry. The ice in it has 
 evaporated. 
 
 All animal respiration also furnishes water to the atmosphere. 
 This is readily shown in winter, when the water vapor of the breath 
 condenses, and so becomes visible, in the cold atmosphere. The water 
 breathed out is not seen in summer, or in a warm room, because it 
 condenses at lower temperatures only. Plants also breathe out 
 moisture, and the amount given off from thrifty growing vegetation 
 is very great. Water vapor is also given off by active volcanoes 
 (p. 368). 
 
566 PHYSIOGRAPHY 
 
 On the whole, water surfaces (oceans, lakes, rivers, etc.) yield 
 more water vapor than equal areas of land surface. The oceans 
 must be looked upon as the ultimate source of most of the water 
 vapor. But for this great reservoir, the waters of the land would 
 presently be exhausted. On the whole, the ocean receives water 
 from rivers, springs, and rains about as fast as it loses it by evapora- 
 tion, so that the amount of water in the ocean remains nearly con- 
 stant from year to year. 
 
 On the average, 30 to 40 inches of rain-water are estimated to 
 fall from the air each year on land; that is, enough to make a layer 
 30 to 40 inches deep if spread over all the land. The amount of water 
 evaporated each year must be about the same as the amount which 
 is precipitated. If the precipitation on the oceans is equal to that 
 on the lands, square mile for square mile, and if all were taken from 
 the oceans and not returned, the oceans would be dried up in less 
 than 4000 years, or, according to the larger figure (40 inches), in less 
 than 3000 years. If this amount of water were evaporated from the 
 lakes of the earth, it would probably exhaust them in less than one 
 year. 
 
 The energy necessary to evaporate this amount of water is very 
 great. Assuming that the average amount of rainfall is 60 inches 
 instead of 30, Strachey has estimated that the energy necessary to 
 evaporate this amount of water and lift it 3000 feet (the average 
 height from which rain falls) would be equal to 300,000 million horse- 
 power constantly in operation. 
 
 Rate of evaporation. Fig. 558 shows the amount of evaporation 
 in inches of water which there would be from a free water surface 
 in various parts of the United States. The evaporation is seen to 
 be highest in the warmer and drier parts of the country. 
 
 Several conditions affect the rate of evaporation. The principal 
 ones are (1) the amount of water vapor in the atmosphere, (2) the 
 temperature of the atmosphere, and (3) the strength of the wind. 
 
 (1) The greater the amount of water vapor in the atmosphere, 
 the less readily does new vapor form and rise into it. The pressure 
 of the water vapor above the surface from which evaporation is 
 taking place seems to be the controlling factor. If it is sufficiently 
 great there will be no evaporation, at least in the sense that there 
 will be no increase of water vapor in the air. Such evaporation 
 as takes place will be balanced by condensation on the evaporating 
 surface. 
 
THE MOISTURE OF THE AIR 
 
 507 
 
568 PHYSIOGRAPHY 
 
 (2) Other things being equal, the warmer the water surface the 
 faster the evaporation. This is illustrated by familiar experiences. 
 Water on a hot stove evaporates sooner than water in a cool place, 
 and water in the sun evaporates, in general, much more rapidly 
 than in the shade. 
 
 (3) The stronger the wind the more rapid the evaporation. 
 The reason appears to be as follows: When the air is still, the space 
 just above a body of water or a moist surface becomes well charged 
 with water vapor, and this tends to retard evaporation; but when 
 the air is in movement, the water vapor is carried away about as 
 fast as it is formed, so that new and often drier air continually 
 comes in over the surface from which evaporation is taking place. 
 If the water vapor formed were moved away as rapidly by some 
 other means, evaporation w r ould go on just as readily as when the 
 wind blows. 
 
 (4) Evaporation is also influenced by pressure of the air, being 
 diminished slightly by increase of pressure. 
 
 The function of the atmosphere in evaporation. The air influ- 
 ences evaporation by its movement, as just noted, and also because 
 it affects the temperature above the land and water (p. 526) ; but 
 evaporation is not dependent on the air. It would go on in a vacuum 
 at a given temperature rather more rapidly than in air at the same 
 temperature. 
 
 Evaporation takes up heat. Evaporation cools the surface from 
 which it takes place. If the hand be moistened, it feels cool as it 
 dries, and the faster the evaporation, as when the wind blows, the 
 more distinct is the cooling. This is why moist clothing seems cooler 
 in the wind than in still air, even when the temperature is the same. 
 It takes about 1000 times as much heat to evaporate a given amount 
 of water as it would take to raise its temperature 1 F. The evap- 
 oration from forested regions in moist tropical lands is so great that 
 the temperature there is often much lower than \vould be expected 
 from the insolation. 
 
 Amount of water vapor in the air. The amount of water vapor 
 in the air varies greatly from place to place, and even in the same 
 place from time to time. Various attempts have been made to 
 estimate the amount in the air at one time, but the results are far 
 apart. Its amount has been estimated as high as 1 per cent, of 
 the weight of the atmosphere (p. 512). This would be equivalent 
 to about 4 inches of water if it were precipitated. This is probably 
 
THE MOISTURE OF THE AIR 
 
 569 
 
 much above the average amount, though very warm air may con- 
 tain locally much more than 1 per cent, of water. 
 
 The following table (p. 570) presents an estimate of the amount 
 of water vapor which the lower part of the atmosphere is capable 
 of holding under different conditions of temperature. Since only 
 about one twenty-fifth of the water vapor is above 30,000 feet, this 
 table shows about twenty-four twenty-fifths of all the atmosphere 
 may hold at these temperatures. 
 
 Some idea of its amount is gained in another way. One cubic foot 
 of air at F. is capable of containing one half grain of water vapor, 
 at 60 F., 5 grains, and at 80 F., 11 grains. The weight of air in a 
 room 40X40X15 feet, at a temperature of 60 F. and under ordi- 
 nary pressure, is about 1800 pounds. The weight of water it is 
 capable of containing is nearly 20 pounds. This would nearly fill 
 a common water-pail. 
 
 Distribution of water vapor. So soon as the water vapor passes 
 into the air it is distributed, partly by winds and partly by diffu- 
 sion. Evaporation at one point, therefore, tends to moisten the 
 air everywhere, though first and most in the region where the evapo- 
 ration takes place. 
 
 The amount of water vapor in the air diminishes rapidly upward, 
 largely because of the low temperature, as shown in the following 
 table: 
 
 Altitude. 
 
 Water vapor. 
 
 Air density. 
 
 Feet. 
 
 
 
 
 
 1.00 
 
 1.00 
 
 13,000 + 
 
 0.24 
 
 0.61 
 
 30,000- 
 
 0.04 
 
 0.32 
 
 Atmospheric moisture and atmospheric movements. Since 
 water vapor makes the air lighter, and since movements result when 
 the air of one place becomes lighter than that of another, inequality 
 in the amount of moisture in the air in different places is a cause of 
 atmospheric movement. 
 
 Saturation. The amount of water vapor in the air at any place 
 at any time depends on the temperature and on the available sup- 
 ply of water. The warmer the air the more the moisture which it 
 can hold. 
 
 When there is all the water vapor in the air which is possible 
 at a given temperature, the air is said to be saturated. It is cus- 
 
570 
 
 PHYSIOGRAPHY 
 
 ternary to speak of the air as being saturated; yet it is in reality 
 not the air which is saturated, but the space which the air occupies. 
 The amount of water vapor necessary to saturate a given space 
 depends on the temperature of the space, and is essentially the same 
 whether air is present or not. It is also sometimes said that the 
 water vapor is saturated. In spite of its inaccuracy, the expression 
 saturation of air is in such common use that it is likely to be retained. 
 Humidity and Dew-point. The amount of moisture which the 
 ah* contains is its absolute humidity. The percentage of moisture 
 which air contains at any temperature, in comparison with what 
 it might contain at that temperature, is known as its relative humidity 
 (Fig. 559). When air contains half the moisture which it might 
 contain it is said to have a relative humidity of 50. When it is 
 saturated with moisture, its humidity is 100. Air is commonly said 
 to be "dry" when its relative humidity is low, and "moist" when 
 its relative humidity is high. The average relative humidity of air 
 over the land is probably about 60 per cent., and that over the ocean 
 about 85 per cent., so that the amount of water actually in the atmos- 
 phere is less than that which might be calculated from the table. 
 The part of our country which is productive, agriculturally, without 
 irrigation, is chiefly where the relative humidity is more than 65. 
 
 Height of column of 
 
 Depth of water which would be in saturated air below the levelo 
 indicated, for the following dew-points at sea-level. 
 
 air above the 
 
 
 ground. 
 
 
 
 
 
 
 80 F. 
 
 70 F. 
 
 60 F. 
 
 50 F. 
 
 Feet. 
 
 Inches. 
 
 Inches. 
 
 Inches. 
 
 Inches. 
 
 6,000 
 
 1.3 
 
 1.0 
 
 0.7 
 
 0.5 
 
 12,000 
 
 2.1 
 
 1.5 
 
 1.1 
 
 0.8 
 
 18,000 
 
 2.5 
 
 1.8 
 
 1.3 
 
 0.9 
 
 24,000 
 
 2.7 
 
 2.0 
 
 1.4 
 
 1.0 
 
 30,000 
 
 2.8 
 
 2.1 
 
 1.5 
 
 1.1 
 
 The relative humidity of the air in dry regions is much greater 
 than is popularly supposed. It is rarely so low as half that of regions 
 moist enough to be productive. Thus at Yuma, Ariz., the average 
 relative humidity for the year is 42.9 per cent., with a mean monthly 
 minimum of 34.7 per cent. The corresponding figures for Santa Fe 
 are 44.8 and 28.7; for Pueblo, 46.2 and 37.6. In Death Valley, 
 Cal., the average relative humidity for five months was 23. 
 
 Any reduction of temperature of saturated air causes some 
 of its moisture to be condensed. The temperature at which air 
 
THE MOISTURE OF THE AIR 
 
 571 
 
572 
 
 PHYSIOGRAPHY 
 
 begins to allow its water vapor to condense is the dew-point. Air 
 which is saturated is therefore at the dew-point. It will be seen 
 that the dew-point is not a fixed temperature, but is influenced by 
 the amount of water vapor in the air. When this amount is large, 
 
 Mdt 
 
 Upon 
 
 50' 
 
 50" 
 
 45 
 40 
 
 45 
 45 
 
 20 
 
 15 
 
 I 
 
 Find Velocity 
 
 20ml 
 
 B 
 
 eo 
 
 70 
 60 
 
 
 
 
 
 
 
 
 
 r 
 
 
 
 
 
 
 
 
 
 
 
 
 ^~ 
 
 1 * 
 
 ito 
 
 
 e<# 
 
 70* 
 60* 
 
 
 
 
 
 
 
 
 
 
 ** 
 
 s^ 
 
 
 
 
 
 
 
 
 ^ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ". 
 
 r= 
 
 I 
 
 1 II 
 
 
 
 ^-* 
 
 f*^~ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 c 
 
 FIG. 560. Graph showing the relations between temperature, wind velocity, 
 and humidity, at Blue Hill Observatory, Massachusetts. (Cox, U. S. 
 Weather Bureau.) 
 
 the temperature of the dew-point is relatively high; when the amount 
 is small, the temperature of the dew-point is relatively low. 
 
 Air may be brought to the dew-point in various ways: (1) It 
 may be carried (by wind) where the temperature is lower, as to 
 a higher latitude or altitude; (2) it may be cooled by having cooler 
 air brought to it, as by a cold wind; (3) it may be cooled by radia- 
 tion, or (4) by expansion. 
 
 Condensation. When the temperature of condensation is above 
 32, the vapor condenses into visible water, which usually takes 
 the form of little droplets. If the temperature of condensation is 
 less than 32, the water crystallizes as it condenses, and takes the 
 form of ice particles. 
 
 Condensation and temperature. When the water vapor of the 
 air is condensed, an amount of heat equal to that absorbed in its 
 evaporation is set free. This is why rising moist air is not cooled 
 
THE MOISTURE OF THE AIR 573 
 
 so rapidly as rising dry air (p. 537). Dry air is cooled about 1 F. 
 for every 183 feet of rise, but saturated air at 68 F. must rise nearly 
 twice as much to be cooled 1 F. This slower rate of cooling is 
 because of the heat set free by the condensation of moisture. 
 
 Dew and frost. It sometimes happens that the temperature of 
 the surface of the land, or of the objects upon it, becomes lower 
 than the temperature of the surrounding air. This is especially 
 likely to be the case in the clear nights of late summer and autumn. 
 If the temperature of the grass blades, for example, becomes lower 
 than the dew-point of the surrounding air at night, moisture from 
 the surrounding air will be condensed on them. Such moisture is 
 dew. Dew does not fall, but condenses on the surface of solid ob- 
 jects. A good illustration of dew is often furnished by the moisture 
 which gathers on the outside of a pitcher of ice-water on a summer's 
 day. The temperature of the pitcher is below the dew-point of its 
 surroundings, and moisture from the air therefore condenses on it. 
 Dew forms on clear nights rather than cloudy ones, because the heat 
 of the day is radiated more readily from the land and the bottom of 
 the air when there are no clouds. Dew is formed on still nights 
 more than on windy ones, because the wind tends to move away the 
 air which is" approaching its dew-point, supplying other air in its 
 place, and the incoming air is often warmer than that moved on. 
 
 When the temperature of the dew-point is below 32 F., the 
 moisture which condenses on solid objects condenses as frost in- 
 stead of dew. Frost is not frozen dew, any more than snow is 
 frozen rain. It stands in the same relation to dew that snow does 
 to rain. In the autumn, frost is more likely to occur in valleys and 
 on low flats than on adjacent hills, because the colder air settles to 
 the lower levels. 
 
 Dew, and sometimes frost, may form on the under sides of ob- 
 jects. If a pan be placed bottom up on the ground, there will be 
 dew on the inside of it in the morning as often as on the outside. 
 There is often dew on the under side of a flat stone when there is 
 none on its top. Even in a desert, a rubber blanket spread on 
 the ground at night will often be wet on the under side in the 
 morning. The explanation is as follows: The air in the ground has 
 some moisture. During the day, when the sun shines, this air is 
 warmed. At night, the air above cools much more quickly than 
 the air in the ground. The cooler heavier air above then sinks 
 into the ground, displacing and crowding up the warmer air below 
 
574 
 
 PHYSIOGRAPHY 
 
 with its water vapor. On reaching the cool pan or the cool rubber 
 blanket, some of the moisture is condensed. If the air in the 
 ground had more moisture than that above ground, water vapor 
 would pass up from below, even if the air with which it is associated 
 did not. IK the daytime the rising moisture would not condense 
 on the pan or blanket, because they would be warmer than the 
 water vapor from below, if the sun were shining; but at night 
 their temperature may be low enough to cause condensation. 
 
 Clouds and fog. The water droplets and the ice particles con- 
 densed from the water vapor of the air take the form of clouds 
 
 FIG. 561. Fog streaming in from the Pacific Ocean. 
 (U. S. Weather Bureau.) 
 
 Coast of California. 
 
 if the condensation takes place without precipitation above th- 
 bottom of the atmosphere, and the form of fog (above 32 F.) or 
 frost (below 32 F.) if in the lower part of the atmosphere. Fogs 
 and air-frost are the same as clouds, except that the former are 
 lower. Fog is, indeed, but a cloud resting on the surface of the 
 land. If moisture condenses and the particles remain suspended 
 in the air about the top of a mountain, there is, to the observer on 
 the plain or in the valley below, a cloud about the mountain; but 
 if the observer were to climb up into the cloud, it would then appear 
 to be fog. Fogs are often formed when the warmer air over a lake 
 in autumn blows over the colder land, or when the air over warmer 
 
THE MOISTURE OF THE AIR 
 
 575 
 
 water from one part of the ocean (e.g., a warm ocean current) blows 
 over colder water. They also often form in valleys at night, espe- 
 cially in autumn, when the night temperatures are much lower than 
 
 FIG. 562. Morning fog over vallevs. Mount Tamalpais, Cal. 
 (U. S. Weather Bureau.) 
 
 those of the day. The cooler air settles in the valleys, which are 
 therefore more likely to have fogs than the uplands are. 
 
 Fogs occasionally lead to shipwreck on sea, and interrupt busi- 
 ness operations on land. A persistent and dense fog in London, 
 December 10 to 17, 1905, was estimated to have cost the city 
 
 FIG. 563. Fog waves. Coast of California. (U. S. Weather Bureau.) 
 
 $1,750,000 per day in one way and another, largely through suspen- 
 sion of business. Such estimates are, however, to be taken with 
 reserve ; since much of the suspended business is transacted later. 
 
576 PHYSIOGRAPHY 
 
 A heavy fog facilitated Washington's retreat to New York after 
 the battle of Long Island. 
 
 The droplets of water in clouds and fogs must be very small to 
 remain suspended in the air. It has been estimated that they are 
 
 FIG. 564. Fog rising and turning to cloud. Mount Tamalpais, Cal. 
 (U. S. Weather Bureau.) 
 
 often about 1/3000 of an inch in diameter, but there is doubtless 
 great variation. 
 
 Clouds also affect temperature by hindering radiation. In gen- 
 eral, cloudiness lowers the summer temperatures of intermediate 
 latitudes, raises their winter temperatures, and gives them a higher 
 average temperature. 
 
 Forms of clouds. Clouds assume many forms. Among the 
 more common are the cumulus, the stratus, the nimbus, and the 
 cirrus clouds. Between these more distinct forms there are all 
 gradations, giving rise to the names cirro-cumulus, cirro-stratus, 
 cumulo-stratus, etc. 
 
 Cumulus clouds are thick clouds, the upper surfaces of which 
 are more or less dome-shaped, with irregular and fleecy protuber- 
 ances. Their bases are nearly horizontal. They appear to be 
 formed by ascending convection currents, and their plane bases seem 
 to mark the level at which condensation takes place as the air rises. 
 They appear especially in clear, hot weather, and most commonly 
 begin to form in mid or late forenoon, after insolation has estab- 
 
THE MOISTURE OF THE AIR 
 
 577 
 
 lished convection currents. They grow as the heat of the day 
 increases, and normally attain their greatest size at about the hour 
 of maximum heat. As evening approaches, they commonly grow 
 smaller. They are frequently dissipated before sundown, but 
 
 FIG. 565. 
 
 FIG. 566. 
 
 FIG. 565. Cumulus (wool-pack) clouds. (Photo, from Cloud Chart, Hydro- 
 graphic Office, Dept. of the Navy.) 
 
 FIG. 566. Cumulus clouds of the fair-weather type. (U. S. Weather 
 Bureau.) 
 
 FIG. 567. 
 
 FIG. 568. 
 
 FIG. 567. Spring cumulus clouds of the rain type. (U. S. Weathei 
 
 Bureau.) 
 FIG. 568. Cumulus clouds at Santa F6, New Mexico. (U. S. Weather 
 
 Bureau.) 
 
 sometimes pass into other forms of cloud (Figs. 565-573 and Fig. 
 576). 
 
 Stratus clouds are horizontal sheets of lifted fog. When the 
 sheet is broken by wind or mountains, it is sometimes called fracto- 
 stratus. 
 
578 
 
 PHYSIOGRAPHY 
 
 Nimbus or rain-clouds consist of thick layers of dark clouds 
 without definite shape, and with ragged edges, from which con- 
 tinued rain or snow generally falls (Fig. 573). 
 
 FIG. 569. FIG. 570. 
 
 FIG. 569. Cumulus clouds; thunder-heads in process of active growth. 
 
 (U. S. Weather Bureau.) 
 FIG. 570. Tumbled cumulus clouds. (U. S. Weather Bureau.) 
 
 FIG. 571. FIG. 572. 
 
 FIG. 571. Cumulus clouds, broken and wind-torn. (U. S. Weather Bureau.) 
 FIG. 572. Alto-cumulus clouds, wave form. (U. S. Weather Bureau.) 
 
 Cirrus clouds are detached, delicate, and fibrous. They are 
 often described as having the form of feathers. They are generally 
 white, and sometimes arranged in belts. They are usually high and 
 thin, and often of particles of snow or ice (Figs. 574-576). 
 
THE MOISTURE OF THE AIR 
 
 579 
 
 Precipitation. The condensation of the water vapor of the air 
 leads to rain, snow, or hail, if the products of condensation fall. 
 Whether precipitation really takes place after the formation of 
 clouds depends on many conditions. To give rain or snow, the 
 
 FIG. 573. FIG. 574. 
 
 FlG. 573. Cumulo-nimbus clouds. (From Cloud Chart, Hydrographic 
 
 Office, Dept. of the Navy.) 
 FIG. 574. Cirrus clouds. (From Cloud Chart, Hydrographic Office, Dept. 
 
 of the Navy.) 
 
 FIG. 575. FIG. 576. 
 
 FIG. 575. Cirro-stratus clouds. (U. S. Weather Bureau.) 
 FIG. 576. Cirro-cumulus clouds; mackerel sky. (U. S. Weather Bureau.) 
 
 particles of water or snow in the cloud must be heavy enough to 
 fall; and if they are to reach the bottom of the atmosphere, they 
 must not pass through air which is dry enough and warm enough 
 to evaporate them before they reach the bottom of the atmosphere. 
 
580 PHYSIOGRAPHY 
 
 Whether precipitation takes the form of rain or snow depends not 
 only on the temperature of condensation, but also on the tempera- 
 ture of the air over the place where the precipitation takes place. 
 Precipitation which starts as snow may become water before it 
 reaches the bottom of the air. It often snows on a mountain when 
 it rains in the valley below. Precipitation which starts as water 
 rarely freezes as it descends, though some hail may be regarded as 
 frozen rain. 
 
 Since condensation follows cooling, and since precipitation fol- 
 lows condensation, sufficient cooling (below dew-point) of the air 
 may cause precipitation. It follows that there may be rain (or 
 snow) (1) when air is blown up a cold mountain-side, (2) when it is 
 blown poleward (or, in general, from a warmer to a cooler place) 
 without rising, (3) when it rises by convection, both because (a) it 
 is cooled by being brought to cooler air, and (6) because it expands ; 
 (4) when cooler air is brought to warmer air. Rains due to (1) are 
 not rare in mountain regions, and rains due to (3) are common 
 where convection currents are strong, as in the region of tropical 
 calms, where precipitation occurs almost daily during the hottest 
 part of the day. 
 
 The distribution of rainfall is dependent, in large measure, on the 
 winds, and will be considered later. 
 
 Rain-making. Various attempts have been made to produce 
 rain by means which may be called artificial. The methods em- 
 ployed have been various, but the results have been uniformly 
 unsuccessful. The method most tried has been that of producing 
 explosions of one sort or another in the air well above the land- 
 If there were cloud particles in abundance in the air, such dis- 
 turbances might perhaps have the effect of causing them to unite 
 and so to become large enough to fall; but the amount of rain- 
 fall which can be thus produced, under the most favorable 
 conditions, is probably too small to be of consequence. Other 
 methods which have been tried or suggested seem equally futile. 
 
 Summary. The air is constantly taking up moisture from all 
 moist surfaces. This moisture, in the form of invisible vapor, is 
 diffused and blown everywhere. When it reaches a temperature 
 which is low enough (the dew-point), the moisture is condensed. 
 If it condenses in the upper air, it may fall as rain or snow, or it 
 may remain suspended in the air in the form of a cloud and be 
 evaporated again. If it condenses on the surface of solid objects 
 
THE MOISTURE OF THE AIR 581 
 
 at the bottom of the atmosphere, it forms dew or frost. Water 
 vapor is thus in constant circulation, and all land life depends upon 
 it. Some of the water which is precipitated out of the atmosphere 
 falls on the surface from which it was evaporated, but much of it 
 falls in places far distant from those whence it was evaporated. 
 
CHAPTER XVI 
 ATMOSPHERIC PRESSURE 
 
 THAT the air is substantial and has weight is shown by the 
 familiar phenomena cited on page 506. Its downward pressure 
 or weight has already been stated to be, on the average, nearly 
 15 pounds to the square inch at sea-level. Differences in atmos- 
 pheric pressure are the primary cause of atmospheric movements, or 
 winds, and winds are of so much significance, in one way and an- 
 other, that it is convenient to have some standard method of 
 measuring and recording atmospheric pressures. 
 
 The pressure of the atmosphere is measured by the barometer. 
 
 The barometer. The principle of the ordinary barometer is 
 as follows: A tube more than 30 inches long, closed at one end, is 
 filled with mercury, and the tube is then placed, open end down, 
 in a dish of mercury (Fig. 577). The mercury in the tube will 
 sink until its upper surface reaches a level about 30 inches above 
 the level of the mercury in the dish, if the place of the experiment 
 be near sea-level. The mercury remains at this level in the tube 
 because the pressure of the air on the mercury in the dish is suf- 
 ficient to balance the downward pressure, or weight, of the mercury 
 in the tube. Since the pressure of the air at sea-level holds the 
 mercury in the tube up about 30 inches or 760 millimetres, the 
 pressure of the air at sea-level is said to be 30 inches or 760 milli- 
 metres. If the atmospheric pressure becomes less, the mercury 
 in the tube falls, and if the atmospheric pressure becomes greater, 
 the mercury in the tube rises. 
 
 At elevations above sea-level the pressure becomes less be- 
 cause more of the air is left below, and the higher the ascent, the 
 less the pressure, as shown in the following table. 
 
 582 
 
ATMOSPHERIC PRESSURE 
 
 583 
 
 Altitude above Barometric pressure 
 
 sea-level in feet. in inches. 
 
 30 
 
 1,800 28 
 
 3,800 26 
 
 5,900 24 
 
 8,200 22 
 
 10,600 ; 20 
 
 13,200 18 
 
 16,000 16 
 
 Altitude above sea-level may be measured by means of baro- 
 metric pressure; but since mercurial barometers are not conve- 
 niently carried and are easily broken, another 
 form of barometer, the aneroid barometer, has 
 been devised for this purpose. 
 
 Air pressures unequal. The general facts set 
 forth in previous chapters make it clear that the 
 pressure of the atmosphere must vary from point 
 to point, and from time to time at the same point. 
 Some of the reasons are as follows: 
 
 1. The temperature of the surface on which 
 the air rests is unequal, and increase of tempera- 
 ture makes the air lighter. Not only this, but the 
 temperature in a given place varies from time to 
 time. It follows that the pressure of the air at a 
 given place varies from time to time. 
 
 2. A cubic foot of dry air at a temperature of 
 68 weighs 523.72 grains under a pressure of 30 
 inches. A cubic foot of saturated air under the 
 same conditions weighs 4.26 grains less (p. 564). 
 On the whole, the amount of moisture in the air is 
 greater in warm regions than in cold ones, and 
 greater over the sea and moist lands than over 
 dry regions. Since the amount of moisture in the 
 air at a given place varies from time to time, the 
 pressure is being constantly disturbed. 
 
 If temperature and moisture were the only 
 
 IG. 577. Dia- 
 gram to illus- 
 trate the princi- 
 ple of the bar- 
 o m e t e r. The 
 pressure of the 
 air at A main- 
 tains the mer- 
 cury at / in the 
 tube when there 
 is no air in the 
 tube above B. 
 
 factors controlling air pressure, it should be least in low latitudes, 
 where it is warmest and where there is abundant moisture. In 
 other words, it should be least where the isotherms are highest, 
 
584 PHYSIOGRAPHY 
 
 especially over moist regions, and greatest in cold regions, where 
 the air is relatively dry. Since the distribution of atmospheric 
 pressure does not correspond with these general rules, as we 
 shall see, and since changes of pressure in a given region take 
 place independently of changes in temperature and moisture in 
 that region, it follows that factors other than temperature and 
 moisture influence atmospheric pressure. 
 
 Representation of Pressure on Maps and Charts 
 
 Isobars. Lines may be drawn on the surface of the earth con- 
 necting points where the atmospheric pressures are equal. Such 
 lines are isobars. A map showing lines of equal pressure is known 
 as an isobaric map or chart . An isobaric chart for the year, that is, 
 an annual isobaric chart, shows isobars connecting points having 
 the same average pressure throughout the year. There may be 
 isobaric charts for the several seasons and for the several months, 
 and there may be charts for any shorter period. The daily weather 
 maps are daily isobaric charts. 
 
 Fig. 578 represents an isobaric chart for the year. The figures 
 on the lines indicate the average pressure for the year in inches. 
 The isobars of 30 inches or more are full lines; those of less than 
 30 inches are dotted lines. A few suggestions will help in the in- 
 terpretation of the map. In the southern hemisphere, the isobar 
 of 30 inches encloses a belt extending almost around the world. 
 It is interrupted only in the vicinity of Australia. Every point 
 within the area enclosed by this isobar has an average atmospheric 
 pressure of more than 30 inches. Every point within the isobar 
 of 30.10 inches has an average annual pressure of more than 30.10 
 inches, while every point between the isobars of 30.00 and 30.10 
 has an average annual pressure of between 30.00 and 30.10 inches, 
 etc. Between the two adjacent isobars of 29.90 in the equatorial 
 part of the Atlantic, the pressure is less than 29.90, but not so low 
 as 29.80. If the pressure sank to the latter figure, there would have 
 been isobars of 29.80 inches. 
 
 It will be noted that the pressure within the areas enclosed by 
 the isobars of 30.10 in the South Atlantic is more than 30.10, while 
 the pressure between the adjacent 29.90 isobars of the mid-Atlantic 
 is less than 29.90. In explanation of the difference, it is to be noted 
 that as the 30. 10 isobar is approached from without, the pressure is 
 increasing; and that as the 29.90 isobars are approached from pole- 
 
ATMOSPHERIC PRESSURE 
 
 585 
 
586 
 
 PHYSIOGRAPHY 
 
 ward, the pressure is decreasing. By the application of this principle 
 the interpretation will be seen. 
 
 In the interpretation of isobaric charts another point should be 
 understood. The pressure of the atmosphere diminishes with in- 
 creasing elevation, as indicated in a general way by the table on 
 page 583. It is shown in greater detail in the following table, 
 which shows the height of a column of air, at different temperatures, 
 corresponding to 0.1 of an inch of pressure: 
 
 HEIGHT OF AN AIR COLUMN CORRESPONDING TO 0.1 OF AN INCH BAROMETRIC 
 PRESSURE, AT VARIOUS TEMPERATURES. 
 
 Air Pressure in 
 Inches. 
 
 Average Temperature in Degrees Fahrenheit. 
 
 20 
 
 30 
 
 40 
 
 50 
 
 60 
 
 70 
 
 80 
 
 22 
 
 Feet. 
 116 
 111 
 106 
 102 
 98 
 94 
 91 
 88 
 85 
 
 Feet. 
 119 
 114 
 109 
 105 
 101 
 97 
 93 
 90 
 87 
 
 Feet. 
 122 
 116 
 111 
 107 
 103 
 99 
 95 
 92 
 89 
 
 Feet- 
 124 
 119 
 114 
 109 
 105 
 101 
 98 
 94 
 91 
 
 Feet. 
 127 
 124 
 116 
 112 
 107 
 103 
 100 
 96 
 93 
 
 Feet. 
 130 
 124 
 121 
 114 
 110 
 106 
 102 
 98 
 95 
 
 Feet. 
 132 
 126 
 121 
 116 
 112 
 108 
 104 
 100 
 97 
 
 23 
 
 24 
 
 25.. 
 
 26 
 
 27 
 
 28 
 
 29 
 
 30 
 
 
 If, for example, one ascends 95 feet from sea-level, where the 
 temperature is 70 F. and the pressure 30 inches, the pressure of 
 the atmosphere is reduced by 0.1 of an inch. At a level where the 
 pressure is but 28 inches (1800 feet above sea-level; see p. 583), 102 
 feet of ascent would be necessary to reduce the pressure 0.1 of an 
 inch. 
 
 It will be recalled that the temperatures shown oti an isothermal 
 chart are not those actually observed, but that allowance is made 
 for altitude above sea-level. Similarly, the pressures shown on an 
 isobaric chart are not those actually observed on the land. They 
 are the pressures which would exist if there were no elevations 
 above sea-level. Before being recorded on an isobaric chart, the 
 observed atmospheric pressure of a place 95 feet above sea-level, 
 when the temperature is 70 F., has 0.1 of an inch added to it if the 
 observed pressure was 30 inches. If the temperature were lower, 
 0.1 of an inch would be added for a lesser height, since colder air is 
 heavier. Thus at a temperature of 40 F., 89 feet of rise 'makes a 
 difference of 0. 1 of an inch in the pressure of the atmosphere. 
 
ATMOSPHERIC PRESSURE 
 
 587 
 
 Isobaric surfaces. An isothermal surface connects places having 
 the same temperature. So an isobaric surface connects places hav- 
 ing the same pressure, that is, the same amount of air above. If, 
 for example, one place at sea-level has a pressure of 30 inches, and 
 another a pressure of 29.80 inches, the isobaric surface of 30 inches 
 would lie beneath sea-level where the pressure is but 29.80 at sea- 
 level. If the temperature of the place be 70, it would be necessary 
 to descend about 190 feet below sea-level, at the place where the 
 
 FIG. 579. A series of isobaric lines showing diminishing pressure toward the 
 
 center. 
 
 29.70" 
 ~ -28.80-"' 
 
 --29.90- ' 
 -30.90 ' 
 
 FIG. 580. Section through the area represented in Fig. 579, showing the 
 position of isobaric surfaces. As the pressure toward the center of the 
 area shown in Fig. 579 diminishes, the isobaric surface bends downward. 
 It will be seen that isobaric lines are the lines where isobaric surfaces 
 cut sea-level. " 
 
 pressure is 29.80 inches at the surface, to find the level where the 
 pressure is 30 inches. If the observed pressure at another place at 
 sea-level be 30.10 inches, the isobaric surface of 30 inches would 
 rise above sea-level there. These relations are shown in Figs. 579 
 and 580. The former is a series of isobaric lines, with pressures 
 varying from 30.00 to 29.70 inches; the latter is a vertical sec- 
 tion through such an area, to show the isobaric surfaces. From 
 these figures it is seen that the isobaric lines (Fig. 579) are the lines 
 where the isobaric surfaces cut the plane of sea-level. 
 
588 PHYSIOGRAPHY 
 
 If a surface of water had the form shown in Fig. 580, the water 
 from the higher parts would flow to the lower parts until the sur- 
 face became level. The air, which is more fluid than water, behaves 
 in a similar way, and moves down every isobaric surface which 
 has slope. Such movements are winds. When the isobaric slope 
 is great, or, in other words, when the isobaric gradient is high, the 
 wind is strong; when the isobaric gradient is low, the wind is 
 gentle; and when there is no isobaric gradient, that is, when 
 the isobaric surface is level, there is no wind. The strong wind is 
 strong for much the same reason that a swift river is swift; the 
 gentle wind is gentle for much the same reason that a slow river 
 is sluggish. 
 
 Isobaric charts have their highest value in showing the direc- 
 tion and the strength of winds, and winds are determined by iso- 
 baric surfaces. In order to know about the winds of a given place, 
 we must compare the pressures of adjacent areas at the same level. 
 For example, it is not the difference in pressure between the top of 
 Pike's Peak and Denver, as measured at the two places by a barom- 
 eter, which is of consequence in determining winds between these 
 places, but it is the pressure at the top of Pike's Peak, as compared 
 with the pressure at the same elevation over Denver, which is sig- 
 nificant. In Fig. 581 it is the relation of the pressures at A and B, 
 not that between A and D, which is significant. If the isobaric 
 surface at A extends as a plane to B, there will be no wind between 
 the two places, because the isobaric surface has no gradient. 
 
 To determine what the winds are to be, therefore, we must com- 
 pare pressures at the same level. This is why all isobars are reduced 
 to sea-level, on isobaric charts. 
 
 The courses of isotars. Returning now to Fig. 578, several 
 points are readily seen. (1) The isobars in general have an east- 
 
 FlG. 581. It is the atmospheric pressure at the same level in adjacent areas 
 which determines movements of air. 
 
 west course, though many of them are irregular; (2) they are in 
 general higher in low latitudes than in high latitudes; (3) they are 
 highest in the latitudes just outside the tropics; (4) they are more 
 
ATMOSPHERIC PRESSURE 589 
 
 regular in the southern hemisphere than in the northern; and (5) 
 they are, on the whole, more irregular on the land than on the sea. 
 
 Isobars and parallels. Though many of the isobars are very 
 irregular, their general courses are east-west, and none of them 
 have a north-south course for any considerable distance. In this 
 respect they correspond in a general way with isotherms (Fig. 538). 
 Furthermore, the extra- tropical belts of high pressure have an east- 
 west course, and are therefore essentially parallel to the parallels. 
 
 We have now to inquire why the isobars follow, in a general 
 way, the parallels. 
 
 It has already been seen that isotherms tend to follow parallels. 
 Is it the latitude, or the distribution of temperature which is largely 
 determined by latitude, which influences the pressure, and therefore 
 determines the position of the isobars? Or is there some other 
 cause which controls or influences their position? 
 
 Low latitudes have higher temperatures than high latitudes; and 
 increase of temperature expands the air, and so makes it lighter. 
 If, therefore, temperature controls the position of isobars, they 
 should be lowest at the equator and highest at the poles. Fig. 578 
 shows not only that this is not the case, but that pressures are dis- 
 tributed in apparent defiance of temperature. The isobars are 
 highest neither where it is coldest nor where it is warmest; they 
 are highest neither in the lowest nor in the highest latitudes. It 
 is clear, therefore, that neither latitude nor temperature, nor both 
 together, control the position of isobars. 
 
 It does not follow, however, that these factors have no effect 
 on atmospheric pressure; and if the principles thus far developed 
 be correct, atmospheric temperature must affect atmospheric pressure. 
 The only inference, therefore, which we are warranted in making 
 at this stage is that latitude and temperature are not the chief factors 
 which determine the distribution of atmospheric pressure, and therefore 
 of isobars. It will be seen in the sequel, however, that temperature 
 is really the fundamental factor, though its effect is, in part, indirect. 
 
 Relation of isobars to land and water. Let us see if further 
 study of the isobaric charts will throw additional light on the dis- 
 tribution of atmospheric pressure. 
 
 The isobars are much more regular in the southern hemisphere, 
 where there is much water, than in the northern hemisphere, where 
 there is less water and more land. In this respect they have some 
 relation to isotherms. (Compare Fig. 538.) 
 
590 PHYSIOGRAPHY 
 
 The map (Fig. 578) also shows that the high-pressure belt in either 
 hemisphere centering about latitudes a little above 30 is somewhat 
 regular in the southern hemisphere, where water is abundant, but 
 very irregular in the northern hemisphere, where there is more land. 
 It is suggested therefore that the distribution of land and water may 
 influence the position of isobars. It will be remembered that this 
 was one of the factors influencing the position of isotherms, because 
 the land is warmer than the sea in the same latitude in summer, and 
 cooler in winter, and anything which influences temperature should 
 influence pressure also. 
 
 If temperature influences the position of the isobars, this influ- 
 ence should appear on seasonal or monthly isobaric charts. 
 
 Isobars and Temperature. The isobaric map for January (Fig. 
 582) shows that the high-pressure belt is much expanded in iho 
 northern hemisphere (winter), especially on the land (compare Fig. 
 578), and much contracted in the southern hemisphere (summer). 
 Since the pressure is high (above 30 inches) over a greater area in 
 the hemisphere which has winter, the map suggests that the low tern - 
 perature of this hemisphere at this season may be a cause of tho 
 widened area of high pressure. 
 
 This inference may be tested further from this chart. The belt 
 of high pressure in the northern hemisphere (Fig. 582) is much wider 
 on land than on the sea. Since the land is cooler than the sea durinj; 
 January, the inference that high pressure goes with low tempera- 
 ture seems to be supported. In the southern hemisphere, Janu- 
 ary is a summer month, and the land is warmer than the sea. If 
 increasing temperature causes low pressure, the pressure should there 
 be lower on the land than in the sea. The map shows this to be 
 the case. This chart therefore seems to show that high tempera- 
 ture tends to reduce the pressure, for (1) the width of the high-pressure 
 belt is greater in the hemisphere which has winter; (2) the width 
 of the high-pressure belt is greater on the cooler land than on the 
 less cool sea in the hemisphere which has winter; and (3) the pres- 
 sure over the land is less than that over the sea in the hemisphere 
 where the land is warmer than the sea. 
 
 This inference may also be tested by the isobaric chart for July 
 (Fig. 583). At that time of year, the high-pressure belt in the 
 southern hemisphere (winter) should be expanded, especially on 
 land, while that in the northern hemisphere should be contracted, 
 especially on the land. Fig. 583 shows this to be the case. We 
 
ATMOSPHERIC PRESSURE 
 
 591 
 
 - 
 
 A 3t 
 
592 PHYSIOGRAPHY 
 
 therefore return with increased confidence to the conclusion that 
 high temperature reduces the pressure, while low temperature increases 
 it. The charts furnish much more evidence in support of the same 
 conclusions. Some of them are the following: 
 
 1. Figs. 582 and 583 show that the atmospheric pressure changes 
 from season to season in the same place. Thus in January the 
 pressure over the larger part of the United States exceeds SO inches, 
 while in July it falls short of 30 inches. Similarly the pressure in 
 southern Africa exceeds 30 inches in July (^winter), and falls short 
 of it in January (summer). The pressure over much of Asia exceeds 
 30 inches in January, and falls short of it in July. Other illustra- 
 tions of the same sort may be found on the maps. In many cases, 
 therefore, increased temperature goes urith decreased pressure, s shewn 
 by Figs. 582 and 83. 
 
 2. It will be seen from Figs. 582 and 583 that the difference 
 between the pressure in January and July is greater in Asia than 
 elsewhere, being, at the maximum, nearly an inch. In North America 
 and southern Africa it is about 0.40 of an inch, while in Europe and 
 South America it is still less. The seasonal range of pressure is 
 greater on large land areas than on small ones. This is in keeping, 
 it will be noticed, with seasonal changes of temperature (compare 
 Figs. 539 and 540), and is another confirmation of the close rela- 
 tionship between isobars and isotherms. 
 
 3. Again, it is to be noted that the center of the high-pressure 
 belt in the northern hemisphere in January is in latitude SG cr a 
 little less, with great expansions of the belt to the northward on 
 land. The centre of the high-pressure belt in the southern hemi- 
 sphere at the same time is in latitude about 35. In July, en the 
 other hand, the centre of the high-pressure belt in the northern 
 hemisphere is in latitude about 3o, and in the southern, in latitude 
 about 30. That is, the centres of the high-pressure belts shift in har- 
 mony with the apparent motion of the sun. 
 
 4. It is to be noted, also, that the seasonal variation of pressure 
 on the sea is not, in general, so great as that on land. The seasonal 
 change of temperature is also less on the sea (compare Figs. 39 
 and 540). 
 
 5. The high-pressure (more than 30 inches) belt in each hemi- 
 sphere is not only greatly contracted in the summer (July in the 
 northern hemisphere and January in the southern, Figs. 582 and 5S3), 
 but it is interrupted in each hemisphere on land. This suggests 
 
ATMOSPHERIC PRESSURE 
 
 593 
 
594 PHYSIOGRAPHY 
 
 that the relations of sea and "land influence pressure. Since sea 
 and land influence temperature, their influence on pressure may 
 be only a result of their influence on temperature. 
 
 A relationship between temperature and isobars is clear, but 
 it is also clear that temperature does not afford a full explanation 
 of the distribution of pressures as shown by the isobars. The expla- 
 nation of the high pressures outside the tropics, and the low pres- 
 sures in high latitudes, a feature which appears on all the charts, 
 is not found in temperature. 
 
 Isobars and humidity. We have seen (p. 564) that water vapor 
 makes the air lighter. Are the isobars lowest over the oceans in 
 warm latitudes, where the air contains on the average most moisture? 
 Figs. 578, 582, and 583 show that this is not the case. It is rea- 
 sonable to conclude, therefore, that the amount of moisture in the 
 air is not the chief factor controlling the isobars, though atmospheric 
 moisture must influence atmospheric pressure. 
 
 Inequalities of temperature and moisture in the air are the only 
 factors thus far studied which might affect the isobars; and since 
 they do not explain the most striking feature in the distribution of 
 atmospheric pressure, namely, the high pressures in low latitudes, 
 we conclude that something besides temperature and moisture must 
 be involved in their explanation. 
 
 The high-pressure belts. The explanation of the high pressure 
 in low latitudes rather than in high, and the explanation of the 
 highest pressures just outside the tropics, is not found on the isobaric 
 charts. These larger features of pressure-distribution are probably 
 to be explained by the general circulation of the atmosphere under 
 the influence of rotation. Several factors bear upon this point. 
 
 1. In the equatorial zone, the air is heated and expanded. As 
 it rises by expansion, it must flow to north and south. If the 
 expansion affected the atmosphere all the way from bottom to 
 top, there would be outflow from the top of the atmosphere in 
 the equatorial zone in either direction, for the same reason that 
 outflow would take place from the top of a mound or ridge of water 
 if such existed. But the expansion of the air by heating is chiefly 
 in the lower part of the air. As the lower ah* expands, it pushes up 
 the air above it. The pressure of the air at the bottom, before out- 
 flow takes place above, is not diminished, but the pressure at a point 
 above, say at the upper limit of the effective heating, is increased, 
 because a larger part of the air is now crowded up above that level 
 
ATMOSPHERIC PRESSURE 
 
 595 
 
 This is illustrated by Figs. 584 and 585. The former shows the 
 crowding of the air above the zone of heating, and the latter the 
 resulting isobaric slopes. Except at the bottom of the atmosphere, 
 the isobaric surfaces slope downward on either hand from the equa- 
 
 FIG. 584. Expansion of the lower air as a result of heating, crowds the air 
 above, and so increases its density and pressure, as compared with the 
 density and pressure of air at the same level outside the heated area. 
 
 torial zone, and air always flows down an isobaric surface. Over 
 the heated equatorial zone, therefore, the expanded air rises, and 
 at some level above the bottom of the atmosphere it flows pole- 
 ward down the isobaric surface (Fig. 586). This is the case in spite 
 
 Level 
 
 FIG. 585. The condition of things represented in Fig. 584 gives rise to move- 
 ments of air. 
 
 of the low pressure at the bottom of the atmosphere in equatorial 
 latitudes. 
 
 When some of the air flows out poleward from the equatorial 
 belt, the pressure at the bottom of the equatorial belt is diminished, 
 because the amount of air above is diminished. At the same time 
 the pressure on both sides of the equatorial belt is increased, because 
 the amount of air is there increased. Furthermore, when the air 
 of the equatorial belt expands, it pushes laterally as well as upward, 
 and so tends to compress the air outside the belt where the expan- 
 sion takes place. 
 
 Both the outward flow and the outward crowding of the air 
 in the equatorial belt tend to increase the pressure of the air out- 
 side the zone of principal heating, but they do not make it ap- 
 parent why the zones of greatest pressure should be in latitude 
 30 or a little above. 
 
596 
 
 PHYSIOGRAPHY 
 
 2. When the pressure in the equatorial belt is diminished by 
 the outflow of air above, a barometric slope is established toward 
 the equator from either side at the bottom of the atmosphere, as 
 shown by the lower arrows. Fig. 585, even when the barometric slope 
 is from the equator in the upper air. Thus a system of convective 
 circulation is established. In the long run, the outflow of air from 
 the equatorial belt toward the poles will be equaled by the inflow 
 of air from higher latitudes on either side to the equatorial belt. 
 The poleward-flowing upper air descends as it reaches higher and 
 higher latitudes, and it takes the place of the air which moves 
 equatorward. On the whole, the amount of ascending poleward- 
 flowing air from the equatorial zone, equals the amount of descend- 
 
 60 
 
 40 
 
 ALTITUDE OF 2000 FEET 
 20 8EA LEVEL 20 40 
 
 30 INCHES 
 
 FIG. 586. Slope of isobaric surfaces along meridians at various altitudes. 
 
 (After Waldo.) 
 
 ing equatorward-flowing air from high latitudes. There must> 
 therefore, be a vertical plane in the atmosphere in each hemisphere, 
 on the equatorward side of which as much air ascends as descends 
 on the poleward side. This vertical plane should be near latitude 
 30, for this parallel divides the surface of the hemisphere, and 
 therefore the volume of air in each hemisphere, into two nearly 
 equal parts. This is regarded as a cause of the high-pressure belts 
 at 30 in both hemispheres. 
 
 3. Given the high-pressure belts in extra-tropical latitudes, 
 the circulation of the bottom air which follows helps to main- 
 tain them. The air moving poleward from these belts of high pres- 
 sure turns to the right in the northern hemisphere, and to the 
 left in the southern, becoming westerly winds in both hemispheres. 
 In both, this turning causes these winds to crowd on the equator- 
 ward side of their lines of movement. This tends to maintain and 
 
ATMOSPHERIC PRESSURE 597 
 
 increase the pressure in the high-pressure belts, by crowding on 
 their poleward sides. 
 
 Given the extra-tropical belts of high pressure, the numerous 
 irregularities and changes of pressure from season to season, as 
 shown by the isobaric charts, may be explained chiefly by varia- 
 tions in temperature. 
 
 Permanent areas of low pressure. Fig. 578 shows areas of 
 low pressure in the North Pacific and North Atlantic oceans. These 
 areas of low pressure are still more pronounced on the January 
 chart (Fig. 582), and but feebly marked on the July chart (Fig. 
 583). No corresponding areas of low pressure are known in the 
 southern hemispheres. No explanation of these areas of low 
 pressure is here attempted. 
 
 Temporary and local variations of pressure. There are many 
 variations of pressure not shown on seasonal or even on monthly 
 isobaric charts, though they appear on daily weather maps. These 
 will be studied in the next chapter. There are even variations of 
 pressure which do not appear on the daily maps. Chief of them 
 are the daily variations, presumably caused by the daily varia- 
 tions of temperature. Thus there are daily maxima at about 10 A.M. 
 and 10 P.M., and daily minima at about 4 P.M. and 4 A.M. These 
 daily changes range from 0.01 to 0.15 of an inch, the range being 
 greatest in low latitudes. No satisfactory explanation of these 
 variations has been given. 
 
CHAPTER XVII 
 GENERAL CIRCULATION OF THE ATMOSPHERE 
 
 Prevailing and Periodic Winds 
 
 INEQUALITIES of atmospheric pressure involve atmospheric move- 
 ments. Since atmospheric pressures are unequal, and since proc- 
 esses are constantly in operation which renew the inequalities, 
 movements are continuous. Unequal insolation is the most im- 
 portant factor in disturbing the equilibrium of the air, and so in 
 generating air movements, and in determining their initial direc- 
 tion; but the rotation of the earth has much influence in directing 
 them, once they are started. Since the greater insolation is always 
 in the same general zone, and since the rotation of the earth is 
 always in the same direction, the ah* movements generated and 
 directed by insolation and rotation are systematic, and result in 
 a general circulation of the atmosphere. 
 
 It is to be borne in mind that the movement of air is always 
 from a region of greater pressure to one of less pressure at the same 
 level, or, in other words, always down a barometric or isobaric slope. 
 The familiar saying that "the wind bloweth where it listeth" is 
 true only in the sense that the air always listeth to blow down the 
 steepest accessible isobaric gradient, and that where there is no 
 gradient, it listeth not to blow. 
 
 The General Effect of Unequal Insolation 
 
 If the air were in equilibrium over the whole earth at a uniform, 
 low temperature, and if it could then be heated by the sun for a 
 time without involving horizontal movement, the effect would be 
 to raise its surface over all areas where its temperature was in- 
 creased, and to raise it most where it was heated most, that is, in 
 the low latitudes (Fig. 586). As indicated in the last chapter 
 (p. 595), the result would be the establishment of a barometric 
 
 598 
 
GENERAL CIRCULATION OF THE ATMOSPHERE 599 
 
 gradient from the equatorial region toward the poles (Fig. 586), 
 and this is the condition necessary for poleward movements of 
 air. 
 
 Since the air in low latitudes is always being heated more 
 effectively than that in higher latitudes, 1 movement should be essen- 
 tially constant, above the bottom of the air, from the equatorial 
 zone to the polar zones in both hemispheres. These poleward 
 movements of air lessen the pressure at the bottom of the atmos- 
 phere in low latitudes, because air has moved away from that zone. 
 As the pressure is thus lessened in the equatorial region, a baro- 
 metric gradient is established toward the equator at the bottom of 
 the atmosphere (Fig. 587), and air must then come in from higher 
 latitudes. Here, then, we have two elements of a general circu- 
 lation: a poleward movement in the upper air, and an equator- 
 ward movement in the lower air, and the causes which generate 
 these movements are constantly in operation. 
 
 It should perhaps be noted that quite apart from circulatory 
 movements, there would be lateral crowding by the expanding air 
 of low latitudes (Fig. 584). In so far 
 as this is effective, it would reduce 
 the mass of air above any point of 
 the surface where the air was ex- 
 panding. It would also tend to in- 
 crease the amount of air over areas 
 poleward from the zone of heating, 
 and so would tend to establish 
 equatonvard gradients in the lower 
 part of the air. The result would be 
 to increase the equatorward isobaric FlQ 587 ._ Dia g ram showing the 
 slope at the bottom of the atmos- general system of circulation 
 i which would be established by 
 
 unequal heating, as a result of 
 
 From unequal heating alone, differences in latitude. 
 
 therefore, there is a constant ten- 
 dency to the movement of ah- (1) from low latitudes toward the 
 poles above the bottom of the atmosphere, and (2) a compensa- 
 tory movement from the higher latitudes toward the equator. 
 These are the most fundamental facts in the general circulation of 
 
 1 High latitudes sometimes receive more heat per day than low latitudes 
 (see p. 525), but the air of high latitudes is never so effectively heated, 
 because of the abundance of ice, snow, ice-cold water, and frozen ground. 
 
600 PHYSIOGRAPHY 
 
 the atmosphere. They involve vertical as well as horizontal move- 
 ments of air. The vertical movements are (1) upward in low lati- 
 tudes, where the air (a) expands upward, and (b) is crowded up- 
 ward by the cooler and heavier air which flows in below, and (2) 
 downward in higher latitudes. The system of circulation which 
 would be established by the greater heating of the low latitudes, 
 taken by itself, is somewhat as shown in Fig. 587. 
 
 The general poleward movement of air from low latitudes 
 seems to be clearly established by observation, but its return to 
 low latitudes is much less clearly indicated in observed winds. 
 Of its return there can be no question, but how and where it is 
 effected is not well understood, for outside the low altitudes of the 
 low latitudes (the trade-wind zones) no persistent equatorward 
 movements of air are recorded. Much air moves equatorward in 
 the aperiodic atmospheric disturbances, and perhaps the return is 
 chiefly effected through them. These aperiodic movements will be 
 studied in the next chapter. 
 
 It is to be noted that the isobaric gradients at the bottom of 
 the atmosphere in low latitudes do not correspond with those in 
 the upper air (Fig. 586) ; yet these apparently inharmonious gradi- 
 ents co-exist. The reasons for each have been given; their co- 
 existence means that the tendency to the poleward slope is so 
 strong that, except in the lower part of the atmosphere, it is not over- 
 come by the causes which develop the equatorward gradient at 
 the bottom of the atmosphere. 
 
 But for the influence of rotation and the unequal heating of land 
 and water areas in the same latitude, the atmospheric movements 
 just outlined would tend to follow meridians. Rotation affects the 
 course of the atmospheric movements in more ways than one. It 
 not only deflects all currents to the right in the northern hemi- 
 sphere, and to the left in the southern, but it appears to be 
 responsible, in part at least, for the concentration of the high 
 pressures of extra-tropical latitudes into belts near the tropics 
 (p. 596) ; and these belts of high pressure have an important influ- 
 ence on the course of circulation at the bottom of the atmosphere, 
 and interfere with the simplicity of circulation outlined above. 
 
 Effect of the Extra-tropical Belts of High Pressure 
 
 In each high-pressure belt (Fig. 578) the isobaric surfaces are 
 bowed up in the lower part of the atmosphere (Fig. 586), and from 
 
GENERAL CIRCULATION OF THE ATMOSPHERE 601 
 
 each there is a barometric gradient both to north and south. From 
 each of these belts, therefore, there should be a flow of air both 
 southward and northward at the bottom of the atmosphere. If no 
 other factors were involved, the movements of the lower air should 
 
 60 
 
 FIG. 588. Diagram representing the general movements which would take 
 place in the lower air if there were no rotation. 
 
 be those shown in Fig. 588; and if forces were in operation to con- 
 stantly renew the high-pressure belt, these movements of air would 
 be constant. At the center of the high-pressure belt, there would 
 be little horizontal movement of the air. The narrow zone in this 
 position is the zone of tropical calms. 
 
 It will be observed that the poleward flow in the lower part of 
 the atmosphere from the high-pressure belts would be in much the 
 same direction as the poleward flow of upper air from the equa- 
 torial zone, while the equatorward flow of lower air from the high- 
 pressure belts would be opposed in direction to the flow in the upper 
 air of the same latitude. 
 
 After the poleward gradients in the larger part of the atmos- 
 phere (Fig. 586), the barometric slopes in the lower air from the 
 high-pressure belts are perhaps the most important fact in the 
 general circulation of the atmosphere. 
 
 The High-latitude Areas of Low Pressure 
 
 The permanent areas of low pressure over the northern oceans 
 (Figs. 578, 582, and 583) constitute another permanent factor in the 
 
602 PHYSIOGRAPHY 
 
 atmospheric circulation. Their influence is less commonly recog- 
 nized than that of the high-pressure belts, but it is perhaps of more 
 than minor importance. To these areas there must be a constant 
 inflow of air, and from them it rises and flows out above, thus 
 modifying the general course of the circulation, and helping to 
 destroy its simplicity. It is perhaps significant that the great 
 centers of glaciation in the glacial period lay on the continents to 
 the east of these areas of permanent low pressure. 
 
 According to the outline given above, the atmospheric circula- 
 tion in one hemisphere appears to be measurably independent of 
 that in the other. This, however, is less true than would appear 
 from the statements already made. The average pressure for the 
 northern hemisphere for January has been estimated at 29.99 inches, 
 and that for the southern hemisphere at the same time 29.79 inches. 
 The average pressures for July are estimated at 29.87 inches in the 
 northern hemisphere and 29.91 inches for the southern. It has 
 been calculated that, to bring about the condition which exists in 
 January, some 32,000,000 tons of air must have been shifted from 
 the southern hemisphere into the northern since the preceding 
 summer. This transfer is probably effected because the low tem- 
 perature of the extensive land areas in the northern hemisphere so 
 reduces the temperature and increases the density of the air over 
 great areas in that hemisphere, that the north-poleward gradient 
 in the upper air is increased, and the crest of the barometric sur- 
 face (Fig. 586) shifted south of the equator. In other words, the 
 wind equator and the thermal equator are then south of the geographic 
 equator. The shifting of the thermal equator, and therefore of the 
 wind equator, is shown in Figs. 539 and 540, respectively. The 
 corresponding shifting of the wind zones is illustrated by Fig. 589. 
 
 These three factors, namely, (1) the poleward gradients in 
 the upper air of low latitudes, (2) the gradients in the lower air 
 from the high-pressure belts in extra-tropical latitudes, and (3) the 
 gradients in the lower air toward the areas of low pressure in 
 high latitudes, are the principal ones, named in the order. of their 
 importance, in the general circulation of the atmosphere. 
 
 Direction of Winds 
 
 Once wind is started, its direction may be influenced by various 
 factors. Chief among them is the rotation of the earth, which 
 affects the course of all winds except such as blow in the plane of 
 
GENERAL CIRCULATION OF THE ATMOSPHERE 603 
 
 the equator. The farther they, go, the more are their directions 
 changed. 
 
 A generalized diagram of the observed winds of the lower air is 
 shown in Fig. 590. This figure represents the winds blowing out 
 
 iUfeTRpPIOAt-B^LTV^V \ ^*S. 
 Sj^UARY'tlMiT ~" >r;U ~ 
 
 FIG. 589. Diagram illustrating the FIG. 590. Generalized diagram of wind 
 shifting of wind zones. (After directions at the bottom of the atmos- 
 Davis.) phere. 
 
 from the extra-tropical belts of high pressure, and following more 
 or less systematic courses. The poleward winds from the high- 
 pressure belts are turned toward the east in both hemispheres, and 
 so become westerly winds (southwesterly in the northern hemisphere 
 and northwesterly in the southern). The winds blowing toward the 
 equator in the lower air from the belts of high pressure become 
 easterly (northeasterly in the northern hemisphere and south- 
 easterly in the southern) and are known as trade-winds. The 
 zone along the thermal equator where the northeasterly and south- 
 easterly trades meet, and where ascending currents of air are more 
 pronounced than horizontal movements, is known as the zone of 
 equatorial calms. The position of the zone of calms shifts a little 
 with the sun, its center remaining near the thermal equator. (Com- 
 pare Figs. 539 and 540.) 
 
 The trade-winds are remarkably persistent, and have long been 
 known and utilized by navigators. 
 
 The westerly winds of middle latitudes and the trades of low 
 
604 
 
 PHYSIOGRAPHY 
 
 latitudes are the prevailing winds at the bottom of the atmosphere, 
 and are sometimes called the planetary winds. 
 
 The explairiation of the deviation of the winds from meridional 
 courses, always to the right in the northern hemisphere nd always 
 to the left in the southern, is the same as that underlying the change 
 in the direction of the swinging pendulum, and is illustrated by 
 Fig. 591. This figure may be taken to represent the northern 
 hemisphere as seen from above the North Pole. The curved 
 arrows show the direction of rotation. The arrow at n represents 
 a wind starting poleward. The arrows o, p, and q represent suc- 
 cessive directions of the wind as it advances. Their departure 
 
 FIG. 591. Diagram illustrating deflection of winds to the right in the north- 
 ern hemisphere. The deflection is to the left in the southern hemi- 
 sphere, for the same reason. 
 
 from meridians is to the right, and the departure becomes more 
 pronounced as the latitude becomes higher. The arrow w repre- 
 sents a wind blowing westward, or contrary to the direction of 
 rotation. Since the motion of the air in the direction of the 
 arrow is much less rapid than the rotation of the earth in the 
 opposite direction, the arrows x, y, and z represent progressions 
 backward. Similarly the arrow s, near the pole, represents a wind 
 blowing southward, and the arrows t, u, and v represent the suc- 
 6essive directions which such a wind would have, the departures 
 from the meridians being still to the right. The arrow e repre- 
 sents a wind starting eastward, and the arrows /, g, and h, the 
 successive directions of the wind. The wind here progresses 
 forward, because its direction corresponds with the direction 
 of rotation. A corresponding diagram might be made for the 
 
GENERAL CIRCULATION OF THE ATMOSPHERE 605 
 
 southern hemisphere which would snow, in a similar way, the 
 deflection of winds to the left of meridians. 
 
 By referring to Fig. 586, it will be seen that the trade-winds can- 
 not have great depth. While they are pronounced at the surface, 
 they must cease at some relatively slight elevation above, for the 
 configuration of the isobaric surfaces changes. As a matter of 
 observation, the trade-winds have been observed to cease at an 
 elevation of about 10,000 feet on Teneriffe (Canary Islands, Lat. 
 28). Their upper limit has also been noted on various mountains 
 in South America and on the Hawaiian Islands, and is not far from 
 the above figure. 
 
 The westerly winds, on the other hand, have much greater depth. 
 Figs. 592 and 593 show the directions of winds in the United States 
 (1) at the bottom of the atmosphere, and (2) in the upper air, as 
 shown by the movements of the upper clouds. The movements of the 
 lower part of the air are -very different in the two cases, but the 
 movements indicated by the upper clouds are to the eastward in both. 
 
 The Circumpolar Whirl 
 
 The circulation in each hemisphere is often looked upon as a 
 great eddy centering at the pole. If this were the true view of the 
 case, it would account for the low pressure in high latitudes and 
 the high pressure in low latitudes, and the pressure should decrease 
 steadily to each pole. 
 
 Land and Water Circulation 
 
 While the winds at the bottom of the atmosphere tend to fall 
 into the general system shown in Fig. 590, the simplicity of the 
 system is interfered with by various disturbing influences which 
 modify the system of planetary winds. Chief of these disturbing 
 factors is the unequal heating of the atmosphere over land and 
 water. This not only interferes with the direction of planetary winds, 
 but is itself the cause of winds. 
 
 Monsoons and land- and sea-breezes have already been cited as 
 illustrations of the effects of the unequal heating of land and water. 
 
 The monsoon influence is probably much stronger than is com- 
 monly recognized, for it overcomes the prevailing winds on a large 
 scale. Thus in winter, Eurasia is, on the average, a centre of air 
 dispersion in the lower air (Fig. 582), while in summer, air flows in 
 
606 
 
 PHYSIOGRAPHY 
 
 FIG. 592. Chart showing the direction of air movements at the bottom of the 
 atmosphere (upper figure), at the horizon of the lower clouds (middle 
 figure), and at the level of the upper clouds (lower figure), at a time when 
 the pressure is high about Lake Superior. The winds are westerly: in the 
 upper air, without reference to inequalities of pressure in the lower air. 
 (U. S. Weather Bureau.) 
 
GENERAL CIRCULATION OF THE ATMOSPHERE 607 
 
 FIG. 593. Figure showing the movements of the air when atmospheric pres- 
 sure is low about Lake Superior. It will be noted that the movements 
 in the upper air (lowest figure) are from the west as in the preceding case. 
 
608 
 
 PHYSIOGRAPHY 
 
 toward it, though most of its area is in the zone of the westerlies. 
 The same influence is probably of great importance over and about 
 every large land area, but it is only where it opposes and overcomes 
 the prevailing wind that it is popularly recognized. 
 
 India is usually cited as affording the best illustration of the 
 monsoon influence. This country is in the latitude of the northern 
 
 * * / / i v\ \ r*> 
 
 * * > . \ ^ \ \ / 
 
 * ' << * \ * 
 
 ^t/ / ,/T i * <: I 
 
 FIG. 594. The isobars of India for FIG. 595. Figure showing the direc- 
 January. (After Bartholomew.) tion of winds in India in winter. 
 
 (After Koppen.) 
 
 trades, where easterly (northeasterly) winds should prevail. In 
 Fig. 594, the gradient is from northeast to southwest, and the direc- 
 tion of the wind (Fig. 595) is in harmony with the planetary system; 
 
 FIG. 596. The isobars of India for 
 August. (After Bartholomew.) 
 
 FIG. 597. The winds of India in mid- 
 summer. (After Koppen.) 
 
 but in Fig. 596 the isobaric gradient is to the northward, because 
 the land is warmer than the sea and the winds blow in that direc- 
 tion (Fig. 597). That is, the planetary (northeast) wind is overcome 
 
GENERAL CIRCULATION OF THE ATMOSPHERE 609 
 
 during the hot season by the winds which result from the seasonal 
 change of temperature which establishes a seasonal gradient. At 
 the same season, the low pressure north of India, developed by the 
 heat of summer, counteracts the high pressure normal to this lati- 
 tude, and the prevailing wind is displaced by seasonal winds blow- 
 ing toward the area of low pressure. Figs. 598 and 599 show the 
 
 FIG. 598. Isotherms of India for 
 January. (After Buchan.) 
 
 FIG. 599. Isotherms of India for 
 August. (After Buchan.) 
 
 isotherms for the same region at the corresponding seasons, and 
 make clear the relation between pressure and temperature. 
 
 When the monsoon blows with the prevailing wind, as in western 
 India in winter, the prevailing wind is strengthened; if the two 
 
 FIG. 600. Isobars and winds in 
 Spain and Portugal, month of 
 January. (After Hann.) 
 
 FIG. 601. Isobars and winds in 
 Spain and Portugal, month of 
 July. (After Hann.) 
 
 tend to blow in opposite directions, as in western India in summer, 
 the stronger prevails. Spain, in the zone of westerly winds, affords 
 an excellent example of the same thing. Figs. 600 and 601 show 
 
610 
 
 PHYSIOGRAPHY 
 
 the conditions in winter and summer. In winter the isotherms over 
 the plateau are low, and the isobars high, and the winds blow out 
 from its center instead of toward it. In summer the case is 
 reversed. 
 
 The general principle of the monsoon makes itself felt about the 
 Great Lakes. At Chicago, which is in the zone of southwesterly 
 winds, northeast winds predominate in spring, because the lake is 
 then much cooler than the land, and the winds set toward the land 
 and overcome the prevailing winds (Fig. 602). Similar diagrams 
 
 January, 1904 
 
 April, 1904 
 
 Septenber, 1904 
 
 FIG. 602. Diagram showing the direction and velocity of winds in Chicago 
 during January, April, July, and September, 1904. The time during 
 which the wind blew from any given direction is shown, relatively, by 
 the length of the lines. The relative average velocity is shown by the 
 width of the lines. The monsoon influence of the lake is seen in the 
 preponderance of northeast winds in April. (Cox, U. S. Weather Bureau.] 
 
 for a station fifty miles from the lake would show less wind from the 
 northeast in April. The same thing is illustrated by Fig. 603, which 
 shows wind "roses" for Chicago and Key West. The former has a 
 prolongation to the northeast, indicating primarily the landward 
 winds in spring, though the principal direction of wind otherwise 
 is southwest. Key West is in the zone of the trade-winds, and 
 the easterly winds greatly predominate over all others. 
 
 The principle involved in the daily land- and sea-breezes along 
 coasts (p. 561) is the same 1 as that of the monsoon, but the resulting 
 
GENERAL CIRCULATION OF THE ATMOSPHERE 611 
 
 winds are more local. They are not usually felt very far from shore, 
 and do not extend to great heights. At Coney Island, the sea-breeze 
 has been found to be limited to a height of about 500 feet at times 
 when it has been determined. At slightly higher levels the air- 
 currents were those of the prevailing winds. At some other places sea- 
 breezes have been known to extend up 1300 feet. 
 
 On the coast of Massachusetts the sea-breeze sometimes starts 
 as early as eight o'clock in the morning, though more commonly not 
 till an hour or two later. At first it advances inland at the rate of 
 3 to 8 miles per hour, and later more slowly. It penetrates inland 
 
 Fia. 603. Wind "roses " for Chicago and Key West, 1902. The shaded parts 
 of the diagrams shows the relative duration of the periods when the wind 
 blew from different directions. The greater the distance from the cross- 
 ing-point of the radiating lines, the longer the period. The influence of 
 the lake, giving rise to lake breezes and to winds of monsoon character, 
 is conspicuous at Chicago. Key West is in the zone of trade-winds. 
 (Cox, U. S. Weather Bureau.) 
 
 10 to 20 miles, and sometimes gives rise to thunder-storms. On 
 the coast of southern California, the land- and sea-breezes persist 
 throughout the year, being much stronger in summer than in 
 winter. Land-breezes are generally less well developed than sea- 
 breezes. 
 
 Breezes corresponding to land- and sea-breezes are often felt about 
 large lakes. 
 
 The sea-breeze is of consequence, not only by lowering the land 
 temperature in hot weather, but by bringing pure air to the land. 
 This is of much importance along densely populated coasts. The 
 explanation of the 'sea-breezes has already been suggested (p. 561). 
 
612 PHYSIOGRAPHY 
 
 The unequal heating of high and low lands in the same latitude 
 also causes slight and temporary departures from the normal planet- 
 ary circulation (p. 562). 
 
 Besides the planetary winds, the seasonal winds, and minor periodic 
 winds, whose times of coming and going are more or less regular, 
 there are numerous winds which blow at irregular times, and whose 
 coming cannot be foretold long in advance. These irregular winds 
 are the chief cause of the uncertain elements of the weather. Some 
 of them are due to unequal temperatures, some to unequal amounts 
 of atmospheric moisture, and some to other causes. 
 
 Illustrations of aperiodic winds due to unequal temperature are 
 whirlwinds and tornadoes, both of which are due to strong convec- 
 tion currents generated by excessive local heating, and some larger 
 whirls of air, especially tropical cyclones. These will be referred to 
 in the next chapter. 
 
 Again, just as \vaves of water generated by the wind are felt 
 far beyond the place where they were generated, and long after the 
 wind ceases to blow, so local disturbances, leading to the flow of 
 air from one place to another, make themselves felt far beyond the 
 place of disturbance. Movements therefore generate movements. 
 
 Summary. We may now recall the chief points thus far studied 
 in connection with atmospheric circulation. They are as follows: 
 
 (1) Above the lower part of the atmosphere there is a pole- 
 ward movement of the air from low latitudes. 
 
 (2) There must be a compensatory movement of air from high 
 latitudes to low; but outside the extra-tropical belts of high pres- 
 sure, this movement is not well defined. 
 
 (3) The extra-tropical high-pressure belts are the zones from 
 which the dominant planetary winds at the bottom of the atmos- 
 phere start. 
 
 (a) These planetary winds tend to blow poleward and equator- 
 ward in each hemisphere, from the belts of high pressure. 
 
 (6) They are deflected to the right in the northern hemisphere 
 and to the left in the southern hemisphere, by the rotation of the 
 earth, thus establishing the double trade-wind zone, with the equa- 
 torial calms in the centre, and two zones of westerly winds, with 
 tropical calms on the equatorward margin of each. 
 
 (4) The simplicity of the system of planetary winds is interfered 
 with by the great inequalities of temperature between land and 
 sea in the same latitude. The isobaric gradients established by 
 
GENERAL CIRCULATION OF THE ATMOSPHERE 613 
 
 unequal heating may be higher than those which direct the planet- 
 ary winds. In such cases the planetary winds are overcome by 
 seasonal winds, such as the monsoons, or by daily breezes, such as 
 land- and sea-breezes, and mountain and valley breezes. 
 
 Gradient, velocity, and directions of wind. The slope of an 
 isobaric surface is its gradient. Gradient is differently expressed 
 in different countries. In England the barometric gradient is 
 said to be 1 when the difference of pressure is 0.01 of an inch in 
 17 miles. In the United States, barometric gradient is commonly 
 denned as the difference in pressure at the same level between two 
 points which are distant from each other the length of 1 of lati- 
 tude. Thus if two places 5 apart in latitude have a difference of 
 pressure of 0.5 inch, the gradient is 0.10 of an inch. Stated 
 mathematically, 30-29.50= .50-7-5 = 0.10. 
 
 The greater the gradient, the greater the velocity of the wind. 
 On the isobaric chart, high gradient is expressed by the crowding 
 of isobaric lines. The crowding of such lines, therefore, means 
 high winds. A gradient of 0.10 inch means a wind of about 30 
 miles an hour, and a gradient of 0.20 means a wind of about 55 
 miles an hour. These figures presume a plane surface. The actual 
 velocity at the bottom of the atmosphere is much modified by 
 the shape of the surface. The rougher the surface, the less the 
 velocity. Observations have shown that the velocity of the wind 
 at the height of low buildings (say 40 to 80 feet) on land is only 
 about one-fourth as great as that at an elevation of 40 feet on 
 the sea; while at a height of 100 to 150 feet, the velocity is 
 half that over the sea at an elevation of 40 feet. 
 
 In general, the average velocity of winds is greatest in latitude 
 50 or thereabouts. The average velocity for the United States 
 has been estimated at about 9.5 miles per hour, and for Europe, 
 10.3. The velocity is greater over the sea than over the land, 
 largely because it is checked on land by friction with the uneven 
 surface, with vegetation, buildings etc. 1 It is also greater in the 
 upper air than in the lower, for the same reason. The following 
 table gives the velocity of the wind at various levels above the 
 bottom. It is based on observations on the movement of clouds 
 at Blue Hill Observatory, near Boston. 
 
 1 Hchr.l-oltz has calculated that if the whole body of air were set in 
 motion at the uniform rate of 20 miles per hour, it would take nearly 43,000 
 years to slow it down to 10 miles as a result of friction. 
 
614 PHYSIOGRAPHY 
 
 COMPUTED EASTERLY OR WESTERLY WIND VELOCITIES ALONG A MERIDIAN. 
 
 
 E=easterly, W = westerly. Velocity of wind 
 in miles per hour at various altitudes. 
 
 Increase in easterly 
 velocities with eack 
 (about) 3300 feet 
 
 Latitude. 
 
 
 in altitude. 
 
 
 Sea-level. 
 
 About 
 3300 feet. 
 
 About 
 13,200 feet. 
 
 Miles per hour. 
 
 N. Lat. 75 
 
 E. 2.7 
 
 W. 0.2 
 
 W. 9.2 
 
 W. +3.0 
 
 70 
 
 E. 2.0 
 
 W. 2.0 
 
 W. 14.3 
 
 W. 4.1 
 
 65 
 
 W. 0.1 
 
 W. 4.9 
 
 W. 19.3 
 
 W. 4.8 
 
 60 
 
 W. 2.4 
 
 W. 7.6 
 
 W. 23.1 
 
 W. 5.2 
 
 55 
 
 W. 3.4 
 
 W. 8.7 
 
 W. 24.5 
 
 W. 5.3 
 
 50 
 
 W. 3.3 
 
 W. 8.7 
 
 W. 24.9 
 
 W 5.4 
 
 45 
 
 W. 3.0 
 
 W. 8.5 
 
 W. 25.0 
 
 W. 5.5 
 
 40 
 
 W. 1.6 
 
 W. 7.2 
 
 W. 24.0 
 
 W. 5.6 
 
 35 
 
 E. 0.7 
 
 W. 5.0 
 
 W. 22.4 
 
 W. 5.8 
 
 30 
 
 E. 5.3 
 
 W. 0.6 
 
 W. 18.2 
 
 W. 5.9 
 
 25 
 
 E. 8.9 
 
 E. 3.1 
 
 W. 14.4 
 
 W. 5.8 
 
 20 
 
 E. 9.4 
 
 E. 3.8 
 
 W. 13.0 
 
 W. 5.6 
 
 N. Lat. 15 
 
 E. 7.8 
 
 E. 4.3 
 
 W. 6.1 
 
 W. 3.5 
 
 Equator 
 
 
 
 
 
 S. Lat. 15 
 
 E. 15.6 
 
 E. 10.5 
 
 W. 4.8 
 
 W. 5.1 
 
 20 
 
 E. 13.0 
 
 E. 8.2 
 
 W. 6.4 
 
 W. 4.8 
 
 25 
 
 E. 6.4 
 
 E. 1.7 
 
 W. 12.5 
 
 W. 4.7 
 
 30 
 
 W. 2.4 
 
 W. 7.0 
 
 W. 21.0 
 
 W. 4.7 
 
 35 
 
 W. 7.7 
 
 W. 12.3 
 
 W. 26.1 
 
 W. 4.6 
 
 40 
 
 W. 11.6 
 
 W. 16.2 
 
 W. 30.0 
 
 W. 4.6 
 
 45 
 
 W.- 14.9 
 
 W. 19.5 
 
 W. 33.3 
 
 W. 4.6 
 
 50 
 
 W. 17.1 
 
 W. 21.7 
 
 W. 35.7 
 
 W. 4.6 
 
 55 
 
 W. 17.0 
 
 W. 21.6 
 
 W. 35.6 
 
 W. 4.7 
 
 a Lat. 60 
 
 W. 13.6 
 
 W. 18.2 
 
 W. 32.2 
 
 W. +4.7 
 
 This table shows that there is an increase of velocity with in- 
 crease of altitude, ranging from 1 to 2 miles per hour for 1000 feet. 
 According to the table, the trade-winds do not reach up to alti- 
 tudes of 13,200 feet, for at this altitude all winds are represented 
 as blowing to the east. The table also represents them as extend- 
 ing farther from the equator, especially in the northern hemi- 
 sphere, in low altitudes, than higher up. 
 
 THE GENERAL, CIRCULATION AND PRECIPITATION 
 
 Rainfall is of the utmost importance to most land life, both 
 plant and animal. This is shown, in a general way, by the ab- 
 sence of forests and the meagreness of herbaceous vegetation in 
 arid regions; and wherever plant life is scanty, animal life is also 
 relatively scarce. Human industries, too, are much affected by 
 
GENERAL CIRCULATION OF THE ATMOSPHERE 615 
 
 the amount and distribution of the rainfall, as shown by the fact 
 that no arid region supports a dense population. Nevada, almost 
 all of which receives less than 10 inches of rain per year, had, in 
 1900, only one inhabitant for each two and a half square miles. 
 Only 3.4 per cent, of the population of the United States lives in the 
 third of the country where the rainfall is less than 20 inches per 
 year. The best of soil is unproductive unless adequately watered. 
 Twenty inches of rain per year is generally considered to be the 
 minimum for general agricultural purposes, but something depends 
 on the latitude and something on the seasonal distribution of the 
 rain. The warmer the climate, the more the rainfall needed, be- 
 cause of the greater evaporation ; and the aggregate amount neces- 
 sary is less if it falls when the growing crops need it most. If 
 rainfall could be ideally distributed, the half of 20 inches would 
 probably be adequate in the middle latitudes of the United States. 
 Rain or snow falling at times when plants are not growing is, 
 however, not worthless, for some of it remains underground, and 
 is available for plants at a later time. The secret of the successful 
 "dry farming," which is just now attracting much attention, con- 
 sists in so treating the soil that the water which falls during all 
 parts of the year is retained in the soil and subsoil till the growing 
 season. 
 
 Land so situated that it may be irrigated is not directly depend- 
 ent on rain and snow; but the water used in irrigation is derived 
 from rainfall, though the fall is often far from the place where the 
 water is used. Great as the results of irrigation are likely to be 
 in our own country, it will never make more than a fraction of the 
 arid land valuable for agricultural purposes, because the amount of 
 water available is limited. 
 
 The distribution of rainfall is largely influenced by the winds, 
 which bear moisture from the places where it is evaporated, to the 
 places where the temperature favors its condensation and precipi- 
 tation. Prevailing winds, periodic winds, and aperiodic winds all 
 play their part in determining where rain falls, how much falls, 
 and at what times of the year. The vertical movements of the air, 
 too, have something to do with rainfall, and in some places are 
 more important than the horizontal movements to which the name 
 winds is usually restricted. 
 
 To know the rainfall (or snowfall) of any given region, it is 
 needful to know (1) what winds affect it, (2) the topography of 
 
616 
 
 PHYSIOGRAPHY 
 
 the surface over which the winds have already blown before reach- 
 ing it, and (3) the topographic situation and relations of the place 
 itself. 
 
 Rainfall in the zones of the trades. In the trade-wind zones 
 the winds are blowing from higher to lower latitudes, and therefore, 
 on the whole, from cooler to warmer latitudes. As the air is 
 warmed, it is capable of taking more moisture. So long as the 
 trades blow on the sea, therefore, they would not ordinarily give 
 rain. Where they blow over low lands, which in this latitude are 
 warmer than the sea, they take moisture rather than give up what 
 
 FIG. 604. Map showing the precipitation for the world. 
 
 they have. On the sea and on low lands, therefore, the trade- 
 winds are " dry " winds. Sahara and considerable parts of Australia 
 are essentially desert, largely because of the drying influence of 
 the trades. 
 
 If, however, the air of the trades is forced up over mountains, 
 it is cooled, and some of its moisture may be condensed and may 
 fall as rain or snow. The windward sides of high mountains in the 
 trade-wind zone should therefore have heavy rainfall. An illustration 
 is afforded by the east side of the Andes Mountains in tropical 
 latitudes, where the rainfall is heavy (Fig. 604). Another illus- 
 tration is afforded by the volcanic cones of the Hawaiian Islands. 
 The trade-winds yield little rain to their lower slopes, but forced up 
 
GENERAL CIRCULATION OF THE ATMOSPHERE 617 
 
 over the mountains, they yield abundant moisture in the cooler 
 altitudes. The level of the rainfall is readily seen by the change 
 in vegetation. 
 
 After the air of the trades passes over a mountain range, it 
 descends, and is warmed both by contact with the warm surface 
 and by compression. It therefore takes up moisture. The lee- 
 ward sides of mountains in the trade-wind zones should therefore 
 be regions of little precipitation. The west slope of the Andes 
 Mountains is a case in point (Fig. 604). A high mountain range 
 on the east side of a continent in the zone of the trades would tend 
 to make all the area to the west of it dry, unless it also is affected 
 by high mountains. 
 
 Since the trade-winds shift a little with the seasons (Fig. 589), 
 the tracts which receive rain from them also shift. Tracts which 
 have trade-wind rains at one season, but not at another, often 
 have wet and dry seasons, and, in general, the dry season corre- 
 sponds with the time of the trade-winds. 
 
 In the zone of equatorial calms, also called doldrums, the tem- 
 perature is high, and the air, warmed by the sun daily, expands 
 and is crowded upward by the cooler air which comes in from the 
 xones of the trades. As it rises, the air expands and cools, and 
 often gives up some of its moisture. In this zone, therefore, there 
 are likely to be daily rains from cumulus clouds at the time of day 
 when the upward currents are strongest, that is, in the afternoon. 
 Since the doldrums shift north and south a few degrees yearly with 
 the shifting of the thermal equator, a place near the equator which 
 receives the daily rains during one season, may be without them 
 at another time of the year. 
 
 In the zone of tropical calms (p. 601), air is descending rather 
 than rising, and so yields little or no rain. Like the equatorial 
 calms, these extra-tropical calms shift north and south a little with 
 the sun. They are, on the whole, the driest latitudes of the globe, 
 crossing Sahara, Arabia, Australia, and the southern part of South 
 America. 
 
 Rainfall in the zones of the prevailing westerlies. The prin- 
 ciples which apply to the trade-wind zones apply also in the zones of 
 the westerly winds. These winds are, on the whole, blowing from 
 lower to higher latitudes, and so are being gradually cooled. They 
 might therefore yield some moisture, even at sea-level or on low 
 land, and especially on land in the winter season. The heat of the 
 
618 PHYSIOGRAPHY 
 
 land in summer often prevents condensation and precipitation until 
 the air has moved far to poleward. When such winds cross moun- 
 tains, they yield moisture to their windward slopes and summits, 
 and become dry on the leeward slopes. 
 
 Planetary winds alone considered, a high mountain range on 
 the west side of a continent in the zones of westerly winds would 
 make all the low land to the east of it dry. 
 
 An application of these principles will help us to understand 
 the rainfall of the United States, so far as it is dependent on planet- 
 ary winds. 
 
 Our prevailing winds for almost all the country are from the 
 southwest. Coming on to the land from the Pacific in the winter, 
 these winds reach the cooler land, and yield moisture, even at low 
 levels. This gives the low lands of California their wet season. 
 As the winds blow over the high mountains back from the coast, 
 they yield more moisture, so that all the area west of the crest of 
 the first high range is well supplied with rain and snow in the winter 
 season. As the winds blow beyond the Sierras and Cascade Moun- 
 tains, the air descends and becomes warmer, and therefore dry. 
 East of these mountains lie the semi-arid lands of eastern Oregon 
 and Washington, and the Great Basin with its Great Salt Lake. 
 
 When these winds reach the higher parts of the Rocky Mountains, 
 which are often higher than the mountains farther west, they again 
 yield some moisture. But farther east, all the way to the Atlantic, 
 these winds, taken by themselves, would remain dry, for they cross 
 no more high mountains, and they do not generally go far enough 
 north to reach a temperature as low as that of the mountains they 
 have passed. For some distance east of the mountains the rain- 
 fall is very deficient; but east of central Kansas and Nebraska the 
 lands are well supplied with moisture. Southeast of a line run- 
 ning from about Galveston to Cleveland, the land might be sup- 
 plied with moisture by the southwesterly winds from the Gulf, but 
 there is abundant rainfall far to the west of this line. It is therefore 
 clear that some factor other than the westerly winds is involved in the 
 precipitation. This factor is the aperiodic cyclonic winds, to be 
 studied in the next chapter. Passing over the country from east to 
 west, the cyclones cause moist air to flow northward from the Gulf to 
 higher and cooler latitudes. This change in latitude, together with 
 the cooling of the air as it rises in the cyclone, is the cause of the 
 precipitation which redeems the central and eastern mrte of fVip 
 
GENERAL CIRCULATION OF THE ATMOSPHERE 619 
 
 United States from the aridity which affects the belt next east of 
 the Rockies. 
 
 The winds which blow from the Pacific to the continent in sum- 
 mer have a somewhat different effect upon rainfall, though the 
 principles involved are the same. At this time of year, the winds 
 which blow from the Pacific to the low lands of central and southern 
 California find a temperature on the land higher than their own 
 These winds are therefore dry in this region, and give to much of 
 California its dry season. Blowing inland, these winds reach moun- 
 tains so high that the temperature is low enough to give rise to con- 
 densation and precipitation. 
 
 Farther north the case is somewhat different. In Washington, 
 for example, the mountains near the coast are high enough to occa- 
 sion precipitation even in summer. In Alaska, where some of the 
 mountains are always covered with snow, precipitation is heavy in 
 the summer, and at high altitudes it often falls as snow instead of 
 rain. 
 
 Monsoon winds likewise yield moisture when they blow from 
 warmer to cooler regions. In general they blow toward warmer 
 regions, and so should be dry winds; but once started they are 
 sometimes forced up over high mountains, and precipitation follows. 
 The heaviest recorded rainfall on the southern slopes of the 
 Himalayas, is due to monsoon winds. Numerous famines in India 
 have followed the failure of the monsoon rains. The famine of 
 1876-78 affected 58,000,000 people directly, and is estimated to 
 have cost 5,000,000 lives. As in the case of the planetary winds, it 
 is the windward sides of the mountains which receive the heavy 
 precipitation from the monsoons. It is clear, therefore, that the 
 windward sides of high mountains are places of heavy rain- and 
 snowfall. 
 
 Land and sea (or lake) breezes (daily) rarely yield much rain, 
 though they often give rise to fogs when they blow from the warmer 
 water to the cooler land. Such fogs may occasionally be seen, as 
 at Chicago in the late autumn or early winter, sometimes ad- 
 vancing over the land with a wall-like front, varying from a few 
 feet to many scores of feet in height. 
 
 Valley breezes sometimes give rise to heavy showers, as already 
 noted. 
 
CHAPTER XVin 
 WEATHER MAPS 
 
 Aperiodic Changes of Pressure 
 
 FIG. 605 is a weather map for the United States for January 12, 
 1899. Like other weather maps, it shows (1) the distribution of 
 atmospheric pressure, (2) the direction of the winds in various parts 
 of the country, (3) the condition of the air with reference to cloudi- 
 ness, rainfall, snowfall, etc., at all points, and (4) the temperature. 
 
 1. Isobars. The full lines of the weather map are isobars. 
 The map shows a range of pressure from 30.6+ inches in the area 
 centering about the Hudson River Valley, to 29.5 in the area cen- 
 tering in North Dakota. The pressure is high (over 30 inches) in 
 the eastern half of the country, and low (less than 30 inches) in 
 the western interior, and high again, but not very high, in an area 
 near the Pacific coast. 
 
 The isobar of 30.6, in the eastern part of the United States, is 
 a closed line. On either side of it is the isobar of 30.5. Since the 
 pressure rises as the isobar of 30.6 is approached from either side, 
 it is inferred to continue to rise after this isobar is passed. The 
 area inside it is therefore inferred to have a pressure of more than 
 30.6 inches, but not so much as 30.7 inches, else another isobar would 
 have been represented. 
 
 Similarly, all points between the isobars of 30.6 and 30.5 have 
 pressures intermediate between those indicated by those figures. 
 The pressure is higher near the former isobar, and less near the latter. 
 
 The center of this high-pressure area is marked "high." " High" 
 on the weather map means an area where the pressure is distinctly 
 higher than that of its surroundings, and generally exceeds 30 inches, 
 and the word is placed in the center cf such an area. The move- 
 ments of air about a high are an anticyclone. 
 
 620 
 
WEATHER MAPS 
 
 621 
 
 To the west of this "high," ; the pressure decreases steadily to 
 North Dakota, where there is a center of low pressure, marked "low." 
 "Low" means an area in which the pressure is less than 30 inches, 
 and on the map the word is placed at the point in such an area where 
 the pressure is lowest. The movements of air about a "low" con- 
 stitute a cyclone. A cyclone is one type, and in middle latitudes 
 the most important type, of a storm. 
 
 The isobar of 29.5 about the low in North Dakota is a closed 
 line. Since the pressure is becoming less as this line is approached, 
 
 S jmfcoll Indicate : QClear 9 Partly Cloudy Cloudy H Rain Snow 
 
 FIG. 605. Weather map of the United States for January 12, 1899. The 
 full lines are isobars, the dotted lines isotherms. (U. S. Weather 
 Bureau.) 
 
 it is inferred that the pressure at all points within this isobar is less 
 than 29.5, though nowhere so low as ,29.4. At all points between 
 the isobars of 29.5 and 29.6, the pressure is between these figures. 
 West of the "low" the pressure increases. The pressure in the 
 high near the Pacific coast is not so great as that in the high over 
 the Hudson Valley. 
 
 Most weather maps show both lows and highs, or at least 
 one of each. This means that there is generally at least one area 
 of high pressure (anticyclone) and one of low pressure (cyclone) at 
 the same time within the area of the United States. Since this is 
 
622 PHYSIOGRAPHY 
 
 the case, the atmospheric pressures are generally somewhat unequal 
 in different parts of the country. 
 
 Weather maps are made by the Weather Bureau, a branch of the 
 national Department of Agriculture. They are prepared in vari- 
 ous central offices of the country. To these offices the facts con- 
 cerning the pressure and the temperature of the air, the direction 
 and velocity of the wind, the cloudiness, and the precipitation, are 
 telegraphed daily from numerous points or "stations" established 
 and maintained by the Government. 
 
 2. Wind. Wherever barometric pressures are unequal, iso- 
 baric surfaces are uneven. They are depressed in cyclones, and 
 elevated in anticyclones. As a result, there must be winds from 
 anticyclones to cyclones. On January 12, 1899 (Fig. 605), winds 
 must have been blowing out from the highs in the east and 
 west respectively, and toward the low in the northwest, on the 
 day when the pressures were as indicated on the map. The arrows 
 on the map show the direction of the winds, which blew as the 
 arrows fly, as reported from the various stations. 
 
 It will be seen that winds do not blow straight out from the 
 anticyclonic centres, nor straight in toward the cyclonic centers. 
 They doubtless start straight out from each high, but they are 
 deflected toward the right, as most of the arrows about the anti- 
 cyclones show. Similarly, the winds which blow toward the cy- 
 clonic centers do not blow straight toward them, but are deflected a 
 little to the right, as most of the arrows about the lows show. 
 In the southern hemisphere the deflections would be to the left. 
 Fig. 606 shows the theoretic circulation about highs and lows: 
 A, northern hemisphere ; B, southern hemisphere. 
 
 It will be noted that two arrows in the western high (Fig. 605) 
 are directed toward the center of the high-pressure area. They 
 probably mean that there are subordinate centres of lesser pressure 
 within the general area of the anticyclone, and the winds blow 
 toward them. If this is the case, the subordinate lows are too 
 weak to be shown by the isobars, which represent differences of 0. 1 
 inch. 
 
 Something as to the strength of the winds at various points may 
 be inferred from the map. The distance from the center of the 
 high in the east, Fig. 605, to Lake Michigan is about 800 miles. 
 The difference in pressure is about 0.5 inch. The gradient is 
 therefore about 1 (English system). This means a wind- velocity 
 
WEATHER MAPS 
 
 623 
 
 of about 12 miles per hour a fresh breeze between these 
 points. The velocity of the wind blowing from Michigan to North 
 Dakota is about the same. The velocity of the wind from Texas 
 to North Dakota is much less. In general, where isobars are 
 crowded, the gradient is high and the winds strong. Where they 
 are widely separated, the gradient is low and the air-flow gentle. 
 The winds in cyclonic storms occasionally attain a velocity of 40 
 to 60 miles an hour; but the average is much less, and the cyclonic 
 
 FIG. 606. Diagram showing the deflection of air currents about highs and 
 lows. A, northern hemisphere ; B, southern hemisphere. 
 
 (not tornadic, p. 667) wind which is violent enough to be destruc- 
 tive is rare. 
 
 The circulation of air about a cyclone is vertical as well as 
 horizontal: the air currents move in toward the center of the 
 storm, and spirally up at the same time. This upward movement 
 is of great consequence in its effect on precipitation. The upward 
 and outward course of the air movement in the cyclone is shown 
 in Fig. 607, which represents a vertical section of a cyclone, and 
 shows that the outflow above is chiefly to the eastward, the direc- 
 tion in which prevailing winds blow. 
 
 3. Cloudiness, precipitation, etc. On the weather maps the 
 open circle on the shaft of an arrow indicates clear skies; the 
 
624 PHYSIOGRAPHY 
 
 half-blackened circle shows that the sky is partly cloudy; while the 
 black circle (Texas, Wyoming, etc.) indicates general cloudiness. 
 Where R appears on the' arrow, it means that rain is falling, as, 
 for example, in Iowa and Alabama. Where S appears in the 
 same position, it shows that snow is falling, as in north-western 
 Minnesota, Virginia, and Maryland. 
 
 This weather map shows that more or less precipitation accom- 
 panies this cyclone, and the examination of a series of weather 
 maps will show that cyclones are vary often attended by rain or 
 snow. Whether the precipitation takes the form of rain or snow 
 depends on the temperature. 
 
 4. Temperature. The dotted lines of the weather map are iso- 
 therms. The isotherm of 50 F. (Fig. 605) crosses the Gulf States. 
 South of it the temperature is above 50, but not so high as 60, 
 within the area of this map. The isotherm of 40 is more irregular. 
 It extends from Georgia to New Mexico, but between these points 
 it turns north into Nebraska. All points between this isotherm 
 and that of 50 have a temperature intermediate between 40 and 
 50. 
 
 The isotherm of 30 is still more irregular. Dubuque, la., 
 Chicago, Cleveland, Charlotte, N. C., and Norfolk, Va., have 
 about the same temperature. An isotherm of 30 also extends from 
 Idaho to Naw Mexico by a crooked course, while a third isotherm of 
 30 appears about the low. Two isotherms of 30 are therefore 
 next each other on the map, one in the area to the east, and the 
 other in the area to the southwest of the low. 
 
 The temperature between these isotherms is to be interpreted as 
 follows: As the low is approached from the east, say from New 
 York, the temperature rises. In the middle of Lake Superior the 
 temperature is 20, and at Duluth 30. The next isotherm to the 
 west, instead of being 40, is 30, and the one still farther west is 20' 
 This arrangement of isotherms means that the temperature west of 
 the isotherm of 30 passing through Duluth is warmer than 30> 
 but not so warm as 40; while farther west the temperature again 
 becomes cooler, reaching 30 in the eastern part of North Dakota. 
 
 On the whole, the isotherms show two pronounced features: 
 (1) they have little relation to parallels, for places in the same lati- 
 tude may have very different temperatures, and places far apart in 
 latitude may have the same temperature; and (2) the isotherms 
 show a pronounced disposition to bend poleward where the isobars 
 
WEATHER MAPS 
 
 625 
 
 \ 
 
 \ 
 
 \\ 
 
 \ 
 
 \ 
 
 \ 
 
 \\ 
 
 \ 
 
 \ 
 
 \ 
 
 \ 
 
 \ 
 
 \ 
 
 \ 
 
 \ 
 
 \ 
 
 \ 
 
 \ 
 
 *S 
 
 73 
 
 II 
 fi w 
 
 J5.S 
 
 o 
 
 d 
 
 3 1 
 
 g 
 
 ^H 
 
626 
 
 PHYSIOGRAPHY 
 
 oooooooooooooooooo 
 
 r--01O^KMrfO<P<DC^<010*nMrHO 
 
 O lOCOIOtOtOIOtQIOCMCMOJCMCMCMCgcgcvICV] 
 
 c 
 
 t 
 
 r"" 
 
 I 
 t 
 
 t 
 
 c " 
 
 1,,, ^ 
 
 1 
 
 i 
 
 
 
 
 " f 
 
 
 1 
 
 > s 
 
 
 a '..,_ 
 " 
 
 i 
 
 
 \ 
 
 
 
 i 
 
 fc. 
 t 
 
 } 
 
 \ 
 
 oooooooooooo 
 
 - 
 
 r~oio^>ncvi 
 
 ooo 
 
 ncvir-4 
 
 2SSSS8 
 
 . g S S 2 S 8 
 
 o o o 
 
 ess 
 
 \ 
 
 I I I 
 
 S g S 8 
 
 9 
 
 o 
 
WEATHER MAPS 627 
 
 indicate low pressure, and equatorward where the pressure is 
 high. 
 
 Fig. 608 shows, by graphs, four of the weather elements for the 
 year at Chicago. The figure shows that the winds are strongest in 
 cold weather, that the proportion of sunshine is highest in mid- 
 summer, while precipitation is greatest in the early summer. The 
 same weather elements at other localities would give somewhat 
 different graphs, and in some cases they would be very different 
 (see Figs. 662 to 673). 
 
 All the weather maps which follow show some relationship be- 
 tween isobars and isotherms. In general the isotherms curve 
 southward (equatorward) about a high, and northward (poleward 
 about a low. To this general rule there are some exceptions. 
 
 The temperature, the pressure, the winds, the cloudiness, the 
 rain, etc., are the elements of the weather. All these things being 
 shown on the above map, it is appropriately called a weather map. 
 
 The lows and highs are sometimes much more pronounced than 
 those shown in Fig. 605. In Fig. 609 the low is more pronounced, 
 the pressure ranging from 29.0 at the center, to 30.1 in the east 
 and to 30.5 in the west. So great a range of pressure as shown by 
 this map is not of common occurrence. The isobars are closer to- 
 gether in this figure than in the preceding, and therefore indicate 
 stronger winds. The approximate velocity of the wind at various 
 points may be calculated from the map. The direction of the 
 winds about the low is the same as in Fig. 605. Cloudy skies 
 prevail in the southeastern part of the low, and snow is falling at 
 some points (Montreal, Duluth). The map also shows great ranges 
 of temperature in areas not far apart. Thus there is a temperature 
 of 30 F. at Sault Ste. Marie, and a temperature of - 10 at Winni- 
 peg, but little farther north; while Montreal has a temperature 
 above that of Santa Fe~. As in preceding illustrations, the low 
 temperature goes with high pressure, and the higher temperature 
 with low pressure. 
 
 Fig. 610 shows a large and less symmetrical low. The winds 
 blow toward it, but are deflected to the right of its centre. Cloudi- 
 ness prevails over a great area about the cyclone, and snow and 
 rain are falling at some points. 
 
 The low of this map dominates almost the whole country. 
 Measuring from the 30-inch isobar on the east to the 30-inch isobar 
 on the west, the cyclone is about 1800 miles across. The isotherms 
 
628 
 
 PHYSIOGRAPHY 
 
 bend northward on the south side of this low, while they curve 
 southward about the high north of Montana. Fig. 611 shows an 
 
 85 80 75 r 70 J 65' 
 
 115 110 J 105' 100- 95 90 85 
 
 FIG. 609. Weather map for January 16, 1901. (U. S. Weather Bureau.) 
 
 125 120= 115 110 105 100 95 90 85 80 75 70 65" 
 
 115" HO' 105" 10W 95" 90 85 80 75 70 
 
 FIG 610. \Veather map showing a large asymmetrical low, March 2, 1904. 
 (U. S. Weather Bureau.) 
 
 elongate cyclone, one diameter of which is very long; and Fig. 612 
 shows its transformation the succeeding day. The isotherms of Fig. 
 
WEATHER MAPS 
 
 629 
 
 612 show few peculiarities, save in the northwest (Nebraska, Wyo- 
 ming, Montana), where the temperature drops from 20 near Rapid 
 
 FIG. 611. Weather map showing a large elliptical cyclone, January 22, 1900 
 (U. S. Weather Bureau.) 
 
 6 100 95" 90" 85 80" 75 70 65 
 
 115 110 105 100 95 90 85 80 75 70 
 
 FIG. 612. Weather map for January 23, 1906, showing great changes in the 
 cyclone of the preceding day. (U. S. Weather Bureau.) 
 
 City, South Dakota, to 30 at Q'Appelle, in Alberta, a difference 
 far greater than can be accounted for by the difference in latitude. 
 
630 
 
 Q'Appelle, it will be seen, is southeast of a high where the fall 
 of temperature is pronounced, as shown by the crowding of the 
 isotherms in South Dakota. The rapid change of temperature is in 
 a region where the wind is strong, and from the northwest. 
 
 Around all the preceding cyclones some precipitation is indi- 
 cated, while around most of the anticyclones there is an absence 
 of precipitation. The chief reason for rainfall or snowfall about 
 
 FIG. 613. Weather map for December 9, 1898, showing a high of great area. 
 (U. S. Weather Bureau.) 
 
 a low is as follows: The inflowing air produces an upward spiral 
 current, and the rising air expands and is cooled (p. 537), and so 
 gives up some of its moisture. In the southeast quadrant of the 
 cyclone, additional precipitation results from the fact that the 
 air entering the cyclone is passing from warmer to cooler latitudes. 
 This is perhaps one reason why the precipitation about a cyclone 
 is greatest in this quadrant. The right-handed movement of the 
 air in the northern hemisphere tends to shift the center of prin- 
 cipal precipitation somewhat to the east of south of the center 
 of the cyclone. 
 
 In the anticyclone there is a descending spiral movement of 
 The descending air comes from an altitude which is colder 
 
 ar. 
 
 than that at the bottom of the atmosphere, and hence brings a 
 low temperature. Since the air is compressed and warmed as it 
 
WEATHER MAPS 
 
 631 
 
 descends, the winds from anticyclones generally bring clear weather. 
 The downward- and outward-moving air may, however, so mingle 
 
 115 110 105 100 95 90 85 80 75 70 
 
 FIG. 614. Weather map for September 24, 1903. The shaded area in this 
 and succeeding maps represents precipitation. (U. S. Weather Bureau.) 
 
 125 120 115 110 105 100 95 90" 85 
 
 115 110 105 100 95 90 85 80 75' 70 
 
 FIG. 615. Weather map for September 25, 1903. (U. S. Weather Bureau.) 
 with the warm air about it as to cause some of the moisture of the 
 latter to condense, giving rise to clouds, or even to precipitation. 
 
632 
 
 PHYSIOGRAPHY 
 
 Highs of great area, as well as lows of great area, sometimes 
 occur. Fig. 613 shows a high or anticyclone some 2200 miles 
 across, with a great range of pressure. The isotherms of this 
 chart stand in very definite relations to the isobars, low tem- 
 peratures going with high pressures. Denver, in the high, is 
 about 30 colder than the southern part of Maine, 3 farther north, 
 in a low. 
 
 Movements of Cyclones and Anticyclones. The highs and 
 lows do not remain in the same place from day to day. This is 
 shown by Figs. 614-620, as well as by the other weather maps 
 which follow, showing the weather of successive days. 
 
 In Fig. 614 there is (1) a low over t!:e Gulf of St. Lawrence; 
 (2) a high central over Iowa; (3) a low over British Columbia; 
 (4) a high in Oregon. 
 
 The map of the succeeding day (Fig. 615) shows (1) that the 
 low of the St. Lawrence Gulf has moved to the east; (2) that the 
 high of the interior has moved to West Virginia; (3) that the low 
 
 FIG. 616. Weather map for September 26, 1903. (u. o. \Veau.er 
 
 which was over British Columbia has moved to Dakota; while (4) 
 the high of the Oregon coast remains about where it was. 
 
 The map of the succeeding day (Fig. 616) shows (1) that the 
 high of the Virginias has moved on, but not so far as on the pre- 
 
WEATHER MAPS 
 
 633 
 
 I25 J 120' 115" 110' 105' 100' 95' 90 85 J 80 75 J 70' 65' 
 
 115 110" 105 100" 95 90 85 
 
 FIG. 617. Weather map for September 27, 1903. The symbol which appears 
 in central Arkansas and western Tennessee indicates a thunder-storm at 
 or near the point where the symbol occurs, during the twelve hours pre- 
 ceding the issue of the weather map. (U. S. Weather Bureau.) 
 
 115 110 105 100 95 90 85 80" 75 70 
 
 FIG. 618. Weather map for September 28, 1903. (U. S. Weather Bureau.) 
 
634 
 
 PHYSIOGRAPHY 
 
 ceding day; (2) that the low which was over JNorth Dakota is 
 now north of Lake Superior; (3) that the high of Oregon has 
 moved east to Idaho and Montana; and (4) that a weak low has 
 developed in Indian Territory. 
 
 The map of the 27th (Fig. 617) shows (1) that the high which 
 was over the Virginias has disappeared, presumably to the east; 
 (2) that the low which was north of Lake Superior is now north of 
 Lake Ontario; (3) that the high of Montana has moved southeast 
 to Kansas; (4) that the weak low in Oklahoma and Indian Ter- 
 ritory has disappeared; and (5) that another low has appeared 
 in southern California. The succeeding map (Fig. 618) shows that 
 all the highs and lows of the preceding map have advanced in a 
 general easterly direction. Fig. 619 shows that the two lows of 
 
 FIG. 619. Weather map for September 29, 1903. (U. S. Weather Bureau.) 
 
 Fig. 618 near the Pacific have unitsd, the southerly one having 
 moved over to the more northerly a not uncommon occurrence. 
 Fig. 620 shows the progress of this low as well as of other highs 
 and lows, and a great rain area about the central low. 
 
 While the highs and lows of these maps have all moved in a 
 general easterly direction, the highs moved rather more to the 
 south of east than the lows. The direction of the progress of the 
 highs and lows shown by these maps is the normal one, though 
 
WEATHER MAPS 
 
 635 
 
 individual cyclones and anticyclones depart notably from the nor- 
 mal. The average direction of the cyclone in our middle latitudes 
 is about N. 80 E., or 10 north of east. The anticyclones have a 
 somewhat more southerly course. 
 
 From the study of these maps not only the fact of movement, 
 but the rate of movement of the highs and lows, may be cal- 
 culated. Thus, from the 25th to the 26th (Figs. 615 and 616), 
 the low of British Columbia moved about 1200 miles. From 
 the 26th to the 27th, and again from the 27th to the 28th, the same 
 
 115 110' 105' 
 
 FIG. 620. Weather map for September 30, 1903. (U. S. Weather Bureau.) 
 
 storm moved between 600 and 700 miles, while from the 28th to 
 the 29th the movement was about 800 miles. The average veloc- 
 ity of cyclones in the United States is a little less than 29 miles per 
 hour (about 700 miles per day); that of anticyclones somewhat 
 less. 
 
 It is not to be understood that the rate of progress of the storm 
 is the same as the velocity of the wind. The velocity of the wind 
 depends on the isobaric gradients. A weak cyclone, that is, a cy- 
 clone in which differences of pressure are not great (Fig. 620), 
 gives rise to weak winds, even though the center of the storm 
 moves rapidly. A strong cyclone, that is, one in which the dif- 
 
636 PHYSIOGRAPHY 
 
 ferences of pressure are great (Fig. 609) . gives origin to strong winds, 
 even though the cyclone itself moves forward slowly. 
 
 Figs. 621 and 622 show the progress of lows and highs, or 
 cyclones and anticyclones, from December 24 to December 25, 
 1904. The course of the low central over Oregon on the 24th, is 
 indicated by the arrows on the map of the 25th. Figs. 623-626 
 show the movement of cyclones and anticyclones for four consecu- 
 tive days in February, 1903, and especially the course of a low from 
 Arizona (Fig. 623) to Maine (Fig. 626). Figs. 627 and 628 show 
 similar features for November 26 and 27, 1898. The progress 
 of highs and lows shown on these maps (Figs. 614-628) represents 
 the general course of movement of most similar atmospheric dis- 
 turbances. 
 
 The mean tracks of cyclones and anticyclones for the United 
 States are shown in Fig. 629. The heavier lines show the average 
 paths of anticyclones, and the lighter the tracks of cyclones. 
 Some anticyclones enter the United States from the Pacific, while 
 others originate on the land north and northwest of Montana. 
 The anticyclones take either a northerly or a southerly route across 
 the continent. The former extends through the Great Lakes region 
 to southern New England, while the latter reaches the Atlantic or 
 the South Atlantic coast. Anticyclones entering from the Pacific 
 may take either of these courses, and those originating in the north- 
 west may do the same, as shown by the figure. 
 
 The cyclones originate, or first appear, in various places. More 
 of them originate near the places where anticyclones are generated 
 than in any other place; but not a few originate in Colorado, the 
 Great Basin, in Texas, and elsewhere. Those originating in the 
 northwest usually pass through the Great Lakes region to northern 
 New England. Those originating farther south may follow a 
 southerly course to the Atlantic, or may pass to the northward. 
 Tropical cyclones, to be mentioned later, sometimes reach the 
 Gulf of Mexico from lower latitudes, and follow the coast thence 
 to the northeast. 
 
 Still another set of lines in Fig. 629, marked 1 day, 2 days, 3 
 days, and 4 days, show the average rate of daily progress of the 
 storms which come in from the northwest on successive days. 
 
 Weather maps are sometimes more complicated than those 
 shown in the preceding figures. Fig. 630 is a weather map on 
 which four highs and four lows, some of them feeble, appear. 
 
WEATHER MAPS 
 
 637 
 
638 
 
 PHYSIOGRAPHY 
 
WEATHER MAPS 
 
 639 
 
640 
 
 PHYSIOGRAPHY 
 
WEATHER MAPS 
 
 641 
 
642 
 
 PHYSIOGRAPHY 
 
WEATHER MAPS 
 
 643 
 
644 
 
 PHYSIOGRAPHY 
 
WEATHER MAPS 
 
 645 
 
 The map also gives some idea of the way lows and highs follow 
 each other. The relations of isotherms and isobars are also in- 
 structive. 
 
 It will be readily seen that the passage of a cyclone involves 
 a change in the direction of the wind. Thus in Fig. 623 the wind 
 at Buffalo is easterly, though in the zone of westerly winds. The 
 next day. after the storm centre has moved forward beyond Buf- 
 falo (Fig. 624), the wind is westerly. The easterly wind of an 
 approaching cyclone is generally taken as a sign of rain through- 
 out much of the eastern part of the United States. 
 
 FIG. 629. The heavier lines show the tracks of anticyclones, and the lighter 
 lines the paths of cyclones. Off the South Atlantic coast anticyclones 
 are likely to turn northward. (U. S. Weather Bureau.) 
 
 Cyclones do not affect the air to great heights. Even when 
 the great whirl or eddy is 2000 miles across, as is sometimes the 
 case, its height (depth) is rarely more than 4 or 5 miles. 
 
 Weather conditions of cyclones and anticyclones. During the 
 passage of a cyclone some air is drawn from lower to higher, and 
 therefore from warmer to cooler latitudes. In midsummer this 
 often gives rise to the "hot wave" (Fig. 631), though "hot waves" 
 are not always closely associated with cyclones. Similar winds 
 are known as the sirocco in the western Mediterranean region, and 
 they go by other names elsewhere. 
 
 " Cold waves" often attend the anticyclones. These winds are 
 
646 
 
 PHYSIOGRAPHY 
 
 known as northers in the southern part of the United States and 
 sometimes as blizzards in the northern part, though this name 
 
 '. 2 9! ... .1 ...'19:. . .'05 1 00 5 90 85* 80 75 70' 65' 
 
 110 105 100 95 90 85' 80 75' 
 
 FIG. 630. Weather map for December 8, 1900. (U. S. Weather Bureau.) 
 
 HOT W ATE 
 
 KA7.I1TUV TEMPERATURES 
 JULY 22, 1901. 
 
 FIG. 631. (U. S. Weather Bureau.) 
 
 usually implies heavy snowfall and high wind, as well as low 
 temperature. Fig. 632 shows a map for January 3, 1896, 
 
WEATHER MAPS 
 
 647 
 
 115' 110" 105 100' 95 90" 85* 
 
 FIG. 632. Map showing the minimum temperatures for January 3, 1896. 
 (U. S. Weather Bureau.) 
 
 125 120 115" 110 105 100 95 90 85 80 75" 70 65 
 
 115 110 105" IOC? 95 90 65 80 19" 70 
 
 FIG 633. Map showing the minimum temperatures for January 4, 1836. 
 This figure shows the progress of the cold wave from the preceding day. 
 At this time a freezing temperature has reached the orange groves of 
 Florida. (U. S. Weather Bureau.) 
 
648 PHYSIOGRAPHY 
 
 and Fig. 633 a map for the following day. The high of Mon- 
 tana has advanced to Arkansas and Mississippi, and a free/ing 
 temperature has been carried down to the orange groves of Florida. 
 
 The mistral of southern Europe belongs to the same class as 
 the northers of our country. 
 
 Origin of the cyclones and anticyclones of intermediate 
 latitudes. The origin of cyclones and anticyclones is not well under- 
 stood. Centres of low pressure might be brought about by the 
 excessive heating of certain areas; but this can hardly be the origin 
 of most cyclones of temperate latitudes, for they are more common, 
 stronger, and move faster in winter than in summer. In the winter 
 season they often originate in areas covered with snow, where 
 excessive heating is impossible. Similarly, anticyclones might be 
 conceived to result from the unusual cooling of certain areas; but 
 that this is not their course seems clear from the fact that they 
 sometimes originate in warm regions, and from the further fact 
 that they are not notably more abundant in cold weather than 
 in warm weather. 
 
 The origin of both sorts of disturbances is probably to be referred 
 to atmospheric movements rather than to atmospheric temperatures 
 directly. The cyclones are frequently regarded as eddies in the 
 descending air which started poleward from the equator. While 
 this may be true, it does not appear to be a satisfactory statement 
 concerning the origin of these common air-whirls. 
 
 Tropical cyclones. Cyclones sometimes originate in tropical 
 regions, and follow courses very different from those of the cyclones 
 in temperate latitudes. The cyclones of this class affecting North 
 America usually originate in the West Indies, and are most com- 
 mon in the late summer and early autumn. They follow a north- 
 westerly course until the latitude of Florida is reached. Here they 
 commonly turn to the northward, and later to the northeastward, 
 and have a tendency to follow the Atlantic coast. Figs. 634-637 
 show the course of one of these storms in August (27-30), 1893, 
 and Fig. 638 shows the average path of the tropical cyclones for the 
 months of August, September, and October, for the years 1878 to 
 1900. Storms of this sort are sometimes called hurricanes. 
 
 The tropical cyclones are usually more pronounced than those 
 of temperate latitudes ; that is, the gradient is higher and the winds 
 therefore stronger. They often do great damage along the coast, both 
 to shipping and to the low lands near the water. The storm which 
 worked such devastation to Galveston in September, 1900, is shown 
 
WEATHER MAPS 
 
 649 
 
650 
 
 PHYSIOGRAPHY 
 
WEATHER MAPS 
 
 651 
 
652 
 
 PHYSIOGRAPHY 
 
WEATHER MAPS 
 
 653 
 
 in Fig. 639, which also shows the course of the storm both before 
 and after September 8. The strength of the storm was exceptional, 
 and its course unusual^ as will be seen by comparing Fig. 639 with 
 Fig. 638. The unusual course was probably due to the combined 
 influence of (1) the anticyclone central over New York, which 
 tended to keep the tropical cyclone from advancing in that direc- 
 tion, and (2) the cyclone of the northwest, which favored the 
 movement of the storm in that direction (Fig. 640). Fig. 639 
 
 FIG. 638. Course of West Indian storms for August, September, and Octo- 
 ber, 1878-1900. The lighter lines show the tracks of individual storms, 
 the heavy lines the mean course. (U. S. Weather Bureau.) 
 
 shows that the rate of progress of the storm was very unequal. 
 Thus northwest of Cuba its progress was much slower than it had 
 been to the southeast. Just south of Florida it traveled only one- 
 fourth as far in twelve hours as it traveled in one hour southeast 
 of Cuba. Figs. 640-643 show the position and strength of the 
 storm at four stages of its progress. 
 
 Tropical cyclones do not occur in the South Atlantic, and their 
 point of origin is several degrees north of the equator, usually between 
 10 and 20, in the North Atlantic. Tn the Pacific they occur on 
 both sides of the equator. They come in the later part of the hot 
 season of the latitudes where they occur, and are thought to be 
 
654 
 
 PHYSIOGRAPHY 
 
WEATHER MAPS 
 
 655 
 
 caused by strong convection currents. Their apparently anomalous 
 courses are probably to be explained by the courses of the prevail- 
 ing winds. The lower part of the cyclone is in the horizon of the 
 trades, but the upper part of the great eddy is probably above the 
 trade-wind horizon and under the influence of northerly currents. 
 The effect of these two controls appears to be to carry the storm 
 somewhat to the north of west (in the northern hemisphere) until 
 it escapes the control of the trades altogether, after which it is influ- 
 
 FIG. 640. An early stage of the Galveston storm when it was central over 
 western Cuba. (Nat. Geog. Mag.) 
 
 enced primarily by the southwest winds. The course of the storm 
 after it escapes the trades, following a more northerly course than 
 the prevailing winds, is probably influenced by the temperature of 
 land and sea. 
 
 Storms similar to the West Indian hurricanes occur in the North 
 Pacific, originating in the vicinity of the Philippines, and sweeping 
 the coast of China. These storms are called typhoons. The courses 
 of typhoons are shown in Fig. 644. The Society Islands and the 
 low coral islands of the neighboring Low Archipelago were swept 
 
656 
 
 PHYSIOGRAPHY 
 
 by a destructive storm of this sort on February 7 and 8, 1906. The 
 Hongkong typhoon of Sept. 18, 1906, was estimated to have de- 
 stroyed 5000 lives, and property to the value of $20,000,000. 
 
 Weather predictions. Weather predictions are based on the 
 phenomena illustrated by the weather maps. Take, for example, 
 the map of the 25th of September, 1903 (Fig. 615). Rain accom- 
 panies the cyclone which is central over Dakota. Since this storm 
 
 FIG. 641. A later stage of the storm after its center had reached Galveston. 
 
 (Nat. Geog. Mag.) 
 
 has, for the last twenty-four hours, been moving a little south of 
 east at the rate of about 40 miles an hour, it is fair to presume that 
 it will move in this same general direction at a similar rate for the 
 next twenty-four hours. If, in this time, it advances to the Lake 
 Superior region, it will probably bring with it weather similar to 
 that which it is now giving to the region where it occurs. Hence 
 on the 25th, the day when the weather conditions are shown in 
 Fig. 615, the prediction might be made that rainfall is to be expected 
 in about twenty-four hours in the region about the head of Lake 
 Superior. 
 
WEATHER MAPS 
 
 657 
 
 The map of the 26th (Fig. 616) shows that the course of the storm 
 has changed a little, being slightly to the north of east, the common 
 path of cyclones. That is, after descending a little to the south 
 of east from British Columbia, cyclones are likely to turn to the 
 east, or even a little to the north of east, in the middle longitudes 
 of the United States (Fig. 629). On the 26th the prediction might 
 be made that the low which is central north of Lake Superior (Fig. 616) 
 
 FIG. 642. The same storm after it had become central about Dubuque,and. 
 much weaker. (Nat. Geog. Mag.) 
 
 will move on to the Gulf of St. Lawrence by the succeeding day, 
 and that rain will accompany it. Rain for the region about Lake 
 Huron and the area east of it may, therefore, be predicted. The 
 map for the 27th (Fig. 617) shows that the area of precipitation 
 extends far to the south. The preceding map had shown some 
 cloudiness in this region, but had afforded no warrant for the pre- 
 diction of such an area of cloudiness. Thunder-storms are shown 
 in the southern part of the area of cloudiness. 
 
 Temperature changes as well as changes in precipitation may 
 be predicted. Thus in Fig. 614 the isotherm of 40 bends south- 
 
658 
 
 ward notably in the high central over Iowa. As the high moves 
 east, it will probably carry the low temperature with it. Hence 
 it is safe to predict that the temperature will fall in the area into 
 which the anticyclone is to move. The map of the succeeding day 
 (Fig. 615) shows that the temperature of western Virginia has fallen 
 from about 60 to about 40 along the path of the high, while areas 
 much farther north are warmer. 
 
 The same map (Fig. 615) shows that North Dakota and Alberta 
 have a temperature of 50, that is, a temperature 10 warmer than 
 
 FIG. 643. A still later stage after the center of the storm had reached New 
 England. (Nat. Geog. Mag.) 
 
 that of western Virginia. It will be noted, too, that the relatively high 
 temperature of Dakota, Montana, and Alberta goes with a low. As 
 the low moves eastward, the presumption is that the temperature 
 along its path will become somewhat higher. This is shown by the 
 succeeding map (Fig. 616), which shows a temperature of about 50 
 north of Lake Superior. The same map shows how the isotherm of 40 
 bends to the southward in front of the high which is central over 
 western Montana. On this day Winnipeg has about the same temper- 
 
WEATHER MAPS 
 
 659 
 
 ature as Cheyenne, several hundred miles farther south. As the high 
 of Montana moves eastward, it will be likely to carry cold tempera- 
 ture with it. From this map, therefore, it may be predicted that 
 the temperature in Nebraska, Kansas, Iowa, and Missouri will fall. 
 The next map (Fig. 617) shows that the temperature at Omaha 
 has fallen from 50 to 40, while that of eastern Kansas has fallen 
 from 70 to 40. 
 
 The time at which the precipitation which a given storm may 
 bring to any given place will fall is calculated from the rate at which 
 
 FIG. 644. Typhoon tracks. (After Herbertson.) 
 
 the storm is progressing. Similarly, the time of arrival of a cold 
 wave which an anticyclone is likely to bring is predicted on 
 the basis of the rate of progress which the anticyclone is making. 
 These rates are known in advance by telegraphic reports. Pre- 
 dictions concerning the weather may be made more readily for 
 the central and eastern parts of the United States than for the 
 western part, for the storms have been under observation longer 
 before they reach the central and eastern parts. 
 
 Predictions may also be made as to the strength and direction 
 
660 PHYSIOGRAPHY 
 
 of wind. The principles involved will be readily understood, and 
 the data on which the predictions are based are received by tore- 
 casters the same as data concerning temperature and rainfall. 
 
 Failure of weather predictions. Weather predictions often fail. 
 The reasons are many. Among them may be mentioned the follow- 
 ing: 
 
 1. The cyclones and anticyclones sometimes depart widely from 
 the courses they are expected to take. They may veer so widely 
 from their normal courses as to avoid altogether the places they 
 were predicted to reach. Thus a storm may be in line for St. Paul, 
 to Avhich it is expected to bring rain and a rising temperature; but 
 instead of keeping its normal course, it may turn off to the north- 
 ward, and the rain which was predicted for St. Paul falls farther 
 north. 
 
 2. Storms often change their rate of advance, so that they arrive 
 earlier or later than predicted. Thus, if a storm which has been 
 advancing at the rate of 600 miles in a day suddenly stops or 
 advances but little, it does not bring the changes predicted to the 
 areas into which it was expected to advance. 
 
 3. A third cause of the failure of predictions is found in the 
 fact that storms sometimes appear and disappear without warning. 
 Fig. 616 shows a low of which there had been no indication on the 
 25th, central over Oklahoma and Indian Territory; Fig. 617 shows 
 that this low has disappeared. It occasionally happens that much 
 more pronounced storms, promising great changes of weather, dis- 
 appear. In such cases the predicted weather does not arrive, and 
 the failure is charged to the forecaster. 
 
 4. Predictions are sometimes based on insufficient data. It will 
 be noticed that on some weather maps the letter M appears in 
 various circles. This means that reports from the station where the 
 M appears are missing. If many reports are missing, the map is 
 correspondingly imperfect, but the forecaster must use such data as 
 he has, as well as he may, and issue a map. 
 
 5. Storms sometimes change their characters. Thus from the 
 map of January 20, 1895 (Fig. 645), it could not be foreseen that the 
 cyclone central in Colorado would develop. the pronounced charac- 
 teristics which appear on the map of the following day (Fig. 646). 
 
 6. In some situations storms are subject to many freaks. This 
 is the case, for example, at Chicago. The frequently erratic be- 
 havior of storms here is probably due to the influence of the lake, 
 
WEATHER MAPS 
 
 661 
 
662 
 
 PHYSIOGRAPHY 
 
WEATHER MAPS 663 
 
 which modifies temperature and air currents. No other of our 
 Great Lakes has so great extension in a north-south direction, and 
 no other therefore presents so broad a front to the prevailing 
 winds. 
 
 Forecasters, like other men, are fallible, but when they have tc 
 work with so many indeterminate elements, it is not strange that 
 they sometimes make mistakes, and one mistake is likely to be 
 remembered longer than many correct prognostications. 
 
 Property saved by predictions of storms, frosts, floods, etc. In 
 spite of all shortcomings, the warnings of storms, floods, cold 
 waves, etc., sent out by the Weather Bureau, have resulted in great 
 benefit to various interests. The value of this service of the Weather 
 Bureau is not always duly appreciated, and much less is heard of it 
 than would have been heard of the losses which would have been 
 incurred in the absence of the warnings. Unfortunately, it is not 
 always possible to devise protection against the evils of which the 
 Weather Bureau gives warnings. 
 
 It has been estimated that property valued at 15,000,000 was 
 saved in 1897 by warnings of impending floods. While this was 
 exceptional, considerable sums are saved each year in this way. In 
 1903-4 the estimated value of the saving was $1,000,000. 
 
 Shipping interests are served by storm warnings. Thus, in Sep- 
 tember, 1903, vessels valued at $585,000 were temporarily held in 
 ports along the coast of Florida by storm warnings. 
 
 Agricultural interests are also served by warnings of storms and 
 of "cold waves," and especially of frosts. Warnings led to the pro- 
 tection of $1,000,000 worth of fruit about Jacksonville, Fla., in 
 1901, with an estimated saving of half this amount. Other warn- 
 ings of cold in 1901 are estimated to have been the means of saving 
 $3,400,000 worth of property. Fruit- and truck-farming are the 
 phases of agricultural work most effectively served in this way. 
 
 Special Types of Storms 
 
 Thunder-storms. Thunder-storms are of common occurrence 
 in the United States. They are most common in warm regions 
 that is, either in low latitudes, or in the summer season of middle 
 latitudes. Not only this, but they are most common on days which 
 are unusually warm, and during the warmer parts of these days. 
 They are, however, not confined to the summer or to the warm 
 part of the day, for there are occasional thunder-storms in the winter 
 
664 
 
 PHYSIOGRAPHY 
 
 in middle latitudes, and even in high latitudes, and there are thunder- 
 storms at night. Peary reports a thunder-storm in North Greenland 
 in midwinter. 
 
 The first indication of a thunder-storm is usually a large cumulus 
 cloud (Fig. 647) which, in the zone of the westerly winds, generally 
 
 
 FIG. 647. Ascending currents and FIG. 648. Air-Currents in thunder- 
 cumulus clouds preparatory to storm. (After Ferrel.) 
 thunder-storm. (After Ferrel.) 
 
 appears in the west. The cumulus cloud (or thunder-head) which 
 yields the rain, like all cumulus clouds, is generated by an ascending 
 current of moist air. It moves eastward, and seems to rise as it 
 approaches the observer, but the rise is apparent only. As the 
 cloud reaches the place of the observer, there is usually a sharp 
 breeze, or "thunder-squall," rushing out before it. Shortly after 
 the squall the rain begins to fall. The rainfall is often heavy and 
 the drops large; but the downpour does not usually last more than 
 an hour, and often much less. Sometimes, however, a second 
 thunder-storm follows close upon the first (Fig. 648), thus prolong- 
 ing the period of rainfall. When a thunder-storm has moved on to 
 the east, the air is usually cooler and fresher, and the barometer 
 distinctly higher. 
 
 As water vapor condenses in the air, the water particles 
 become charged with electricity. The charge of the individual 
 droplets increases as they increase in size, and the lightning is due 
 to the discharge of the electricity from one part of a cloud to an- 
 other, or from one cloud to another cloud, or from cloud to ground. 
 Rain and snow bring down much electricity to the land and water. 
 
WEATHER MAPS 665 
 
 The flash of lightning is followed by thunder, the noise being 
 due to the vibrations in the air resulting from the disturbance 
 caused by the electrical discharge. The thunder has been com- 
 pared to the noise which follows any other violent disturbance in 
 the air, such as the explosion of a rocket or the cracking of a whip 
 (Davis). Rolling thunder may follow a prolonged flash of lightning, 
 or it may be due to a succession of flashes but slightly separated 
 from one another, or sometimes to the echoing of the thunder from 
 hills and mountains. 
 
 In temperate latitudes the thunder-storms usually occur during 
 the passage of cyclones, though they do not accompany all cyclones. 
 They are more common on the south sides of cyclones than else- 
 where, and they often occur at a considerable distance from the 
 centre of the storm. In middle latitudes, thunder-storms, like 
 cyclones, move in a general way from west to east; while in the zone 
 of trade-winds they move from east to west. In both cases they 
 move with the prevailing winds. 
 
 The forward movement of a thunder-storm is commonly 20 to 
 50 miles an hour. They often spread, and become weaker as they 
 move forward (Fig. 649) , and do not usually run long courses before 
 
 FIG. 649. Vertical section of a thunder-storm which is moving toward the 
 right. (After Koppen.) 
 
 disappearing. The period of a thunder-storm is usually much 
 shorter than that of the cyclone which it accompanies. 
 
 It sometimes happens that lightning at a great distance illu- 
 minates the clouds over a region where the lightning itself cannot 
 be seen. Where the clouds seen from a given point are thus illu- 
 minated by lightning which is itself invisible, the lightning of the 
 clouds is called heat lightning. The heat lightning is simply a re- 
 flection of lightning. It is more likely to occur in hot weather 
 
666 
 
 PHYSIOGRAPHY 
 
 than at other times, because lightning is more common at such 
 times. 
 
 Rainbows sometimes accompany or follow thunder-storms. 
 They are always seen opposite the sun, and hence are seen in the 
 
 west in the morning, and in the 
 east in the afternoon or evening. 
 They are usually seen just after 
 the passage of a thunder-storm, 
 while a little rain is still falling, 
 but after the sun has appeared. 
 They are seen on looking in the 
 direction opposite the sun; that 
 is they are in the east in the even- 
 ing, and in the west in the morn- 
 ing. The rainbow is due to the effects of the drops of water in the 
 atmosphere on the sun's rays. A bow is also seen through water 
 spray, such as that at a great fall, even when no rain is falling. 
 Whirlwinds. Distinct ascending whirls of air are often seen 
 on hot days. They are especially well seen in dusty regions, for 
 
 FIG. 650. Shape of thunder-storm 
 in ground-plan, illustrating growth 
 and change as it progresses. 
 (After Waldo.) 
 
 \ 
 
 FIG 
 
 . 651. Graph showing the relative frequency of thunder-storms in Chicago 
 in different months. (Cox, U. S. Weather Bureau.) 
 
 there the dust is swept up,' making the w r hirl distinctly visible. 
 They are often seen in dusty roads, plowed fields, etc., but are 
 seen at their best in deserts. From a given point in the Mojave 
 Desert of California, as many as eight or ten of these whirls, some 
 of them rather conspicuous and imposing, may sometimes be seen at 
 one time from a single point on a hot summer day. The whirl- 
 
WEATHER MAPS 667 
 
 winds are probably caused by the excessive heating of the air at 
 some point, and this excessive heating gives rise to a sharp con- 
 vection current. It moves on for a time with the prevailing wind, 
 but soon plays out. 
 
 In humid regions the whirlwinds do not usually appear to ex- 
 tend up to any considerable height, but in desert regions they 
 often reach heights of 1000 feet or more, as shown by the whirling 
 columns of dust. The rise is sometimes so great that the air is 
 expanded and cooled enough to cause condensation of even the 
 small amount of moisture contained in the desert air. Sharp 
 showers may then occur. Showers of this sort are likely to be of 
 short duration, but the rainfall is sometimes very heavy. If ex- 
 ceptionally heavy, such rains are known as cloudbursts. In such 
 a storm in the summer of 1898, rain enough fell in a few minutes, 
 in the vicinity of Bagdad, in the Mojave Desert of California, to 
 occasion serious washouts along the railroad for several miles. 
 A cloudburst at Clifton, S. C., June 6, 1903, caused the loss of 
 more than fifty lives, and property damage to the estimated 
 extent of $3,500,000. 
 
 Tornadoes. When a convection current is very strong, but 
 has very small diameter, the whirl sometimes becomes so intense 
 as to cause great destruction. A whirling storm of this sort is 
 known as a tornado. Tornadoes, like thunder-storms and whirl- 
 winds, are phenomena of hot weather. They occur in the United 
 States in the warm season, appearing earlier in the South and 
 later in the North. 
 
 The tornado may be looked upon as a concentrated cyclone 
 or an intensified whirlwind. The pressure in the center of the 
 tornado is usually much lower than in the center of a cyclone. In 
 a strong tornado the pressure at the center may be a fourth less 
 than that of its surroundings. Herein lies the explanation of one 
 phase of the destructive action of a tornado. During the passage 
 of a tornado the pressure may be reduced from the normal amount, 
 14.7 Ibs. per square inch, or 2117 Ibs. per square foot, to three- 
 fourths of this, or to 11 Ibs. per square inch or 1584 Ibs. per square 
 foot. If such a tornado passes over a closed building in which the 
 air pressure is normal (2117 Ibs. per square foot), the pressure on 
 the outside becomes 1584 Ibs. The walls are therefore pushed out 
 with a force of 533 Ibs. per square foot, and unless they are very 
 strong, they will collapse outward, as if the building had ex- 
 
668 
 
 PHYSIOGRAPHY 
 
 ploded. Often it is only the weakest part, such as a window, 
 which yields. 
 
 Not only is the pressure at the center low, but the area of low 
 pressure is very small. While a cyclone may be 1000 miles or more 
 across, a tornado may be no more than one-eighth of a mile across, 
 or even less. The result is that the pressure gradient in a tornado 
 is very much higher than in a cyclone, and the winds are violent. 
 Their velocities, estimated by the size and weight of the objects 
 moved, have been thought to reach 400 or 500 miles per hour. 
 With this velocity, or even a velocity which is much less, the de- 
 struction is great. Trees are overturned, buildings unroofed or 
 even blown down, and bridges hurled from their foundations. 
 
 A tornado is often heralded by a funnel-shaped cloud (Fig. 
 652), the point of which may be far above the ground. As the 
 funnel moves forward, its lower end may rise or fall. The tornado 
 
 FIG. 652. Funnel-shaped cloud of a tornado. Solomon, Kan. 
 (U. S. Weather Bureau.) 
 
 becomes especially destructive where the funnel sinks so as to ap- 
 proach or touch the ground. The cloud is due primarily to the 
 condensation of the moisture in the sharp convection current, and 
 the funnel shape is due to the expanding and spreading of the air 
 as it rises. 
 
 The tornado is, of all storms, the most destructive, but it usu- 
 
WEATHER MAPS 
 
 669 
 
 ally has a very narrow track, and does not commonly work destruc- 
 tion for a very great distance. After a short course it generally 
 plays out, or rises so high as to cease to be destructive. 
 
 One of the most destructive, though not one of the most violent, 
 tornadoes of recent times was that at St. Louis, May 27. 1896. 
 It was an incident of a thunder-storm in the southeastern part of 
 a cyclonic area central some distance northwest of St. Louis. 
 
 The humidity at St. Louis was exceptionally high, about 94. 
 At noon the barometer at St. Louis stood at 29.87, the tempera- 
 ture was 80 F., and the velocity of the wind 12 miles per hour. 
 By 1.45 the temperature had risen to 86. At 2 o'clock the barom- 
 
 j f 1 f# 2 46810X112 46 j 10 
 
 FIG. 653. Thermograph (at left) and barograph (at right); traces at St. Louis 
 during the tornado of May 27, 1896. (U. S. Weather Bureau.) 
 
 eter began to fall rapidly, and by 6 P.M. it had dropped to 29.59. 
 Meanwhile the wind had become shifting, and shortly before 6 
 o'clock had attained a velocity of 45 miles per hour, and by 6 
 o'clock the temperature had fallen to 77. 
 
 During the earlier part of the afternoon, cumulus clouds had 
 been abundant, but by 4.30 they had settled into a stratus cloud. 
 Soon after 5.00, thunder and lightning occurred, and rain began to 
 fall at 5.43. 
 
 At 6.04 there was a marked increase in the violence of the 
 wind, which shifted its direction rapidly. The barometer rose to 
 29.67, but fell almost instantly to 29.57, then rose to 29.67 in less 
 
670 
 
 PHYSIOGRAPHY 
 
 than five minutes, falling again .31 inch to 29.36 in fifteen minutes, 
 and then rose almost instantly to 29.76. Sharp oscillations of 
 barometric pressure occurred until 10 P.M. The wind probably 
 attained a maximum velocity of 120 miles per hour at 6.18, with 
 numerous and rapid changes in velocity and direction. The rain- 
 fall accompanying the storm was extremely severe, more than 2\ 
 inches falling. The electrical display was brilliant. 
 
 The destruction began at about 6.10 P.M. and lasted for several 
 minutes. The forward motion of the storm was at the rate of 
 about 36 miles per hour. The width of the belt of destruction was 
 about l\ miles where it entered the city, but it was constricted to 
 less than a mile farther on. 
 
 One of the extraordinary features of the storm was the fact 
 that its base was about 30 feet above the surface. Trees were 
 
 FIG. 654.- 
 
 - Weather map for the morning of the day (March 27, 1890) of the 
 Louisville tornado. (U. S. Weather Bureau.) 
 
 twisted off at this level, and the principal destruction of houses 
 was above the first floor. Evidences of great heat w r ere visible 
 after the storm, as shown in scarred branches and twigs, a phe- 
 nomenon which has been noted in some other tornadoes. 
 
 As in other tornadoes, the wind played many curious freaks. 
 Single stones and bricks were picked out of walls, w 7 hile the walls 
 
WEATHER MAPS 
 
 671 
 
 remained standing. In one case a span of horses attached to a 
 loaded wagon were taken away, though the wagon was not over- 
 turned. The most extraordinary recorded instance of violence 
 was in East St. Louis, where, at the approach to the bridge, a plank 
 2"X8" "was driven into ... a steel girder with such velocity 
 that it punched a hole in the webbing and remained sticking in the 
 
 FIG. 655. Weather map for the evening of March 27, 1890, at the time of 
 the Louisville tornado. (U. S. Weather Bureau.) 
 
 girder." The destruction of property in and about St. Louis was 
 estimated at about $13,000,000. 
 
 The interpretation put upon the storm was that "tornadic 
 action was developed successively at different points in the track 
 of the general storm," which was a thunder-storm belonging to the 
 class of thunder-storms "which move broadside in a southeasterly 
 direction." 
 
 A more violent tornado was that at Louisville on the 27th of 
 March, 1890, just before nine o'clock in the evening. Its rate of 
 advance was nearly 40 miles per hour, but its diameter was so 
 slight, about 300 yards, that it took but about three-fourths of a 
 minute for the storm to pass a point. It was accompanied by 
 "a most terrific electric display." Many weak buildings were 
 wrecked, the walls falling toward the center of the storm. Sev- 
 
672 
 
 PHYSIOGRAPHY 
 
 enty-six persons were killed and about 200 injured in Louisville 
 alone, and the loss of property was estimated at about $2,500,000. 
 
 FIG. 656. Track of tornado in the outskirts of Chicago, May 25, 1896. 
 (Cox, U. S. Weather Bureau.) 
 
 FIG. 657. General view of the wreckage caused by the tornado at Rochester, 
 Minnesota, August 21, 1883. 
 
 The path of the storm was traced for 75 miles, and throughout 
 this distance its width was nearly uniform. At least five tornadoes 
 occurred in Kentucky the same night. 
 
WEATHER MAPS 
 
 673 
 
 Fig. 656 shows the track of a tornado in the suburbs of Chicago 
 on May 25, 1896. 
 
 Fia. 658. Wreckage of the Union Station Power-house at St. Louis, May 
 27, 1896. (U. S. Weather Bureau.) 
 
 FIG. 659. Trees twisted off by tor- FIG. 660. Straws driven into dry 
 nadic winds. (U. S. Weather wood by tornadic winds. (U. b. 
 Bureau.) Weather Bureau.) 
 
 Waterspouts. Waterspouts are virtually tornadoes at sea. 
 When the base of the upward spiral movement is as low down as 
 the surface of the water, sea-water may be drawn up to some slight 
 
674 
 
 PHYSIOGRAPHY 
 
 extent by the ascending current. The lesser atmospheric pres- 
 sure in the centre of the whirl will occasion the rise of the water to 
 some extent at that point, and the upward current of air may catch 
 it and carry it upward. The larger part of the water in a water- 
 spout is, however, due to the condensation of the water vapor in 
 the air, and not to the uplift of water from the sea. 
 
 Foehn winds, Chinook winds, etc. When warm, moist air is 
 forced up over mountains, it precipitates some of its moisture. The 
 precipitation sets free heat, so that the air is cooled much less than 
 it would be otherwise. Beyond the crest of the mountains it de- 
 
 FIG. 661. Distribution of tornadoes in the United States, 1794-1881. 
 
 scends, and is warmed in the process. It is warmed much more 
 (often twice as much) in the descent than it was cooled in the 
 ascent, because moisture is not condensed during the descent (p. 
 572). It may therefore descend as a hot wind. Such winds are 
 known as Foehn winds in Switzerland and as Chinook winds in the 
 United States, especially just east of the Rockies. 
 
 These winds may be beneficial or harmful. Thus the Chinook 
 winds temper the rigorous winters of certain parts of the Northwest- 
 ern States and the Canadian provinces east of the mountains. 
 They frequently evaporate a foot or more of snow in a very few 
 hours. Such winds are sometimes called snow-eaters. These 
 winds make winter grazing possible over large areas. In Alberta 
 
WEATHER MAPS 675 
 
 the Chinook has been declared to be "the grand characteristic of the 
 climate as a whole, that on which the weather hinges." These 
 winds sometimes develop with great suddenness. At Fort Assini- 
 boine, Montana, on January 19, 1892, the temperature rose 43 F., 
 from 5.5 to 37.5, in fifteen minutes, under the influence of the 
 Chinook wind. In other cases the temperature has been known to 
 rise 80 F. in six or eight hours. The Chinook winds of summer are 
 sometimes so hot and drying as to wither vegetation, and some- 
 times to destroy crops completely. 
 
CHAPTER XIX 
 CLIMATE 
 
 IN the preceding discussions of temperature, rainfall, winds, and 
 weather, much has been said or implied concerning climate. The 
 principal points involved may be here summarized and applied to 
 the principal zones of the earth. 
 
 / > Definition. Climate is the average succession of weather condi- 
 \ ) tions for a considerable period of time. The summer climate of a 
 place is shown by the weather of many summers, not by the weather 
 of one. So with the climate of autumn or winter or spring. The 
 average weather conditions for 10 years would give some approxi- 
 mation to the true climate, those for 25 years would give a closer 
 approximation, and those for 50 or 100 years would be still better. 
 The distinction between climate and weather is correctly recognized 
 by such expressions as these : The winter climate of Chicago is cold 
 and windy, but the winter weather of Chicago in 1905-6 was mild. 
 // Climate is otherwise defined as "the sum total of meteorological 
 " conditions in so far as they affect animal or vegetable life." Ac- 
 cording to this conception of climate, those meteorological elements 
 which have most influence on life are most important in climate 
 (Hann) . 
 
 The principal elements of climate are (1) temperature and (2) 
 humidity, which includes (a) relative humidity (p. 570), (b) absolute 
 humidity, (c) degree of cloudiness, and (d) precipitation. A climate 
 may be described as warm or cold, dry or moist. In common 
 speech, other elements of climate are often neglected, but there are 
 others of importance, especially (3) wind. 
 
 Of these elements, temperature is, on the whole, the most impor- 
 tant, but from some points of view, relative humidity and precipita- 
 tion are hardly less important. 
 
 In characterizing the climate of a region, account is taken not 
 only of the average temperature of the year and of the several 
 
 676 
 
CLIMATE 677 
 
 seasons, but also of the temperature of exceptional seasons and of 
 the extremes of temperature during the season. These extremes are 
 considered not merely for their effects on averages, but also on 
 their own account. Sensible temperature, as distinct from absolute 
 temperature, is also to be taken into account. Moist air of a given 
 degree of heat seems much warmer than dry air of the same temper- 
 ature when the temperature is high, and much colder when the tem- 
 perature is low. Sunstroke is much more common wtjere the rela- 
 tive humidity is high than where it is low. Sunstrokes are rare, for 
 example, in the arid West, even with temperatures considerably 
 above those of Chicago or New York. Sudden changes of tempera- 
 ture are also less injurious where the relative humidity is low than 
 where it is high. Air of a given temperature seems much cooler 
 when in motion than when quiet. 
 
 Sim'.larly, climate takes account not only of the average amount 
 of yearly precipitation, but of the variations of precipitation from 
 year to year and from season to season, of its average distribution 
 throughout the year and of departures from this average, and of the 
 proportions \vhirh fall as rain and snow respectively. 
 
 The other elements of climate are considered in the same way, 
 their variations and extremes as well as averages being taken into 
 account. 
 
 Uniformity and variability. If the range of temperature is 
 small, the distribution of precipitation somewhat equal, and the 
 winds reasonably constant in direction and strength, the climate is 
 uniform. If, on the other hand, the variations of these climatic ele- 
 ments are great, either in a year or in successive years, the climate 
 is variable. The climate of the middle and northern latitudes of the 
 United States, for example, is variable, (1) because the annual 
 range of temperature is great, (2) because the range varies from 
 year to year, (3) because corresponding seasons have very different 
 temperatures, (4) because changes of temperature may be very 
 sudden, and (5) because the amount and distribution of rainfall vary 
 notably and irregularly from year to year, and from season to season. 
 It is the variability of the weather which makes weather predictions 
 important, and variable weather makes a variable climate. 
 
 Figs. 662-670 represent certain elements of variability. Fig. 
 662 shows the annual range of temperature in eight places: Duluth, 
 Chicago, Memphis, and New Orleans in the upper part; and Denver, 
 Chicago, New York, and San Francisco in the lower part It will be 
 
678 
 
 PHYSIOGRAPHY 
 
 seen that the range is greater in the higher latitudes. It is 
 120 at Duluth, 108 at Chicago, 87 at Memphis, and 70 at New 
 Orleans. Denver, Chicago, and New York do not differ widely in 
 annual range of temperature, but the range at San Francisco, where 
 the prevailing wind is from the sea, is notably less. Though near 
 
 130 
 
 120 
 110 
 
 100 
 90 
 80 
 70 
 60 
 50 
 40 
 
 130 
 
 120 
 
 110 
 
 100 
 
 90 
 
 80 
 
 70 
 
 60 
 
 50 
 
 40 
 
 75 1880 1885 1890 1895 1900 1905 
 
 
 
 
 
 
 
 
 1 
 
 
 1 
 
 s 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 130 
 120 
 110 
 100 
 90 
 80 
 70 
 60 
 50 
 40 
 
 130 
 120 
 110 
 100 
 90 
 80 
 70 
 60 
 
 
 
 
 
 
 
 
 / 
 
 \ 
 
 
 \ 
 
 
 s 
 
 ( 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 -, 
 
 
 
 j 
 
 'v 
 
 .- 
 
 
 
 
 ( 
 
 
 ** 
 
 
 . 
 
 
 
 
 ... 
 
 
 
 
 
 
 
 \ 
 
 J 
 
 
 
 
 
 
 
 5 
 
 ,' 
 
 
 
 ^ 
 
 ,' 
 
 
 
 
 
 
 I 
 
 
 
 
 
 
 
 --^ 
 
 
 
 ^ 
 
 // 
 
 s\ 
 
 
 
 
 s 
 
 
 
 
 s. 
 
 / 
 
 
 
 
 
 
 
 
 / 
 
 ^^ 
 
 
 ^, 
 
 
 / 
 
 
 
 
 
 s 
 
 s 
 
 / 
 
 s 
 
 2 
 
 \ \ 
 
 ' 
 
 
 
 
 . 
 
 
 
 X 1 
 
 
 
 \ 
 
 f 
 
 
 
 
 
 
 s 
 
 
 
 
 / 
 
 s 
 
 / 
 
 ~~^ 
 
 f 
 
 \l 
 
 / 
 
 
 
 \ 
 
 
 s 
 
 
 
 \ 
 
 / 
 
 
 
 
 \ 
 
 / 
 
 
 
 r ' 
 
 < 
 
 
 s 
 
 
 ^~ 
 
 
 
 
 
 
 
 
 1 1 
 
 ', 
 
 
 \ 
 
 s 
 
 
 
 
 \ 
 
 / 
 
 
 
 
 s 
 
 
 ' 
 
 --. 
 
 
 v t 
 
 
 
 
 
 
 
 
 ' 
 
 \ 
 
 
 
 J 
 
 1 
 
 ,' 
 
 i 
 
 
 
 
 
 
 
 
 
 
 - 
 
 
 
 7 
 
 
 
 
 "" 
 
 
 
 
 
 
 ,' 
 
 
 J 
 
 
 
 ' 
 
 
 
 ! 
 
 
 
 / 
 
 
 
 i ~. 
 
 
 
 
 --H 
 
 
 
 
 
 
 
 \ 
 
 
 
 
 
 / 
 
 
 
 "' 
 
 \ 
 
 
 v 
 
 
 ' 
 
 . 
 
 "-_ 
 
 ? 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 > 
 
 
 
 / 
 
 J 
 
 
 
 
 
 * 
 
 
 \ 
 
 
 
 ? 
 
 
 
 
 
 
 
 
 
 
 
 
 
 J 
 
 1 N 
 
 
 
 ,- 
 
 "^ 
 
 
 
 
 . 
 
 
 
 
 1 
 
 V 
 
 / 
 
 >^ 
 
 
 
 
 
 
 
 
 _- 
 
 
 
 
 
 *\ 
 
 / 
 
 
 \ 
 
 
 
 
 
 
 
 ' 
 
 \ 
 
 / 
 
 k v 
 
 
 
 j 
 
 ^ 
 
 
 
 1 
 
 ! 
 
 r' 
 
 ^ 
 
 ' 
 
 
 \ 
 
 
 
 t 
 
 k , 
 
 1 
 
 
 ^, 
 
 
 
 
 ^ 
 
 ^, 
 
 / 
 
 
 
 
 
 
 
 
 
 ' 
 
 >^ 
 
 1 
 
 
 
 '-, 
 
 
 
 
 ^ 
 
 ^ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 *" 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 \ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 s 
 
 
 k ^ 
 
 
 I 
 
 
 , 
 
 s 
 
 
 
 
 / 
 
 i 
 
 
 
 
 
 
 
 
 
 
 t 
 
 V- 
 
 
 ^ 
 
 ^ 
 
 
 
 
 
 
 
 
 
 \ 
 
 1 
 
 
 
 / 
 
 / 
 
 
 j 
 
 
 
 * 
 
 
 
 
 
 / 
 
 V 
 
 1 
 
 \ 
 
 2 
 
 
 s 
 
 
 
 - 
 
 
 
 s 
 
 
 
 v> 
 
 ft 
 
 
 
 
 
 
 \ v 
 
 ,, 
 
 
 / 
 
 y 
 
 \ 
 
 / 
 
 s 
 
 // 
 
 \ 
 
 1 
 
 
 
 
 \ 
 
 rf 
 
 
 
 ^ 
 
 
 
 
 
 s\ 
 
 t/ 
 
 
 
 
 
 
 [ 
 
 
 >< 
 
 r ' 
 
 i 
 
 ,' 
 
 
 
 
 
 1 , 
 
 
 
 
 
 
 
 
 A 
 
 j 
 
 f 
 
 
 
 
 2 
 
 
 
 
 
 ' 
 
 
 
 
 
 
 
 
 ' 
 
 
 
 
 
 / 
 
 l v 
 
 
 
 
 
 .. 
 
 
 i 
 
 f 
 
 
 
 
 
 ^ 
 
 
 , ; 1 
 
 > 
 
 
 , 
 
 z 
 
 
 
 
 / 
 
 
 
 
 
 
 
 
 
 r *' 
 
 > 
 
 
 
 
 
 !| 
 
 
 
 
 
 
 s 
 
 ' 
 
 
 
 
 
 
 '-' 
 
 s 
 
 ,- 
 
 
 
 
 
 
 ., ; 
 
 
 
 
 ! 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 / 
 
 
 s 
 
 
 f, 
 
 \ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ^ 
 
 f 
 
 "^ 
 
 
 / 
 
 \ 
 
 
 
 
 > 
 
 
 A 
 
 
 
 \ 
 
 
 / 
 
 
 =3 
 
 
 \ 
 
 / 
 
 
 = 
 
 
 ^ 
 
 9 
 
 ^ 
 
 
 / 
 
 
 50 
 
 40 
 
 / 
 
 
 \ 
 
 I 
 
 < 
 
 1 
 
 
 
 
 
 \ 
 
 f 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 \ 
 
 / 
 
 
 
 
 
 ^ 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 FIG. 662. The upper part of the figure, reading from the top down, shows 
 the annual range of temperature in degrees Fahrenheit at Duluth (aver- 
 age, 120), Chicago (average 108), Memphis (average 87), and New 
 Orleans (average 70), four places in about the same longitude, but in 
 different latitudes. The lower part of the figure, reading from the top 
 down, shows the an ual ranges of temperature in Denver (average 113), 
 Chicago, New York (average 94), and San Francisco (average 53),' 
 four places in similar latitudes but in different positions with reference 
 to the sea. (Cox, U. S. Weather Bureau.) 
 
 the sea, New York has a much greater range of temperature than 
 San Francisco, because the prevailing winds are from the land. 
 
 Fig. 663 shows the average winter temperature by years for 
 Chicago. It will be seen that the average temperature of some 
 winters is 7 higher than that of other years. Fig. 664 shows the 
 temperature of Chicago during two Januaries, when the weather was 
 
CLIMATE 
 
 679 
 
 abnormally warm. The average temperature for 1880 was 40, and 
 that for 1906, 33. Fig. 665 shows that the average temperature 
 for January at the same place is about 23. It also shows the 
 
 40 
 35 
 
 I8f 
 
 5-6 
 
 
 / 
 
 
 18! 
 
 0-1 
 
 
 
 
 18! 
 
 5-(i 
 
 
 t 
 
 
 liH 
 
 0-1 
 
 
 
 1'MM 
 / 
 
 30 
 25 
 
 
 
 s 
 
 X 
 
 
 
 
 \ 
 
 X 
 
 V 
 
 
 
 
 3 
 
 / 
 
 
 
 S 
 
 
 / | 
 
 Fio. 663. Average winter temperatures at Chicago in degrees Fahrenheit 
 1885-1905. (Cox, U. S. Weather Bureau.) 
 
 mean monthly temperatures for several cities, the range being 
 greater with increasing latitude, and with increasing distance from 
 the sea in the direction whence the wind blows. Fig. 666 shows the 
 
 10 
 
 15 
 
 20 
 
 25 
 
 30 
 
 50 
 
 45 
 40 
 35 
 30 
 
 ^ 
 
 
 
 -V 
 
 I 7 
 
 
 r- 
 
 
 p 
 
 
 
 
 
 4^ 
 
 I 
 
 =- 
 
 
 ^ 
 
 * 
 
 
 4 
 
 -V 
 
 1 
 
 1 
 
 E 
 
 
 
 
 
 i 
 
 25 
 20 
 15 
 10 
 
 
 
 
 
 
 
 _L_ 
 
 -y 
 
 -/- 
 
 
 
 
 
 
 i 
 
 =i 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 FIG. 664. Mean daily temperatures during two warm Januaries at Chicago. 
 The dotted line represents January, 1880, and the full line January, 1906. 
 The numbers at the left are degrees Fahrenheit. The average tempera- 
 ture for the former month was 40 F., and for the latter 33. The aver- 
 age temperature for January in Chicago is 23. (Cox, TJ. S. Weather 
 Bureau.) 
 
 variations in daily range which are sometimes possible in middle 
 latitudes far from the sea. Fig. 667 shows the variation in snowfall 
 in one locality for a period of twenty years, while Figs. 668 and 
 
680 
 
 PHYSIOGRAPHY 
 
 
 r T 
 
 
 
 
 
 
 
 
 f 
 
 ) r 
 
 I T 
 
 60 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 70 
 
 
 
 
 
 
 2 
 
 
 
 
 
 
 
 
 
 
 
 
 ? 
 
 
 -^s. 
 
 
 
 
 60 1 
 
 
 
 
 
 V 
 
 
 
 S 
 
 \ 
 
 
 
 
 
 
 
 ^_ - , 
 
 '- 
 
 "" 
 
 
 
 \ 
 
 
 s 
 
 
 ^_ 
 
 - 
 
 1 
 
 ,/ 
 
 
 
 
 
 1 
 
 
 v v 
 
 
 
 
 
 V 
 
 
 
 
 
 
 I 
 
 
 
 
 
 2 
 
 
 
 
 
 
 
 \ 
 
 
 
 
 ,' 
 
 y 
 
 
 
 
 
 
 
 1 
 
 \\ 
 
 30 
 
 , 
 
 'V 
 
 
 
 
 
 
 
 
 
 \J> 
 
 
 fc 
 
 2 
 
 
 
 
 
 
 
 
 
 \ 
 
 ?n 
 
 ** 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Chicago, 111. 48.5 
 
 80 
 70 
 60 
 50 
 40 
 30 
 20 
 
 _____ San Francisco, Cal. 55.7 
 
 ______ .._ Denver, Colo. 49.4 
 
 _ ____ Boston, L'ass. 48.9 
 
 K J J A S G N D 
 
 \ 
 
 \\ 
 
 Chicago, 111. 46.5 
 
 Harquette, Xich. 40.6 
 
 Uenphis, Tenn. 61.1 
 
 Kev; Orleans, La. 68.8 
 
 FIG. 665. Mean monthly temperatures in degrees Fahrenheit at Chicago, 
 San Francisco, Denver, and Boston, four places in similar latitudes; and 
 for Chicago, Marquette, Memphis, and New Orleans, places in different 
 latitudes in about the same longitude. Averages for the year are shown 
 below the graphs. (Cox, U. S. Weather Bureau.) 
 
 FIG. 666. 
 
 55 ; ; ; 
 
 : ; : : 68 
 
 1 iiii INI TUMI iii i ; 40 
 
 o HBBBBBBBI 
 
 
 
 FIG. 667. 
 
 FIG. 666. The range of temperature at Chicago for February 9, 1900 when 
 the range was 53 F. (from 9 to 62), and for March 24, 1891, when the 
 temperature did not vary, being 32 throughout the day. The curves 
 are illustrative of the great variation in daily range in "temperate " 
 latitudes. (Cox, U. S. Weather Bureau.) 
 
 FIG. 667. Total snowfall at Chicago, in inches, by winters, 1885-1905, 
 showing the great variations. (Cox, U. S. Weather Bureau.) 
 
CLIMATE 
 
 681 
 
 669, and Figs. 670 and 671, show the great differences in the 
 amount of snow at corresponding dates in successive years. 
 
 DEPTH OF SNOW OH GROUND 
 JANUARY 31, 1905. 
 
 FIG. 668. The numbers at the ends of the lines indicate the depth of snow 
 in inches. (U. S. Weather Bureau.) 
 
 FIG. 669. (U. S. Weather Bureau.) 
 
 It will be seen that a variable climate varies in different ways. 
 A climate which is regularly dry during one season of the year and 
 
682 
 
 PHYSIOGRAPHY 
 
 wet during another, is variable within the year with reference to pre- 
 cipitation, even though the range of temperature is not great; the 
 
 FIG. 670. (U. S. Weather Bureau.) 
 
 FIG. 671. (U. S. Weather Bureau.) 
 
 climate of such a region may, however, be very constant from year 
 to year. Such a climate is found on the borders of the equatorial 
 
CLIMATE 683 
 
 calms, which shift a little north and south with the apparent shift- 
 ing of the sun. A narrow belt on each border of the calm zone is 
 therefore alternately in the calms and in the zone of the trades. 
 At the former times it has plentiful rain, while at the latter it is 
 generally dry. 
 
 A region which is hot at one time of the year and cold at another 
 is variable within the year with respect to temperature. In such 
 regions, too, one winter or summer may be much cooler or warmer 
 than the next, giving a variation from year to year rather than from 
 season to season. A climate which is variable with respect to 
 temperature is not necessarily variable with respect to moisture. 
 Commonly, however, variations in temperature and moisture go 
 together. 
 
 In some regions the winds shift regularly from season to season, 
 as where monsoons blow (Figs. 595 and 597), and again along the 
 poleward borders of the trades. The climates of such places are 
 variable within the year with respect to winds, and this makes them 
 variable also with respect to some of the other elements of climate. 
 Considered from year to year, the climates of such regions may be 
 uniform. 
 
 These illustrations are sufficient to show that the meaning of 
 "variable climate " is itself variable. 
 
 Classification of Climates 
 
 As in the case of many other topics, climates may be classified 
 in various ways, and each classification helps to emphasize some im- 
 portant point. One classification has already been suggested, 
 namely, uniform and variable. Another classification has reference 
 primarily to the amount of heat received from the sun. On this 
 basis the earth is subdivided into climatic zones, the borders of 
 which are parallels. The climatic zones based on insolation rep- 
 resent solar climate. Solar climate is so much modified by various 
 factors other than insolation, that climatic zones bounded by lines 
 other than parallels have been suggested. The effect of land and 
 water on temperature has already been noted. So important is 
 this effect that climates are also classified into oceanic and conti- 
 nental, and continental climates, in turn, are subdivided on the 
 basis of (1) distance from the sea, (2) altitude, and (3) topographic 
 relations. The controlling element in most of these classifications 
 is temperature. 
 
684 
 
 PHYSIOGRAPHY 
 
 Climatic Zones 
 
 On the basis of climate, the earth is divided into several zones- 
 Those commonly recognized are (1) the torrid zone, which is cen- 
 tered about the equator, (2) the temperate zones, which occupy 
 the extra-tropical latitudes, and (3) the frigid zones, which lie 
 about the poles. Better names for these zones are the tropical, 
 
 jj 
 
 Portland 
 Ore. 
 
 Boston 
 Cairo, 111. 
 
 Winne- 
 mucca, 
 Nev. 
 
 v, 
 -*r 
 
 - . 
 
 s 
 
 5 
 
 r'x 
 
 ** 
 \ 
 
 \\, 
 
 I 
 
 \ 
 \ 
 
 s 
 
 s 
 
 7 
 2 
 
 
 
 1 
 
 1 
 
 /.-. 
 
 S 
 
 / 
 
 / 
 
 - 5 
 
 f 
 / 
 
 ->- 
 
 100 
 90 
 60 
 70 
 60 
 60 
 40 
 
 30 N.W. 
 ... Ceylon 
 20 Siberia 
 10 S- W. 
 Ceylon 
 Malabar 
 
 T 
 
 V 
 
 ~-*- 
 ^ 
 
 A 
 
 f- 
 
 ~ 
 / 
 
 11 
 
 -J 
 f- 
 
 j 
 
 / 
 
 -4 
 4, 
 
 f- 
 
 / 
 / 
 
 \ 
 
 J 
 
 / 
 
 " 
 
 s' 
 
 
 
 -\- 
 
 -V 
 
 -t 
 
 \ 
 
 ^^ 
 
 / 
 
 / 
 
 ( 
 
 f 
 
 \ 
 
 \ 
 
 1 
 
 I 
 
 / 
 
 f 
 
 V*, 
 
 \ 
 
 I 
 
 x 
 
 5 
 
 IOC 
 90 
 60 
 70 
 60 
 50 
 40 
 30 
 20 
 10 
 
 
 Havre, 
 Mt' 
 Tampa 
 San 
 Dieeo 
 1 Paso 
 
 
 V 
 
 ^ 
 
 ^ 
 
 ^ 
 
 -f- 
 +- 
 
 'r 
 
 x^ 
 
 \ 
 
 \-t 
 / 
 
 ,' 
 
 \ 
 
 ^ 
 
 ~~ " 
 
 j 
 
 -+ 
 
 ^ 
 
 \ 
 
 
 
 100 Coaf.t 
 
 90 
 60 
 70j 
 60 
 
 50 NorthSea 
 40 Germany 
 S. Medit- 
 30 erranean 
 Central 
 20 France 
 Steppes 
 10 of Russia 
 
 
 
 
 2 
 
 \, 
 
 * 
 
 . 
 
 ' 
 ^ 
 
 
 s. 
 
 > 
 
 N 
 
 V. 
 
 s* 
 
 ^ 
 
 / 
 
 / 
 
 ^^ 
 'x^ 
 
 / 
 
 x 
 
 100 
 90 
 60 
 70 
 60 
 50 
 40 
 30 
 20 
 10 
 
 
 Fio. | 672. Graphs showing percentage 
 of' rainy days in each month for eight 
 
 eight 
 
 stations in the United States. (Cox, 
 U. S. Weather Bureau.) 
 
 FIG. 673. Graphs showing percent- 
 age of rainy days in each month 
 for several stations in foreign 
 lands. 
 
 the intermediate, and the polar zones l respectively, and these 
 terms will be used hereafter. The limits of these zones have 
 been variously denned. One classification defines them by lati- 
 tude, a second by the direction of the winds, and a third by tem- 
 perature. 
 
 Ward, Bull. Am. Geog. Soc., Vol. XXXVII, 1905. 
 
CLIMATE 
 
 685 
 
 Zones defined by latitude. Defined by latitude, the tropical 
 zone is limited poleward by the tropics, and the polar zones are 
 limited, equatorward, by the Arctic and Antarctic circles respec- 
 tively, while the intermediate zones lie between the tropical zone 
 and the polar zones on either hand. 
 
 Stated in other terms, the tropical zone, according to this classifi- 
 cation, is the zone (1) where the sun is vertical at some time during 
 the year (except at the tropics, twice a year); (2) where varia- 
 tions in the length of day and night, are least; (3) where the annual 
 
 FIG. 674. Comparative data, average monthly precipitation. The letters 
 are the initial letters of the months. The places in the upper row, com- 
 mencing at the upper lett corner and reading to the right, are: Port- 
 land, Ore.; Havre, Mont.; Moorhead, Minn.; Marquette, Mich.; and 
 Boston. In the second row: San Francisco; Denver; Omaha; Chicago; 
 and Norfolk, Va. In the third row: Yuma, Ariz.; El Paso. Tex.; Gal- 
 veston; New Orleans; and Jacksonville, Fla. (U. S. Weather Bureau.) 
 
 insolation is greatest and the average temperature highest; (4) 
 where the range of annual insolation is least, and consequently (5) 
 where the annual range of temperature is least. 
 
 The greater insolation of this zone is explained by the lesser 
 obliquity of the sun's rays. At the equator, at noon, they depart- 
 nearly 23^ from vertically at a maximum; at the tropics, they 
 depart nearly 47 from vertically at a maximum; and at inter- 
 mediate latitudes by an intermediate amount. The average in- 
 
686 PHYSIOGRAPHY 
 
 clination of the sun's rays for this zone at noon is between 17 
 and 18, while the average inclination for the intermediate zones is 
 about 45. The sun's rays, therefore, depart much less from ver- 
 ticality in the tropical zone than in the intermediate zones. The 
 slight range of temperature in this zone is explained by two facts, 
 namely, (a) within the tropical zone the days and nights are never 
 very unequal (Fig. 536), and (6) the change in angle of the sun's 
 rays during the year is less than elsewhere. 
 
 The intermediate (temperate) zones are the zones (1) where the 
 sun's rays are never vertical; (2) where the days and nights are 
 very unequal, each ranging from 10J hours at the equatorward 
 limit of the zone to nearly 24 hours at the poleward limit, in 
 the course of the year, but where the sun never appears above the 
 horizon for twenty-four hours consecutively; and (3) where the 
 amount of annual insolation is less, and (4) its annual range greater, 
 than below the tropics. 
 
 The polar zones are the zones (1) where the days and nights are 
 sometimes more than twenty-four hours long. They are the zones 
 (2) of least annual insolation, and (3) of greatest range of insola- 
 tion. 
 
 According to this definition of the zones, the tropical zone is 
 about 47 wide, each of the temperate zones about 43, and each 
 of the polar zones about 232. 
 
 This classification has the merit of simplicity, and it has a 
 definite astronomical basis; but the limits of the zones thus de- 
 fined do not always separate one sort of climate from another. 
 As applied to actual climate, and to the things which climate affects, 
 the subdivision seems arbitrary. Thus, the climate in that part of 
 the intermediate zones near the tropics is essentially like that of 
 the tropical zone, while the climate of that part of the intermediate 
 zones next to the polar zones is not very different from that of the 
 polar zones. On this basis there is far more difference between the 
 climate of the lowest and the highest latitudes of an intermediate 
 zone, than between the climate of the lowest latitudes of an inter- 
 mediate zone and that of the highest latitudes of the tropical zone, 
 or between the climate of the highest latitudes of the intermediate 
 zone and that of the lowest latitudes of the polar zones. 
 
 Zones defined by winds. 1 If climatic zones be defined by 
 
 1 Davis' Elementary Meteorology. 
 
CLIMATE 687 
 
 the direction of prevailing winds, the tropical (or trade-wind) 
 zone is the zone where the trade-winds blow. It extends some- 
 what beyond the tropics, even to latitudes of 30 or 35 on the east- 
 ern sides of the oceans. The intermediate zones lie poleward from 
 the trade-wind zone, and are characterized by prevailing westerly 
 winds and variable climate, but they have no definite poleward 
 boundaries. If definite poleward boundaries must be assigned, 
 they might be placed at the polar circles, though the westerlies 
 prevail beyond them. This definition has the merit of placing 
 the dividing-line between the tropical and intermediate zones 
 where the simple and uniform climate of the trade-wind zone gives 
 place, poleward, to the more complex and variable climate of the 
 zone where the westerlies prevail. This classification has the 
 merit, therefore, of grouping together climates which are really 
 similar. It takes account of climatic elements other than tem- 
 perature. The poleward limits of the intermediate zones has 
 1 much less logical definition. 
 
 The lack of simplicity and of mathematical precision is some- 
 times considered an objection to the definition of zones by the 
 direction of the winds. Thus, the zone of the trades shifts north 
 and south with the seasons. On this basis, therefore, places near 
 the border of the trade-wind zone are sometimes in that zone and 
 sometimes in the zone of the westerly winds. As a matter of fact, 
 this annual shifting of a place from on& zone to another corre- 
 sponds with actual climatic conditions, even if it is not simple. 
 The basis of this definition of zones is therefore quite as rational 
 as that of the preceding, and if it leaves the zones with rather 
 vague boundaries, it is because nature has left them somewhat ill- 
 defined. 
 
 Zones defined by isotherms. If the zones be defined on the 
 basis of temperature, the dividing-lines between zones are iso- 
 therms. One proposed division makes the annual isotherms of 
 68F. the equatorward limits of the intermediate zones, while their 
 polar limits are the isotherms of 50 for the warmest month (Fig. 
 675). On this basis, the tropical zone is narrowed over the eastern 
 sides of the oceans and broadened on the western sides, as a result 
 of the influence of the lands. 
 
 On the whole, this seems a fairly satisfactory basis for the 
 definition of climatic zones, though it lacks the mathematical 
 
688 
 
 PHYSIOGRAPHY 
 
 simplicity and precision of the first, and it fails to take account of 
 some diverse elements of climate recognized by the second. 1 
 
 Each climatic zone has at least two principal subdivisions, a 
 continental and an oceanic. The oceanic climate of any zone 
 prevails where there are extensive areas of water, and the con- 
 tinental climate prevails elsewhere. 
 
 Oceanic climates. Oceanic climates are less variable than 
 continental climates in the matter of temperature. Between the 
 
 FIG. 675. Temperature zones. Degrees F. (After Supan.) 
 
 latitudes of and 40 the diurnal range of temperature is only 2 
 to 3 over the sea. It is far more on land. The annual range of 
 temperature over the sea is also much less than that on land. 
 This is illustrated by Fig. 676, which shows the annual variation 
 on the island of Madeira (curve M) and at Bagdad (curve Bd) 
 in Asia Minor. The former represents a marine climate, the latter 
 a continental climate. In higher latitudes the differences are still 
 greater, as shown by the curves V and N. The former repre- 
 sents the marine climate of Valentia on the southwest coast of 
 Ireland, and the latter the continental climate of eastern Siberia. 
 The sea retards the annual march of temperature more than 
 
 1 The classification of climates on the bases here considered is well 
 discussed by Ward, Bull. Geog. Soc. of Am., 1906, p. 4Q1, 
 
CLIMATE 
 
 689 
 
 the land does (Fig. 677). The springs are therefore relatively 
 colder and the autumns warmer than on land. The humidity of 
 the oceanic climate is greater than that of continental climates. 
 This results in more cloudiness and often in more rainfall, especially 
 in winter. The winds of the sea are, on the whole, stronger than 
 those of the land. The leeward shores of the oceans (the shores 
 to which winds blow) have climates which are essentially oceanic. 
 The more equable temperatures and the greater amount of mois- 
 
 J. P. M. A. M. j. j. A. s. o. N. o. j. 
 
 F. 
 
 AO 
 
 
 
 
 
 
 Bd 
 
 
 
 
 
 
 C. 
 
 
 
 
 
 7 
 
 
 \ 
 
 ^ 
 
 
 
 
 30* 
 20 
 
 00 
 -100 
 -20 
 
 -300 
 
 68 
 
 500 
 
 320 
 
 ko 
 
 
 
 
 / 
 
 
 
 
 \ 
 
 
 
 
 
 
 / 
 
 
 
 N 
 
 M 
 
 -~^ 
 
 \ 
 
 
 
 M 
 
 
 
 x^ 
 
 1 
 
 
 \ 
 
 
 
 
 M 
 
 & 
 
 J 
 
 
 / 
 
 / 
 
 V 
 
 
 v 
 
 N 
 
 ^. 
 
 Bd 
 
 
 
 ^ S 
 
 / 
 
 
 
 
 \ 
 
 
 * 
 
 
 V 
 
 
 
 
 v 
 
 
 
 1 
 
 
 
 
 
 
 \ 
 
 
 
 
 
 j 
 
 
 
 
 
 
 \ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 -4 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 I 
 
 
 
 
 
 
 
 
 
 
 
 
 I 
 
 
 
 
 
 
 
 
 I 
 
 
 
 / 
 
 
 
 
 
 
 
 
 \ 
 
 
 
 ^> 
 
 N 
 
 
 
 
 
 
 
 
 
 * 
 
 N 
 
 
 
 
 
 
 
 
 
 
 
 
 J. F. M. A. M. J. J. A. S. 0. "N. D. J. 
 
 FIG. 676. Graphs to illustrate oceanic and continental climates in different 
 latitudes. Af= Madeira, Bd= Bagdad, F = Valentia, and JV=Eastern 
 Siberia. (After Angot.) 
 
 ture on such coasts have important effects on vegetation and on 
 animal life. These effects go beyond the mere facts of life and 
 death, and even beyond degrees of thrift. For example, wheat 
 has less protein in a marine climate than in a continental. Again, 
 starch decreases and gluten increases with increase of temperature. 1 
 Potatoes grown in the arid West, where the necessary (but no un- 
 necessary) water is supplied by irrigation, produce more nutritious 
 
 1 Hann's Handbook of Climatology. 
 
690 
 
 PHYSIOGRAPHY 
 
 matter than those grown in moister climates. These are but illus- 
 trations of general facts. 
 
 Continental climates. In contrast with marine climates, con- 
 tinental climates have greater annual and daily ranges of tem- 
 
 C_T^ L- 
 
 "S 
 
 FIG. 677. Annual march of temperature (in degrees Fahr.) in continental 
 and oceanic climates. The horizontal line represents the annual average. 
 (After Hann.) 
 
 perature, and the seasons lag less than over the sea. In high 
 latitudes the skies are clearer and the winters colder; in low 
 latitudes the winters are warmer than over the sea. The hu- 
 midity and the rainfall are less, and the rain less frequent in the 
 interiors of the continents than over the sea; but its amount and 
 distribution are largely influenced by topography, winds, etc. 
 The air over continents is also dustier than that over the sea. 
 
 The differences between oceanic and continental climates, so 
 far as temperature is concerned, are indicated by the following 
 tables: 
 
 
 January. 
 
 July. 
 
 Year. 
 
 Range. 
 
 Mean temperature of oceans 
 Mean temperature of continents 
 
 17. 9 C. 
 
 7 3 
 
 10. 2 C. 
 22 9 
 
 18. 3 C. 
 15 
 
 7.6C. 
 15 6 
 
 
 
 
 
 
 Latitude. 
 
 0. 
 
 10. 
 
 20. 
 
 30. 
 
 40. 
 
 50. 
 
 Mean temperature of land 
 hemisphere 
 
 44.8 
 
 42.5 
 
 36 4 
 
 26 
 
 15.7 
 
 3.6 
 
 Mean temperature of water 
 hemisphere 
 
 22.2 
 
 21.2 
 
 19.6 
 
 17.4 
 
 12.7 
 
 7 6 
 
 Difference. . 
 
 22 . 6 
 
 21.3 
 
 16.8 
 
 8 6 
 
 3 
 
 -4 
 
 
 
 
 
 
 
 
 Desert climates are an extreme phase of continental climates. 
 Here daily ranges of temperature are great. As a result, winds are 
 
CLIMATE 
 
 691 
 
 high by day, and the air often so dusty as to make travel 
 difficult. The nights are calmer and cooler. As a result of the 
 great daily ranges of temperature and the high winds, rock break- 
 ing, due to changes of temperature (p. 73), and dust and sand 
 transportation by the wind (p. 55), are at a maximum. The 
 dryness is hostile to plants, and therefore to animals. 
 
 Since the littoral (coastal) climate on the windward side of the 
 continent is very like the oceanic climate of the same latitude, 
 
 15* 110* 105 100 9* 90 85 80" 75 
 
 FIG. 678. Figure showing sunshine in the United States in hours per year. 
 The numbers on the lines show the hours per year. (After van Bebber.) 
 
 the west coasts of the continents, in the zones of westerly winds, 
 have oceanic climates, and east coasts have continental climates. 
 In the zone of trade-winds the reverse is the case. 
 
 The climate of the littoral zones is sometimes controlled largely 
 by monsoon winds. So important are the monsoon winds that 
 it is proper to speak of a monsoon climate. Monsoons are gen- 
 erally on shore in summer, and so give summer rains; but locally 
 the monsoons give precipitation in winter. 
 
 Mountain and plateau climates differ from other continental 
 climates, because of (1) the increased insolation and radiation which 
 go with increase of altitude, (2) the lesser absolute humidity, (3) 
 the lower temperature, (4) the greater range of solar temperature, 
 
692 
 
 and (5) the greater frequency of precipitation, up to certain alti- 
 tudes. The difference between soil temperature and air tempera- 
 ture is also greater than at lower levels. 
 
 Mountains, like oceans, have relatively pure air and high winds. 
 They modify general winds and give rise to local winds, such as 
 mountain and valley breezes. They interfere with free horizontal 
 
 62 
 
 67 
 
 72 77 82 87 92 97 102 107 112 117 122 
 
 35 
 
 FIG. 679. Figure showing sunshine in Europe in hours per year. 
 (After Konig.) 
 
 interchange of air, so that pressure and moisture conditions may 
 be quite different on opposite sides of a mountain range. 
 
 Climatic effect of forests. Forests also exert a modifying in- 
 fluence on continental climates. They lower the summer tempera- 
 ture by increasing the radiating and evaporating surfaces, and by 
 increasing the cloudiness. They increase the relative humidity 
 of the air; but it seems to be uncertain whether they have much 
 effect on precipitation. The data for some regions seem to point 
 to an affirmative answer, and for others to a negative one. In 
 any case, they tend to hold back the water after it falls and to 
 
CLIMATE 693 
 
 retard the melting of snow, so that their general effect on the 
 moisture of the region is much the same as it would be if the precipi. 
 tation were increased. Forests also afford protection against winds. 
 With these distinctions in mind, we may study briefly the cli- 
 mates of the several zones. 
 
 THE CLIMATES OF THE SEVERAL ZONES 
 The Tropical Zone 
 
 The leading characteristic of the climate of this zone is its 
 relatively high temperature. Uniformity of winds, temperature, 
 and humidity are especially characteristic of the oceanic climate of 
 this zone. On land the variation of all the principal elements of 
 climate is greater. 
 
 The prevailing winds of the tropical zone are easterly, north- 
 easterly in the northern part of the zone and southeasterly in the 
 southern part, with a zone of calms (the doldrums) between 
 (p. 603). So long as these winds blow over the sea or over low 
 lands, they are, in general, dry winds (p. 616). Many lands in 
 their path, notably Sahara and parts of Australia, are desert; 
 but, where they blow over mountains or plateaus, they yield mois- 
 ture to them, especially to their windward sides (p. 616). The 
 abundant rainfall on the east slope of the Andes, on the table- 
 land of Brazil, and on the higher parts of the Hawaiian Islands, 
 are illustrations. Even in the Sahara there are mountains which 
 occasion precipitation enough to support forests, but the descend- 
 ing streams soon disappear in the surrounding desert. For reasons 
 which have already been pointed out (p. 617), the leeward sides 
 of mountains in the zone of the trades get little moisture from the 
 trade-winds. In many cases the water which falls on the moun- 
 tains, and flows thence to the plains below, may be utilized there 
 for irrigating the lands not favored by precipitation. 
 
 Monsoon winds are often pronounced in the tropical zone. 
 They sometimes displace or greatly modify the trades, and become 
 the controlling element in the precipitation, often giving rain to 
 regions which would otherwise be dry. It is from the southwest 
 monsoon that India and Farther India receive their heavy rains, 
 concentrated in a wet season when the monsoon blows from sea 
 to land. Since this monsoon blows during the warm season, rain 
 
694 PHYSIOGRAPHY 
 
 must fall at that time (July to September). Like other winds, 
 monsoons which blow from the sea over low lands yield much less 
 rain than those which blow over high lands. Monsoon winds often 
 relieve the dryness which would otherwise exist on the west sides of 
 mountains in the trade-wind zone. 
 
 The tropical zone is not entirely dependent on winds for its 
 rainfall. Rainfall and cloudiness increase toward the center of 
 the tropical zone, while the strength of the winds decreases. In 
 the doldrums (p. 617) convection currents give daily (afternoon) 
 rains, thus interrupting the generally arid low lands of the tropical 
 zone by a middle or sub-equatorial belt of more abundant rain- 
 fall. In this belt flourish the forests of the Amazon and of middle 
 Africa. Since the belt of calms shifts a little with the sun, the 
 zone of daily rains also shifts. A place which is in the belt of daily 
 rains at one time of the year may be in the path of the trades at 
 another, and may therefore have alternating wet and dry seasons. 
 
 Near the poleward margin of the tropical zone also there is 
 likely to be variation in precipitation, for places in such positions 
 find themselves in the zone of dry trades in the summer and in 
 the zone of variable winds in the winter. In neither case, how- 
 ever, are they favorably situated, so far as winds are concerned, 
 for heavy rainfall, and have abundant moisture only where topog- 
 raphy or some other special factor favors. The most pronounced 
 desert of the whole earth, the Sahara, lies near the poleward border 
 of the trade-wind zone. 
 
 Along the coasts of tropical lands the temperature is modified 
 by the daily sea-breezes as well as by the monsoons. Because of 
 the direction of the prevailing winds, the climates of the eastern 
 coasts of continents in tropical latitudes are affected by the sea 
 more than those of western coasts. 
 
 The range of temperature in the tropical deserts is very great. 
 The annual temperature of Sahara is about 80 F. The temperature 
 of the warmest month averages about 90 F. and that of the cold- 
 est about 70 F. The average annual range is therefore relatively 
 slight; but the yearly extremes are far greater, for it sometimes 
 reaches 120 F. (Fig. 540), and sometimes drops to 50. Great as 
 this range is, it is far less than that of most inland places in the 
 intermediate zones, where extreme ranges of 120 are not uncom- 
 mon. The daily range of temperature in tropical deserts is great. 
 
CLIMATE 695 
 
 Climate of Intermediate Zones 
 
 The average temperature of the intermediate zones is lower 
 than that of the tropical zone, its annual range greater, and its 
 daily range, on the average, less. 
 
 The lower average temperature of the intermediate zones 
 results from the fact that they receive less heat per square mile 
 than lower latitudes, where the rays are more nearly vertical. 
 The range of temperature from season to season is greater, because 
 of (1) the greater inequality of day and night, and (2) the greater 
 range in the angle of the sun's rays. In latitude 45 there are, at 
 the maximum (summer solstice), about 15^ hours of sunshine (and 
 heating), and 8J hours of night (cooling with no insolation), while 
 at a minimum (winter solstice) there are but 8 hours of sunshine 
 with 15 hours of night. Not only this, but when the days are 
 longest the sun's rays are most nearly vertical, so that the heating 
 power per hour is greatest when the days are longest, and least 
 when they are shortest (Fig. 536). The result is that the summers, 
 even in the latitude of 45, may be very hot, while the winters, 
 except near windward coasts, are very cold. The annual range 
 includes summer heat which, at its maximum, is not less than that 
 of the tropical zone, and winter cold which is often frigid. The 
 annual range is greater in the higher latitudes of this zone than 
 in the lower. These great extremes of annual temperature, as well 
 as the sudden changes of temperature and of humidity which 
 accompany the passage of cyclones and anticyclones, make the 
 term "temperate" singularly inappropriate for the climate of the 
 intermediate zones. These last statements are especially true of 
 the Intermediate Zone of the Northern Hemisphere. They have 
 less application to the Southern Hemisphere, where the area of 
 land in middle latitudes is not so great. 
 
 Figs. 662 and 666 show different phases of the variations to 
 which the temperature of intermediate latitudes is subject. 
 
 The effects of the cold winters and the hot summers of tem- 
 perate latitudes make themselves felt in the temperature of the 
 springs and autumns respectively, as already pointed out, but this 
 point is of so much importance, climatically, that it will now be 
 stated with greater fullness. 
 
 During the winter the ground is cooled, and in the higher 
 latitudes of the intermediate zones the water in the ground is 
 frozen to the depth of several feet. The frozen water in the ground 
 
696 PHYSIOGRAPHY 
 
 may be looked on as "stored-up" cold. Even where the rock is 
 dry, it becomes cold in winter, and has much the same effect as 
 ice on the air of the succeeding spring. Snow accumulated on the 
 land in winter, and ice formed on lakes and ponds, have the same 
 effect. When the lengthening days of spring come, with their less 
 oblique rays of the sun, the snow and the frozen water in the 
 ground must be melted and the soil and rock warmed, before 
 the air in contact with it can fully respond to the increased insola- 
 tion; for so long as the ground is cold, the air next to it cannot 
 become very warm. In intermediate latitudes, therefore, the 
 spring is retarded because of the snow, the ice, and the cold soil 
 and rock. A simple analogy may be suggested. If a fire were 
 built in a stove covered with ice, it would take notably longer for 
 the stove to warm the room than if the fire were built in a stove 
 already warm; for the ice must be melted before the stove can 
 bring the air about it to a high temperature. 
 
 In the early autumn, on the other hand, the ground is warm 
 from the heat of the summer, some of which has been absorbed 
 and retained by the soil, the rock, etc. Warmth has been stored 
 up, and the warmed ground helps to warm the air, so that, as the 
 days shorten and the rays of the sun become more oblique, the 
 temperature does not fall as fast as it would were the ground not 
 warm. Thus, the effects of the summer's heat hold over into the 
 autumn, as the effects of the winter's cold hold over into the spring. 
 The result is that September is much warmer than March in middle 
 latitudes, though insolation is essentially the same in the one 
 month as in the other. Cold is more effectively stored up than 
 heat, so that spring lags more than autumn. 
 
 The lagging effect produced by the storing up of heat and cold 
 is shown by a comparison of the average monthly temperatures 
 of individual places. The monthly averages for Chicago are as 
 follows: 
 
 January, 23 F. July, 75 F. 
 
 February, 27 F. August, . 72 F. 
 
 March, 35 F. September, 65 F. 
 
 April, 50 F. October, 53 F. 
 
 May, 61 F. November, 39 F. 
 
 June, 70 F. December, 30 F. 
 
 These figures show that the average temperature of Chicago 
 
CLIMATE 697 
 
 for March is less than that for November. The insolation of No- 
 vember is far less than that of March is, indeed, but little less 
 than that of January, the coldest month of the year. This notable 
 lagging of the temperature behind that normal to the insolation 
 is the result of the storing up of heat and cold. Again, the insola- 
 tion of May is but little less than that of July, but the stored-up 
 cold from the preceding winter prevents the insolation during May 
 from having as much effect as it does in July, after the cold stored 
 up in the preceding winter has been more largely dissipated. 
 
 The lagging of the seasons is greater in the higher latitudes of 
 the intermediate zones than in the lower, and the effects are more 
 pronounced inland than near coasts, and are less on west coasts 
 than on east ones. These and other significant facts are shown 
 by the tables on page 698. 
 
 Westerly winds prevail in the intermediate latitudes (p. 603), 
 and many of the distinctive features of the climate of these zones, 
 both as regards temperature and moisture, are determined by 
 these winds. As they blow from sea to land, as from the Pacific 
 to the American coasts, they are nearly saturated with moisture. 
 Where they blow over land which is warmer than the sea (low 
 lands in summer), they become dry winds, because they take up 
 moisture; but when they blow over land which has a temperature 
 lower than their own (most lands in winter and mountains at 
 most times), some of their moisture is condensed and rain (or 
 snow) falls. The windward slopes of high mountains in these zones 
 are therefore well supplied with moisture, while plains to the lee 
 of such mountains are likely to be dry. This is the explanation 
 of the general aridity of the regions east of the Sierra Nevada 
 and the Rocky Mountains. Though these regions have little rain- 
 fall, that which falls in the mountains is coming to be utilized to 
 some extent in irrigating the valley lands adjacent. 
 
 Middle latitudes are fortunately not dependent entirely on 
 the westerly winds for their rainfall. Cyclonic storms often fur- 
 nish a sufficient supply of moisture where the westerlies are dry 
 (p. 618). Thus east of the 98th meridian in the United States 
 the rainfall is generally adequate for agriculture, though most of 
 it is not supplied by the westerly winds from the Pacific. 
 
 The cyclone and the anticyclone are important factors in the 
 temperature as well as the precipitation of the intermediate zones. 
 They give us our greatest annual extremes of heat (cyclones in 
 
698 
 
 MONTHLY AVERAGE TEMPERATURES (FAHRENHEIT) FOR FOUR PLACES 
 IN LATITUDE 33 TO 35 
 
 
 Los Angeles. 
 
 Santa F<5. 
 
 Vicksburg. 
 
 Charleston. 
 
 January 
 
 53 
 
 40 
 
 48 
 
 50 
 
 February 
 
 55 
 
 46 
 
 51 
 
 52 
 
 March 
 
 58 
 
 55 
 
 57 
 
 57 
 
 April 
 
 60 
 
 64 
 
 66 
 
 65 
 
 May 
 
 63 
 
 72 
 
 73 
 
 71 
 
 June 
 
 67 
 
 84 
 
 81 
 
 77 
 
 July 
 
 70 
 
 88 
 
 83 
 
 81 
 
 August 
 
 73 
 
 85 
 
 82 
 
 80 
 
 September 
 
 71 
 
 76 
 
 76 
 
 70 
 
 October 
 
 64 
 
 66 
 
 67 
 
 66 
 
 November 
 
 60 
 
 52 
 
 57 
 
 55 
 
 December 
 
 53 
 
 44 
 
 50 
 
 50 
 
 Range 
 
 20 
 
 48 
 
 35 
 
 31 
 
 MONTHLY AVERAGE TEMPERATURES FOR FOUR PLACES IN LATITUDE 41 
 
 TO 43 
 
 
 Coast of 
 North Carolina. 
 
 Long. 100. 
 
 Chicago. 
 
 New York. 
 
 January 
 
 46 
 
 23 
 
 23 
 
 30 
 
 February 
 
 46 
 
 28 
 
 27 
 
 31 
 
 March 
 
 50 
 
 39 
 
 35 
 
 36 
 
 April 
 
 51 
 
 55 
 
 50 
 
 47 
 
 May 
 
 54 
 
 64 
 
 61 
 
 59 
 
 June 
 
 55 
 
 75 
 
 70 
 
 70 
 
 July 
 
 60 
 
 81 
 
 74 
 
 73 
 
 August 
 
 60 
 
 77 
 
 71 
 
 71 
 
 September 
 
 55 
 
 67 
 
 65 
 
 C5 
 
 October 
 
 53 
 
 56 
 
 53 
 
 55 
 
 November 
 
 50 
 
 41 
 
 39 
 
 42 
 
 December 
 
 46 
 
 31 
 
 30 
 
 34 
 
 Range 
 
 14 
 
 58 
 
 51 
 
 43 
 
 MONTHLY AVERAGE TEMPERATURES IN THREE PLACES IN LATITUDE 48 TO 50 
 
 
 Seattle. 
 
 Winnipeg. 
 
 Mouth of the 
 St. Lawrence. 
 
 January. ... 
 
 35 
 
 -6 
 
 7 
 
 February 
 
 37 
 
 -1 
 
 12 
 
 March 
 
 44 
 
 14 
 
 22 
 
 April 
 
 49 
 
 38 
 
 33 
 
 May 
 
 56 
 
 50 
 
 44 
 
 June 
 
 60 
 
 63 
 
 55 
 
 July 
 
 64 
 
 67 
 
 60 
 
 August 
 
 65 
 
 65 
 
 60 
 
 September 
 
 59 
 
 53 
 
 50 
 
 October 
 
 52 
 
 40 
 
 40 
 
 November 
 
 45 
 
 20 
 
 30 
 
 December 
 
 40 
 
 4 
 
 15 
 
 Range 
 
 30 
 
 73 
 
 53 
 
CLIMATE 
 
 699 
 
 summer) and cold (anticyclones in winter). They also occasion 
 the great aperiodic changes of temperature which recur at short 
 intervals ("spells" of heat and cold), as well as the sudden changes 
 of weather, and so are an element of the variable climate of these 
 zones. Figs. 668 to 671 represent one phase of the variations to 
 which the continental climate of the intermediate zone is subject. 
 
 The climates of the northern and southern intermediate zones 
 are very unlike, because of the greater expanse of land in the former. 
 The climate of the southern zone is essentially oceanic, for the 
 lands are limited. The prevalence of water reduces the extremes of 
 temperature. Compared with the corresponding zone of the 
 northern hemisphere, the cool summers are one of the most notable 
 characteristics of this zone. Cloudiness and humidity are also 
 prevalent, except in the lee of mountains. These characteristics 
 make the climate relatively inhospitable, and the lands of the 
 southern hemisphere in latitudes corresponding to those of London 
 and New York are usually unproductive, more because of the cold 
 summers, than because of the low temperatures of winter. A com- 
 parison of the annual and seasonal temperatures of southerly lati- 
 tudes, say in latitudes of 30 to 40, with those of the correspond- 
 ing latitudes in the northern hemisphere is instructive (Figs. 
 538, 539, and 540). 
 
 The temperature of January and July for the northern and 
 southern hemispheres (all latitudes) is shown in the following 
 table : 
 
 
 January. 
 
 July. 
 
 Difference. 
 
 Mean. 
 
 Northern hemisphere 
 
 8.0C. 
 
 22.5 C. 
 
 14. .5 C. 
 
 15. 2 C. 
 
 Southern hemisphere. . . 
 
 15.5 
 
 12.4 
 
 5.1 
 
 14.9 
 
 Earth as a whole 
 
 12.7 
 
 17.4 
 
 4.7 
 
 15.0 
 
 
 
 
 
 
 The oceanic climate of the north intermediate zone is com- 
 parable to that of the southern zone. The prevailing westerly 
 winds tend to carry the oceanic climate over onto the western 
 borders of the continents. Hence the mild climate of the western 
 coasts of both North America and Europe. On both these coasts 
 the range of temperature, like that of the tropical zone, is rela- 
 tively low. Blowing over the cooler land, the oceanic winds give 
 abundant moisture and often much cloudiness and fog, especially 
 in the higher latitudes. In the Americas, the tempering and moist- 
 
700 PHYSIOGRAPHY 
 
 ening effect of the winds is limited to a relatively narrow belt, for 
 in crossing the high mountains the air loses both the warmth and 
 the moisture it brought from the ocean. Beyond the mountains 
 therefore the direct effect of the ocean is little felt. The mountains 
 separate climates of notably different types. 
 
 In western Europe, on the other hand, there are no high moun- 
 tains facing the ocean for any considerable distance, and the 
 moist climate, without great extremes of temperature, which char- 
 acterizes the coast, passes gradually into the continental climate 
 of the interior, with its drier air and clearer skies. 
 
 The continental interiors of the intermediate zones have much 
 greater ranges of temperature than the western coasts, and the 
 ranges become greater with increasing distance from the ocean, and 
 with increasing latitude (Figs. 665 and 666). In Siberia, for 
 example, in high latitudes and far from a western coast, are found 
 the greatest annual ranges of temperature known (Fig. 547). 
 
 In these zones the prevailing westerly winds are interrupted 
 by storms throughout the year, and the winds are stronger and 
 moister in winter than in summer. 
 
 The interiors of the continents in this zone receive their pre- 
 cipitation largely from cyclonic winds, and the climate has the 
 variability which always characterizes the lands of cyclones and 
 anticyclones. 
 
 The eastern borders of the continents are in contrast with the 
 western. On the former, continental rather than oceanic climates 
 prevail. The differences are brought into effective contrast be- 
 tween Vancouver and Labrador, or between England and Kam- 
 chatka, on opposite sides of a continental area; or between Lab- 
 rador and England, or Kamchatka and Vancouver, on opposite 
 sides of an ocean (see Figs. 538 to 540). The contrast is greater 
 on the opposite sides of the Atlantic, than on opposite sides of 
 North America, (1) because the tempering effect of the Gulf Stream 
 on western Europe is greater than that of the Japan Current on 
 western America, and (2) because the Atlantic opens more broad 'y 
 to the cold Arctic Ocean, allowing more ice-water to pass down the 
 eastern coast of North America than along the corresponding coast 
 of Asia. 
 
 Climate of the Polar Zoties 
 
 The distribution of the sun's heat is more unequal in the 
 polar zones than in lower latitudes (p. 525).- At the poles there is 
 
CLIMATE 701 
 
 half a year of continuous night and half a year of continuous day. 
 Between the poles and the polar circles, the inequality of heat 
 distribution is somewhat less than at the poles, but still great. 
 The seasonal range of insolation is greater here than in other lati- 
 tudes, but, in spite of this fact, the annual range of temperature 
 is less than in some other latitudes. This is because much of the 
 surface is covered with snow or ice (or ice-cold water), and the heat 
 received cannot bring the temperature of the surface above 32 F., 
 so long as snow and ice remain. Where these conditions exist, 
 the summer temperature of the air is raised but little above the 
 freezing-point. Where the land is free from snow during the warm 
 season, on the other hand, the annual range of temperature is 
 great. The diurnal range of temperature is, on the average, not 
 so great as in lower latitudes. 
 
 Precipitation in the polar zones is not usually heavy, and much 
 of it falls as snow. Where the surface is continually covered with 
 snow or ice, the precipitation is generally heaviest in the summer. 
 The winds are then more heavily laden with moisture, and blow- 
 ing over the surface of snow and ice, the air is chilled to the point 
 of precipitation. Because of the low temperature of winter, the 
 air of that season contains but little water vapor, and so gives but 
 little rain or snow. 
 
 Even in very high latitudes, such as that of North Greenland, 
 anomalous conditions of weather sometimes occur, and show how 
 variable the climate may be. In the winter of 1894-5 a thunder- 
 storm with rain occurred in latitude 78 45'; and even in the 
 midst of the long winter night, a temperature above the freezing- 
 point has been known to occur in the same region. 
 
 These extraordinary phenomena are doubtless the result of 
 extraordinary movements of air in ways and for reasons not 
 well understood. 
 
 Rainfall and Agriculture, etc. 
 
 The relation of rainfall to agriculture has already been men- 
 tioned (p. 615). It should be added that the beneficial effects 
 of precipitation depend not only on its amount and distribution 
 through the year, but also on its rate of fall. Besides the damage 
 they occasion through floods, heavy downpours of rain are much 
 less advantageous to crops than slow rains. Some figures on this 
 point are shown in the following table. 
 
702 
 
 PHYSIOGRAPHY 
 
 ANALYSIS OF RAINFALL, AT IOWA CITY IN 1889 AND 1890 
 
 
 1889. 
 
 1890. 
 
 Total rainfall 
 
 724 mm. 
 
 687 mm. 
 
 Washing and flooding rains 
 Insignificant rains 
 
 330 " 
 36 " 
 
 149 " 
 29 " 
 
 Total utilizable rains 
 
 358 " 
 
 509 " 
 
 
 
 
 Much may be done toward utilizing semi-arid lands by selec- 
 tion of the crops to be raised, with especial reference to the tem- 
 perature and the moisture of the region to be tilled. 
 
 Harm calls attention to the fact that in Jamaica and the Barba- 
 does the sugar crop may be calculated with approximate accuracy 
 from the amount of precipitation. In South Australia, land which 
 has 8 to 10 inches of precipitation w T ill support 8 or 9 sheep to the 
 square mile. In New South Wales, 4 inches more of rainfall will 
 allow the land to support 96 sheep per square mile; an increase of 
 7 inches more (20 inches in all) will allow an equal area of land to 
 support 640 sheep. In Argentina, with 34 inches of precipitation, 
 land will maintain 2630 sheep per square mile. These figures 
 do not take account of possible differences of soil. 
 
 From the human point of view, w r inds are an important element 
 of climate. Calms are enervating and winds stimulating. Hy- 
 gienically, winds are of great importance where population is dense. 
 
 Climate and life. The distribution of life is controlled very 
 largely by climate. The dry deserts of low latitudes, the deserts 
 in the lee of lofty mountains, and the snow deserts of polar 
 regions, are essentially climatic. Where rainfall is adequate and 
 where temperature favors, life abounds wherever there is a proper 
 soil; and even the accumulation of a proper soil is largely in- 
 fluenced by climate. The best soil, inherently, is worthless where 
 water is wanting, or where the temperature is too low for plant 
 life. 
 
 Of Australia it has been said that "Land without rain is worth 
 nothing; and land in an Australian climate, with less than 10 inches 
 a year, is worth next to nothing. Rain-water, without land, if 
 the water can be stored in a reservoir and sent along a canal, is 
 worth a great deal." 1 
 
 1 Wills, cited by Harm. 
 
CLIMATE 703 
 
 Changes of Climate 
 
 Within historic time. There appears to be little basis for the 
 popular notion, especially among elderly people, that the climate 
 is changing. There seems to be a natural disposition to exaggerate 
 the striking features of notable seasons. Thus winters of heavy 
 snow or of intense cold come, in time, to be the only winters which 
 are distinctly remembered. Exceptional seasons thus come to 
 stand for the normal winters of the past. Another reason for the 
 notion that climate is changing appears, in many cases, to be 
 that those who entertain this view have changed their place 
 of residence, so that the comparison is unconsciously made be- 
 tween the climate of New York, for example, and that of Iowa, 
 climates which are somewhat different. Actual records of climate, 
 covering as much as a century for some parts of our country, 
 do not afford any basis for the conclusion that the climate is 
 changing materially. 
 
 Fluctuations of rainfall, temperature, etc., do occur in rela- 
 tively short cycles. Thus there seems to be a faintly marked 
 weather cycle of about eleven years, corresponding to the sun-spot 
 cycle; but it is not yet clear that this periodic change is general or 
 persistent. Hann says that the only thing which can be con- 
 sidered as proved is that there are traces of a parallelism in the 
 march of certain meteorological elements and that of the sun-spot 
 period. 1 
 
 A longer cycle of about thirty-five years is indicated for Europe, 
 where records have been kept longer than in our own country. This 
 conclusion is based on the weather data of more than two centuries. 
 Within the cycle of this duration there may be said to be two 
 focal periods of a few years each, one when the rainfall is above 
 and the temperature below the average, and the other when the 
 rainfall is below and the temperature above. These two focal 
 periods are not, however, symmetrically placed in the cycle. Thus 
 the period of minimum rainfall may occur five years after the 
 period of maximum rainfall, or it may occur thirty years after it. 
 In view of this great irregularity, it may be doubted whether the 
 cycle is to be regaided as based on laws which are universally 
 applicable. 
 
 1 Hann, Handbook of Climatology. 
 
704 PHYSIOGRAPHY 
 
 The focal periods, as set forth by Bruckner, are as follows: 
 
 Wet and 
 cool. 
 
 Interval 
 between. 
 
 Dry and 
 warm 
 
 Interval 
 between. 
 
 167 1-1675 i 
 
 
 1681-1685 N 
 
 
 ) 
 
 25 
 
 I 
 
 45 
 
 1696-1700 
 
 
 1726-1730 j 
 
 
 I 
 
 45 
 
 I 
 
 30 
 
 1741-1745 f 
 
 
 1756-1760 { 
 
 
 I 
 
 25 
 
 
 30 
 
 1766-1770 f 
 
 
 1786-1790 < 
 
 
 I 
 
 50 
 
 ( 
 
 r 
 
 37 
 
 1816-1820 ( 
 
 
 1820-1830 
 
 
 I 
 
 35 
 
 I 
 
 38 
 
 1851-1855 J 
 
 
 1861-1865 J 
 
 
 The reader may judge for himself as to the adequacy of the 
 basis of the thirty-five-year period. 
 
 Variations of climate reflect themselves in the movements of 
 the glaciers. This has been observed especially in connection with 
 the glaciers of the Alps. The glaciers advance after (commonly 
 some years after) the periods of maximum precipitation and mini- 
 mum temperature, and retreat after the opposite conditions are 
 most pronounced. 
 
 Certain historic facts have been interpreted to indicate changes 
 of climate in some regions since the beginning of the historic period. 
 Thus regions once populous are now too arid to support an abun- 
 dant population. This is the case in southwestern Asia and north- 
 ern Africa, where the ruins of aqueducts and irrigating canals 
 exist where there are now no adequate sources of water. 
 
 In geologic time. Going still further back, there is abundant 
 evidence of profound changes of climate in the course cf the earth's 
 history. There have been at least three periods and perhaps more, 
 widely separated in time, when glaciation took place where glaciers 
 do not now exist. During one of these periods of cold there were 
 extensive glaciers in low latitudes, even in regions which now have 
 tropical and sub-tropical climates (India, Australia, South Africa). 
 The first of these periods of exceptional cold made its appearance 
 early in the earth's history (at the beginning of the Paleozoic era, 
 or perhaps even before), and the last (the late glacial period) has 
 but just passed. 
 
 Warm climates, on the other hand, have persisted for long 
 periods in polar regions, even down to relatively recent times. 
 
CLIMATE 705 
 
 Thus Greenland enjoyed a warm climate not long (geologically) 
 before the development of its present ice-sheet. The data now 
 known seem to indicate that the climate of the present time is 
 cooler than that which has existed throughout the larger part of 
 the earth's history. 
 
 Repeated changes in humidity seem to be as clearly indicated 
 as changes in temperature. Arid climates have existed at various 
 periods of the earth's history, in regions which have moist climates 
 at the present time (e.g., New York and Ohio), and moist climates 
 have been enjoyed by regions which are now essentially desert 
 (e.g., Arizona). The aridity in the one case is indicated by salt 
 and gypsum deposits, and the humidity in the other by conclusive 
 evidence of luxuriant plant life in regions which are now desert. 
 
 In some cases the causes of these changes were doubtless local, 
 and due to changes in topography; but in others this explanation 
 is not applicable. It seems clear, therefore, that causes have 
 long been in operation which bring about variations both in tem- 
 perature and humidity. These causes have sometimes been thought 
 to be (1) geographic, and due to the changes in the relations of land 
 and water, or to changes in the topography of the land; (2) as- 
 tronomic, due to changes in the shape of the earth's orbit, the pre- 
 cession of the equinoxes, etc.; and (3) atmospheric, due to changes 
 in the constitution of the atmosphere. Still other causes have been 
 conjectured. As the facts concerning these changes accumulate, 
 they seem to be pointing to the third of these lines of explanation 
 as the most plausible. It cannot be said, however, that final con- 
 clusions have been reached. 1 
 
 1 On this point see Chamberlin and Salisbury's Earth History. 
 
PART IV 
 THE OCEAN 
 
 CHAPTER XX 
 
 GENERAL CONCEPTIONS 
 
 THE ocean occupies the great depressions in the earth's surface 
 (p. 5). The area of the depressed segments is about twice as great 
 as that of the elevated segments; but since the water more than 
 fills them, it spreads out over some 10,000,000 square miles of 
 the continental platforms. The result is that the ocean water 
 covers nearly three-fourths (about T W) of the earth's surface. 
 All the oceans are connected at the surface, and are therefore in 
 some sense one, though different names are applied to different 
 parts; but the ocean basins proper are measurably distinct. 
 
 Although the depressions in which most of the ocean water lies 
 are called basins, they have little resemblance to the homely vessel 
 which this name suggests. This is readily seen by the construc- 
 tion of a diagram. An arc three feet long, with a radius of about 
 four feet, represents approximately the eighth of a circle. If such 
 a curve be drawn on the blackboard, it may be taken to represent 
 the width of the Atlantic Ocean between the United States and 
 Europe. If the top of the chalk-line be taken to represent the sur- 
 face of the ocean, another line representing the bottom of the ocean 
 could not be drawn below it with an ordinary crayon, without ex- 
 aggerating the depth of the water. 
 
 Fig. 680 may help to give us some conception of the real shape 
 of an ocean basin. It is, in general, convex upward, but locally, 
 especially where it joins the continental platforms, it is often con- 
 cave upward. Fig. 681 shows the belts where the bottom is con- 
 cave upward. These belts are usually 100 to 300 miles in width. 
 
 706 
 
GENERAL CONCEPTIONS 
 
 707 
 
 The sea-level. The surface of the sea is in sharp contrast 
 with that of the land, in that the former seems 
 to be level. We are accustomed to speak of it as 
 if it were free from all unevennesses, and it is the 
 datum plane from which all elevations on land 
 are measured. It is, therefore, of importance to 
 understand what the sea-level really is. 
 
 In the first place, it is a curved surface, and 
 its curvature is approximately that of an oblate 
 and slightly imperfect spheroid (p. 482). But its 
 surface corresponds only approximately to that of 
 a spheroid, for the land-masses which rise above 
 the ocean basins and which culminate in moun- 
 tains attract the waters of the sea to themselves, 
 and so act somewhat against the principal at- 
 traction of gravitation, which tends to draw all 
 objects toward the center of the earth. The 
 Andes Mountains, for example, rise far above the 
 sea close at hand, and the water adjacent to 
 them is pulled up somewhat above the normal 
 spheroid level by their attractive force. It has 
 been estimated that the sea-water at the mouth 
 of the Indus on the coast of India is 300 feet 
 higher (that is, 300 feet farther from the center of 
 the earth) than that about the island of Ceylon 
 at the southern end of the peninsula. This dis- 
 tortion of sea-level is due to the attraction of the 
 Himalaya Mountains and the adjacent high lands, 
 All land-masses act in the same way, and the 
 distortion is greater the greater the mass of land 
 above sea-level close to the shore. 
 
 The sea-level, therefore, does not correspond 
 exactly with the curvature of a spheroid. Further- 
 more, the heights and masses of mountains vary 
 from age to age, so that their distorting effects 
 vary somewhat in long periods of time. If it, is 
 desired to record the elevation of a place, for 
 example in California, in a way which will be per- 
 manently accurate, it should be recorded not only that it is, say, 
 500 feet above sea-level, but that, for example, it was 500 feet 
 
708 
 
 PHYSIOGRAPHY 
 
GENERAL CONCEPTIONS 709 
 
 above sea-level in latitude 40 on the coast of California on 
 January 1, 1900. 
 
 Apart from the more or less permanent distortions of the sur- 
 face of the sea, due to the attraction of land-masses, there are tem- 
 porary and slight inequalities of level, which will be considered 
 later. 
 
 What the physical geography of the sea includes. The physi- 
 cal geography of the sea includes many things. Among them are 
 (1) the distribution of its waters, (2) its depth at all points, (3) 
 the topography of its bottom, (4) the composition of the water, 
 (5) its color, (6) its temperature at all points at the surface and 
 beneath it, (7) its movements, (8) its life, and (9) the material of 
 its bottom. 
 
 The physical geography of the sea has become known, so far as 
 it is now known, in various ways. The distribution of its waters 
 has been made clear by outlining the areas of the land. The char- 
 acter of its waters is determined by chemical analysis. The move- 
 ments of its waters are studied in various ways. Some of them, 
 such as waves, may be studied from the shore; others, such as the 
 currents, are less readily observed, but have become known, so 
 far as they are known, (1) by their effects in changing the courses 
 of sailing-vessels, (2) by observations on the course of objects 
 floating in the water, (3) by their effects on temperature, and in 
 various other ways. 
 
 Most of our knowledge concerning the depth of the ocean, its 
 temperature, its life, the material and the topography of its bot- 
 tom, and much concerning its movements, has been gained 
 through expeditions which have been sent out from time to time 
 to study these especial problems. The expeditions which have 
 contributed to our knowledge of the ocean have been fitted out 
 by governments in some cases, by societies in others, and by in- 
 dividuals or combinations of individuals in still others. The 
 expedition which was carried out on the most elaborate scale was 
 that of the Challenger, 1872-6, under the auspices of the British 
 Government. This vessel made extended explorations in the 
 Atlantic, the Pacific, and the Southern oceans (Fig. 682). The 
 results of the observations made during the voyage of the Chal- 
 lenger, and inferences from them, have been published in a great 
 series of large volumes which give us our most detailed knowledge 
 of the sea. Numerous other lesser expeditions have made less 
 
710 
 
 PHYSIOGRAPHY 
 
 voluminous but still valuable contributions to our knowledge of 
 the ocean. Here may be mentioned the work of the U. S. S. 
 Mercury (Barbadoes to Sierra Leone, 1871), H. M. S. Lightning and 
 H. M. S. Porcupine (Faroe Islands to the Mediterranean, 1868-70), 
 the German frigate Gazelle (1874-6; , and the U. S. Coast and 
 Geodetic Survey steamer Blake (Gulf of Mexico, Caribbean Sea, 
 east coast of the United States, 1877-80), and the work of the 
 Coast and Geodetic Survey on the Gulf Stream (1845-59.) The 
 
 
 120- 140' IW IW IW IW 1M- IW 
 
 100 HO' 140 
 
 FIG. 682. The course of H.M.S. Challenger, shown by broken line on 
 
 oceanic areas. 
 
 expedition of Nansen into the Arctic region and the numerous 
 expeditions of the last years into the Antarctic regions have given 
 us much information concerning the waters of high latitudes. 
 Some indication of the manner in which these expeditions do some 
 of their work will appear in the following pages. 
 
 Distribution of the ocean waters. The distribution of the 
 ocean waters has been outlined in a general way in connection 
 with the distribution of the land (Chap. I.). The ocean encircles 
 the earth in latitude 60 S., and from this encircling sea three 
 great bodies of water, the Atlantic, the Pacific, and the Indian 
 oceans respectively, extend northward. South of latitude 60 S. 
 lies an elevated tract, Antarctica. It will be recalled that in the 
 northern hemisphere the land makes an almost complete circuit 
 in latitude 60 to 70, whence it radiates southward in three (or 
 
GENERAL CONCEPTIONS 711 
 
 two) great arms, and that north of the encircling land lies the Arctic 
 Ocean. The waters south of latitude 40 S. are often called the 
 Southern Ocean. The Arctic Ocean is about one-thirty-seventh of 
 the sea area, the Indian Ocean about one-eighth, the Southern Ocean 
 one-fourth, the Atlantic one-fifth, and the Pacific three-eighths. 
 
 The unequal distribution of land and water in the northern 
 and southern hemispheres has an important influence on their 
 climates, as already stated. 
 
 Depth. The average depth of the ocean is estimated to be 
 about 2J miles, or between 12,000 and 13,000 feet. The average 
 depth of the Pacific is estimated at 2f miles; that of the Atlantic 
 at 2^ miles; and that of the Indian and Southern oceans at 2J 
 miles. The average depth of the Arctic Ocean is not known, but 
 Nansen found a depth of more than 12,000 feet off the continental 
 shelf of Eurasia. The greatest known depth of ocean water is 
 nearly six miles. This depth slightly exceeds the height of the 
 highest mountain above sea-level. There are numerous places 
 where the depth of the ocean exceeds four miles, and the area of 
 ^ery deep water is very much greater than the area of very high 
 land. The tracts which are notably below the average depth of 
 the ocean are often known as deeps. 
 
 The greatest known depth of water, 31,614 feet, is in the Pacific, 
 aear the Ladrone Islands. Another area of almost equal depth 
 (30,930 feet) occurs in the Aldrich Deep northeast of New Zealand. 
 A depth of nearly 28,000 feet is found in the Tuscarora deep east 
 of Japan, and a depth of about 25,000 feet (nearly five miles) off 
 the coast of Chile, in latitude 24 to 25 S. 
 
 None of these great depths is in the midst of the Pacific. Some 
 of them are close to continental shores, and the others are in regions 
 of abundant islands, and in surroundings where the water is not 
 very deep. Most of them are in the western part of the ocean. 
 In all cases the slopes down to these great depths are steep, as 
 submarine slopes go, and the deeps have a pronounced tendency to 
 elongation parallel to the nearest coasts or to adjacent submarine 
 ridges, or to ridges the crests of which rise into islands. 
 
 The only area in the Atlantic where comparable depths are 
 known is north of Porto Rico in the Blake Deep (lat. 20 N., long. 
 65 to 68), where a maximum depth of 27,366 feet has been 
 sounded. This deep, too, is elongate, has steep slopes, and is 
 parallel to the great ridge of which Porto Rico is a part, and near 
 
712 
 
 PHYSIOGRAPHY 
 
 which it lies. In few other places in the Atlantic does the depth 
 reach 20,000 feet. 
 
 The Indian Ocean is not known to have depths much exceed- 
 ing 20,000 feet, and the deepest known place in the Southern Ocean 
 is still less. 
 
 The depth of the ocean becomes known by soundings. Sound- 
 ings are made from ships, by reeling out a heavy metallic ball 
 attached to a fine steel wire. (Why not a rope?) The ball is so 
 fastened to the line as to be detached when it reaches the bottom 
 (Fig. 683), because it is much simpler to leave it 
 than to draw it up again. A sounding of 3000 
 fathoms may be made in about an hour. 
 
 There is a more or less wide-spread notion that 
 the deeper waters of the sea are so dense that 
 weights will not sink readily, and that deep-sea 
 sounding is attended with difficulty on this ac- 
 count. This is incorrect, for water is but slightly 
 compressible, and the water in the deepest part 
 of the ocean is but little heavier (probably not a 
 twentieth), volume for volume, than that at the 
 surface. There are difficulties connected with deep 
 soundings, but their cause is not the great density 
 of the deep water. 
 
 Volume. The average depth and the area of 
 the oceans being known, the volume of water 
 which they contain may be calculated. It is found 
 to be nearly fifteen times the volume of land. If 
 all the material of the land were carried to the 
 sea and deposited in its basin, it would raise the 
 level of the water about 650 feet. If the surface 
 of the lithosphere were brought to a common level 
 by planing down all continental platforms and 
 building up the deep parts of the ocean basins, 
 the ocean water would cover the whole of the earth to a depth of 
 about 9000 feet, or nearly two miles. 
 
 Mass. The mass (weight) of the sea is only about five times the 
 mass of the land above the sea, since water is much lighter than an 
 equal volume of rock. The mass of the sea is about 265 times 
 the mass of the air which surrounds it, and about j^g- of the mass 
 of the solid part of the earth. 
 
 FIG. 683. The 
 sounding-line. 
 (Challen ger 
 Report.) 
 
GENERAL CONCEPTIONS 
 
 713 
 
 Topography of the bottom. The larger part of the sea bottom 
 is so nearly flat that if the water were removed the eye would 
 scarcely detect its departure from flatness. Its topography is 
 therefore very different from that of the land. As already indi- 
 cated, the agent which does most to roughen the surface of the 
 land is running water, and rivers do not flow on the bottom of 
 the sea. The most notable difference between the topography 
 of the sea bottom and that of the land is due to their absence. 
 
 In spite of the prevailing flatness of the sea bottom, its relief 
 is not less than that of the land. Its irregularities of bottom are 
 of several types. These are (1) vokanic cones, often built up from 
 the bottom of the deep sea to the surface of the water, and even 
 far above it (Chap. VII); (2) relatively steep slopes or scarps, such 
 as those at the junction of the continental platforms with the deep 
 sea basins, and such as occur about some of the pronounced deeps; 
 (3) valley-like depressions, found especially in the shallow waters 
 about the borders of the continents; (4) pronounced swells which 
 may be compared to the mountain ridges of the land; and (5) 
 broad, plateau-like areas rising notably above their surroundings, 
 over which the water is relatively shallow. The general con- 
 figuration of the bottom of the Atlantic is indicated by Fig. 5. 
 
 1. Volcanic cones are wide-spread, but are more numerous in the 
 Pacific Ocean than elsewhere, and more numerous in its deeper 
 western part than in its shal- 
 lower eastern part. Though 
 
 such cones seem to rise ab- 
 ruptly, their slopes are really 
 much less steep than they seem. 
 The summits of volcanic islands 
 rarely have a slope of more 
 than 30, and their lower parts 
 rarely more than 6 to 10. 
 Below the sea the slope is still 
 gentler, rarely more than 3, or 
 1 mile in 20. Figs. 684, 685, 
 and 686 show slopes corre- 
 sponding to 1 mile in 5, 1 in 10, 
 and 1 in 20. 
 
 2. Though the slopes of the bottom at the edges of the con- 
 tinental shelves and about the deeps are steep, as slopes in the 
 
 FIG. 684. Diagram illustrating a 
 slope corresponding to 1:5. 
 
 FIG. 685. Diagram illustrating a 
 slope corresponding to 1 : 10. 
 
 FIG. 686. Diagram illustrating a 
 slope corresponding to 1 : 20. 
 
714 PHYSIOGRAPHY 
 
 ocean bottom go, they are much less steep than many slopes on 
 the land. A slope of 1 mile in 8 is rare, and a slope of 1 mile in 
 20 (Fig. 686) can hardly be said to be common. The last would 
 make a steep railway grade. Slopes of 1 in 60 are higher than the 
 average steep slope at the edge of the continental shelves. Even 
 up most of thece "steep" slopes, therefore, railway trains could be 
 run without change of grade. 
 
 In rare instances, slopes which would be regarded as very steep, 
 even on land, are found on the sea bottom. " Thus in the Mediter- 
 ranean Sea, differences of 1500 feet are said to have been found 
 between the bow and stern soundings. Such slopes or scarps 
 are doubtless the result of faulting (p. 406). 
 
 3. Many continental shelves are affected by valleys which have 
 the general characteristics of river valleys. Many of them seem 
 to be continuations of valleys now in existence on the land. Thus, 
 the Hudson, the Delaware, the Susquehanna, the St. Lawrence, 
 the Saguenay, and other valleys have submerged continuations 
 beneath the sea. The valley of the Hudson is continuous out to 
 the edge of the continental shelf, where for 20 miles it becomes 
 pronounced, with a maximum depth of 2400 feet below its sur- 
 roundings, and 2844 feet below sea-level. Elsewhere it is shallow. 
 The others are not so deep. The submerged continuations of the 
 Delaware and the Susquehanna on the continental shelf are less 
 than 100 feet below their surroundings, but those of the Saguenay 
 and St. Lawrence, both of which extend out to the edge of the 
 continental shelf, are much deeper. 
 
 Other submerged valleys, as some of those on the Pacific coast 
 of the United States, do not seem to be the continuations of exist- 
 ing land valleys. Some of these valleys are hundreds of miles 
 long and a thousand feet or more (maximum) deep. Such val- 
 leys are commonly believed to have been formed by rivers when 
 the sea did not cover the areas where they exist (p. 173). 
 
 4. Examples of mountain-like swells are furnished by Cuba and 
 the adjacent islands, which are really the crests of a great moun- 
 tain system rising from deep water. 
 
 5. The plateau type of elevation is exemplified by the dolphin 
 ridge of the Atlantic (Fig. 5). This broad, low "ridge," over 
 which the water is less than 12,000 and sometimes as little as 
 5000 feet deep, traverses the Atlantic lengthwise, as far south as 
 latitude 40 S., and divides its basin into two troughs, the one 
 
GENERAL CONCEPTIONS 715 
 
 to the east and the other to the west, where the water is some- 
 what deeper. In the southern Pacific, volcanic islands often rise 
 from submerged plateaus. 
 
 From the foregoing it will be seen that great irregularities are 
 found on the sea bottom as on the land, but that the many minor 
 unevennesses of the land, especially those developed by running 
 water, wind, glaciers, etc., find no analogies on the ocean's bed, 
 except in shallow water. 
 
CHAPTER XXI 
 COMPOSITION OF SEA-WATER 
 
 THE most distinctive characteristic of sea-water is its saltness ; 
 but besides common salt, it contains dissolved mineral matters of 
 many other sorts. One hundred pounds of sea-water contain 
 nearly 3J (3.44) pounds of mineral matter. Of this mineral mat- 
 ter, common salt makes up more than three-fourths (about 77.758%). 
 The other important minerals are magnesium chloride (10.878%), 
 magnesium sulphate (4.737%), calcium sulphate (3.600%), potas- 
 sium sulphate (2.465%), and calcium carbonate (.345%). Very 
 many others occur in very small quantities. These mineral matters 
 in the sea-water make it somewhat heavier than fresh water. 
 If the weight of fresh water be taken as 1, the average weight of 
 salt water is 1.026. 
 
 A cubic mile of fresh water weighs about 4,205,650,000 tons of 
 2240 pounds each, while a cubic mile of normally salt water weighs 
 4,314,996,900 tons The mineral matter in a cubic mile of sea-water 
 weighs about 151,025,000 tons. This, it will be seen, exceeds the 
 difference between the weight of a cubic mile of fresh water and a 
 cubic mile of sea-water. It follows, therefore, that a cubic mile 
 of sea-water does not weigh the same as a cubic mile of fresh water 
 plus the weight of the salts in the former. The reason is that 
 when mineral matter is dissolved, it increases the volume of the 
 water, bat not by an amount equal to the volume of the mineral 
 matter dissolved. If all the salts were taken out of the sea-water 
 and removed from the ocean basins, the level of the sea would be 
 drawn down more than 100 feet. If all the salts of the sea were 
 taken out of solution and laid down as a layer of solid mineral 
 matter on the ocean bottom, they would make a layer about 175 
 feet thick, and would raise the surface of the water (then with- 
 out the salt) about 75 feet. If all the mineral matter now in solu- 
 
 716 
 
COMPOSITION OF SEA-WATER 717 
 
 tion in the sea were taken out of it, its aggregate volume would 
 be equal to about one-fifth of the volume of all lands now above 
 the sea-level. 
 
 The mineral matter in solution. ^Mineral matter is being 
 constantly brought to the sea by rivers. The rivers are largely 
 fed by springs, and the spring water, while underground, dissolves 
 various sorts of mineral matter from the rocks, as we have seen 
 (p. 96). The rivers are probably the chief source of the mineral 
 matter in the sea, but the sea-water also dissolves mineral matter 
 from the rocks beneath it. The amount of mineral matter carried 
 in solution to the sea by rivers each year is estimated to be nearly 
 half a cubic mile. 
 
 The rivers do not carry mineral matters to the sea in the pro- 
 portions in which they exist in the sea-water. Of the minerals 
 dissolved in river water, calcium carbonate is by far the most 
 important, being nearly as abundant as all the rest, while com- 
 mon salt is one of the minor constituents, so small in amount as 
 not to be detected by the taste. Yet the amount of the latter in 
 the sea-water is more than 200 times that of the former. This 
 great difference is one of the things to be explained. 
 
 It is to be noted that the mineral matters which are most 
 abundant in the sea are not those which are most abundant in 
 the rocks of the land. Those minerals of land rocks which are 
 most soluble, such as calcium carbonate, get into river water, and 
 thence to the sea, in greater quantity than those which are less 
 soluble. Many minerals in the sea-water do not exist as such in 
 the common rocks of the land, but are made by the combination 
 of matter in the rocks with gases (especially CO 2 ) in the air. Thus 
 many volcanic rocks contain calcium in complex combinations. 
 These complex compounds are broken up, and the calcium unites 
 with the carbonic-acid gas of the air to form calcium carbonate. 
 This is one of the prolific sources of this material carried by 
 rivers to the sea. Again, common rocks do not contain salt, but 
 some of them, such as granite, contain sodium, one of the ele- 
 ments of salt. When the sodium unites with chlorine, the result 
 is salt. It takes much rock to yield a little salt. The great quan- 
 tities of salt in the sea, therefore, must mean the decay of much 
 greater quantities of rock. 
 
 Some mineral substances in the sea, on the other hand, are 
 derived directly by solution from the rock. This is true of much 
 
718 PHYSIOGRAPHY 
 
 of the lime carbonate, which is the dissolved substance of lime- 
 stone. 
 
 Withdrawal of mineral matter from the sea. Some of the 
 substances in solution in the sea-water are extracted from the 
 water by animals to make their shells, tests, etc. Most shells are 
 made of calcium carbonate, but animals appear to be able to use 
 the sulphate of calcium also in the making of their shells, trans- 
 forming it into calcium carbonate. In spite of the abundant supply 
 of calcium carbonate, therefore, the amount of this substance in 
 the sea is relatively small, because animals, and some sea-plants as 
 well, take it out to make shells and other hard parts about as fast 
 as it is brought in. Silica also, though found in sea-water in small 
 quantities only, is extracted by some animals and plants, as calcium 
 carbonate is by others. Salt, on the other hand, is not used by 
 any of the animals or plants of the sea, and so remains in solu- 
 tion, and most of all that has ever gone to the sea appears to b 
 there still. 
 
 A suggestion as to the age of the ocean. At the rate at 
 which salt is now being taken from the land to the sea by rivers, 
 it would take some 370,000,000 years for the salt of the sea to 
 have accumulated. It is by no means certain, however, that salt 
 has always been carried in at the present rate, and it is certain 
 that some of the salt which has been carried to the sea has been 
 taken out again to make the great salt beds which occur in various 
 parts of the earth. While, therefore, 370,000,000 years is not to 
 be taken as the age of the ocean, it may give us some hint of the 
 length of time during which it has been in existence. 
 
 Gases in sea-water. Besides the solids in solution in sea- 
 water, there are numerous gases. The most abundant are those 
 which exist in the air in abundance, namely, nitrogen, oxygen, and 
 carbonic-acid gas. The amounts of these gases in solution vary 
 from place to place and from time to time, but the averages of 
 many analyses show that, of the total amount of gases in sea-water, 
 nitrogen makes up about ^1\%, oxygen about 33J%, and car- 
 bonic-acid gas about 16J%. In the aggregate, the amount of oxygen 
 dissolved in the ocean water is rather more than 3 -J of the amount 
 hi the air; that of the nitrogen about TTBTT; while that of the car- 
 bonic-acid gas is about 18 times the amount in the air. 
 
 The gases in the sea-water are dissolved chiefly from the atmos- 
 phere, in proportions determined by the pressure of each gas, by 
 
COMPOSITION OF SEA-WATER 719 
 
 its solubility, and by the temperature of the water. Gases are more 
 soluble in cold water than in warm, and carbonic-acid gas is more 
 soluble than oxygen, and this in turn is more soluble than nitrogen. 
 Once dissolved at the surface, these gases are distributed through 
 the ocean water by the movements of the water and by diffusion. 
 Carbonic-acid gas is also furnished to the sea-water in abundance 
 by the animals which live in the sea, and it issues from submarine 
 volcanic vents. 
 
 The oxygen of the water is being constantly consumed by the 
 animals which live in the sea, and its supply is being as constantly 
 renewed by solution from the air. The amount in the sea-water 
 decreases with increasing depth, and its paucity at great depths 
 is probably one of the reasons why animal life is not more abun- 
 dant there. Though constantly diffused downward, diffusion is a 
 very slow process. The nitrogen of the water is little used, and 
 the same nitrogen probably stays in solution from year to year 
 and from age to age. The carbonic-acid gas of the sea is con- 
 sumed by some of the plants of the sea, and some of that exhaled 
 by the marine animals and volcanic vents escapes into the air. This 
 is one of the sources of the carbonic-acid gas of the air. 
 
 The gases dissolved in the water do not greatly affect its volume, 
 though they increase it slightly. 
 
 Salinity, density, and movement. The waters of different 
 parts of the earth contain slightly different amounts of salt and 
 other mineral matters. The variation comes about in different 
 ways: (1) Evaporation is more rapid at some points than at 
 others. Since the salts are left behind when sea-water evaporates, 
 the water becomes more saline where evaporation is great. The 
 greater the amount of mineral matter in solution, the greater the 
 density of the water. (2) Where rainfall is great, the water is 
 freshened and so made lighter. (3) Where rivers enter the sea, 
 they bring in fresh water, which, mingling with the salt water, 
 makes it lighter. 
 
 In all the above ways the salinity of the sea-water at the top 
 of the ocean is being continually altered. Every alteration in 
 salinity changes the density of the water, and unequal density 
 causes movement. When the surface water becomes more dense 
 than that beneath, it sinks, and the lighter water comes in over it 
 from all sides. When the surface water becomes less dense than 
 the surrounding water at the same level, the heavier water dis- 
 
720 PHYSIOGRAPHY 
 
 places the lighter, causing it to spread out on the surface, for the 
 same reason that oil spreads on water. Since variations in the 
 salinity of water are being constantly produced, motion due to 
 inequalities of density resulting from inequalities of salinity, is 
 also constant. Movements brought about in this way are partly 
 vertical and partly horizontal. They are usually so slow as to be 
 imperceptible, and may appropriately be called creep. 
 
 Density of Water under Certain Conditions 
 
 Pure water at 39.6 F 1 .00 
 
 " " 212 F. 95 
 
 Surface sea-water at 60 F 1 .024 to 1.03 
 
 Sea-water five miles down 1 06 
 
 Salinity and color. The water of the sea is generally blue or 
 green, but its color varies from point to point and from time to 
 time. It seems to be indicated by numerous observations that 
 the blue color of sea-water is intensified by increase of salinity. 
 The Gulf Stream is distinctly bluer than the less salty cold current 
 off Labrador, and inland seas, such as the Mediterranean, which 
 are more salty than the open ocean, are of deeper blue. The cold 
 and less salty waters of high latitudes are often distinctly green. 
 Many of the variations of color are due to the solid matter in sus- 
 pension in the water. Microscopic animals and plants, and the 
 sediment washed or blown out from the land or furnished by ex- 
 plosive volcanoes beneath the sea, all contribute to the observed 
 variations. 
 
CHAPTER XXn 
 THE TEMPERATURE OF THE SEA 
 
 THE temperature of the sea is to be considered both horizontally 
 and vertically. In other words, account must be taken of the 
 temperature both at the surface and beneath it. 
 
 Temperature of the surface. In general, the temperature of 
 the surface of the ocean water decreases from the equator toward 
 the poles, the same as the temperature of the land (Fig. 538). 
 It varies from about 80 F. in the equatorial regions, to about 28 F. 
 in the polar regions. When the temperature sinks below the latter 
 figure, the sea-water freezes, and the temperature of the surface 
 of the ice may sink as low as the temperature of the air above it; 
 but the temperature or the water immediately beneath the ice does 
 not sink much below 28 F. The decrease of temperature with 
 increase of latitude is by no means regular, as shown by the isother- 
 mal charts. In Figs. 539 and 540, for example, the isothermal 
 lines over the ocean are far from parallel with the parallels of lati- 
 tude. 
 
 The notable departures of the ocean isotherms from the parallels 
 of latitude are due to various causes. In the open ocean they are 
 due chiefly to currents in the ocean water. Some of these cur- 
 rents are of water which is flowing into waters warmer than them- 
 selves, and some are of water flowing into waters cooler than them- 
 selves. The former are known as cold currents, and the latter as 
 warm currents. A cold current deflects an isotherm equatorward, 
 and a warm current deflects it poleward. Fig. 539 furnishes a 
 good illustration of the effect of a warm current in the North 
 Atlantic on the position of the isotherms. 
 
 There are other reasons why the temperature of the surface 
 water of the ocean does not decrease steadily from equator to 
 poles Thus rivers entering the sea are sometimes (especially in 
 
 721 
 
722 PHYSIOGRAPHY 
 
 summer) warmer and sometimes (especially in winter) colder than 
 the sea-water where they enter. Rivers, therefore, help to pro- 
 duce irregularities of surface temperatures. Enclosed or partially 
 enclosed arms of the sea in low latitudes are generally warmer 
 than the open ocean in the same latitude, and in such situations 
 the highest temperatures of the sea are found. Thus the surface 
 temperature of the Red Sea is sometimes 90 or even 100. 
 
 Temperature and movement. Since water expands on being 
 warmed, warm water is lighter than cold, if both are equally salt. 
 Unequal surface temperatures therefore cause movement of the 
 surface waters. The tendency of the resulting movement is to 
 cause the colder waters of the higher latitudes to displace the 
 warmer waters of the same level in lower latitudes, while the 
 warmer waters of the latter zone spread widely over the sur- 
 face. The movement is, therefore, circulatory. The movements 
 due to this cause are always slow, but since the surface temperature 
 is constantly kept unequal by unequal heating, by the inflow of 
 rivers, and by melting ice, there must be constant movement of 
 the surface waters as a result of the constantly renewed inequality 
 of temperature. 
 
 The surface of the sea is subject to both seasonal and daily 
 changes of temperature. Both are much less than the corre- 
 sponding changes on land in the same latitude (p. 530). 
 
 Temperature beneath the surface. Except where the surface 
 waters of the sea are at or near the freezing-point, the temperature 
 becomes cooler with increasing depth. Even where the surface 
 water is warmest, its temperature at the depth of a few hundred 
 fathoms (rarely more than 800, and generally much less) is below 
 40 F., and that at the bottom still colder. The following table 
 shows the average temperature of the sea at various depths : 
 
 Depth. Temperature. 
 
 600 feet 60.7 
 
 1,200 " 50.0 
 
 3,000 " 40.1 
 
 6,000 " 36.5 
 
 13,200 " 35.2 
 
 It is estimated that not more than one-fifth of the water of the 
 ocean has a temperature as high as 40 F., while its average tem- 
 perature is probably below 39 F. At the bottom of the deep sea 
 
THE TEMPERATURE OF THE SEA 
 
 723 
 
 I4'C 
 
 \ 
 
 5 '_ 
 
 the temperature is generally below 35 F. The only parts of the 
 ocean-bottom where the temperature is as high as 
 40 F. are the areas of shallow water and the en- 
 closed seas of relatively low latitudes. Such areas 
 do not constitute more than 8% of the area of 
 the sea. Fig. 687 represents a temperature curve 
 for the South Atlantic, and is fairly typical for the 
 ocean in general. Fig. 688 shows a similar curve 
 for the North Atlantic, and Figs. 689 to 691 show 
 the temperature curves for other places. 
 ^v The tempera- 
 
 ture of the sea 
 does not every- 
 where decrease 
 IsooFms, steadily from the 
 FIG. 687. A temperature curve in the South Atlan- surface down, for 
 tic; latitude 35 59' S., longitude 1 34' E. (Chal- th more or 
 
 lenger Report.) 
 
 less well-defined 
 
 currents beneath 
 the surface, some- 
 times warmer and 
 sometimes colder 
 than their sur- 
 roundings. Such 
 currents introduce 
 
 lenger Report.) 
 
 D 
 
 isooFms 
 
 Fms. 
 
 irregularities into 
 the temperature 
 curves (Fig. 692). 
 The tempera- 
 
 FIG. 689. Temperature curve for the South At- f f} 
 
 lantic, where the water is affected by the Antarr- ' 
 tic current; latitude 42 32' S., longitude 56 27' 
 W. (Challenger Report.) 
 
 parts of enclosed 
 seas in low lati- 
 tudes present striking contrasts to the temperatures of the 
 
724 
 
 PHYSIOGRAPHY 
 
 deeper parts of the open sea. Thus the temperature of the Red 
 Sea decreases from 90 F. or more at the surface, to 70 F. at a 
 depth of 1200 feet, and then remains nearly constant to the bottom 
 at 3600 feet. In the Mediterranean, the temperature falls from 
 about 75 F. at the surface to 55 F. at a depth of 750 feet, and 
 then remains essentially constant to the bottom, 13,000 feet, while 
 the temperature of the ocean outside falls to 37 F. in its deeper 
 parts. The high temperature of the deep waters of these enclosed 
 seas is due to the submerged barriers which partially shut them 
 off from the ocean, and do not allow the colder and therefore denser 
 waters outside to flow in and displace the warmer and lighter 
 waters below the top of the basin (Fig. 693). The temperature 
 of the bottom of enclosed seas is, in general, approximately the 
 
 temperature of the adjacent open- 
 sea water at the level correspond- 
 ing to the top of the submerged 
 barrier. 
 
 The phenomena of these and 
 other similarly situated basins 
 show that the sun shining on an 
 enclosed body of water may, in 
 the course of time, warm it to the 
 
 K 
 
 50 C 
 
 25 li 
 
 5 . 
 
 25- 
 
 20*_ 
 
 K>*_ 
 
 5'. 
 
 \ 
 
 Fms. 
 
 G. 690. -Temperature curve for the equatorial 
 Atlantic; latitude 9' N., longitude 30 18' W. 
 (Challenger Report.) 
 
 bottom, even where 
 the water is deep, 
 "Soo Fms. in low latitudes at 
 
 FIG. 691. Temperature curve for the equatorial least The low tern- 
 Pacific: latitude 40' N., longitude 148 41' E. , 
 (Challenger Report.) perature of the body 
 
 of the ocean is not, 
 therefore, due to the inability of the sun to heat it. 
 
 The reasons for the low temperature of the great body of the 
 sea-water are readily understood. 1. The heat of the sun has 
 little direct effect below some such depth as 200 to 300 feet, and 
 none at all below 600 feet. Taken alone, this does not account 
 
THE TEMPERATURE OP THE SEA 
 
 725 
 
 for the low temperature of the body of the ocean, as the phenomena 
 of the enclosed basins show, but it is one of the elements of the prob- 
 
 1000 
 
 1500 
 
 Fio. 692. Temperatures in the South Atlantic between Falkland Islands 
 Rio de la Plata, Tristan d'Acunha Islands, and Cape of Good Hope. 
 (Challenger Report.) 
 
 lem. The sun has been shining long enough to have warmed the 
 ocean to its bottom, even by the slow (in water) process of con- 
 duction. 2. The bottom of the sea, though warmed by conduc- 
 tion from the lithosphere below, is warmed with extreme slowness. 
 
 Indian Ocean 
 
 Bed Sea 
 
 FIG. 693. -^Diagrammatic section of Red Sea and the adjacent part of the 
 Indian Ocean, to illustrate the effect of a barrier on the temperature of 
 the waters. The temperature is expressed in degrees Fahrenheit. The 
 numbers at the left show depth in fathoms. 
 
 As its temperature rises, the expanded water is crowded up by the 
 colder, heavier water which gets beneath it. 3. The cold waters 
 of the surface, whether produced by contact with the cold air or 
 by the melting of ice and snow, tend constantly to sink. The supply 
 of ice-water from the polar regions, especially from the Antarctic 
 region, is very great, and though that which comes from the land 
 is fresh, and therefore lighter than sea-water at the outset, it soon 
 becomes salt by diffusion and by mixing with salt water. This 
 enormous supply of ice-water is the great cause of the low average 
 temperature of the sea. Without the polar ice-caps, the average 
 temperature of the ocean would in time be raised perceptibly. 
 
726 PHYSIOGRAPHY 
 
 If the ice-caps were melted, it would probably go far toward re- 
 storing the genial climates of earlier times in high latitudes, when 
 temperate and even subtropical plants and animals lived in Green- 
 land and in Antarctica. 
 
 The temperature below the surface is ascertained by a ther- 
 mometer constructed for this especial purpose. Its chief peculiar- 
 ity is such construction as will enable it (1) to withstand the 
 great pressure of deep water without having the position of the 
 mercury in the tube influenced seriously by it, and (2) to register 
 the temperature of any desired depth. Since the pressure in the 
 sea increases a ton per square inch for a little less than a mile of 
 descent, it will be seen that the ordinary thermometer used in the 
 atmosphere would not be serviceable. A satisfactory thermometer 
 was not devised until 1869, just before the departure of the Chal- 
 lenger expedition. 
 
 The ice of the sea has been referred to in other connections 
 (pp. 210 and 269). 
 
 The movement of floating ice is controlled partly by the winds, 
 and partly by the movements of the water in which the ice is 
 floating. 
 
CHAPTER XXIII 
 THE MOVEMENTS OF SEA-WATER 
 
 Causes of Movement 
 
 WE have seen that inequalities of density in sea-water arise 
 chiefly from (1) unequal salinity, and (2) unequal temperature, and 
 that these inequalities taken by themselves insure a constant, 
 though slow, circulation of the waters of the sea. There are other 
 causes, also, which produce movement. Chief among them are 
 (3) inequalities of level, (4) the wind, and (5) the differential at- 
 traction of heavenly bodies, especially the moon and the sun. 
 There are also (6) occasional causes, such as earthquakes, volcanic 
 explosions in the sea, landslides on coasts, etc., which produce 
 temporary and sometimes disastrous movements. The effect of 
 
 (1) and (2) have been noticed already (pp. 719 and 722). 
 Movements due to inequalities of level. The inequalities of 
 
 level which produce movement are brought about by (1) the in- 
 flow of land waters, which raise the surface at the point of inflow; 
 
 (2) winds, which tend to pile up the waters along the shores against 
 which they blow; (3) unequal rainfall, which tends to raise the 
 surface where it falls; (4) unequal evaporation, which tends to 
 lower the surface where it is excessive; and (5) variations in at- 
 mospheric pressure, the surface of the water being slightly de- 
 pressed where atmospheric pressure is excessive. 
 
 All such inequalities of level in the surface of the sea cause 
 movement. The movement begins as soon as the inequality of 
 surface appears, and before it becomes considerable. The result 
 is that the movements due to differences of level are generally 
 slow. So far as they are due to rainfall, to inequalities of evapo- 
 ration and to atmospheric pressure, they are generally impercep- 
 tible. The movement occasioned by*- the inflow of rivers is more 
 noticeable, and is often distinctly '.felt for some distance off 
 
 727 
 
728 PHYSIOGRAPHY 
 
 shore. When waters are piled up against a shore by winds, there 
 is sooner or later a return movement which tends to make the 
 surface level again. During a storm on the coast of India in 
 1864 (Oct. 5), the water was raised 24 feet at Calcutta, drowning 
 i8,000 people. The raising of the surface of the water was the 
 most destructive element in the storm at Galveston, already re- 
 ferred to (p. 654). 
 
 Since the several causes producing inequalities of level are in 
 constant operation, it follows that movements due to inequalities 
 of level are constant. 
 
 It will be recalled that there are other inequalities of level, 
 due to the attraction of land masses (p. 708). These inequalities 
 are in some sense permanent, and therefore do not produce cir- 
 culation of the sea-water. 
 
 Movements due to the wind. Winds not only produce tem- 
 porary inequalities of level, as indicated above, but they affect the 
 water in other and important ways. Their most familiar effect 
 is in the generation of waves, but they also drag the water along 
 beneath themselves. If a floating solid is dragged rapidly through 
 the water, the water about it moves, and that immediately beneath 
 the object moves faster than that farther below. If the object 
 were so light as to sink but little into the water, there would still 
 be movement of the water beneath because of the friction at the 
 plane of contact. The air, though not a solid, acts in the same 
 way. When it moves rapidly over water, it drags the surface 
 water along with it. 
 
 Since winds are always blowing, the movements to which they 
 give rise are always taking place. When winds have a more or 
 less constant direction, as in the zone of trades, there must be a 
 somewhat constant movement of the surface water in the same 
 direction. A constant movement in one direction necessarily 
 involves a return movement, that is a circulation, of the sea- 
 water. 
 
 Movements due to differential attraction of sun and moon. 
 Another cause of movement in sea-water is found in the attrac- 
 tion of bodies outside the earth, especially the sun and moon. 
 Bodies attract each other in proportion to their masses, and in- 
 versely as the squares of their distances; that is, a body t\vice 
 as massive as another has twice the attractive force at the same 
 distance, and if one of two bodies of a given mass be twice as far 
 
THE MOVEMENTS OF SEA-WATER 729 
 
 from a third body as the other is, their attractive forces on the 
 third are to each other as J: 1 ( 2 : 1). 
 
 The side of the earth towards the moon is nearer to the moon 
 than the center of the earth is, and so is attracted more strongly 
 than the center. The opposite side is attracted less strongly. 
 Because of these inequalities of attraction, the mobile waters on 
 the earth's surface are disturbed. The attraction of the sun pro- 
 duces similar but less pronounced effects. These inequalities of 
 attraction of moon and sun on different parts of the earth are the 
 cause of the movements known as tides. 
 
 Movements due to occasional causes. The aperiodic and 
 more or less accidental causes of movement which belong in this 
 class often produce violent wave movements which last but a 
 short time. Illustrations of their nature and effects have already 
 been given in connection with earthquakes. Landslides along the 
 coast, submarine volcanic eruptions, etc., also occasion violent but 
 temporary movements of the water of the sea. 
 
 Types of Movement 
 
 The general types of movement resulting from these various 
 causes are (1) waves, with their accompanying undertow, and shore 
 or littoral currents, (2) ocean currents, (3) drift (slow ill-defined 
 currents), (4) tides, and (5) creep. The first two and the fourth 
 are more obvious than the third and fifth, and the importance of 
 the less obvious is often overlooked. 
 
 Waves 
 
 The nature and the work of waves has already been outlined 
 (p. 318). Since the water in waves does not commonly move 
 forward, waves do not involve a general circulation of the ocean 
 water. In addition to what has been said concerning the work of 
 waves, it may be added that in the aggregate the sea destroys 
 more land by erosion than it makes by deposition, so that if noth- 
 ing else interposed, the sea, by its continued gnawing at the shores, 
 would ultimately destroy all land. It has already been pointed 
 out that wave erosion tends to do away with large irregularities of 
 coasts, though not with small ones (p. 320). In the long run, too, 
 deposition along coasts tends to develop regularity of outline, but 
 temporarily the coasts are often made very irregular (p. 328). 
 On the whole, the final effect of coastal activities is to make the 
 
730 PHYSIOGRAPHY 
 
 coasts more regular if they have any considerable irregularity at 
 the outset. 
 
 Currents 
 
 Experience and observation have shown that there are more 
 or less distinct currents in various parts of the ocean. This be- 
 came known first through the effect of the moving water on the 
 courses of sailing vessels, and it was later confirmed in various other 
 ways, as by following the course of floating bottles. 
 
 The better-known currents are at the surface, extending down 
 to depths of several hundred feet; but there are also currents 
 beneath the surface, as shown by variations of temperature and 
 by some other phenomena. Ocean currents are much less well 
 defined than currents of running water on the land, because the 
 former flow through a liquid, while the latter flow over a solid 
 bed between solid banks. These currents of the ocean are the 
 most distinct feature, though not the only one, of the oceanic 
 circulation, which has already been referred to. 
 
 Fig. 694 shows the general course of movement of the surface 
 waters of the seas. The figure represents a large part of the sur- 
 face water as involved in movement. There is a westward move- 
 ment of surface water in low latitudes in both the Atlantic and 
 the Pacific oceans. These are the equatorial currents or drifts, as 
 they are sometimes called. In each ocean the drift is double, and 
 a narrow counter-current moves eastward between them. The 
 equatorial current of the Atlantic is divided as South America is ap- 
 proached, one part being deflected to the southwest and the other 
 to the northwest. A part of the latter flows through the Caribbean 
 Sea and into the Gulf of Mexico. From this Gulf a distinct cur- 
 rent issues through the narrow passageway between Cuba and 
 Florida. This is the Gulf Stream. It is fed partly by the water 
 which enters the Gulf from the equatorial drift, and partly by the 
 large amount of water which enters the Gulf from the land, thus 
 tending to raise the level of the water. The stream issuing from 
 the Gulf has a velocity of more than four miles per hour where it 
 is swiftest. 
 
 Escaping from its narrow passageway between Florida and 
 Cuba, the Gulf Stream becomes wider and deeper. The current 
 tends to drag along the mobile water beneath and beside it, and as 
 more water becomes involved in the movement, the rate of progress 
 
THE MOVEMENTS OF SEA-WATER 
 
 731 
 
732 PHYSIOGRAPHY 
 
 becomes slower, and in the open ocean the rate of movement is 
 perhaps no more than 10 to 15 miles per day. As the current be- 
 comes slow, its boundaries become less well defined. In the open 
 ocean it is detected by its temperature, its color, its life, etc., more 
 readily than by its motion. 
 
 After leaving the Gulf, the Gulf Stream manifests a pronounced 
 tendency to turn to the eastward. Following this tendency, it 
 crosses the Atlantic, approaching the coast of Europe in a latitude 
 higher than that where it leaves America. Here it divides and 
 spreads widely. Long before this point is reached, it has ceased 
 to be a definite stream, and is to be looked upon rather as a general, 
 wide-spread drift of water. 
 
 That part of the equatorial drift which is turned southward 
 on the coast of South America first follows the coast of that conti- 
 nent, but soon shows a tendency to turn to the left (Fig. 694). 
 
 The equatorial drifts of the Pacific follow similar courses. The 
 part corresponding to the Gulf Stream of the Atlantic is known 
 as the Japan Current. The Indian Ocean has a south equatorial 
 drift only, and its course corresponds to that of the southern part 
 of the corresponding drifts of the other oceans. 
 
 All currents or drifts moving poleward from the equatorial 
 region consist of warm water moving into cooler water, and they 
 are known as warm currents. 
 
 The poleward movement of warm currents necessitates a re- 
 turn equatorward movement, and this movement is strengthened 
 by the inequalities of temperature in high and low latitudes. The 
 cold waters moving equatorward are deflected to the right in the 
 northern hemisphere and to the left in the southern, and the tend- 
 ency of this deflection is to concentrate them on the eastern coasts 
 of the continents in both hemispheres. 
 
 The equatorward currents start from latitudes where ice 
 abounds. They are cold, but not so salt (in summer) as normal 
 sea-water. By virtue of their temperature, therefore, they would 
 be denser than average sea-water ; but by virtue of their deficiency 
 of salt, they tend to be less dense than normal sea-water. As they 
 flow equatorward, they become warmer and more salt, and finally 
 attain such a degree of salinity that they sink and continue their 
 courses toward the equator as cold undercurrents. On the other 
 hand, the poleward (warm) currents start in low latitudes as sur- 
 face currents, kept at the surface by their high temperature in 
 
THE MOVEMENTS OF SEA-WATER 733 
 
 spite of their slight excess of salt. But in their poleward jour- 
 ney, they may sink beneath the cooler though fresher water, 
 and continue as warm undercurrents. Undercurrents of both 
 the types cited have been detected. 
 
 Cause of ocean currents. The equatorial drifts in the Atlantic 
 and Pacific oceans correspond somewhat closely, both in position 
 and direction, with the trade-winds. It is now generally believed 
 that winds which are constant in direction will cause a general 
 movement of the surface waters beneath them. It therefore seems 
 rational to infer that the equatorial currents or drifts are generated 
 by the trade-winds. The effect of the westward-moving equa- 
 torial currents is to bank up waters on the east coasts of the 
 continents, especially South America. Some of this water moves 
 eastward between the main west-bound currents, and constitutes 
 the narrow counter current of the equatorial calms. This current 
 of warm water is felt on the coast of Africa. 
 
 In extra-tropical latitudes the winds are less constant, and 
 therefore less effective in generating currents. But in regions of 
 strong monsoon winds, as about India, the drift of the surface 
 waters changes with the shifting winds, thus showing the com- 
 petency of winds to generate surface movements. 
 
 Were the ocean universal, the westward drift of the equatorial 
 waters under the influence of the trade-winds would doubtless 
 correspond with the trade-winds themselves; that is, they would 
 encircle the earth. But where the waters of this equatorial drift 
 reach a continent, as, for example, South America, they are de- 
 flected from their westerly course. 
 
 After the moving waters pass out of the control of the trade- 
 winds, they are directed (1) by the continental borders, (2) by the 
 configuration of the ocean bottom, (3) by the prevailing winds of the 
 latitudes which they reach, and (4) by the rotation of the earth. 
 Their courses are therefore determined partly by the causes which 
 generate them, and partly by other causes which direct them. 
 
 Another factor which is of importance in the development of 
 ocean currents is inequalities of temperature. This alone would 
 not give rise to distinct currents, but movements thus generated 
 (p. 722) may be concentrated and directed so as to emphasize the 
 currents generated by the winds. 
 
 Climatic effects of ocean currents. The air over a warm 
 ocean current is warmed by contact with the warm water. In 
 
734 PHYSIOGRAPHY 
 
 middle latitudes the prevailing westerly winds carry the warmed 
 air over to the coasts of the continents to leeward, giving them, in 
 winter, temperatures higher than they would otherwise have, and 
 giving them, at the same time, an abundant supply of moisture. 
 The winter temperature of the west coast of northern Europe is 
 much less severe than it would be but for the Gulf Stream. 
 
 The amount of heat which the Gulf Stream carries northward 
 from low latitudes has been estimated by Croll to be "one-fourth 
 of all the heat received from the sun by the North Atlantic, from 
 the tropic of Cancer up to the Arctic Circle." Its benefit, so far 
 as the land is concerned, is primarily to Europe. 
 
 The similar warm current in the North Pacific lessens the 
 severity of the winter climate of the northern part of western 
 North America. Similar results would be seen in the southern 
 hemisphere, were there land so situated as to feel the effects of the 
 corresponding currents in the southern oceans. 
 
 One other atmospheric effect of currents should perhaps be 
 mentioned. When the wind blows over a warm current, such as 
 the Gulf Stream, it is warmed and takes up a goodly supply of 
 moisture. On blowing from the current over colder water, its tem- 
 perature is lowered, and some of its moisture may be condensed. 
 The result is often a fog. Fogs are common along the leeward side 
 of the Gulf Stream, in latitudes where the adjacent land or water 
 is much cooler than the current itself. Fogs are more abundant 
 in the latitude of Newfoundland than farther south, because the 
 difference in the temperature of the Gulf Stream and its surround- 
 ings is here greater than farther south. Fogs also occur about 
 the Gulf Stream when there is no wind. This appears to be due 
 to the chilling of the warmer air by proximity to cooler air above 
 or on either side. 
 
 Fogs, often grading into mist or into clouds which yield rain, 
 are rather common in the northwestern parts of North America 
 and Europe, especially in the cold season. 
 
 Gradation al effects of ocean currents. Currents have rela- 
 tively little effect on the ocean bottom, and almost none on coasts, 
 because they rarely touch either. Where the water is shallow, how- 
 ever, as between Florida and Cuba, the Gulf Stream reaches and 
 scours its bottom effectively, somewhat as a great river might. 
 Since ocean currents do little eroding, except locally, they carry but 
 little debris. They do, however, transport considerable quantities 
 
THE MOVEMENTS OF SEA-WATER 735 
 
 of matter of organic origin. The waters, especially of warm cur- 
 rents, teem with minute organisms, and these organisms, or their 
 shells after the organisms are dead, are often carried far, and 
 finally scattered over the bottom of the ocean. 
 
 Historical suggestions. The currents of the Atlantic played 
 an important part in the early history of America. Once Iceland 
 was colonized by the Northmen, the currents southwest from the 
 Arctic insured the early discovery of North America. The south 
 equatorial current carried the Portuguese, bound for India, in 
 1500, to the shores of South America. 
 
 Tides 
 
 The level of the ocean water rises and falls twice every day, 
 or, more exactly, every 24 hours and 52 minutes. This periodic 
 rise and fall of the water constitutes the tides. The tide rises 
 (flood-tide) for about six hours, when it is high, and then falls (ebb- 
 tide) for about six hours, when it is low. 1 The tide often "comes 
 in" as a series of waves, the water after each wave failing to sink 
 to its former level. In other cases it rises quickly, without dis- 
 tinct waves. 
 
 Tides are not perceptible in the open ocean, for there is nothing 
 there to mark the slight rise of water; but they are readily seen 
 wherever there is an island on the shores of which the rise and 
 fall may be measured. The rise in the open sea has been esti- 
 mated to be two to three feet. Along coasts, the variation in 
 the water level between high and low tides is generally several feet. 
 In bays which open broadly to the sea but narrow toward their 
 heads, the range is sometimes 20 or 30 feet, or in rare cases 50 feet 
 or more, as in the Bay of Fundy. Where the tide runs in among 
 islands or passes through narrow straits, it often gives rise to dis- 
 tinct currents which scour the channels through which they flow. 
 
 The tide sometimes runs up a broad open river. As it ad- 
 vances up the channel, its progress is retarded by the shallowness 
 of the water, and its front may become a steep and often wall- 
 like wave. Such a wave is called a bore. The bore is felt in the 
 Severn and the Wye of England, in the Seine of France, in the 
 Petit-Codiac of Canada, in the Hugli of India, and the Tsien-Tang- 
 Kiang of China. In the last-named river the waves are some- 
 
 1 The ebb-tide is usually somewhat longer than the flood-tide. 
 
736 PHYSIOGRAPHY 
 
 times 25 feet high, and are disastrous to navigation. On one 
 occasion, Captain Moore estimated that 1J million tons of water 
 went by a point in the river in a minute in the bore wave. Trading 
 ships at Calcutta formerly hastened to the middle of the stream 
 for safety on the approach of the bore. 
 
 Bores do not appear, even in the rivers subject to them, with 
 every high tide. Favoring winds seem to be an important factor 
 in their development, and they are stronger in spring tides (p. 744) 
 than at other times. 
 
 High tides make themselves felt, though not as bores, up the 
 Hudson River to Troy, where the range of the tide is more than 
 two feet, and up the Delaware nearly to Trenton, though the salt 
 water does not run up so far. The sea-tide raises the sea-level at 
 the debouchures of these streams, and so dams back their waters. 
 The tide runs 70 miles up the St. Johns River in New Brunswick, 
 and is felt where the elevation of the river is 14 feet above mean 
 sea-level. The tide runs up the estuary of the St. Lawrence 283 
 miles to Three Rivers, near Montreal. 
 
 Tides are imperceptible in small lakes and feeble in large lakes 
 and enclosed seas. In Lake Michigan, for example, there is a tide 
 of about two inches. Tides are feeble in all bodies of water con- 
 nected with the open sea by a narrow passageway. Thus, at 
 Galveston in the Gulf of Mexico, the range of the tide is less than 
 one foot. 
 
 In many harbors, especially where the water is shallow, the rise 
 and fall are enough to have an important effect on navigation. 
 Vessels arriving at such harbors at low tide are often obliged to 
 wait until high tide before entering. Tidal currents or races are 
 sometimes so strong as to interfere with navigation. The race 
 through Hell Gate near New York City is a case in point. 
 
 The periodicity and the cause of tides. The time between 
 successive high or successive low tides is about half the time of the 
 apparent revolution of the moon around the earth. It appears to 
 have been this fact which suggested a connection between the tides 
 and the apparent motions of the moon, a connection which was 
 known, or at any rate suspected, some 2000 years ago, though 
 not fully understood until the time of Newton, about 200 years ago. 
 
 The law of attraction between heavenly bodies has already 
 been stated (p. 729). Without attempting to give a detailed 
 explanation of the tides, the essential principles involved may be 
 
THE MOVEMENTS OF SEA-WATER 737 
 
 readily understood. 1 We may consider first the tide produced 
 by the moon. 
 
 If a weight be attached to a string and whirled, the string is 
 put under tension. The weight constantly tends to move forward 
 in a straight line, but it is prevented from doing so by the string. 
 The tendency of the weight to depart from the circle in which the 
 string holds it, is often called centrifugal force, though it is only 
 inertia. The pull of the string which holds the weight is a cen- 
 tripetal force. The taut string therefore is affected by two opposite 
 and equal forces. 
 
 The motion of the moon about the earth is not unlike the 
 motion of the weight at the end of the string in the above illustra- 
 tion. In place of the string there is the attraction of gravitation, 
 and the moon goes about the earth at such a rate that her centrif- 
 ugal tendency is just balanced by the attraction of the earth. The 
 center about which the moon revolves is not, however, the center 
 of the earth but their common center of gravity. Since the earth 
 is about 80 times as massive as the moon, the center of gravity of the 
 two bodies is much nearer the center of the earth than the center 
 of the moon. It is, in fact, 1000 miles below the surface of the 
 earth, and 3000 miles from its center (Fig. 695). Both the moon 
 and the earth revolve about this common center, as they travel 
 together about the sun. The earth's center describes a circle 
 with a radius of 3000 miles about the common center of gravity, 
 while the moon describes a circle with a radius of about 237,000 
 miles about the same point. The concepticn may be made more 
 definite by conceiving two very unequal weights at the opposite 
 ends of a stiff but extremely light rod. These weights may be 
 such that the two will be balanced, if the point corresponding to 
 g, Fig. 695, is supported. If now the couple (E and M) be ro- 
 tated, the center of E will rotate about g in a small circle, while 
 the center of M will rotate about it in a much larger circle, each 
 in about 28 days. 
 
 The earth and the moon attract each other and would fall 
 together, but for the centrifugal tendency developed by the revo- 
 lution. The distance of the two bodies from each other is de- 
 termined by the balance between (1) their mutual attractions 
 
 1 For further accounts of the tides, see the astronomies referred to on 
 p. 505. 
 
738 PHYSIOGRAPHY 
 
 on the one hand, and (2) their centrifugal tendencies on the other. 
 This balance is perfect at the center of the earth and at the center 
 of the moon. But on the side of the earth nearest the moon the 
 attraction is somewhat stronger than at the center of the earth, 
 and overbalances the centrifugal tendency. The attraction there- 
 fore tends to make the earth bulge up under the moon. On the oppo- 
 site side of the earth the attraction is weaker than at the center, 
 
 FIG. 695. Diagram showing the position of the center of gravity, g, of the 
 earth-moon system. 
 
 and is overbalanced by the centrifugal tendency. Here, also, there 
 is therefore a tendency for the earth to bulge out, as a result of the 
 differential attraction of the moon. If the mass of the earth were 
 fluid, these bulgings or tides would be sensible. But the solid part 
 of the earth is essentially rigid, and there is not time for its parts 
 to yield sensibly to the strain set up by the differential attraction 
 of the moon, before rotation carries them forward where there is 
 no tendency to bulging. The surface waters, however, are mobile 
 and respond to the distorting effect of the moon's differential at- 
 traction, with the result that the water is bulged up on opposite 
 sides of the earth, so as to produce a slight elongation of the earth's 
 diameter in the direction of the moon. These bulges of water are 
 the high tides, and between them the tides are low. 
 
 The fact of the differential attraction of the moon may be 
 stated in other terms. The distance of the center of the moon 
 from the center of the earth is about 240,000 miles. The side of 
 the earth nearest to the moon is therefore about 236,000 miles 
 from the center of the moon, while the side farthest away is about 
 244,000 miles distant. 
 
 If the mass of the moon be taken as 1, the average pull of the 
 
 moon on the earth is represented by the fraction The 
 
 ^ 
 
 fraction which represents the moon's pull on the side of the earth 
 
THE MOVEMENTS OF SEA-WATER 739 
 
 nearest the moon is 2g 2 , and the fraction which represents 
 
 the pull on the opposite side is The solid part of the 
 
 earth acts essentially as a unit, since its parts are not free to move 
 on one another. The effect of the attractive force of the moon 
 on the solid part of the earth is therefore essentially the same as 
 it would be if it were all exerted on its center, that is, the same as 
 
 the average pull of the moon on the earth, 2 . 
 
 ^40000 
 
 Since the waters on the surface are readily mobile, they re- 
 spond to the differential attraction of the moon. The waters on 
 the side nearest the moon, being pulled with a force stronger than 
 the average pull on the solid part of the earth, are bulged up a 
 little. The force of the moon's pull here is represented by the 
 
 fracti n 236> and 236^-240^ ^presents the moon's 
 tide-producing force on the side of the earth nearest to it. The 
 waters on the opposite side of the earth are farther from the moon 
 than the center of the earth is, and so are pulled less strongly 
 than the latter, and are allowed to bulge out. Mathematically 
 the force on the side of the earth most distant from the moon is 
 
 ., and the tide-producing force there is 
 
 The result is a rise of water on opposite sides of the earth at the 
 same time. These are the high tides. Midway between the places 
 where the tides are high the water is correspondingly lowered and 
 the tides are low. 
 
 It will be seen from the above figures that the tide-producing 
 force on the side of the earth away from the moon is slightly less 
 than that on the side nearest the moon: 
 
 i JL <_!_ J_ 
 
 2400002 244000 2 ^236000 2 240000 2 ' 
 
 The explanation of the tides is sometimes so troublesome 
 that another statement of their cause is added: 
 
 "Let E (Fig. 696) represent the center of the earth, and M the 
 moon. (The distance of the moon is greatly minimized.) Con- 
 sider the tendency of the moon to displace the particle P on the 
 surface of the earth. Let EB represent the acceleration of A/' on 
 
740 PHYSIOGRAPHY 
 
 E (the solid earth) in direction and amount. In the same units 
 let PA represent the acceleration of M on P in direction and amount. 
 Since P and M are nearer together than E and M, it follows that 
 PA is greater than EB. 
 
 Let the acceleration PA be resolved into two components so 
 that one of them shall be equal and parallel to EB. It is PK in 
 the figure. The other component is found by using PA as a di- 
 agonal and PK as a side, and completing the parallelogram. It is 
 PQ in the figure. By the law of the parallelogram of forces PA 
 is exactly equivalent to PK and PQ, and conversely. By the 
 preliminary theorem, EB and PK being parallel and equal do not 
 
 FlG. 696. Diagram to illustrate the cause of tides. Explanation in text. 
 (From Moulton's Introduction to Astronomy. By permission of The 
 Macmillan Company.) 
 
 tend to change the relative positions of E and P, and therefore 
 cause no tide. The remaining acceleration PQ cannot be paired 
 with any other, and is the tide-raising acceleration. 
 
 The part of the figure with accents is drawn from precisely the 
 same principles. P'K' is parallel and equal to EB, and P'Q' is 
 the tide-raising acceleration. 
 
 Suppose figures are constructed for points all the way around 
 the earth. The lines representing the tide-raising accelerations 
 will be as given in Fig. 697. The method of drawing them is the 
 
 FIG. 697. Diagram to illustrate tides. (From Moulton's Introduction to 
 Astronomy. By permission of The Macmillan Company.) 
 
 geometrical counterpart of the rigorous mathematical treatment 
 of the subject, and may be relied upon as giving the full explana- 
 tion of the reason for the tides." * 
 
 1 Moulton's Introduction to Astronomy. 
 
THE MOVEMENTS OP SEA-WATER 741 
 
 Tides if the ocean were universal. If the earth were com- 
 pletely covered with a deep ocean, its surface would have two 
 extensive tidal bulges or waves at the same time. The highest 
 part of each would be a point, the one directly under the 'moon, 
 and the other directly opposite it. Each wave would be hemi- 
 spherical, and their borders would meet in a great circle, where the 
 tide would be low. This circle may be conceived of as the trough 
 of the tidal wave. 
 
 The period of the earth's rotation is shorter than that of the 
 revolution of the moon about the earth. The result is that rota- 
 tion tends to carry the high tides on beyond the position which 
 the moon would give them. The moon tends to hold them back, 
 and so they seem to travel about the surface of the earth in a 
 direction opposed to its rotation. The tides are therefore said 
 to lag. 
 
 Theoretically, successive high tides are 180 (12 hours) apart, 
 and rotation of the earth alone considered, high tides at any 
 place should recur every 12 hours. The longer period (12 hrs. 
 26 min.) is the result of the forward movement of the moon in its 
 orbit about the earth (Fig. 698). 
 
 There are two points, the tidal poles, where the tide does not 
 rise and fall. When the moon is vertical at the equator, it will be 
 seen that the highest point of the high tide should be on the equator 
 continuously, and that the great circle marking the position of 
 the low tide will pass through the geographic poles. The poles 
 will therefore have low tide continuously so long as the moon is 
 vertical at the equator. Whatever the latitude where the moon is 
 vertical, there will be a point, the tidal pole, 90 from the latitude 
 where the moon is vertical, where there would be no rise and fall 
 of the tide. Since the latitude where the moon is vertical varies 
 from time to time, the position of the tidal poles varies. 
 
 The simplicity of the tidal movements outlined above is inter- 
 fered with by many things, especially by (1) the continents, which 
 stop the tidal wave, and (2) the shallowness of water in many 
 places. The tidal wave travels more slowly in shallow water than 
 in deep, for the same reason that other waves do. Since tides 
 are retarded most in this way near continents and islands, their 
 advance is here most irregular. Irregular tidal waves often inter- 
 fere with one another. 
 
 Solar tides. The sun also attracts the earth and tends to 
 
742 
 
 PHYSIOGRAPHY 
 
 cause tides If there were no moon we should still have tides 
 produced by the sun. 
 
 FIG. 698. Diagram illustrating the motion of the moon about the earth 
 The larger circles represent the earth, and the smaller the moon on the 
 line which represents its orbit. 
 
 In spite of its great distance from the earth (about 93,000,000 
 miles), the sun, because of its great size, attracts the earth much 
 
THE MOVEMENTS OF SEA-WATER 
 
 743 
 
 more strongly than the moon does. If the moon's attraction 
 
 were the stronger, the earth would revolve about the moon instead 
 
 of the sun. But in spite of its greater attraction, the tide-producing 
 
 force of the sun is less than that of the moon. It is not difficult 
 
 to calculate their relative attractions. If the moon's mass be taken 
 
 as 1, the mass of the sun is 26,648,000. Mass 
 
 alone considered, the sun should attract the 
 
 earth 26,648,000 times as strongly as the 
 
 moon. The sun is about 389 times as far 
 
 from the earth as the moon is. Distance 
 
 alone considered, its pull should therefore be 
 
 1/389 2 (=1/151321) of that of the moon. 
 
 1/151321 X 26 ,648,000= 175 approximately. 
 
 That is, the sun pulls the earth with 175 
 
 times the force that the moon does. 
 
 It has been seen that the tide produced 
 
 by the moon is due to the difference between 
 the pull of the moon on the center of the earth 
 and on the parts nearest to and farthest from 
 it. Any tide which the sun produces must 
 also be due to the difference between its pull 
 on the center of the earth and on the sides 
 nearest and farthest from it. 
 
 The sides of the earth nearest to the sun 
 and farthest from it are 4000 miles nearer 
 to and farther from the sun than the center 
 of the earth is; but 4000 miles is a very 
 much smaller part of 93,000,000 miles than 
 it is of 240,000 miles. Hence the difference 
 between the attractive force of the sun on the center and on the 
 side of the earth nearest it is much less than the difference in the 
 attractive force of the moon on the same points. In other words, 
 the differential pull of the sun is less than the differential pull of the 
 moon. The moon's tides are therefore higher than the sun's. 
 Their ratio is 0.0342:0.0151. If the sun were as near the earth as 
 the moon is, its tidal effect would be millions of times greater than 
 now, and perhaps sufficient to disrupt the earth. 
 
 Some of the tides are the result of the combined influence of 
 the moon and the sun, but since the lunar tides are the stronger, 
 the solar tides serve merely to modify them. The solar influence 
 
 FIG. 699. Diagram 
 to illustrate the lag- 
 ging of the tides. 
 (After Comstock.) 
 
744 PHYSIOGRAPHY 
 
 strengthens the tides when sun and moon work together, and weak- 
 ens the tides when they work against each other. 
 
 Spring tides and neap tides. When the sun and the moon 
 stand in the relation to each other and to the earth shown in Fig. 
 700 (New moon), each tends to make high tides at the same points. 
 When the relations are those shown in Fig. 701 (Full moon), the 
 result is the same. At these times, and each occurs once a month, 
 
 
 FIG. 700. Diagram to illustrate the relative positions of earth, moon, and 
 sun at the time of full moon. Spring tide. 
 
 FIG. 701. Diagram to illustrate the relative positions of earth, moon, and 
 sun at the time of full moon. Spring tide. 
 
 the high tides are higher, and the low tides lower, than at other 
 times. The tides of such times are called Spring Tides. Spring 
 tides therefore have no relation to the spring season. 
 
 When the earth, moon, and sun sustain the relative positions 
 shown in Fig. 702, and this occurs twice each month, the tidal 
 influences of the sun and the moon are opposed, and the result is 
 that the high tides are not so high, or the low tides so low, as under 
 other conditions. The tides of such times are known as Neap 
 Tides. 
 
 Other variations in the height of high tides. There are sev- 
 eral other causes of variation in the height of high tides. Two of 
 these causes show themselves daily, tw ; o have monthly periods, 
 and one an annual period. 
 
 The two successive high tides of a given place are often of un- 
 equal height. One daily variation in the height of high tides is due 
 to the fact that the high tide on the side of the earth away from the 
 
THE MOVEMENTS OF SEA-WATER 
 
 745 
 
 moon is slightly lower than that on the side next the moon, for 
 
 1 _L 1 1 
 
 236000 2 240000 2 > 240000 2 ~ 2440002' The difference '*> however, 
 
 
 
 M 
 O 
 
 FIG. 702. Diagram to illustrate the relative positions of sun, moon, and 
 earth at the time of neap tides. 
 
 slight, and in the presence of larger variations is not commonly 
 noticed. 
 
 Again, if the high tide on one side of the earth is highest at A 
 (Fig. 703), the highest point in the high tide on the opposite side 
 would be at B, if the ocean were universal and of uniform depth. 
 From A on the one side, and from B on the other, the height of 
 
 .--M 
 
 FIG. 703. Diagram showing why successive high tides are often unequal. 
 
 the high tide diminishes in -all directions. The point A' has high 
 tide at the same time that A 'and B have, but the tide at A' is not 
 so high as that' at A. Twelve?' hours (and twenty-six minutes) 
 
746 
 
 PHYSIOGRAPHY 
 
 later, point A will have the same position relative to the moon that 
 A' now has, because of the rotation of the earth and the revolution 
 of the moon. The high tide which will occur at A when this point 
 shall have reached the position A' will not be so high as that which 
 it had when in the position A. Similarly the high tide which the 
 point A' will have when it reaches the position A, will be higher 
 than the preceding high tide at the same place. The amount of 
 daily variation due to this cause is often considerable. Locally at 
 least it is several feet. It is to be noted that it would not occur 
 when the moon is vertical at the equator, for then all points on the 
 same parallel stand in the same relation to the highest part of the 
 tidal wave. 
 
 Moon Farthest 
 North of Equator 
 
 Moon Over 
 the Equator 
 
 Moon Farthest 
 South of Equator 
 
 FIG. 704. Diurnal inequality of the tides at San Francisco. The space 
 between the vertical lines represents a day. The several crests of 
 the curves represent high tides and the troughs low tides. 
 
 The monthly variations in the height of the high tide are less 
 notable. One is due to the variation in the distance of the moon 
 from the earth. This distance decreases from its maximum for 
 about two weeks, and then increases from its minimum for about 
 the same length of time. The variation in the distance of the 
 moon from the earth makes a slight- difference in the height of 
 the tides, the high tides being highest and the low tides lowest 
 when the moon is nearest. Another monthly variation at any 
 given place, is due to the fact that the moon is vertical in differ- 
 ent latitudes at different times. In this particular, its monthly 
 range is comparable to the annual range of the sun. 
 
 The distance of the earth from the sun also varies during each 
 year, and this variation has its appropriate effect, small though 
 it is, on the height of the solar tides, and so on the height of the 
 observed tides. Other variations in the distance of the sun from 
 the earth occur in much longer periods of time, but they need not 
 
THE MOVEMENTS OF SEA-WATER 
 
 747 
 
 >o 
 g 
 
748 PHYSIOGRAPHY 
 
 be considered here. The variations for any given place, produced 
 by the apparent annual motions of the sun, are trivial. 
 
 The highest high tides in any given place should occur, theo- 
 retically, when the sun and the moon work together (spring tides), 
 at that time of day when the moon is most nearly in the zenith, 
 at that time of the month when the moon is nearest to the earth, 
 and at that time of year when the sun is nearest to the earth. 
 
 Cotidal lines. If a line were drawn on the ocean surface, 
 connecting all points which have the crest of the same high tide 
 at the same time, it would be a cotidal line. Any line connecting 
 points having the trough of the same tidal wave at the same time 
 would also be a cotidal line; or in general, any line connecting 
 points having the same phase of the same high tide at the same 
 time, is a cotidal line. If the ocean were universal and equally 
 deep, the cotidal lines would be the halves of great circles; but the 
 continents and the shallow waters, the islands and the straits, cause 
 many irregularities in them. The tide runs ahead, relatively, 
 where the water is deep and lags when it is shallow. 
 
 Rate of movement. Theoretically the tide should move for- 
 ward, from east to west, so as to complete a circuit in 24 hours 
 and 52 minutes. This would give it great velocity in low latitudes, 
 a velocity nearly equal to that of the rotation of the earth. This 
 velocity is nearly reached in the deep and open sea, but nowhere 
 else. 
 
 Effects of tides on shores. Since tides commonly rise in a 
 series of waves, they affect shores much as wind waves do. The 
 erosion effected by tidal currents among islands, and through 
 straits has been referred to. Tidal scour often keeps thorough- 
 fares open through tidal marshes, to which the tide has access 
 through bays. Illustrations are found on the coast of New Jersey. 
 Tidal scour also sometimes maintains deep waterways in bays to 
 the great advantage of navigation. 
 
 
CHAPTER XXIV 
 THE LIFE OF THE SEA 
 
 THE sea teems with plants and animals. The latter abound at 
 and near the surface nearly everywhere; they abound at the bot- 
 tom in shallow water, and they occur, though far less abundantly, 
 at the bottom of even the deep sea. In the great body of water 
 intermediate between the uppermost 100 fathoms and the bottom, 
 life is nearly absent. It has been estimated that the amount of life 
 in the sea exceeds that of the land, square mile for square mile; 
 but there is probably no one level in the sea where life is so abun- 
 dant as on the surface of the fertile parts of the land. Murray has 
 estimated that the weight of the lime carbonate of the shells of 
 organisms in the uppermost 100 fathoms of sea-water is some- 
 thing like 16 tons per square mile. This is far less than the weight 
 of the plant and animal life per square mile on land in fertile 
 regions. 
 
 The abundance of life in the sea-water may be shown in an- 
 other way. If a bucket of water be dipped up from the surface of 
 the ocean almost anywhere, it will be found to contain hundreds 
 or even thousands of minute organisms, though most of them are 
 too small to be visible to the unaided eye. 
 
 The distribution of the plant life of the sea differs somewhat 
 from that of the animal life. Plant life is plentiful at the surface 
 nearly everywhere, and at the bottom, down to the depth of about 
 50 fathoms. Where conditions are favorable, it occurs somewhat 
 sparingly down to depths of nearly 200 fathoms or so; but below 
 some such depth it is absent, probably because of the absence of 
 sunlight. Animal life is abundant where plant life is, and also to 
 considerably greater depths, beside being found to some extent 
 over the whole of the ocean's bed. 
 
 The most important physical factors which influence the dis- 
 
 749 
 
750 PHYSIOGRAPHY 
 
 tribution of the various types of sea life are (1) temperature, and 
 (2) depth of water. Other less important factors are (3) the 
 clearness, (4) the degree of saltness, (5) the quietness or roughness, 
 and (6) the presence or absence of ice. The relations of various 
 types of life to one another are also important. Some are depend- 
 ent on others for food, some are hostile to others, and some are 
 rivals for the same sorts of food. 
 
 The manner in which most of these factors influence the dis- 
 tribution of life will be readily understood from analogy with the 
 factors which control the distribution of land life. One factor, 
 however, which finds no analogy in connection with the distribu- 
 tion of land life, is the depth of the water. Land life is restricted 
 practically to the surface of the land, while sea life has a wide ver- 
 tical range. The depth of the water affects the distribution only 
 of those plants and animals which rest on the bottom; it has little 
 effect on the range of those which float or swim near the surface. 
 
 The most important influence of depth appears to be in con- 
 nection with the penetration of light and with the supply of 
 oxygen. Light is so rapidly absorbed by the water that vision 
 is virtually cut off at a depth of some 50 fathoms, though a little 
 light penetrates to somewhat greater depths. But in the great body 
 of the ocean darkness reigns. No form of plant life which depends 
 directly on sunlight can live in darkness. This includes all forms 
 of green plants, and some others. At the bottom, too, the water 
 is not stirred, and any oxygen it contains must pass down from 
 the surface after being dissolved there. As it is consumed below 
 by the animals, the supply is renewed by diffusion, an extremely 
 slow process. 
 
 Since the several factors which influence the distribution of 
 sea life vary widely, the distribution of various types of life also 
 varies widely. Some animals, such as coral polyps, are restricted 
 to warm regions where the water is shallow, clear, and normally 
 salt, while others, such as narwhales, seals, etc., are found only 
 in cold waters. Still others range through great differences of 
 temperature. 
 
 The plant life of the sea varies less with latitude than the plant 
 life of the land, and less than the animal life of the sea. 
 
 The life of the sea is in strong contrast in many ways with that 
 of the land. Thus most plants with which we are familiar on 
 land are fixed in position, while many of the plants of the sea 
 
THE LIFE OF THE SEA 751 
 
 float. Most animals on the land are free to move about, while a 
 very considerable proportion of sea animals, such as coral polyps, 
 barnacles, crinoids, etc., are fixed through most of their lives. 
 Many others, though not fixed, move about but little, either lying 
 on the bottom or burrowing into it. Some, on the other hand, 
 such as many of those in the surface waters (pelagic life), appear 
 to be always in motion. 
 
 The pressure of the water at the bottom of the ocean is very 
 great, but the animals living there withstand it, because their 
 tissues are full of liquids under equally high pressure, and these 
 high internal pressures counterbalance the external pressure. If 
 an animal from the bottom of the deep sea were brought suddenly 
 to the surface it would explode. This has indeed happened in 
 raising animals from the deep sea, even when the raising was by 
 no means instantaneous. 
 
 The deep-sea animals have some notable peculiarities. Some 
 are blind, but some have eyes, implying sight and therefore light. 
 It has been conjectured that the phosphorescence of the animals 
 themselves supplied the light. Some of the deep-sea animals also 
 are ornamented, a fact which seems to have no rational explanation, 
 unless the ornamentation is seen. 
 
 All the great groups of animal life are represented in the sea- 
 water. Even warm-blooded mammals (whales, narwhales, seals, 
 walruses, etc.) abound in the frigid waters among icebergs and 
 ice-floes. Some of these animals, like the seals and walruses, do 
 not spend all their time in .the water, but frequently crawl up on 
 the floes of ice to warm themselves and sleep in the sun. From 
 this highest class of animals (mammals) down to the lowest, every 
 important subdivision of the animal kingdom is represented, though 
 no birds spend all their time in the water. The variations of plant 
 life are also great, though the higher forms, such as we are most 
 familiar with on land, are wanting. 
 
 It is to be noted not only that the range of marine plants and 
 animals is great, but that the largest living animals, the whales, 
 are marine. Many of the marine plants, too, are of great size. 
 Some seaweeds are six inches in diameter, and some are hundreds 
 of feet long, exceeding in length the hei~ht of the tallest trees. 
 They are, however, not so bulky as large trees, and the amount 
 of solid matter which the largest seaweed contains is far less 
 than that of the largest tree. 
 
752 PHYSIOGRAPHY 
 
 The life of the sea is important in many ways. Many of the 
 animals, fish, oysters, clams, crabs, lobsters, etc., are used for food. 
 The total value of food products derived from the sea is probably 
 not less than $500,000,000 per year. Other animals furnish other 
 articles of commerce; for example, the seal furnishes fur and oil; 
 the whale, oil and whalebone; the walrus, exceptionally strong 
 leather, etc. Corals and sponges, the products of animal life, are 
 also articles of commerce. 
 
 Many of the animals of the sea have shells or other hard parts. 
 These hard parts accumulate on the bottom of the sea when the 
 animals are through with them, and this is a chief source of the 
 sediments of the sea bottom. When the shells, etc., accumulate 
 with little admixture of other material, they may in time be solidi- 
 fied by cementation and form limestone. Most of the limestone 
 now found on land was formed hi this way beneath the sea, when 
 the sea covered the areas where it now occurs. The animals which 
 make the heavier shells or other secretions of calcium carbonate 
 live chiefly in shallow water, and the seas in which the limestones 
 of the land were formed were generally shallow. 
 
CHAPTER XXV 
 
 MATERIALS OF THE SEA BOTTOM 
 
 Dredging. The material on the bottom of the sea has been 
 made known by dredging. Various forms of apparatus have been 
 used to bring up matter from the bottom. One, known as the 
 Cup Lead, is shown in Fig. 706. B is a hollow inverted cone on a 
 spike. Above the cone is a sliding disc, D, somewhat larger than 
 the base of the cone. This piece of apparatus is let down and the 
 cone sinks into the soft sediment and is filled with it. On being 
 raised, the disc shuts down and prevents the escape 
 of the contents of the cup, and also the access of 
 new matter from higher levels. 
 
 Fig. 707 shows a dredge. The flaring strip of 
 metal E is dragged along the bottom, and directs 
 the surface sediment into the sack. Swabs are 
 attached below to entangle animals missed by the 
 dredge. 
 
 The bottom of the sea is generally covered with 
 sediment which is, for the most part, in a loose 
 or uncemented condition. This sediment has 
 come from various sources. Some of it was carried 
 to the sea by rivers, some of it was worn from 
 the shores by the waves, some of it was blown out 
 from the land, some of it is made up of the shells, 
 etc., of the organisms which live in the water, and 
 some of it is composed of fine debris thrown out F _ __, 
 from submarine volcanoes. Cosmic ("shooting- cup lead. 
 
 star") dust is also an element, though a very (Challenger Re- 
 
 J port.) 
 
 minor one. 
 
 The sediments derived from the land came from the disinte- 
 gration of land rock. In their present state, however, they are to 
 be looked upon as rock in the making, for all sediments in the sea 
 
 753 
 
754 
 
 PHYSIOGRAPHY 
 
 Dredge Rope, 
 wivel. 
 
 Dredge Chain. 
 
 may become solid rock by cementation, and cementation is now 
 taking place at many points in the bottom of the sea. Locally, it 
 takes place as fast as the sediments accumulate. 
 
 Physically, the materials of the sea bottom may be grouped 
 
 into several classes, namely: 
 gravel, sand, mud, shells, coral, 
 etc., and ooze. 
 
 Gravel is found chiefly 
 along the borders of the land 
 out to depths of a few fathoms, 
 or at most a few scores of 
 fathoms. Gravel and bowlders, 
 carried out by icebergs, are 
 occasionally found at great 
 depths and far from land. 
 Sand also is generally confined 
 to relatively shallow water, 
 but it occurs out to depths 
 beyond that reached by gravel, 
 but rarely out to 100 fathoms. 
 Mud is much more wide- 
 spread. While it frequently 
 occurs in shallow water, it also 
 extends out to a depth far be- 
 yond that reached by gravel 
 and sand. In general, land- 
 derived mud is not washed 
 out far from the land, but m 
 exceptional cases, as off the 
 mouths of rivers, it is carried hundreds or even a thousand 
 miles. 
 
 Ooze is the name applied to the loose materials of the sea bottom 
 composed primarily of the minute shells and tests of organisms 
 which live in the water. Many of these organisms live near the 
 surface of the water, and their shells, etc., sink when they die. 
 The distinctive names of the oozes are derived from the names of 
 the organisms which contributed most to them. Thus, foramini- 
 feral ooze is the ooze in which shells of foraminifera are abundant, 
 diatom ooze is ooze in which tests of diatoms predominate, etc. 
 Foraminiferal ooze has a composition very similar to that of 
 
 FIG. 707. The dredge. (Challenger 
 Report.) 
 
MATERIALS OF THE SEA BOTTOM 755 
 
 chalk. Other oozes, such as diatom ooze, radiolarian ooze, etc., 
 are made up largely of silica. 
 
 In the deepest part of the ocean, below the depth of some 
 2200 fathoms, the surface is covered with red clay. The origin of 
 this clay has long been in question, but it is probably made up of 
 material derived from many sources. A considerable part was 
 doubtless derived from the fine material ejected from volcanoes; 
 another part was probably carried out from the land by the wind, 
 a part was probably derived from the shells and tests of animals 
 which lived in the ocean, and cosmic dust doubtless enters into 
 its composition. The materials from all these sources, so far as 
 they enter into the composition of the red clay, are probably only 
 the insoluble parts of the original material. 
 
 On the lands there is rock (conglomerate) composed of ce- 
 mented gravel, rock (sandstone) composed of cemented sand, rock 
 (shale) composed of cemented mud, and rock (limestone) com- 
 posed of material derived from shells, corals, etc. None of these 
 correspond to the deep-sea oozes, and none correspond to the red 
 clay of the abysmal depths. In the lands, therefore, there are 
 varieties of rock corresponding to all the sediments now making 
 in the shallow water of the sea, but, so far as known, none corre- 
 sponding to those of the deep waters. This suggests that the lands 
 have been at some tune beneath the sea, a conclusion which is 
 borne out by the finding of the shells of marine species imbedded 
 in the sandstone, shale, etc., of the land; but it also indicates that, 
 so far as now known, no part of the present continents ivas ever at 
 the bottom of the deep ocean. 
 
CHAPTER XXVI 
 RELATION OF THE SEA TO THE REST OF THE EARTH 
 
 THE ocean has an important influence on the rest of the earth. 
 This is felt in various ways, some of which have already been 
 noted. By way of summary they may here be brought to- 
 gether. 
 
 1. Waves affect the coast-line; they wear away the land in 
 some places and build new land in others. On the whole, destruc- 
 tion exceeds construction, so far as the land is concerned, so' that 
 the tendency of the ocean is to extend itself at the expense o the 
 land. 
 
 2. Oceans modify the climate of the land, affecting both tem- 
 perature and precipitation. The general influence on tempera- 
 ture arises from the fact that water is heated and cooled more 
 slowly than land is. The air over the sea, therefore, has a lesser 
 range of temperature than that over the land, and blowing to the 
 iand tends to carry the temperature of the sea, as well as abun- 
 dant moisture, over to it. Winds from the ocean therefore temper 
 the climate of the land both in summer and winter. The warm 
 currents enhance the general effect of the sea in this respect. The 
 climatic effect of the sea on the land is felt especially on the west 
 sides of the continents, in the temperate zones, because of the 
 westerly winds, and on the east sides of the continents in the zone 
 of easterly winds. The cold currents of the sea have much less 
 effect than warm ones on the climate of the land, because they 
 tend to hug the east sides of the continents, so far as they stay 
 at the surface; and in the latitudes where they occur the winds 
 blow from them to the sea rather than to land. 
 
 3. The ocean is the great source of the water for rain and snow, 
 and its precipitation from the atmosphere furnishes the conditions 
 necessary for life on the land. 
 
 4. Through its effects on rainfall, snowfall, and temperature, 
 the ocean has an important effect on the degradation of the land. 
 
 756 
 
RELATION OF THE SEA TO THE REST OF THE EARTH 757 
 
 The total amount of rainfall for the earth is not accurately 
 known. If it is as much as three feet per year, on the average, 
 for the whole earth, and if all this were derived directly from the 
 ocean, an amount equal to all the water in the ocean would be 
 evaporated in about 3000 years. Since most of the water evap- 
 orated from the ocean falls again into the sea, or runs to it in 
 rivers, or issues beneath it as springs, the amount of the ocean 
 water is not, so far as known, growing less. 
 
 5. The ocean affords an enormous harvest of foodstuff annually, 
 and many thousands of people depend on this harvest for their 
 livelihood. Fisheries were among the earliest industries. 
 
 6. The ocean also plays an important part in the commerce 
 of the world by serving as a great highway. The obstacle which 
 the oceans long seemed to interpose to quick communication be- 
 tween continents separated by them, has been overcome during the 
 last half century, and several cables now connect Europe and 
 America, so that all the important news of either continent is 
 known in the other almost as soon as it is at home. The Pacific, too, 
 is bridged by cables, though their number is small. 
 
 Some conception of the role which the ocean plays in the af- 
 fairs of the earth may perhaps be gained by picturing the con- 
 ditions which would exist if there were no oceans. 
 
 REFERENCES 
 
 1. WILD, Thalassa: Marcus Ward & Co., London. 
 
 2. MAURY, Physical Geography of the Sea. 
 
 3. CHAMBERLIN AND SALISBURY, Chapter VI, Geologic Processes, and 
 other text-books of Geology. 
 
 4. AGASSIZ, Three Cruises of the Blake: Hough ton Mifflin <fc Co. 
 
 5. THOMSON, The Depths of the Sea, and Voyage of the Challenger: 
 Macmillan. 
 
 6. Challenger Reports, especially Narrative, Vol. I, and Summary, First 
 Part: Eyre & Spottiswoode, London. 
 
 7. MURRAY, Important articles on oceanography in Geog. Jour., Vol. 
 XII, pp. 113-137, and Vol. XIV, pp. 34-50, and 426-441; and Scot. Geog. 
 Mag., Vol. XV, pp. 505-522. 
 
 8. BIGSBEE, Deep-sea Soundings and Dredgings: U. S. Coast Surv* 
 Washington. 
 
 9. TANNER, Deep-sea Exploration: U. S. Fish Commission, Washington 
 
 10. BELKNAP, Deep-sea Soundings in the North Pacific: U. S. Hydro- 
 graphic Office, Washington. 
 
 11. FLINT, Oceanography of the Pacific: Smithsonian Institution. 
 12! Jour, of Geol., Vol. XIII, pp. 469-484. 
 
INDEX 
 
 Abert Lake, 40 
 
 Abnormal temperature, 546 
 
 Abrasion by ground-water, 105 
 
 by the wind, 70 
 Absolute humidity, 570 
 
 temperature, 677 
 Accidents to streams, 173 
 Aconcagua, 445 
 Agassiz, A., 474, 757 
 Aggradation, 44 
 Agonic lines, 477 
 Agriculture and rainfall, 701 
 Air, chemical work of, 71 
 Aitkin, J., 290 
 Alaska, coast of, 458 
 Alden, W. C., (and Salisbury), R. D., 
 
 78 
 
 Aldrich deep, 711 
 Alluvial cone, 182 
 
 fan, 183 
 
 plains, 183, 187 
 fertility of, 191 
 
 terraces, 203 
 Alpine glaciers, 219 
 Alps, glaciers of, 233 
 Alps Mountains, 441 
 
 section of, 39 
 
 Altitude and temperature, 537, 545 
 Anchor-ice, 214 
 Andes Mountains, 441 
 Aneroid barometer, 583 
 Annual isobars, 585 
 
 maximum temperatures, 550 
 
 minimum temperatures, 551 
 
 parallax, 488 
 
 ran^e of temperature, 549, 552 
 Antarctic Circle, 502 
 Antarctica, ice-cap of, 216, 240 
 Antecedent streams, 177 
 Anticyclones, 620 
 
 movements of, 632 
 
 origin of, 648 
 
 tracks of, in United States, 636, 645 
 
 Ants, effect on surface, 78 
 Aperiodic changes of pressure, 620 
 Aphelion, 489 
 
 Appalachian Mountains, 39, 41 
 Aral Sea, 297 
 Arctic Circle, 502 
 
 Areas of low pressure in high lati- 
 tudes, 597, 601 
 Argon, 512 
 
 Aridity, effect on temperature, 545 
 Artesian wells, 94, 95 
 Ashes, volcanic, 368 
 Asteroids, 505 
 Astronomic latitude, 491 
 Atlantic, equatorial temperature 
 
 curves for, 723, 724 
 Atmosphere, a mixture of gases, 513 
 
 carbonic-acid gas of, 514 
 
 constitution of, 512 
 
 density and altitude, 507 
 
 general conception of, 506 
 . heating of, 521 
 
 height of, 508 
 
 history of, 510 
 
 impurities of, 513 
 
 mass of, 510 
 
 oxygen of, 514 
 
 relation to rest of earth, 507 
 
 temperature of, 520 
 
 volume of, 510 
 
 water-vapor of, 517 
 
 work of, 55 
 
 Atmospheric moisture, effect on 
 movements, 569 
 
 function of, 564 
 Atmospheric pressure, 582 
 
 temperature, effect on movement, 
 
 561 
 
 Atolls, 470 
 Attraction of sun and moon, cause of 
 
 ocean movement, 728 
 Aurora, 509 
 Autumn, 531 
 
 759 
 
760 
 
 INDEX 
 
 Avalanches, 253 
 Axis of earth, 484 
 inclination of, 497 
 
 Bad lands, 160 
 Balloon ascents, 508 
 Barometer, 582 
 
 aneroid, 583 
 
 Barometric gradient, 613 
 Bars, 327 
 Base-level, 131 
 
 temporary, 139 
 B(ath, springs of, 99 
 Batholiths, 375 
 Bayous, 189 
 Beaches, 324, 325 
 Belknap, G. E., 757 
 Bigsbee, Lieut. Com. C. D., 757 
 Black Hills, 23 
 Blake deep, 711 
 Blizzards, 646 
 Boiling springs, 124 
 Bonney, T. G., 390 
 Bore, tidal, 735 
 Bowlders, 251 
 Braided stream, 185 
 Brigham, Prof. A. P., 337 
 Buttes, 173 
 Bysmalith, 375 
 
 Caldera, 305 
 California, model of, 37 
 
 rainfall in, 619 
 Campbell, M. R., 206 
 Canyon of the Yellowstone, 159 
 Canyons and gorges, 156 
 Carbon dioxide of atmosphere, 512 
 Carbonation, 72 
 Carbonic-acid gas, 512 
 in atmosphere, 514, 517 
 
 effect on temperature, 517 
 in ocean, 719 
 Cascade Mountains, 35 
 Caspian Sea, 297 
 Castle Geyser, 92 
 Catskill Mountains, 440, 446 
 Caucasus Mountains, 442 
 Caverns, 97 
 Central America, earthquakes in, 
 
 428 
 
 Challenger, course of, 710 
 Chamberlin, Prof. T. C., 112, 289, 20, 
 
 291,337, 757 
 
 Changes of level, 43, 398, 400-404 
 Changes of temperature, effect of, 
 
 72 
 
 Charleston earthquake, 413 
 Chemical work of air, 71 
 
 Chemical work of ground-water, 96 
 Chesapeake Bay, drainage about, 174 
 Chile, coast of, 332, 459 
 
 earthquakes in, 416 
 Chimborazo, 445 
 Chinook winds, 673, 675 
 Chittenden, H. M., 113 
 Cinder buttes, 365 
 Cinders, volcanic, 368 
 Circle of illumination, 497 
 Circulation of the atmosphere, 598 
 Circumpolar whirl, 605 
 Cirques, 231, 249 
 Cirrus clouds, 576 
 Cliff glacier, 221, 222 
 Climate, 676 
 changes of, 703 
 continental, 683 
 oceanic, 683 
 of polar zones, 700 
 of tropical zone, 693 
 Climate and life, 702 
 Climates, classification of, 683 
 Climatic changes, causes of, 705 
 effects of ocean currents, 733 
 zones, 684 
 Cloudbursts, 667 
 Cloudiness, precipitation, etc., shown 
 
 on weather maps, 623 
 Clouds, 574 
 
 forms of, 576 
 
 Coal in the United States, distribu- 
 tion of, 455 
 Coastal plains, 17 
 Coast-line irregularities, distribution 
 
 of, 461 
 Coast-lines, affected by diastrophism, 
 
 463 
 
 affected by gradation, 462 
 affected by vulcanism, 464 
 effect on history, 465 
 irregularities of, 457 
 relief of, 460 
 Cold currents, 721 
 Cold waves, 645 
 Colorado, Grand Canyon of, 32, 33, 
 
 157 
 
 Colorado River, delta of, 199 
 Columnar structure, 388 
 Comets, 505 
 
 Compound alluvial fan, 183 
 Condensation of water-vapor, 572 
 Conduction, 526 
 Conglomerate, 49 
 Consequent falls, 168 
 
 streams, 177 
 
 Continental and oceanic tempera- 
 tures, 560 
 climate, 683, 690 
 
INDEX 
 
 761 
 
 Continental glaciers, 234 
 
 changes produced by, 274 
 
 of Europe, 272 
 
 of North America, 270 
 Continental platforms, 10, 11 
 
 shelves, 6, 22 
 
 slopes, 713 
 Continents, grouping of, 11 
 
 outlines of, 457 
 Contour interval, 20 
 
 map, explanation of, 19 
 Contours, 20 
 Convection, 93, 527 
 Convection currents, 529 
 Coon butte, 380, 381 
 Copper ores in the United States, 
 
 distribution of, 454 
 Coral islands, 469 
 
 reefs, 469 
 Cornish, V., 78 
 Corrasion, 129 
 Corrosion, 129 
 Coseismic lines, 425 
 Cotidal lines, 747 
 Counter currents, 733 
 Cowles, H. C., 78 
 Crater, 339 
 
 Crater Lake, 297, 305, 306, 382 
 Craterlets, 410 
 Creep, 108, 133 
 Crevasses of glaciers, 224 
 Crustal deformation, 405 
 
 movements, 392 
 Cumulus clouds, 576 
 Currents, oceanic, 721, 730 
 Cycle of erosion, 153 
 Cyclone, 621 
 
 structure of, 625 
 
 Cyclones, mean tracks of, in United 
 States, 636 
 
 movements of, 632 
 
 origin of, 648 
 
 paths of, 645 
 
 tropical, 648 
 Cyclonic winds, 618 
 
 Daily range of temperature 553, 
 
 559 
 
 Daly, R. A., 206 
 Dana, Prof. J. D., 473, 474 
 Darwin, Charles, 474 
 Davis, B. M., 113 
 
 Davis, Prof. W. M., 205, 290, 51 1, 686 
 Dead Sea, 297 
 Deepening of valleys, 129 
 Deeps, 711 
 Deflection of winds by rotation, 602, 
 
 604 
 Degrees, length of, 494 
 
 Delta of the Colorado River, 199 
 200 
 
 the Danube, 202 
 
 the Mississippi, 181, 197 
 
 the Nile, 201 
 
 the Rhone River, 201 
 
 the St. Clair, 197 
 Delta lakes, 310 
 Delta-land, 330 
 Deltas, 198, 199 
 Density and movement of sea-water 
 
 719 
 
 Deposition by continental glaciers 
 276 
 
 by ground-water, 99 
 
 by running water, 179, 180 
 Depth-limit of valleys, 131 
 Desert climates, 690 
 Dew, 573 
 
 Dew-point, 570, 572 
 Diastrophism , 303, 392 
 
 a cause of change of sea-level, 401 
 
 effects on coast -lines, 463 
 Diatom ooze, 754 
 Dikes, 375 
 
 Diller, J. S., 337, 390, 391 
 Dip compass, 479 
 Direction of winds, 602, 613 
 Distributary streams, 185 
 Divides, permanency of, 140 
 Dodge, Prof. R. E., 206 
 Doldrums, 617 
 Dolphin ridge, 714 
 Drainage changed by glaciation, 285 
 Drainage of the upper Ohio basin 
 
 285 
 
 Dredging, 753 
 Drift, 257 
 
 disposition of, 258 
 
 oceanic, 730, 732 
 
 topography, 260, 264 
 Driftless area, 272 
 Drowning of valleys, 173 
 Drumlins, 265, 278, 279 
 Dry farming, 615 
 Dry winds, 697 
 Dunes, 63 
 
 distribution of, 63 
 
 migration of, 67 
 
 shapes of, 65 
 Dust, 55 
 
 in atmosphere, 518 
 
 distribution of, 62 
 
 sources of, 56 
 Dust-wells, 227, 228 
 Dutton, Major C. E., 206, 391, 434 
 
 Earth, flattened form of, 482, 496 
 motions of, 484, 488 
 
762 
 
 INDEX 
 
 Earth relations, 482 
 
 Earth, size of, 484 
 
 Eartnquake fissures,! 431 t i .; 
 
 Earthquake wave, 424 
 
 Earthquakes, 408 
 
 beneath the sea, 420 
 
 Calif ornian, 419 
 
 causes of, 430 
 
 distribution of, 429 
 
 frequency of, 427 
 
 in Italy, 418 
 
 in Japan, 412 
 
 in Mississippi valley, 417 
 
 in Panama, 428 
 
 strength and destructiveness, 408 
 
 surface changes caused by, 432 
 Earthworms, effect on surface, 78 
 Ebb-tide, 735 
 Eddying streams, 124 
 Elbruz Mountains, 445 
 Elk Mountains, 36, 39 
 Enchanted Mesa, 172 
 Englacial drift, 253 
 Entrenched meanders, 174 
 Eolian sand, 63, 69 
 Epicontinental seas, 7 
 Equator, 484 
 Equatorial calms, 603, 617 
 
 currents, 730 
 
 temperature curves for the At- 
 lantic, 724 
 
 temperature curves for the Pacific, 
 
 724 
 
 Equinoxes, 499 
 Erosion, 129 
 
 by continental glaciers, 274 
 
 by streams, 120 
 
 by wind, 55 
 
 conditions affecting rate of, 155 
 Eskers, 267 
 Evaporation, 80, 565 
 
 affected by temperature, 568 
 
 amount of, 567 
 
 effect of temperature on, 568 
 
 influenced by wind, 568 
 
 rate of, 566 
 Exfoliation, 74 
 Extinct lakes, 333 
 Extra-tropical belts of high pressure, 
 600 
 
 Fairchild, Prof. H. L., 291 
 Fall Line, 22 
 Fault, normal, 406 
 
 reversed, 406 
 Faulted mountain, 39 
 Faulting, 405, 406 
 Faults and folds, 406 
 Fault-scarps, 407 
 
 I Fenneman, Prof. N. M., 337 
 Ferrel, W., 511 
 "Fetch" of waves, 321 
 Filling of lake basins, 302 
 Fiords, 248, 331 
 Fissure eruptions, 371 
 Fissures, 338 
 Flint, J. M., 757 
 Floe-ice, 212 
 
 Flood-plain meanders, 187 
 Floods of rivers, 115, 195 
 
 of Yellow River, 202 
 Flood-tide, 735 
 Flowing wells, 94 
 "Fluvial period" of history, 191 
 Fluvio-glacial deposits, 265 
 Foehn winds, 673 
 Fog, 574, 734 
 Folding, 405 
 Folds and faults, 406 
 Foraminiferal ooze, 754 
 Forests, climatic effect of, 692 
 Form of earth, 482 
 Fossils as evidence of change of level, 
 
 393 
 
 Foucault's pendulum, 486 
 Freezing and thawing, 72 
 Frigid zones, 684 
 Frost, 573 
 Fuji-yama, 445 
 Funnel-shaped cloud, 668 
 
 Galveston storm, 648, 654 
 Gannett, H., 206 
 
 Geikie, Prof. J., 78, 112, 290, 337 
 Geikie, Sir Archibald, 290, 337 
 General circulation and precipitation, 
 
 614 
 
 Geodetic latitude, 491 
 Geographic latitude, 491 
 Geography defined, 4 
 Geology defined, 3 
 Geysers, 90 
 Geyser tube, 91 
 Giant Geyser, 91 
 Gilbert, G. K., 205, 290, 337, 391, 
 
 434 
 Glacial bowlders, 252 
 
 epochs, cause of, 273 
 
 grooves, 249, 250, 263 
 
 lakes, 311 
 
 period, 270 
 
 plains, 436 
 
 stria?, 248, 250, 263 
 Glaciated valleys, 244 
 Glaciation and drainage, 280, 285 
 
 effects of, on human affairs, 287 
 Glacier movement, nature of, 231 
 Glaciers, 219 
 
INDEX 
 
 703 
 
 Glaciers, conditions affecting rate of 
 mo vement, 230 
 
 crevasses of, 224 
 
 deposition by, 255 
 
 effect on shores, 330 
 
 erosion by, 242 
 
 movement of, 229 
 
 size of, 233 
 
 types of, 219 
 
 waste of, 229 
 
 work of, 242 
 Glaciers and ice sheets, ancient, 
 
 270 
 Gold and silver, distribution of, 
 
 454 
 Gradation, 305 
 
 agents of, 44 
 
 effects on coast-lines, 462 
 Gradient of wind, 613 
 Graham Island, 379 
 Grand Canyon of the Colorado, 32, 
 
 33, 157 
 Granite, 52 
 Gravity faults, 406 
 Great Ararat, 445 
 Great Bear Lake, 297 
 Great Lakes, the, 284 
 Great Plains, 23 
 Great Salt Lake, 315 
 Great sea waves, 412 
 Greece, earthquakes of, 420 
 Greely, Gen. A. W., 511 
 Greenland, ice-cap of, 216, 235 
 Ground-ice, 214 
 Ground moraine, 257, 278 
 Ground-swell, 319 
 Ground-water, 83 
 
 abrasion by, 105 
 
 amount of, 86 
 
 deposition by, 99 
 
 descent of, 84 
 
 existence of, 83 
 
 movement of, 86 
 
 solution by, 96 
 
 source of, 83 
 
 surface, 85 
 
 work of, 80, 96 
 Gulf Stream, 544, 730 
 Gullies, 119 
 Gulliver, F. P., 474 
 Gully, growth of, 142 
 
 Heat received in different latitudes, 
 
 524 
 
 Heating of land and water, 530 
 Heilprin, Prof. A., 390, 474 
 Henry Mountains, 373 
 Hanging valleys, 247 
 Hann, Prof. J., 511, 689 
 
 Harrington, M. W., 337 
 Hawaiian volcanoes, 301 
 Hayden, F. V., 511 
 Hayes, Dr. C. W., 206 
 "High." (See Anticyclone.) 
 High-latitude areas of low pressure. 
 
 601 
 
 High-latitude glaciers, 219 
 High-pressure belts, 594 
 High tides, 738 
 
 variation in height of, 744 
 Hill, R. T., 390 
 Himalayas, rainfall in, 619 
 Hitchcock, A. S., 78 
 Hogbacks, 171 
 Hot waves, 645 
 Hovey, E. O., 112, 390 
 Humidity, 570 
 
 absolute, 570 
 
 effect on temperature, 545 
 
 relative, 570 
 Hurricanes, 648 
 Hydration, 105 
 Hydrography denned, 3 
 
 Ice and snow, work of, 207 
 Ice of lakes, 207 
 
 of sea, 210 
 
 in soil, 207 
 Icebergs, 269 
 Ice-cap of Antarctica, 240 
 
 of Greenland, 235 
 Ice-caps, 221, 234 
 Ice cascade, 224 
 Ice columns, 228 
 Ice-fields, 218 
 Ice-foot, 212 
 Ice jams, 213 
 Ice-packs, 212 
 Ice-sheets, 221 
 Ice-shove, 212 
 Igneous rocks, 51 
 Immediate run-off, 83 
 Inclination of earth's axis, effects of, 
 
 497 
 India, earthquakes in, 416 
 
 famines in, 619 
 Inequalities of level, 727 
 Insolation, 522 
 
 in different latitudes, 524 
 Interior plains, 17, 22 
 Intermediate zones, 684, 686 
 
 climate of, 695 
 Intermittent streams, 118 
 Intrusions of lava, 374 
 Iron ores in the United States, dis- 
 tribution of, 455 
 Irrigation, 193, 451 
 Is-abnormal lines, 546 
 
764 
 
 INDEX 
 
 Is-abnormal temperatures for Janu- 
 ary, 547 
 
 for July, 548 
 Islands, origin of, 466 
 Isobaric chart for January, 591 
 
 for July, 593 
 Isobaric charts, 584 
 Isobaric gradient, 587 
 Isobaric surfaces, 587, 596 
 Isobars, 584, 620 
 
 annual, 585 
 
 courses of, 588 
 
 and humidity, 594 
 
 and parallels, 589 
 
 and temperature, 590, 592 
 
 relation of, to land and water, 589 
 Isogonic lines, 477 
 Isothermal chart of United States, 
 annual, 553 
 
 for April, 554 
 
 for January, 554 
 
 for July, 555 
 
 for October, 555 
 Isothermal charts, 539 
 Isothermal surfaces, 546 
 Isotherms, 539 
 
 courses of, 540, 624 
 
 Japan, earthquakes in, 412 
 Jefferson, M. W., 206 
 Jetties of the Mississippi, 128 
 Johnson, L. C., 206 
 Joints, 53 
 
 Jordan, Pres. D. S., 434 
 Jordan craters, 364 
 Judd, Prof. J. W., 390 
 Jupiter, 504 
 
 Kames, 268 
 Karst, 98 
 
 Karst topography, 98 
 Kaskaskia, 189 
 Kemp, Prof. J. F., 337 
 Kenai, 445 
 Kilauea, 363, 364 
 Kilimanjaro, 445 
 King, F. H., 112 
 Krakatoa, 348 
 Kiimmel, H. B., 291 
 
 Laccoliths, 374 
 Lacustrine plains, 333, 437 
 Lag of tides, 741 
 Lake Agassiz, 282, 333 
 
 Agnes, 35 
 
 Baikal, 297 
 
 Balkash, 297 
 
 Bonne ville, 314 
 
 Chad, 297 
 
 lake Chelan, 297 
 
 Como, 297 
 
 Erie, 297 
 
 Garda, 297 
 
 Huron, 297 
 
 Michigan. 297 
 
 Nyassa, 297 
 
 Ontario, 297 
 
 Pepin, 292 
 
 Pontchartrain, 310 
 
 St. Clair, delta of, 197 
 
 Superior, 297 
 
 Tanganyika, 297 
 
 Titicaca, 297 
 
 Victoria Nyanza, 297 
 
 Winnipeg, 297 
 Lake basins, filling of, 302 
 
 origin of, 303 
 Lake water, movements of, 300 
 
 sources of, 301 
 Lakes, 292 
 
 area of, 295 
 
 changes of level of, 300 
 
 changes taking place in, 302 
 
 climatic effects of, 316 
 
 conditions necessary for, 301 
 
 depth of, 295, 296 
 
 distribution of, 293 
 
 economic advantages and disail* 
 vantages of, 316 
 
 fate of, 303 
 
 lowering of outlets, 303 
 
 origin of, 313 
 
 sections of, 298 
 
 topographic position of, 295 
 
 volume of water in, 297 
 Lakes of Red River (La.) Valley, 301 
 Land, materials of, 45 
 Land and water, effect on temper* ,~ 
 ture, 541 
 
 heating of, 530 
 Land-breezes, 561, 610 
 Land degradation, rate of, 154 
 Landslides, 106 
 Landslip Mountain, 106 
 Land-tied islands, 329 
 Land-water, source, 80 
 Lateral moraines, 254 
 
 mode of origin, 257 
 Latitude, 490 
 
 and sun altitude, 502 
 
 length of degrees, 494 
 Lava, 338, 367 
 
 Lava-flows of the northwest, 372 
 Lead and zinc ores in the United 
 
 States, distribution of, 456 
 Le Conte, Prof. J., 473 
 Leuk, springs of, 98 
 Leverett, F., 290 
 
LXDEX 
 
 765 
 
 Life and climate, 702 
 Lightning, 665 
 Limestone, 49 
 
 origin of, 718 
 Limestone sinks, 97 
 Lithosphere, relief of, 9 
 Littoral climate, 691 
 
 currents, 319, 326, 729 
 Load of streams, 122, 128 
 Local variations of pressure, 597 
 Loess, 58, 59, 61 
 Lone Star Geyser, 92 
 Longitude, 491 
 
 and time, 490, 492 
 Louisville tornado, 670, 671 
 "Low." (See Cyclones) 
 Low tides, 738 
 
 Magnetic declination, 477, 478 
 
 inclination, 479 
 . intensity, 479 
 
 meridians, 475, 476 
 Magnetism, terrestrial, 475 
 Malaspina Glacier, 210 
 Mammoth Cave, 97 
 Mammoth Springs, 102 
 Mantle rock, 46 
 Mars, 504 
 
 Maryland vein, 100 
 Marysville buttes, 384 
 Massive rocks, 52 
 Materials of the land, 45 
 Mature topography, 152 
 
 valley, 149 
 
 Maturity of rivers, 150 
 Mauna Kea, 442-443 
 Mauna Loa, 339, 362, 365 
 McGee, W J, 434 
 Mean monthly temperatures, 680 
 Medial moraine, 255 
 Medicinal springs, 90 
 Mercury, 504 
 Meridians, 484, 491 
 Merrill, G. P., 78, 206 
 Mesa, 34, 173 
 Metamprphic rocks, 53 
 Meteorites, 514 
 Meteorology defined, 3 
 Meteors, 509 
 Mill, H. R., 337 
 Milne, J., 434 
 Mineral matter of sea, 716 
 
 amount of, 716 
 
 source of, 717 
 
 withdrawal of, 718 
 Mineral springs, 90 
 Mississippi River, amount of sedi- 
 ment carried by, 121 
 
 delta of, 181, 197 
 
 Mississippi River, flat, 136 
 
 rate 01 erosion, 154 
 Mistral, 648 
 
 Moisture, effect on temperature, 545 
 Moisture of the air, 564 
 Monadnocks, 153, 171 
 Monoclinal folds, 406 
 Monsoon winds, 562, 605 
 
 of Chicago, 609 
 
 of India, 608 
 
 of Spain, 610 
 Mont Pelee, 350 
 Monument Park, 71 
 Moore, W., 511 
 Moraine-dams, 258 
 Motions of earth, 484 
 Moulton, Prof. F. R., 740 
 Mount Hood, 381 
 
 Mazama, 307 
 
 McKinley, 441 
 
 Rainier, 381, 383, 441 
 
 Shasta, 380, 381, 382 
 
 Whitney, 441 
 
 Wrangell, 445 
 Mountain breezes, 562 
 
 chain, 439 
 
 climate, 691 
 
 group, 439 
 
 range, 438 
 
 ridge, 438 
 
 system, 439 
 Mountains, 15, 33, 437 
 
 barriers to animals and plants, 452 
 
 barriers to transportation, 452 
 
 changes taking place in, 443 
 
 distribution of, 439 
 
 effects of, 450 
 
 produced by erosion, 445 
 by faulting, 448 
 by folding, 447 
 by intrusion, 447 
 
 height of, 440 
 
 in history, 38 
 
 in ocean, 442 
 
 origin of, 39, 445 
 Movement of sea-water, 719 
 
 causes of, 727 
 Mud cones, 389 
 
 volcanoes, 388 
 Murdoch, L. H., 337 
 Murray, Sir John, 337, 757 
 
 Nansen, F., 291 
 
 Narrows, 169 
 
 Natural Bridge of Virginia, 163 
 
 Natural bridges, 98, 161 
 
 Natural levees, 187 
 
 Neap tides, 744 
 
 Needle Mountains, 34 
 
766 
 
 INDEX 
 
 Neptune, 504 
 
 New Madrid earthquake, 418 
 
 Niagara Falls, 167 
 
 Nimbus clouds, bt\j 
 
 Nitrogen, 512, 513 
 
 Nitrogenous compounds, 513 
 
 Normal fault, 406 
 
 North Atlantic, temperature curve 
 
 for, 723 
 
 Northern lights, 509 
 Northers, 646 
 Nunatak, 237 
 
 Obsidian, 367 
 Ocean, the, 706 
 
 age of, 718 
 Ocean basins, 708 
 
 continuity of, 11 
 
 Ocean bottom, topography of, 713 
 Ocean currents, 721, 730 
 
 cause of, 733 
 
 climatic effects of, 733 
 
 effect on temperature, 544 
 
 gradational effects of, 734 
 Ocean, depth of, 711 
 Ocean life, 749 
 Ocean, mass of, 712 
 Ocean movements, types of, 729 
 Ocean, reasons for low temperature 
 of, 724 
 
 relation to rest of earth, 756 
 
 temperature of, 721, 722 
 
 volume of, 712 
 
 Ocean water, temperature and move- 
 ment, 722 
 
 Oceanic climates, 683, 688 
 Oceanography, defined, 3 
 Oceans, area of, 6 
 Old age of rivers, 150 
 Ores, distribution of, 452 
 Orizaba, 445 
 Ouachita Mountains, 23 
 Outwash plain, 266 
 Ox-bow lakes, 189 
 Oxygen of atmosphere, 512, 514 
 Ozark Mountains, 23 
 
 Pacific, equatorial temperature 
 
 curves for, 724 
 Paint-pots, 390 
 Palisade Ridge, 385, 387 
 Parallels 4SO 
 Peary, Lieut. R. E., 291 
 Pelee, 58 
 Peneplains, 153 
 Perched bowlder, 262, 264 
 Perihelion, 489 
 Periodic winds, 98 
 Petrifaction, 104 
 
 Petrified tree-trunks, 104 
 Physical geography of the sea, 709 
 Piedmont alluvial plain, 183 
 Piedmont glaciers, 222, 240 
 Piedmont plateau, 28 
 Piracy, 176 
 Plains, 15, 16, 435 
 
 topography of, 24 
 Planetary winds, 604 
 Plateau climate, 691 
 Plateaus, 15, 28, 437 
 
 origin of, 31 
 
 position and area of, 30 
 Polar circles, 502 
 
 zones, 684, 686 
 
 climate of, 700 
 Poles of earth, 484 
 Ponding, 175 
 Popocatepetl, 445 
 
 Population of the United States, 26 
 Pore space, 82 
 Pot-holes, 168, 169 
 Powell, Major J. W., 474 
 Precipitation, 80, 573, 616 
 
 and general circulation, 614 
 
 necessary for agriculture, 615 
 Pressure, aperiodic changes of, 620 
 
 atmospheric, 532 
 
 inequalities of, 583 
 Prevailing winds, 598, 604 
 
 Radiation, 526 
 
 Rafts, 308 
 
 Railway map of the United States, 
 
 453 
 
 Rain, 580 
 Rainfall, 614 
 
 and agriculture, 701 
 
 in California, 619 
 
 distribution of, 615 
 
 and sugar crops, 702 
 
 of the United States, 452, 61& 
 
 zone of trades, 616 
 
 zones of westerlies, 617 
 Rain-making, 580 
 Rain-water, fate of, 81 
 Raised beaches, 394 
 Ravines, 119 
 Reade, T. M., 474 
 Recessional moraines, 277 
 Reconstructed glacier, 222 
 Red Sea, temperature of, 725 
 Red snow, 239 
 Reefs, 325 
 
 Reid, Prof. H. F., 290 
 Regolith, 46 
 
 Rejuvenation of streams, 174 
 Relative- humidity, 570 
 Relief of iithosphere, 9 
 
INDEX 
 
 767 
 
 Relief features of first order, 5, 12 
 
 of the land, 15 
 
 of sea bottom, li> 
 
 of second order, 1 5 
 Relief-map of United States, 27 
 Reversed fault, 406 
 Revolution of earth, 488 
 Rise of land (relative), evidences of. 
 
 393 
 
 River floods, 195 
 River ice, 213 
 River lakes, 308 
 River plains, 436 
 River system, history of, 141 
 Rivers, amount of water in, 114 
 
 effect on shores, 329 
 
 load of, 122 
 Rock, 47 
 Rock-basins, 275 
 Rock-breaking, 73, 111 
 Rock decay, 111 
 Rock flour, 251 
 Rock terraces, 170, 171 
 Rock waste, 46, 47 
 Roots, agents of weathering, 76 
 Rotation of earth, 484 
 Rotation, effect of, 488 
 
 winds deflected by, 602, 604 
 Running water, work of, 114 
 Run-off, 83 
 
 Russell, Prof. I. C., 205, 289, 290, 337, 
 390, 511 
 
 Salinity and movement, 719 
 Salt Lake, 315 
 Salt lakes, 314 
 Sand, 62 
 
 lodgment of, 62 
 
 sources of, 62 
 Sandstone, 49 
 
 San Francisco earthquake, 419 
 San Francisco Mountain, 381, 384, 
 
 386 
 
 Satellites, 505 
 Saturation, 569 
 Saturn, 504 
 
 Scandinavia, changes of level in, 393 
 Sea, temperature of, 721 
 Sea bottom, materials of, 753 
 Sea-breezes, 561, 610 
 Sea caves, 394 
 
 ice of, 726 
 
 Sea cliff, 321, 322, 323, 394 
 Sea-level, 400. 707 
 
 distortion by attraction, 707 
 
 inequalities of , 707,727 
 Seasonal range of temperature. 559 
 Seasons, 531 
 
 change of, 533 
 
 Seasons in different latitudes, 534 
 Sea-water, composition of, 710 
 
 gases in, 718 
 
 movements, 727 
 
 salinity and color, 720 
 Secular changes of level, 392 
 Sedimentary rocks, 48 
 Sedimentation, a cause of change of 
 
 sea-level, 400 
 Seepage, 89 
 Seiches, 300 
 Semi-arid lands, 618 
 Sensible temperature, 677 
 Seward glacier, 234 
 Shale, 49 
 Shaler, Prof. N. S., 112, 205 290, 
 
 337, 434 
 
 Shepard, E. M. Prof., 434 
 "Shooting-star" dust, 753 
 Shooting-stars, 509, 514 
 Shore current, 318, 319 
 
 deposition by, 324 
 Shore-drift, 325 
 Shore ice, 333 
 
 lakes, 311 
 
 terraces, 334 
 Shore lines, 292 
 
 effect of glaciers on, 330 
 
 effect of rivers on, 329 
 
 effect of wind on, 330 
 
 topographic features of. 317 
 Sierra el Late Mount aias, 34 
 Sills, 374, 375 
 
 Sinking of coasts, evidences of, 395 
 Sirocco, 645 
 
 Slichter, Prof. C. S., 112 
 Sliding, 105 
 
 Slumping, 105, 107, 135 
 Snake river, 31 
 Snow, 214, 580 
 Snow and ice, work of, 207 
 Snow-eaters, 674 
 Snowfall, 682 
 
 at Chicago, 680 
 
 in the United States, 556 
 Snow-fields, 215 
 Snowflakes, 81 
 Snow line, 215, 217 
 Soil, 46 
 Solar climate, 683 
 
 system, 504 
 
 tides, 743 
 Solstices, 499 
 
 Solution by ground-water, 96 
 Soufridre, 57, 350 
 Sounding line, 712 
 South Atlantic, temperatures in, 725 
 
 temperature curve for, 725 
 Spanish Peaks, 445 
 
768 
 
 INDEX 
 
 Spits, 327 
 
 Spring, defined, 531 
 Spring tides, 736, 744 
 Springs, 89 
 
 mineral and medicinal, 90 
 
 temperature of, 89 
 Stalactites, 101 
 Stalagmites, 101 
 Standard-time zones, 495 
 St. Louis tornado, 669 
 Stone, G. H., 2^0 
 Storms, 621 
 
 special types of, 663 
 Stratified drift, 287 
 Stratified rocks, 48 
 Stratus clouds, 576 
 Stream water, sources of, 116 
 Streams, accidents to, 173 
 
 antecedent, 177 
 
 consequent, 177 
 
 deposition by, 179 
 
 erosive work of, 120 
 
 load of, 122 
 Stromboli, 341 
 
 Submarine volcanic extrusions, a 
 cause of change of sea-level, 
 
 401 
 
 Submerged forests, 396 
 Submerged valleys, 397, 714 
 Subsoil, 47 
 Summer, defined, 531 
 Summer and winter, differences, 
 
 532 
 
 Summer solstice, 499 
 Sun, varying distance of, 536 
 Sun-cracks, 135 
 
 Sun's distance, effect on tempera- 
 ture. 536 
 Supan, A., 688 
 Surf, 319 
 
 Suspension of sediment in rivers. 
 127 
 
 Tajamulco, 445 
 Talus, 74, 134 
 Talus glacier, 107 
 Tanner, Lieut. Com., Z. L., 757 
 Tarr, Prof. R. 8., 337, 474 
 Taylor, F. B., 291, 337 
 Temperature of air, 520 
 effect on movement, 561 
 vertical movements, 563 
 Temperature and altitude, 537 
 and condensation of water-vapor, 
 
 572 
 Temperature changes with rise of air, 
 
 537 
 
 curve for the North Atlantic, 
 723 
 
 Temperature curve for the South 
 
 Atlantic, 723 
 
 Temperature, effect on evaporation, 
 568 
 
 of springs, 89 
 
 of sea, 721 
 
 on weather map, 624 
 
 zones, 688 
 
 Temperatures in South Atlantic, 725 
 Temporary base-level, 139 
 Terminal moraines, 257, 276 
 Terrestrial magnetism, 475 
 Texas, coast of, 461 
 
 model of, 36 
 Thermal equator, 602 
 Thermometer, 520 
 
 self-registering, 726 
 Thomson, Dr. C. Wyville, 757 
 Thrust-fault, 406 
 "Thunder-squall," 664 
 Thunder-storms, 610, 663, 666 
 Tidal currents, 736 
 Tidal poles, 741 
 Tidal races, 736 
 Tides, 729, 735 
 
 cause of, 736 
 
 effects of on shores, 748 
 
 lag of, 741 
 
 monthly variation of, 746 
 
 periodicity of, 736 
 
 rate of movement, 748 
 Timber jams, 308 
 
 Topographic features of shores, 317 
 Topographic relations, effect on tem- 
 perature, 545 
 
 Topographic unconformity, 332 
 Tornado near Chicago, 672 
 
 at Louisville, 670 
 
 at St. Louis, 669 
 
 at Rochester, Minn., 672 
 Tornadoes, 611, 657 
 Tower, N. S., 206 
 Trade-winds, 603, 617 
 Trade-wind zone, 687 
 Transpoitation by rivers, 127 
 Travertine, 101 
 Tributaries, growth of, 147 
 Tropic of Cancer, 501 
 
 of Capricorn, 501 
 Tropical belts of high pressure, 594 
 
 calms, 601, 617 
 
 cyclones, 611, 648 
 
 zone, 684, 685 
 
 climate of, 693 
 
 Turtle Mountain landslide, 106 
 Tuscarora deep, 711 
 Tyndall, J., 290 
 Typhoons. 655 
 Typhoon tracks, 659 
 
INDEX 
 
 769 
 
 Udden, Prof. J. A., 78 
 
 Undertow, 318, 319, 729 
 
 Unequal heating of land and water, 
 
 effect on wind, 605 
 Unequal insolation, effect on winds, 
 
 598 
 
 Upham, Warren, 291, 337 
 Uranus, 504 
 
 Valley, stages in history of, 148 
 Valley breezes, 562 
 
 flats, 135 
 
 glacier, 219, 223 
 
 train, 266 
 Valleys, 118 
 
 courses of, 144 
 
 deepening of 129 
 
 depth-limit of, 131 
 
 lengthening of, 140 
 
 widening of, 131 
 
 width-limit of, 134 
 Variable climate, 677 
 Veins, 100, 101 
 Velocity of wind, 613 
 Venus, 504 
 Vesuvius, 341 
 Volcanic "ash," 56, 368 
 Volcanic belts, 368 
 
 cinders, 368 
 
 cones, 378 
 
 dust, 56 
 
 eruptions, phenomena of, 366 
 
 gases, 368 
 
 glass, 367 
 
 mountains, 445 
 
 necks, 385 
 
 plugs, 385 
 
 vapors, 368 
 Volcano, 338 
 Volcanoes, active, 341 
 
 distribution, 368 
 
 explosive type, 340 
 
 number, 368 
 
 products of, 367 
 
 quiet type, 340 
 
 topographic effects of, 378 
 Vulcanism, 305, 338 
 
 causes of, 375 
 
 effect on coast-lines, 464 
 
 Walcott, Dr. C. D., 206 
 Waldo, Frank, 511 
 Wall Lake , Iowa, 210 
 Ward, Prof. R- D. C., 511, 684 
 Warm currents, 721 , 732 
 Warped shore-lines, 394 
 Warping, 405 
 
 Washington, rainfall in, 619 
 Watchung Mountains, 387 
 Waterspouts, 673 
 Water surface, 85 
 Water table, 85 
 Water-vapor, 564 
 
 amount of, 568 
 
 of atmosphere, 512, 517 
 
 distribution of, 569 
 
 sources of, 565 
 Wave-cut terrace, 324 
 Waves, 318, 729 
 
 crest of, 318 
 
 deposition by, 324 
 
 erosive work of, 320 
 
 "fetch " of, 321 
 
 length of, 318 
 
 of translation, 319 
 
 period of, 318 
 
 trough of, 318 
 Weather Bureau, 622, 663 
 Weathering, 72, 76, 110, 129 
 
 conditions affecting, 111 
 Weather maps, 620, 627 
 Weather predictions, 656 
 
 failure of, 660 
 
 value of, 663 
 Weed, W. H., 113 
 Westerly winds, 603 
 West Indian storms, 653 
 Whirlwinds, 611, 666 
 Width-limit of valleys, 134 
 Wild, J. J., 757 
 Willis, Bailey, 206, 473 
 Wind, abrasion by, 70 
 Wind-carving, 70 
 Wind, cause of ocean movements, 
 
 728 
 Wind, direction, 613 
 
 in upper air, 606 
 
 effect on evaporation, 568 
 
 effect on shores, 330 
 Wind equator, 602 
 Wind -gap, 177 
 W T ind gradient, 613 
 W T ind, mechanical work of, 55 
 Wind roses, 609 
 Wind shown on weather maps. 
 
 622 
 
 Wind velocities, 613, 614 
 Wind zones, 686 
 Winds, affected by moisture, 569 
 
 deflected by rotation, 602, 604 
 
 due to unequal temperature, 561 
 
 effect on temperature, 543 
 
 gradational effect of, 62 
 Winter, denned, 531 
 Winter and summer, difference*, 
 532 
 
770 
 
 INDEX 
 
 Winter solstice, 499 
 Woonsocket, artesian well at, 94 
 Wyanaotte Cave, 97, 101 
 
 Yellow River of China, 196 
 Yellowstone Canyon, 159 
 Yellowstone Lake, 102 
 Yellowstone National Park, 90, 103 
 
 Young valley, 148 
 Youthful topography, 152 
 Youth of rivers, 15U 
 
 Zones of climate, 684 
 defined by isotherms, 687 
 by latitude, 685 
 by winds, 686 
 
Chamberlin and Salisbury's Geology 
 
 By THOMAS C. CHAMBERLIN AND ROLLIN D. SALISBURY, 
 Professors in the University of Chicago. (American Science Se- 
 ries, Advanced Course.) 3 Volumes 8vo. 
 
 Volume I. Processes and Their Results. xix-f654 pp. $4.00 
 Volumes II and III. Earth History. xxvi-|-692-(-xi-)-624 pp. 
 Vols. II and HI not sold separately. $8.00. 
 
 Chas. D. Walcott, Director of U. S. Geological Survey: \ am 
 impressed with the admirable plan of the work and with the thorough 
 manner in which geological principles and processes and their results 
 have been presented. The text is written in an entertaining style, 
 and is supplemented by admirable illustrations, so that the student 
 cannot fail to obtain a clear idea of the nature and work of geological 
 agencies, of the present status of the science, and of the spirit which 
 actuates the working geologist. 
 
 Israel C. Russell, Professor in the University of Michigan: 
 The work is certainly monumental, and, like Lyell's Principles of 
 Geology, will in the future, I am confident, be recognized as marking 
 the beginning of a new period in the development of the science of 
 the earth. 
 
 R. S. Woodward, Director of the Carnegie Institution: It is 
 admirable for its science, admirable for its literary perfection, and 
 admirable for its unequalled illustrations. 
 
 College Geology 
 
 By T. C. CHAMBERLIN AND R. D. SALISBURY. (American 
 Science Series.) xvi-f-9/8 pp. 8vo. $3.50. 
 
 A one-volume text-book on the lines of the authors' widely 
 known three-volume work. It supplies ample material for the av- 
 erage college course in geology. 
 
 T. C. Hopkins, Professor in Syracuse University: I find it so 
 far in advance of the other text-books I have seen that I decided to 
 use it in my class. 
 
 A. H. Purdue, Professor in thz University of Arkansas: It 
 probably will be the standard college text-book in Geology for 
 several years to come. 
 
 E. N. Lowe, Director of Geological Survey, Jackson, Miss.: It 
 is an unrivaled text-book for a brief course in general geology. 
 
 rT> 34 West 33d Street, Hew York 
 
 & L/\J. 623 So. Wabash Avenue, Chicago 
 
Salisbury's Physiography. Advanced Course 
 
 By ROLLIN D. SALISBURY, Professor in the University of Chi- 
 cago. (American Science Series, Advanced Course.) xx-f-77o 
 pp. 8vo. $3.50. 
 
 While planned for courses in colleges and normal schools, it 
 presupposes no previous study of physical geography. 
 
 W. M. Davis, Harvard University: It is an important addition 
 to college literature, the first book of college grade on physiography. 
 The illustrations are for the most part remarkably successful and 
 appropriate. 
 
 Salisbury's Physiography for High Schools 
 
 By ROLLIN D. SALISBURY, Professor in the University of Chi- 
 cago. (American Science Series, Briefer Course.) viii-f53i 
 pp. Large I2mo. $1.50. 
 
 This is intended to cover the work in physical geography as 
 given in the first or second year of the high school. 
 
 W. H. Hawkes, Ann Arbor (Mich.*) High School: I am of the 
 opinion, after careful examination of the leading Physiographies 
 on the market, that Salisbury's Briefer Course of Physiography is 
 in the lead, in the method of treatment of the subject of Earth 
 Science, its clearness of presentation of detail, and especially in the 
 ways in which the subject matter is discussed in illustrative dia- 
 grams, 
 
 Salisbury's Elementary Physiography 
 
 By ROLLIN D. SALISBURY, Professor in the University of 
 Chicago. (American Science Series, Elementary Course.) 
 X ' + 359PP- 47 plates. Large i2mo. $1.30. 
 
 This is particularly well adapted to first year pupils and short enough 
 for a half-year course if need be. 
 
 C. L. Walton, Lake Visiu High School. Chicago: "It would seem 
 that nothing further could be done in the way of perfecting a text-book in 
 elementary physiography. As we are to have but five months of this study 
 hereafter, I think no other book is so well adapted to o-ir needs as this one." 
 
 HENRY HOT^T & CO 3 * West 33d Street. Hew York 
 
 injlXJVI ~\\J\^ 1 <X ^\J 623 So. Wabash Avenue, Chicago 
 
Elements of Geography 
 
 By ROLLIN D. SALISBURY, HARLAN H. BARROWS 
 and WALTER S. TOWER, of the Department of Geog- 
 raphy, University of Chicago. (American Science Series.) 
 ix+6i6 pp. 7 maps in color. I2mo. $1.50. 
 
 A course in geography which is adapted to courses in colleges, 
 normal schools and in the later years of the high school. It treats 
 physiographic processes and features briefly and develops at greater 
 length the relations of earth, air, and water to life and especially to 
 human affairs. 
 
 J. C. Branner, Stanford University: It is a remarkably well 
 written, well arranged, and well illustrated book, and is fairly burst- 
 ing with information of the most trustworthy kind, and references 
 to the best books and papers on special topics. 
 
 Modern Geography 
 
 The Effects of Physical Features on Living Things 
 
 By ROLLIN D. SALISBURY, HARLAN H. BARROWS, 
 and WALTER S. TOWER, of the Department of Geog- 
 raphy, University of Chicago. ix+4Q6 pp. 7 maps in color. 
 I2mo. $1.25. 
 
 This book is planned for use in the first or second year of the 
 high school. Representing the "new" geography which stresses 
 the geographic factor in human life, it lays a solid foundation of 
 physiography but vitalizes it by constantly showing that broad phy- 
 siographic principles underlie all commercial, industrial and econo- 
 mic activities of mankind. 
 
 Caroline C. Haney, Eastern High School, Detroit, Mich.: 
 What we need in our curriculum is the humanizing of the technical 
 so that the child may co-ordinate his everyday life and interests with 
 his school work. This has been most successfully accomplished in 
 Modern Geography. 
 
 Laboratory Manuals of Geology and 
 Physiography 
 
 By ROLLIN D. SALISBURY, Professor in the University of 
 Chicago, and ARTHUR C. TROWBRIDGE, Professor in 
 the State University of Iowa. I2mo. Paper, 25 cents each. 
 
 I. The Interpretarion of Topographic Maps. 64 pp. 
 
 For beginning courses in physiography. 
 II. Studies in Geology. 68 pp. 
 
 Based on topographic maps and folios for classes in 
 physiographic and structural geology. 
 III. Historical and Structural Geology. 76pp. 
 
 TTTJi\TT>V TTf>T r P XT PO 34 West 33d Street, New York 
 
 lH(lMt I tlULi 1 <X l^LJ. 623 So. Wabash Ave., Chicago 
 
UNIVERSITY OF CALIFORNIA LIBRARY 
 
 Los Angeles 
 This book is DUE on the last date stamped below. 
 
 MOV 9 1951 
 2 1 1954 
 
 NOV 
 
 1 1959 
 
 Form L9-50m-ll,'50 (2554)444 
 
 
 
 The RALPH D. RFED UBRARY 
 
 DEPARTMENT OF GEOLOGY 
 
 UNIVERSITY of CALIFORNIA 
 
 U AXOELES. CALIF.