^■; CORNELL UNIVERSITY LIBRARY FROM Cornell University Library The original of this book is in the Cornell University Library. There are no known copyright restrictions in the United States on the use of the text. http://www.archive.org/cletails/cu31924031505674 THE IGE AGE INr NORTH AMERICA AND ITS BEARINGS UPON THE- ANTIQCJITY OF MAN BY G.' PKBDERICK WRIGHT, D. D., LL. D., F. G. S. A. -Professor in Oherlin Theological Seminary Assistant on the United States Geological Survey Author of "Logic of Christian Evidences^'''' etc. WITH AN APPENDIX ON " THB PROBABLE CAUSE OF GI-AOIATION ' BY WARREN UPHAM, F. G. B. A. Assistant on tfie 'Geological Surveys of New Hampshire^ Minnesota and the United States WITH MANY NEW MAPS AND ILLUSTRATIONS NEW YORK D. APPLETON AND COMPANY LONDON ; CAXTON HOUSE, PATEKNOSTER SQUARE 1889 Copyright, 1889, By D. APPLETON AND COMPANY. ELISHA GRAY CHEVALIER DE L\ LEGION d'hONNEUR INVENTOR OF THE HARMONIC TELEGRAPH, THE TELErHONE AND THE TELAUTOGRAPH WHOSE INTELLIGENT INTEREST IN GLACIAL GEOLOGY AND WHOSE GENEROUS APPRECIATION OF MT WORK HAVE BEEN A CONSTANT INSPIRATION THIS VOLUME IS AFFECTIONATELY DEDICATED PREFACE, The present treatise is the outcome of special studies upon glacial plienomena begun in the summer of 1874, in the eastern part of Massachusetts, the results of which were published in a communication to the Boston Society of Natural History in December, 1876. These first studies pertained to the origin of the gravel-ridges described in this volume under the name of "kames." Fortunately, in the preparation of that paper, I was favored with an interview with Mr. Clarence King, who then gave me the information referred to in the following pages, concerning the terminal moraine south of New England, which has been so fruitful of suggestion to other investigators as well as to myself. Since that time the subject has never been out of mind, and my summer months have all been devoted, under favorable conditions, to the collection of field notes regarding it, and so it has seemed to others, as well as to myself, appropriate that I should endeavor to bring the facts within the reach of the general public. After having become, during the four following sea- sons, familiar with the glacial phenomena over the larger part of New England, I was invited by Professor Lesley to survey, in company with the late Professor H. Carvill Lewis, the boundary of the glaciated area across Pennsyl- vi . PREFACE. vania (our report constitutes Yol. Z of the Second Geologi- cal Survey of that State). The summers of 1882 and 1883 were spent under the auspices of the Western Reserve His- torical Society of Cleveland, Ohio (whose secretary, Judge C. C. Baldwin, had the sagacity to recognize, at that early time in the investigations, the historical bearing of the work), in continuing the survey across Ohio, Kentucky, and Indiana (see my report to that society, 1884, and an article in the "American Journal of Science," July, 1883). Dur- ing the summers of 1884 and 1885 I was employed, as a member of the United States Geological Survey, in tracing the boundary across Illinois, and in reviewing the field in Ohio and western Pennsylvania. The report of this work has not yet been made public, but permission to use the facts has been generously granted by the director of the survey. Major J. W. Powell. The summer of 1886 was spent in Washington Territory, and upon the Muir Glacier in Alaska. The two following seasons were occupied in further exploration of Ohio, Dakota, and other portions of the Northwest. Thus I have personally been over a large part of the field containing the wonderful array of facts of which I am now permitted to write. In the autumn of 1887 I was invited to give a course of lectures upon the Ice Age in North America before the Lowell Institute in Boston, and in the following year be- fore the Peabody Institute in Baltimore. For the informa- tion of the audiences who heard those courses of lectures it is proper to say that the present treatise incorporates all the facts then presented, though in a different form. The volume covers, however, a much wider field than the lect- ures, and is more ample in its treatment of all the topics. But it is not to be supposed that a single person can PBEFAOE. vii adequately survey so large a field. The writer is but one of many investigators who have been busily engaged for the past fifteen years (to say nothing of what had been previously accomplished) in collecting facts concerning the Glacial period in this country. My endeavor has been to make the present volume a pretty complete digest of all these investigations, and in carrying out that aim I have had the generous assistance of the great array of careful and eminent observers who have turned their attention to the subject. So far as possible I have, by their permission, given their results in their own language, and with due credit. I hereby take occasion to erpress my obligations for the help which they have, one and aU, so courteously rendered. The numerous maps accompanying the text have been compiled from the latest data, as indicated in the abundant foot-notes scattered throughout the volume. These render it unnecessary to make here any more specific acknowl- edgment of authorities. The volume is committed to the puhlic in the belief that it will meet a widely felt want. The accumulation of facts for the past decade has been so rapid, and from so many sources, that few persons have been able to keep themselves informed of the progress made. And so great has this progress been that we may now safely assume that future discussions will pertain mainly to the details of the subject. We now know, from actual observation, the limits and prominent characteristics of the glaciated area on this continent. The Glacial age of North America is no longer a theory, but a well-defined and established fact. It will become apparent also that, though the title of the book is the " Ice Age in North America," it is really viii PREFACE. a treatise oh the whole subject of the Glacial period ; for, with the vast field open for investigation on this continent and the amount of attention recently given to its explo- ration, ISTorth America is now by far the most favorable place from which to approach the study of ice-action and ice periods. The last chapters of the volume treat of man's relation to the Ice age on this continent ; and I need not disguise the fact that the bearing of the discoveries upon this ques- tion has all along given zest to my investigations. The facts with regard to this subject also are now so far in hand that they can be properly discussed in a treatise de- signed in part for the general public. "While presenting as fully as is necessary the evidence of man's occupancy of the continent during the great Ice age, and while accepting this as necessitating a considerable extension of man's antiquity as usually estimated, I have not felt called upon in the present discussion to say any- thing about the method of reconciling this fact with the chronology of the human race supposed to be given in the sacred Scriptures ; for I have elsewhere (in my " Studies in Science and Eeligion," W. F. Draper, Andover, 1882, and " Divine Authority of the Bible," Congregational Pub- lication Society, Boston, 1884) said all that it seems at present necessary for me to say upon this point. I will only remark here that I see no reason why these views should seriously disturb the rehgious faith of any believer in the inspiration of the Bible. At all events, it is incum- bent on us to welcome the truth, from whatever source it may come. G. Feedekick Weight. Oberlin, Ohio, Ap-il 15, 1889. OONTEI^TS. CHAPTER I. PAGE What is a Glacier? 1-12 The Appearance of Ice deceptive, 1 ; Discovery of its Eeal Nature, 2 ; Motion of a Glacier, 2 ; Effect of Friction, 3 ; Cause of the Motion, i ; Plasticity of Ice, 5 ; Eelation of Snow to Ice, 6 ; Structure of a Glacier, 7 ; Veins, t ; Fissures, 8 ; Superglacial Streams, 9 ; Lateral and Medial Mo- raines, 9 ; Ice-pillars, 10 ; Terminal Moraines, 10 ; Kames, 11 ; Glacial Scratches and Groovings, 12. CHAPTER II. Existing Glaciers on the Pacific Coast . . . 13-35 In Southern California, 13 ; in Northern California : Mount Shasta, 15 ; in-,Pregon, 19; in Washington Territory: Mount Taooma, 19; British Columbia and Alaska, 23; on the Stiokeen Eiver, 25; in Taku Inlet: Norris Glacier, 27 ; in Lynn Canal : Davidson Glacier, 27 ; north of Cross Sound, 29 ; on Mount St. Elias, 30 ; north of the Alaskan Peninsula, 32. CHAPTER III. A Month with the Muie Glacier ..... 36-66 Purposes of the Expedition, 36 ; Facilities for Observation, 87 ; Descrip- tion of Glacier Bay and its Surroundings, 37 ; Muir Inlet, 39 ; Dimensions of the Muir Glacier, 39 ; Various Characteristics, 43 ; Moraines, 45 ; Indi- rect Evidences of Motion, 46 ; Formation of Icebergs, 47 ; Subglacial Streams, 47 ; Direct Measurement of Velocity, 48 ; Ketreat of the Ice- front, 51 ; Former Extension of the Glacier, 55 ; a Buried Forest, 57 ; Kames and Kettle-holes, 62 ; Transportation and Waste by Water, 63 ; Temperature in August, 64 ; Flora of the Vicinity, 65. CHAPTER IV. Glaciers of Greenland 67-91 Extent of Greenland, 67 ; Nordenskiold's Expedition, 67 ; Amount of Coast Line already explored, 69 ; Number of First-class Glaciers in Dan- X CONTENTS. PAOB ish Greenland, 70 ; Movement of Ice in them, 70 ; Nunataks, 71 ; Former Extension of the Ice, 71 ; Helland's Observations, 72 ; Whymper's De- scription, 75 ; Explorations of Kane and Hayes, 78 ; Humboldt Glacier, 84 ; Glaciers on the Eastern Coast of Greenland, 89 ; Nansen's Expedi- tion, 90. OHAPTEE V. Glaoibes in Othbe Paets op the Woeld .... 92-107 In the Alps, 92 ; in Scandinavia, 94 ; in Spitzbergen, Franz-Josef Land, and Iceland, 95 ; in Asia, 95 ; in South America, 96 ; Darwin's Ac- count, 97 ; in New Zealand, 100 ; on the Antarctic Continent, 100 ; Ice- bergs of the Southern Ocean, 103 ; CroU's Inferences from the Size of these Bergs, 106. CHAPTER VI. Signs of Glaoiation 108-119 Introductory Eemarks, 108 ; Grooves and Scratches, 109 ; Sir Charles Lyell's Observations in Nova Scotia, 112 ; the Ground Moraine, or "Till," 115; Cause of its being unstratified, 116; Sifting Power of "Water, 117 ; Distribution of Bowlders, 118. CHAPTER VII. Boondaet of the Glaciated Aeea in Noeth America . 120-165 Confluent Character of the Ice-sheet, 120 ; Progress of Discovery, 120 ; Elements determining the Amount of Glacial Deposition, 121 ; Counter- acting Influence of Subglacial Streams, 122 ; Influences determining the Amount of Marginal Deposits, 123 ; south of New England, 123 ; Interior Marginal Deposits, 125 ; Long Island a Moraine, 126 ; Marginal Deposits across New Jersey, 126 ; Eelation of Kettle-holes to the Moraines, 129 ; Marginal Deposits in Eastern Pennsylvania, 130 ; Western New York and Pennsylvania, 137; through Ohio, 139; Extension of the Ice into Ken- tucky, 142 ; Course of the Boundary in Indiana and Illinois, 142 ; Bound- ary west of the Mississippi, 144; west of the Eocky Mountains, 148 ; An- cient Glaciers in Southern California, 151 ; Glacial Boundary north of Puget Sound, 155 ; near the Head-waters of the Yukon, 162 ; Summary of Facts regarding the Pacific Coast, 162. CHAPTER Vin. Depth of Ice dtjeing the Glacial Peeiod . . . 166-174 Means of estimating it, 166 ; Mountain Summits covered in New Eng- land, New York, and Pennsylvania, 166 ; Depth estimated from the Distance moved, 171 ; Slope of a Glacier, 173. CHAPTER IX. Terminal Moraines 175-195 Indefiniteness of the Term, 175 ; Prominence of the Moraine south of New England, 176 ; Details respecting the Kettle-holes in this Part of the Mo- OONTENTS. xi PA»£ raine, 177 ; Submerged Portions of the Moraine, 178 ; Moraines of the Mid- dle States, 179 ; President Chamberlin on the Moraines west of the Alle- .ghanies, 179 ; Gilbert on Moraines of the Maumee Valley, 179 ; The Ket- tle Moraine of Wisconsin, 183 ; Battle of the Glaciers in Minnesota and "Wisconsin, 184 ; Moraines in Dakota, 186 ; in Central British America, 189 ; Later Moraines in the White Mountains, 193 ; on the Sierra Nevada and Cascade Mountains, 194. CHAPTER X. Glacial Eeosion and Transportation .... 196-260 Erosive Action of Ice compared with that of Ennning Water, 196; Chemi- cal Action of Water, 199 ; Dagger of exaggerating the Erosive Action of Ice, 200 ; Glacial Erosion least near the Margin, 203 ; Analogy between the Glacial Front and Breakers in the Ocean, 204 ; Transportation of Bowlders on the Surface of a Glacier, 206 ; Details respecting, in Southern New England, 207 ; in Eiehmond, Mass., 209 ; in New Jersey and Pennsyl- vania, 211 ; in Ohio, 212; in Southern Indiana and Illinois, 213 ; in Iowa and Dakota, 213 ; in British America, 214; Elevation of Bowlders in the Ice, 216 ; Explanation of, 218 ; Erosion of Subglaoial Streams, 224 ; Ero- sion estimated by the Amount of Till, 227 ; Attempts at Direct Measure- ment of Glacial Erosion near the Delaware Water-Gap, 230 ; affected by Preglacial Disintegration, 231 ; Evidence of, near Western End of Lake Erie, 232 ; in the Sierra Nevada, 237 ; Summary, 249. CHAPTER XI. Deumlins ... 251- 26V Definition, 251 ; thosein the Vicinity of Boston enumerated, 251 ; Mr. Up- ham's Description, 252 ; Series of, in Bockingham County, N. H., and Es- sex County, Mass. , 254 ; Occurrence of, in the Interior of the Countiy, 255 ; Theoryof their Formation, 257; Irregularities of Glacial Erosion and Dep- -, osition, 264. CHAPTER XII. Preglacial Dbainagb 268-278 Length of Preglacial Time, 268 ; Extent of Preglacial Erosion, 269 ; the ~ Trough of the Ohio, preglacial, 270; Other Preglacial Valleys, 272; Pre- glacial Drainage of the Great Lakes, 274 ; Preglacial Drainage of the Up- per Alleghany Eiver into Lake Erie, 276. CHAPTER XIII. Drainage of the Glacial Period 279-296 Obstruction of Ice- barriers across the Eed Eiver of the North, the St. Law- rence, and the Mohawk, 279 ; Closing Floods of the Glacial Period, 280 ; Terraces produced by these Floods in the Trough of the Mississippi ; in the Minnesota, 281 ; in the Northern Tributaries of the Ohio, 283 ; in the Streams of Northern Pennsylvania, 285 ; Material composing the Gla- cial Terraces, 285 ; such Terraces absent from Streams wholly in the Un- xii CONTENTS. PAGE glaciated Eegion, 285 ; Terraces on the Ohio, 287 ; on Beaver Creek, Pa., 288 ; on the Delaware, 289 ; Kelation of Pot-holes in Grafton, N. H., to Glacial Drainage, 290; Similar Phenomena in Lackawanna County, Pa., 291 ; Eemarkable Evidence of Abnormal Glacial Drainage in Dakota, 292 ; Marginal Drainage in the Northwest, 294. CHAPTER XIV. Kames .... 297-314 In Andover, Mass., 297 ; Definition, 297 ; Geikie's Description, 297; Re- lation of Kames to Terminal Moraines, 299 ; Origin of, 300 ; indicate Lines of Temporary Glacial Drainage, 302 ; Lines of, in New England enumerated, 303 ; Question respecting, in the Connecticut Eiver Valley, 305 ; Possible Extent of Glacial Floods in this Valley, 806 ; Relation of Sandy Plains to Kames, 308 ; Existence of Kames foretold, 809 ; Over- wash Gravel limited in Amount, 310 ; Abnormal Relation of Kames to the Slope, 311 ; Summary, 313. OHAPTEE XV. Glacial Dams, Lakes, and Watebfalls .... 315-358 Rock Basins eroded by Glaciers, 315 ; Theory of the Great Lakes, 316 ; Two Classes of Glacial Dams, 319 ; Kettle-holes, 319 ; Relation of, to Peat-bogs and to Terminal Moraines, 320 ; Lakes formed by Permanent Obstruotion- of Preglaoial Channels, 322 ; the Formation of Waterfalls, 322 ; ^Tempo- rary Lakes formed by Ice Barriers,^ 323 ; Supposed Glacial Dam in the Ohio, at Cincinnati, 326 ; Evidence that the Ice crossed the Ohio, 327 ; Consequences of such an Obstruction, 328 ; Histoiy of its Discovery, 330 ; Theory discussed by the American Association for the Advancement of Science, 331 ; Theory confirmed by the Absence of Terraces in Brush Creek, Ohio, 332; by the Terrace at Bellevue, Pa., 335; Difficulty of Other Explanations, 336 ; the Occurrence of Vegetable Matter in the Terraces of the Monongahela support the Theory, 337 ; Teazes Valley, W. Va., explained by the Theory, 839 ; Various Other Phenomena ex- plained in a Similar Manner, 342; Objections considered, 343 ; Claypole on, 346 ; Glacial Dam across the Mohawk and the St. Lawrence, 351 ; the Lake Ridges of Ohio and New York, 354 ; Glacial Lake in the Red River Eegion of the North: Lake Agassiz, 356 ; Summary, 357. CHAPTER XVI. The Loess 359-371 Extent of the Deposit in China and North America, 359 ; Riohthofen's Theory of Deposition by Wind, 860 ; Difficulties of the Theory, 362 ; Characteristics of, 364 ; Changes of Level necessary, 365 ; Probable Con- nection with Glacial Floods, 367 ; Supplementary Theories, 369. CHAPTER XVII. Flight of Plants and Animals during the Glacial Peeiod 372-391 Peculiar Distribution of Plants in the North Temperate Zone, 372 ; Profes- sor Asa Gray's Solution of the Problem, 374 ; more Detailed Statement by CONTENTS. xiii PAGE Professor Gray, 375 ; Comparison of the Pacific ■with the Atlantic Forests, 376 ; Comparison of both with those of Japan, North China, and Europe, 377; Number of Preglacial Species now found m the Temperate Zone, 380 ; our Trees originated in the High Latitudes, 382 ; the Vicissitudes to which they have been subjected since the Approach of the Glacial Period, 383 ; Open Lines of Emigration in America, 384 ; Peculiar Influences upon the Pacific Coast, 385 ; Extinction of Animals in America by the Glacial Period, 386 ; Alpine Butterflies upon the White Mountains, 388. CHAPTER XVIII. EuEOPB DnEiNG THE Glacial Pbriod .... 893-404 Glaciated Area in Great Britain, 393 ; on the Continent, 395 ; Investiga- tions of Professor Lewis, 396; of Professor Salisbury, 402. CHAPTER XIX. The Cause of the Glaoial Period . ... 405-447 Eecent Astronomical Speculations, 405 ; the Combination of Conditions necessary to produce a Glacier, 405 ; Theories to account for the Glaoial Period, 407 ; Decrease of the Original Heat of the Planet, 407 ; Shifting of the Earth's Axis of Eotation, 407 ; Theory of Progressive Desiccation, 408 ; Effect of Changes in the Distribution of Land and Water, 408 ; Theory of Changes of Level, 410 ; Diflerent Temperatures of Space, 415 ; Mr. CroU's Theory ; the Ellipticity of the Earth's Orbit and the Pre- cession of the Equinoxes, 416 ; Possible Efifeot of these upon the Climate, 417 ; Cause of the Gulf Stream, 419 ; Causes controlling the Distribution of the Heat from the Sun, 427 ; Woeikoff's Objections to CroU's Theory, 429 ; Supposed Evidence of Former Glaoial Periods, 433 ; Deficiency of Evidence, 439 ; probably a Compound Cause, 441 ; produced by the Accu- mulation of Snow over Definite Centers, 441 ; Field for Mathematical In^ vestigation, 444 ; Summary, 445. CHAPTER XX. The Date oe the Glacial Period ..... 448-60-5 Uncertainty of Astronomical Calculations, 448 ; Defect in Lyell's Theory of Dnifoiinitarianism, 449 ; Post-glacial Erosion below Niagara Falls, 452 ; below Falls of St. Anthony, 458 ; in Ohio, 466 ; in Wisconsin and Minne- sota, 470 ; about Lake Michigan, 471 ; Post-glacial Deposition in Kettle- holes, 472 ; the Question of Two Glaoial Epochs, 475 ; Freshness of Inter- glacial Forest-beds, 482 ; Greater Oxidation of Material near the Glacial Boundary, 479 ; Growth of Peat, 484 ; Extent of Forest-beds, 493 ; For- mer Expansion of Lakes Bonneville and Lahontan, 496 ; Eecentness of these Lakes, 499 ; Length of the Glacial Period, 501 ; Summary, 504. CHAPTER XXI. Man and the Glacial Period 506-550 Prominence given by Lyell to the Subject, 506 ; Artificiality of the Imple- ments, 507 ; Professor Haynes on, 509 ; Genuineness of, 512 ; Discoveries xiv GONTENTB. PAGE of Boucher de Perthes in France, 518 ; of other Investigators in England, 514 ; of Dr. Abbott in New Jersey, 515 ; Nature of the Gravel at Trenton, N. J., 520; Mode of Deposition, 523; Series of Events in the Delaware Valley, 526 ; Discoveries in Ohio foretold, 528 ; Discoveries by Dr. Metz in Ohio, 530 ; Deposit at Madisonville, Ohio, described, 531 ; Oressou's Discoveries at Medora, Ind., 533; Winchell's Discoveries in Morrison County, Minn., 537; Miss Babbitt's Discoveries at Little falls, Minn., 538 ; Upham's Discussion of the Deposits in Minnesota, 538. CHAPTER XXII. Man and the Glacial Pbeiod (continued) . . . SSl-ST'l Cresson's Discoveries at Claymont, Del., 551; Prehistoric Development in the Delaware Valley, 556 ; Implements discovered by Aughey and McGee, 558 ; Whitney's Discoveries in California, 558 ; Le Conte on the Quaternary Deposits of California, 559 ; Whitney's Evidence in Detail, 562 ; Criticism on Whitney, 564 ; Conclusion, 567. APPENDIX A. — Peobablb Causes of Glaciation .... 573-595 Theory stated, 573 ; Molten Condition of the Earth's Interior, 574 ; Weight of Ice a Possible Cause of Subsidence, 576 ; Preglacial Elevation of the Continent, 577 ; Changes of Sea-Level during the Glacial Period, 579 ; Amount of Water looked up in the Ice of the Glacial Period, 579 ; Subsid- ence toward the close of the Glacial Period, 580 ; Changes of Level in Europe, 582 ; Eecent Elevation of the Isthmus of Panama, 584 ; Objections to tbe Theory of Croll and Geikie, 585 ; Mountain-building during the Quaternary Epoch, 586 ; Meteorological Causes co-operating, 587 ; Inter- glacial Periods, 588 ; the Lack of Glaciers in Asia, 589 ; Rate of the Earth's Contraction, 590 ; Eapidity of Changes during the Quaternary Epoch, 591 ; Former Periods of Glaciation, 592 ; Eecentness of the Glacial Epoch, 593 ; Summary, 594. B. — Chalmbes on the Glaciation of Eastern Canada 596-599 INDEX . . 601-622 LIST OF ILLUSTRATIONS. p[G, PAGE Side view of the front of Muir Glacier from the surface. Frontispiece. Map of Alaska Facing 32 Map showing the glacial geology of the United States . . Facing ITS Map showing the preglacial drainage of the Great Lakes . Facing 2*79 1. Differential motion of ice ... ... 2 2, 3. Differential motion of ice .... 3 4. Plasticity of ice . . .6 5. Fissures and seracs ... 8 6. 7. Marginal fissures and veins .8 8. Veined structure at the junction of two branches . . .9 9. Mode of formation of ice-pillars . ... . .10 10. Moraines of the Mer de Glace .10 11. Glacial scorings ... 12 12. Mount Lyell, California. (Russell.) . ... 14 13. Mount Shasta, California. (Russell.) 17 14. Mount Tacoma, Washington Territory, looking westward. (Charles S. Fee.) 20 15. Mount Tacoma, Washington Territory, looking eastward. (Charles S. Fee.) ... 21 16. Glacier Station, British Columbia. (Canadian Pacific Railroad.) . . 24 17. Map of southeastern Alaska . . .... 26 18. Norris Glacier, Alaska (Partridge.) . . . 28 19. Davidson Glacier, Alaska . . 29 20. Map of Glacier Bay, Alaska . . .40 21. Front of Muir Glacier, side view. (Charles S. Fee.) . . 42 22. Map of Muir Inlet, Alaska .49 23. Surface of Muir Glacier. (Partridge.) . . .52 24. Formation of kettle-hole, Alaska 54 25. Buried forest. Glacier Bay, Alaska .... . . 68 26. The same 59 27. Central part of front of Muir Glacier ....... 61 28. Ice-pillars. (Russell.). ..,,..,,. 66 XVI LIST OF ILLUSTRATIONS. FIG. PAGE 29. Map of Greenland .... . . .68 30. Map of Frederikshaab Glacier, Greenland. (Dana.) . 73 31. A Greenland glacier and icebergs .76 32. Morteratsch Glacier, Switzerland .93 33. Svartisen Glacier, Norway. (Warner.) 94 34. Iceberg ... . . . . . . 104 35. Floating iceberg .107 36. Scratched stone from till of Boston, Mass. .... . 109 37. Glacial striae, Amherst, Ohio. (Chamberlin.) . . . .110 38. Cut in till, Hamilton, Ohio . 114 39. Cut in till at Darrtown, Ohio 116 40. Glacial map of southern New England . . . . .124 41. Glacial map of New Jersey .127 42. Glacial map of Pennsylvania and southern New York . . . 131 43. Glaciated pebble, Pennsylvania 132 44. The same, side view . . . ... 133 45. Glacial map of Ohio . ... . . . 140 46. Glacial map of southern Indiana ... ... 143 47. Glacial map of southern Illinois 145 48. Section of till In Seattle, Washington Territory \SS 49. Section of modified drift at Point Wilson, Washington Territory . .154 50. Glacial groovings, Victoria, British Columbia 156 51. Glacial map of North America .... . . 163 52. Depth of ice and erosion in eastern Pennsylvania. (Lesley.) . 168 63. Map of kettle-holes near Wood's Holl, Mass. (Koons.) . . . 177 64. View of kettle-moraine. Eagle, Wis. (Chamberlin.) . . 182 65. Map of the Missouri coteau. (Todd.) . . . . 188 56. Glacial map of Central British America. (Dawson.) . . . .190 67. View of the Missouri coteau, British America. (Dawson.) . . . 191 58. Canon of the Colorado. (Newberry.) 196 59. Embossed floor of an ancient glacier, Colorado. (Hayden.) . . . 197 60. facial bowlder, Gilsum, N. H. (Hitchcock.) . . . 206 61. Glaoialed^gebble, Indiana . 222 62. Reverse side of the same 223 63. Ideal section, showing distribution of till .... . 227 64. Ideal section showing subaerial disintegration. (Chamberlin.) . .231 65. Glacial grooves. South Bass Island, Lake Erie 283 66. Tortuous glacial grooves, Kelly's Island, Lake Erie. (Chamberlin.) . 234 67. Section of glacial furrows, Kelly's Island, Lake Erie . . . 237 68. Full view of the same. (Younglove.) . . . . . 238 69. Glacial furrows. South Bass Island, Lake Erie 240 70. Glacial furrows, Gibraltar Island, Lake Erie 242 71. Cross striiB, Middle Bass Island, Lake Erie .... .246 72. Map of drumlins near Boston. (Davis.) .... .251 73. View of Corey's Hill, Brookline, Mass. A typical drumlin. (Davis.) . 252 LI8T OF ILLUSTRATIONS. xvii PIQ. 74. "75. 76. 11. 78. •79. 80. 81. 82. 83. 84. 86. 86. 87. 88. 89. 90. 91. 92. 93. 94. 95. 96. 97. 98. 99. 100. 101. 102. 103. 104. 105. 106. 107. 108. 109. 110. 111. 112. 113. 114. 115. 116. 117. 118. Outline of drumlius in Boston Harbor. (Davis.) Map of drumlins in northeastern Masaachiisetts. (Davis.) Outline of drumlins, central New Yorli. (Davis.) Drumlins in Wisconsin. (Chamberlin.) Drumlins in Goffstown, N. H. (Hitchcock.) PAGE . 253 . 254 266 . 256 . 257 Drumlius in Ireland. (Kinnahan and Close.) 261 Section of preglacial valley of the Cuyahoga River, Ohio. (Claypole.) 274 Glacial terrace, Granville, Ohio .... . . 284 Maps of kames in eastern Massachusetts .... . 298 Section of kame, Dover, N. H. (tTpham.) . . . 299 Sections of kame, Bennington Station, N. H. (Upham.) . . 300 Yiew of kames, Hingham, Mass. (Bouv6.) . .301 Map of the kames of Maine. (Stone.) . . . 304 Section of buried kame, Hanover, N. H. (Upham.) . . 305 Section of kame, Hanover, N. H. (Upham.) ... . 306 Buried kame, Stroudsburg, Pa. (Lewis and Wright.) . 308 Canadian bowlder, Boone county, Ky. . . . 328 Map showing the effect of the glacial dam at Cincinnati. (Claypole) . 329 Map of Paint Creek and Beech Flats, Ohio 333 Section of a deposit, Teazes Valley, W. Va. . . 340 Split Rock, Boone county, Ky. . . ...» . 345 Glacial map of Lake Cuyahoga. (Claypole.) 348 Section of the lake ridges near Sandusky, Ohio 350 Map showing glacial lakelets in northern Ohio. (Claypole.) . . 352 Map of glacial lake, Erie-Ontario. (Claypole.) . . 365 Stratified loess, Nebraska. (Chamberlin.) . . 363 Polar projections .... . . . 373 Glacial map of North America and Europe 392 Glacial map of Europe . 403 Diagram showing eccentricity of the earth's orbit . . 416 Map showing Atlantic Ocean currents 420 Map of July isobars and prevailing winds . . ... 421 Map of January isobars and prevailing winds . ... 424 Bird's-eye view of Niagara River. (Pohlman.) . . 451 Section of strata along the Niagara gorge . . 453 Map of the Niagara River below the Falls . . . 454 Map showing the recession of the Horseshoe Falls. (Pohlman.) . . 457 Map showing the recession of the Falls of St. Anthony . 461 Ideal view of an unglaciated country. (Chamberlin.) . 468 Ideal view of glaciated country. (Chamberlin.) . . 469 Section of kettle-hole, Andover, Mass. ... . 472 Perpendicular section of till containing wood, Oxford, Ohio . 477 Section of till overlying peat, Germantown, Ohio . . 483 Section of kettle-hole, Freehold, Pa .488 Section in till containing wood, Darrtown, Butler county, Ohio . 489 xviii LIST OF ILLUSTRATIONS. Fia. PAGE 119. Map of the ancient lakes Bonneville and Lahontan. (Le Conte.) . 497 120. Devil's Lake, Wis. (Chamberlin.) 605 121. Collection of palaeolithic implements . . . . . 506 122. Reverse side of the same . . 507 123. Argillite implement, Trenton, N. J. (10,985.) (Putnam.). . .516 124. Side view of the same. ■ (Putnam.) . . ... 517 125. Argillite implement, Trenton, N. J. (11,286.) (Putnam.). . .518 126. Black chert implement, Trenton, N. J. (10,986.) (Putnam.) . . 519 127. Section of Trenton gravel, N. J. (Abbott.) ... .621 128. Ideal section of the Delaware River Valley, Trenton, N. J. (Lewis) . 522 129. Black chert implement, Madisonville, Ohio. (40,970.) (Putnam.) . 530 130. Map of the vicinity of Madisonville, Ohio 631 131. Gray flint implement, Medora, Ind. (46,145.) (Putnam.) . . 534 132. Side view of the same. (Putnam.) . . .... 535 133. Section of gravel at Medora, Ind. ... . 536 134. Chert implements, Morrison county, Minn. (Winchell.) . . 537 135. Quartz implement. Little Falls, Minn. (31,323.) (Putnam.) 541 136. Quartz implements, Little Falls, Minn. (31,316.) (Putnam.) . . 542 137. Ideal section. Little Falls, Minn. (Upham.) . . . 544 138. Map of moraines in Minnesota. (Upham.) 646 139. Baltimore and Ohio Railroad cut, Claymont, Del. (Cresson.) . . 552 140. Nearer view of the same 664 141. Argillite implements from the preceding cut. (45,726.) (Putnam.) . 556 142. Lava-stream cut through by rivers in California. (Le Conte.) . . 559 143. Section across Table Mountain, Cal. (Le Conte.) . . . 659 THE ICE AGE IN NORTH AMERICA. CHAPTER I. WHAT IS A GLACIEE? To the ordinary man of science, water is a mineral and ice a rock ; but to the glacialist both are fluids. The appar- ent solidity of ice is an illusion due to the dullness of our senses. The reason why its viscous or semi-fluid character remained unsuspected until a comparatively recent period is due to the fact that the ordinary movement of accessible glaciers was so slow that we could not by observation readily note their rate of progress. The difference between water and other substances is most noticeable in the phenomena connected with solidifi- cation and fusing. Lead melts at 612° Fahr. above zero ; sulphur, at 226° ; water, at 32° ; while mercury becomes liquid at 39° below zero, and some other substances at even lower temperatures. Thus, with reference to its fusing- point, water appears toward the middle of the scale. If, like the fabled salamander, man were able to endure intense degrees of beat, he might, very likely, sustain relations to iron similar to those he now sustains to water. He might then bathe with pleasure in a molten flood, and venture on the thin crust of a glowing mass of metal. The suddenness with which water passes from the solid to the liquid state, and the amount of heat absorbed in tbe process of fusion, involve many important consequences. Down to the freezing-point water may be made to part with its heat by gradual stages, but in the act of freezing it sud- 2 THE ICE AGE IN NORTH AMERICA. denlj gives out an enormous amount of heat ; on the con- trary, when ice melts, a corresponding amount of heat is absorbed in accomplishing the result. To melt a cubic foot of ice requires as much heat as to raise a cubic foot of water 80° C. or 144° Fahr. For our knowledge of the nature of the movements tak- ing place in glaciers, we are largely indebted to the investi- gations of Louis Agassiz and Professor Forbes between the years 1840 and 1842, and later to more detailed investigations of Professor Tyndall and other physicists. The mode of measurement with all these investigators was essentially the same. Stakes were driven across a glacier in a line at right angles to the direction of the movement ; and, by means of a theodolite, accurate notations were taken, from hour to hour and day to day, of any changes in the relative position of the points where the stakes were driven. The uniform re- sult of these observations was that the line of stakes began immediately to curve slowly down near the middle, showing that the motion on the surface was greater near the middle than on the sides. This curve continued to increase as long as the stakes remained standing. Professor Tyndall's observations show also that the most rapid line of motioii d, e~f, g, represent on the surface of a fflacier is not exactly stakes driven across the " ^^^.y^u^j r,™S"angies%oTti^nne' ™ *^® middle ; but that, wherever there S*',??"™!?', represent ^^ ''' ^^^^ ^" *^® glacial currcut, the more sabse^M^^rge. *' " ^apid ^ movement is uniformly on the con- vex side of the channel, so that the curve of the line of most rapid motion is more tortuous than that of the main channel. This conforms to the facts concerning the movement of water in a crooked river-bed, and illustrates again the analogy between the movement of ice and that of water. The most rapid motion observed by Tyndall, in the sum- mer time, in the center of one of the largest of the Alpine a. z c ,€ e / 3 --0---0-- O- a \ 'o— o--^ -.0. l'-o...o.-'- c „• e' Fig, 1.— The letters a, b, WHAT IS A GLACIER f 3 glaciers, was thirty-seven inches per day. Near the sides of the glacier, however, the movement was reduced to two or three inches. The rate of motion during the winter was only about one half that during the summer. A further resemblance of the motion of a glacier to that of a river appears in the fact that the ice near the top moves faster than that near the bottom. At a point in the Mer de Glace where the side of the gla- cier is exposed, presenting a wall of ice about one hundred and fifty feet in height, Professor Tyndall drove three stakes ; one at the summit of the ice, another thirty-five feet from the bottom, and another four feet from the bottom. Upon examination of them, at the end of twenty-four hours, it appeared that, while the top stake had moved forward six inches, the middle one had moved but four and a half inches, and the bottom stake but two and two thirds of an inch. In all these experiments the influence of friction is clearly visible. The ice of the glacier is retarded by the friction of the sides and bottom of Fie, Fig. Z.—a, 6, c, are stakes driven in the vertical wall of the side of a glacier ; a', V, c', are the points occupied at a subsequent date. •The continu- ous lines define the valley occupied by the glacier. The dotted line with the arrow - heads indi- cates the line of most rapid motion in the ice, showing its more sinuous course. the channel through which it moves, so that the most rapid motion is upon the surface, near the middle, the part farthest removed from this retarding influ- ence. A little attention to this last principle will prepare the mind for crediting the observations which more recently have been reported from the large glaciers in Greenland and Alaska, showing a motion fifteen or twenty times that of the Alpine glaciers. As the cross-section of a glacier is in- 4 THE ICE AGE IN NORTH AMERICA. creased, the relative influence of friction in retarding the motion is rapidly diminished. The friction on the sides of a glacier two miles wide is no greater than that upon one a quarter of a mile in width, though the cross-section is eight times as large. A cross-section of the Mer de Glace at Les Moulins is estimated to be one hundred and ninety thousand square yards ; whereas a cross-section of the Muir Glacier, in Alaska, a mile above its mouth, is upward of one million square yards. Though observation shows that ice actually moves as if it were a fluid, the scientific imagination is tasked to the utmost to conceive how such motion can be consistent with other manifest qualities of the material ; for in many conditions ice seems as brittle as glass and as inelastic as granite. The mystery is probably solved, so far as such questions are ever solved, by attention to the facts already referred to concern- ing the behavior of ice at its melting-point. When ice passes into water, an immense amount of heat is absorbed in the process, which yet does not produce any effect upon the thermometer. If a hole be bored in the surface of a melting glacier, and a thermometer inserted, it will stand at 32° Fahr. If the same thermometer be inserted in'the subglacial stream issuing from the icefront, it will stand at the same point. Yet the absolute difference between the heat contained in the particles of ice, and that contained in the particles of water, is 144° Fahr. — so much heat being occupied in keep- ing the substance in a liquid form. Ice is also transparent to the rays of heat as it is to the rays of light. Scoresby amused himself, in the arctic latitudes, by making lenses of ice with which to concentrate the sun's rays and set com. bustible substances on fire. The fusing-point of ice is also modified by pressure. Under pressure the freezing-point of water may be lowered two or three degrees ; but upon the removal of the pressure, the water will instantly become solid. This has been demon- strated in various ways. M. Boussingault, for example, filled a hollow steel cylinder with water, having a bullet loose with- WRAT JS A GLACIER? 5 in it, and plugged tlie aperture up. He then subjected the cylinder to intense cold till the whole was two or three de- grees below the freezing-point of water. But that the water remained liquid was evident from the fact that, upon shaking the cylinder, the bullet inside rattled about as at higher tem- peratures ; while, upon removing the plug so as to relieve the pressure, the whole was instantly converted into solid ice. Various similar experiments have been made in which, upon removal of the plug, the water ejected from the aperture by the expansive power of the cooling water within the cylinder immediately freezes, and forms a projecting column of ice several inches in length. It was at first thought that this projecting column illustrated the plasticity of ice ; but it is now pretty certain that it illustrates, rather, the curious effect of pressure upon the freezing-point of water. The capacity of water at the freezing-point to transform itself, under varying degrees of pressure, from the solid to the liquid state, and vice versa, is illustrated by another ex- periment, ascribed by Professor Tyndall to Mr. Bottomley. A copper wire was looped over a bar of ice about four inches square, and a weight of twelve or thirteen pounds was sus- pended from it. The pressure under the wire caused the ice in immediate contact with it to melt ; but, as the resulting water escaped around the wire, and was relieved from press- ure, it immediately froze, and cemented together again the walls of ice above the wire. In half an hour the wire Jiad cut completely through the bar of ice, and yet the whole breach above it was repaired, and the bar was intact. This capacity of fragments of ice, when near the melting- point, to freeze together when their faces are joined, can be readily observed in a variety of experiments. When two pieces of ice in a basin of warm water are brought together they will immediately adhere. If a cake of ice whose tem- perature is near the melting-point be placed in a mold and subjected to pressure, the first result is to break it into pieces ; but, on contifuuing the pressure, the particles reunite and freeze together into a shape corresponding to that of the THE ICE AGE IN NORTH AMERIOA. mold. This capacity of ice, when near the melting-point, to undergo disintegration, and then to become suddenly re- T^=>r i\/r Pig. a.— a, B, (7, molds ; o, c, e, original forms of the ice ; b, d, f, the forras into which they were molded. congealed, is probably that by which it simulates in its mo- tion the properties of ordinary fluids, while at the same time retaining other properties connecting it with the most brittle of substances. It is thought, by Mr. Croll and others, that when heat passes through a stratum of ice, as it is known to do, it involves a process of transference from one particle of ice to another, in which there are successive melting and freezing of the particles in the progress of the heat, and that finally the mole- cule of ice upon the opposite side, in becoming recongealed, delivers up the unit of heat which had entered the stratum from the other side. But, whatever be the explanation of the process, the facts remain that ice behaves in many re- spects like a fluid, and, on application of pressure, slowly ad- justs itself to its bed or mold in obedience to the force ap- plied, and, if time enough is given, moves wherever a fluid would find its way. Ice jo plastic under pressure and brittle under tension. Snow is one form of ice, and, as every school-boy who makes a snow-ball knows, can by a moderate degree of press- ure be made into compact ice. The reason why snow is white, and ice is blue, is that snow is pulverized, while in ice the particles are brought into closer contact, and the inclosed air is expelled, so that the real color of the substance WffJT IS A GLAGIERf 7 is brought out. Tlie powder of almost any substance differs in color from the compact mass. Glacial ice is compressed snow, and originates wherever the snow-fall is largely in ex- cess of the melting power of the sun and warm currents of air. Any one can observe how amch more compact old snow is than new, and how, under pressure, the lower strata in a snow-bank become in a single season almost like ice. Hence it is easy to see what must be the result where the annual snow-fall is never wholly melted away. In such regions the ice would accumulate without limit, were it not for its semi- fluid character, which permits it to flow off, in lines of least resistance, to lower levels and toward warmer climes. In structure glacial ice is characterized by both veins and fissures — two phenomena, which are produced by opposite causes — the first by pressure, and the second by tension. Glacial ice ordinarily presents a vemed structure. Instead of being homogeneous, it consists of alternate bands of light- colored and blue ice. These bands do not, however, lie in a horizontal position, but are often vertical. Sometimes they run parallel with the movement of the glacier, and sometimes at right angles to the motion ; while, at other times, they are arranged at an angle of forty-five degrees, pointing down the line of motion. From close examination it appears that the veins are always at right angles to the line of greatest pressure. For example, where two branches of a glacier join, and press together from the sides, longitudinal veins are produced below the point of junction. And again, where ice has de- scended a declivity, and is advancing upon a less inclined plane, the increased pressure necessary to push the mass along produces bands at right angles to the line of motion ; thus demonstrating the connection of veins with pressure. The. theory is, that the blue veins in the ice are those from which pressure has expelled the particles of air, thus making it more compact, and giving it its blue color. As already re- marked, snow is white because of the abundant particles of air inclosed within it. Under pressure it can be transformed into blue ice, corresponding to the blue veins alluded to. THE ICE AGE IN NORTH AMERICA. An active glacier is also characterized hj fissures. When- ever the ice-stream reaches a point where its slope is increased even by a very small amount (a change in inclination of two degrees being sufficient), the ice instead of moving in a con- tinuous stream, forms crevasses across the current, which gradually enlarge at the top, until they present a series of long chasms, very difficult for the explorer to traverse. Where there is considerable irregularity in the bottom, and the increased slope extends for some distance, thesie crevasses become very complicated, and the surface presents an ex- panse of towers and domes and pinnacles of ice, often of fan- tastic appearance; but at the bottom these masses are still joined, and on coming down to a i||ft||»_^s^ gentler slope they close up again at the surface for their onward Fig. 5. — c, c, show fissures and seracs where the glacier moves down the steeper portion of its incline \ «, 5, show the vertical structure produced by pressure on the gentler slopes.. march. In addition to the crevasses or fissures, produced by the tension where the ice-stream passes over a steeper incline, a set of marginal fissures extend from the sides of the glacier toward the center, but pointing upward at an angle of about forty-five degrees. These, too, appear to be the result of tension. The motion of the ice in the center, being \ more rapid than that to- ward the sides, produces a line of tension, or strain, ex- tending from the center di- agonally downward toward Fios. 6, 7" illustrate the formation of margm- the sides at an angle of al Assures and veins. forty-five degrees. The pressure upon these masses of ice, whose central point is being wheeled downward by the difEer- FiG. 6. Fig. 7. WHAT JS A GLACIER ? 9 ential motion, produces also a veined structure in the masses themselves, at right angles to these marginal fissures. The surface of a glacier presents many interesting phe- nomena. When the ice-stream is of sufficient size, the sur- face is covered with a network of small streams of water, flowing through blue channels of ice sometimes many yards in depth and width. But these are destined eventually to encounter some crevasse, where a circular shaft, or moulin, as it is called, is formed, opening a way to a subglacial chan- nel, into which the streams plunge Fia. 8. -illustrates the forma- . , 1 1 T T . 1 1,1 tion ot veined structure by With a loud roar, and the accumulated pressure at the junction of », , , , -. , two branches. waters may often be heard rusmng onward hundreds of feet below the surface. During the melting of a glacier, also, in the summer season, the surface of the ice is frequently dotted with bowl-shaped depressions, from one or two inches to many feet in depth, and filled with beautiful clear water. The cause of this can not well be conjectured. In Greenland, Nordenskiold attributed the initial melting to accumulations of meteoric dust which he named kryokonite. Glaciers in mountainous regions are also characterized by lateral and medial moraines. Where the ice stream passes by a mountain-peak, the falling rocks and the avalanches started by streams of water, form along the edge of the gla- cier a continuous line of dShris, which is carried forward by the moving ice, and constitutes what is called a lateral mo- raine. If there be a current of ice on each side of the mount- ain-peak, two of the lateral moraines will become joined bo- low the mountain, and will form what is called a medial moi-aine, which will be carried along the back of the ice as far as the motion continues. As the ice wastes away toward the front, several medial moraines sometimes coalesce. This, as will be seen, is finely shown in some glaciers of Alaska. A medial moraine, when of sufficient thickness, protects 10 THE ICE AGE IN NORTH AMERICA. the ice underneath it from melting ; so that the moraine will often appear to be much larger than it really is : what seems to be a ridge of earthy material being in reality a long ridge of ice, thinly covered with earthy debris, sliding down J,j the slanting sides as the ice slowly wastes I away. Large blocks of stone in the same \. manner protect the ice from melting un- Fi^9.-Mode of formation dcmeath, and are found standing on pe- of ice-piiiais. destals of ice, where the general surface has been lowered sometimes several feet. An interesting feature of these blocks is that when the pedestal fails, the block uniformly falls to- ward the sun, since that is the side on which the melting has proceeded most rapidly. All the material brought down upon the surface of the glacier in the medial moraines is deposited at the front, forming a terminal mo- raine, which will vary in size according to the abundance of material transported by the ice, and in proportion to the length of time during which the front rests at a particular point. But, ordinarily, for a consid- erable distance this mo- raine material near the front will rest upon ex- tonaiVo maocao rv-f 100 ^™' 10.— Mer de Glace. The parallel lines in the leubive mdbbeb ox ice middle are medial moraines. The main ice- wTiioli r\Tt\xT ulmirlTT rviolf Stream on the right pushes the Others to the Wall, WniCn Oniy SlOWiy melt and divides the terminal moraine above g. WHAT IS A OLAOIER? 11 away. It is largely owing to this that a true terminal moraine is made up of knolls and bowl-shaped depressions called Icettle-holes, and of short tortuous ridges of bowlders and gravel. Another result connected with the decay of a glacier is the production of Tcames — this being the Scotch word for sharp, narrow ridges of gravel, corresponding to what are called osars in Sweden and eshers in Ireland. The trend of these ridges is the same as that of the motion of the glacier, and is at right angles to the terminal moraine. Their for- mation can be witnessed on a large scale near the front of the Muir Glacier in Alaska. In certain localities a great amount of sand, gravel, and bowlders becomes spread out over the snrface of the ice at a considerable elevation. Through some changes in the subglacial drainage a stream wears a long tunnel in the ice underneath this deposit, which at length proceeds so far that the roof caves in, and the earthy debris is gradually precipitated to the bottom of the tunnel, thus forming one class of kames. In other places, evidently, water-worn channels in the ice have been silted up by the stream, and then the line of drainage changed, so that, when the supporting walls of ice melted away, another class of kames, with what is called " anticli- nal " stratification, is produced. It should be mentioned also that, after the analogy of a river, a glacier shoves sand and gravel and bowlders under- neath it along its bed ; from which it can easily be seen that a glacier is a powerful eroding agency, rasping down the surface over which it moves, and by the firm grasp in which it holds the sand, gravel, and bowlders^^undemeath it, pro- ducing grooves and scratches and polished surfaces on the rocks below, while these stones themselves will in turn be scratched and polished in a peculiar manner. Wherever the glaciers have receded, so that their bed can be examined, these phenomena, which we reason from the nature of the ease must have been produced, are found actually to occur, and a terminal moraine is sure to contain many pebbles and 12 rSE ICE AGE IN NORTH AMERICA. bowlders bearing marks of the peculiar attrition to which they have been subjected in their motion underneath the ice. The rocks brought along upon the surface of the gla- cier of course are not thus striated, and ordinarily the mate- FiG. 11. — Glacial scorings (after Agassiz). rial of the kames has been so much rolled by water that if the pebbles ever were scratched, the marks have been erased. With this brief account of the physical characteristics of ice, and of the effects produced by its movement in a gla- cier, we are prepared to enter more understandingly upon a survey of the actual facts relating to the past and present extent of the ice-fields over the northern part of North America. Keserving the discussion of theories concerning the cause and date of the glacial period to the latter part of the treatise, we will first consider the facts concerning the glaciers still existing in America, and then briefly, by way of comparison, those concerning glaciers in other portions of the world ; after which we will present in considerable detail the more recent discoveries concerning the extension and work of the great American ice-sheet during the so- called Glacial period. CHAPTEE II. GLACIERS ON THE PACIFIC COAST. Notwithstanding the great height of the Rocky Mount- ains, they are at the present time almost devoid of living glaciers. This lack is doubtless caused by the dryness of the atmosphere, the winds from the Pacific having already, before reaching the interior, yielded their moisture to the solicitations of the lofty peaks of the Sierra E^evada and the Cascade Range. Still, a few small glaciers are found among the summits of the Wind River Mountains of Wyoming, and near the sources of Flathead River in Montana. The country in this vicinity, however, is still so imperfectly ex- plored that many glaciers may have escaped observation.* A vride arid space, of which few who have not traversed the region can have any conception, separates the Rocky Mountains from the Sierra Nevada nearer the Pacific coast. This latter range of mountains is, in some respects, favorably situated for the production of glaciers, since the peaks are lofty, rising in many places upward of 14,000 feet, and there is abundance of snow-fall. Ordinarily, however, there is not breadth enough to the summit of the range to furnish ade- quate snow-fields for the production of first-class glaciers. The most southern collection of glaciers in the Sierra Nevada is found near the thirty-seventh parallel, a little east of the Yosemite Valley, in Tuolumne and Mono coun- ties, California. Here is a remarkable cluster of mount- * See " The Existing Glaciers of the United States," by Israel C. Russell,- in the " Fifth Annual Report of the United States Geological Survey,'' pp. 344- U1. GLACIERS OF THE PAOIFIG COAST, 15 ain-peaks rising upward of 14,000 feet above the sea, and with breadth enough to support numerous snow-iields and glaciers. JSTo less than sixteen glaciers of small size have been noted among these summits, of which those on Mount Dana, Mount Lyell, and in Parker Creek are the principal. None of them, however, are of great size, being in no case over a mile in length, and none of them descending much below the 11,000-foot line.* The continuation of the Sierra Nevada Mountains to the north of California is called the Cascade Range, and is largely composed of volcanic rocks. It is on Mount Shasta, in the extreme northern portion of California, that we next find glaciers of any considerable size. But from this point on, glaciers multiply and continue, in ever-increasing glory, through the Coast Eange of British Columbia and southern Alaska to the islands of the Aleutian Archipelago. The glaciers upon Mount Shasta were iirst described by Mr. Clarence King in 1870. Previous explorers had as- cended the mountain upon the southern side, and reported it as free from glaciers, which are all upon the northern side. The most recent and detailed account of the glaciers on this mountain has been furnished by Mr. Gilbert Thompson, of the United States Geological Survey .f According to Thomp- son, Mount Shasta is a volcanic peak whose altitude above the sea is 14,511 feet. " It stands alone and has no connec- tion with neighboring mountains, none of which within a radius of forty miles attain two thirds its height." The mountain is a conspicuous object to attract attention for over a hundred miles. Five glaciers have been explored upon its northern flank, none of them, however, reaching lower than the 8,000-foot level, and none being more than three miles in length. The lower part of these glaciers is covered with vast quantities of earthy debris, so that it is difficult to tell where the ice-field now ends. It was from these half-buried edges * Russell, " Existing Glaciers," pp. 310-32'?. \ Ibid., pp. 332-334. 16 THE ICE AGE IN NORTH AMERICA. of the ice-front on the flanks of Shasta that Mr. King drew the analogies which first solved the problem of the irregular gravel deposits forming the so-called kames and kettle-holes in JSTew England, as above described.* Mr. King gives a thrilling account of how he at one time started such a move- ment of earth into one of the ice-tunnels, and came near himself falling into the yawning ice-chasm. f The following are the principal portions of Mr. King's clear and vivid description of the glaciers on the north side of Mount Shasta : We reached the rim of the cone, and looked down into a deep gorge lying between the secondary crater and the main mass of Shasta, and saw directly beneath us a fine glacier, which started almost at the very crest of the main mountain, flowing toward us, and curving around the circular base of our cone. Its entire length in view was not less than three miles, its width opposite our station about four thousand feet, the surface here and there terribly broken in "cascades," and presenting all the characteristic features of similar glaciers else- where. The region of the terminal moraine was more extended than is usual in the Alps. The piles of rubbish superimposed upon the end of the ice indicated a much greater thickness of the glacier in former days. After finishing our observations upon the side crater, and spending a night upon the sharp edge of its rim, on the following morning we climbed over the divide to the main cone, and up the extreme summit of Shasta. . . . From the crest I walked out to the northern edge of a prominent spur, and looked down upon the system of three considerable glaciers, the largest about four and a half miles in length, J and two to three miles wide. On the next day we descended upon the south side of the cone, following the ordinary track by which earlier parties have made the climb. From the moment we left the summit we encountered * Russell, "Existing Glaciers," p. 11. t " Proceedings of the Boston Society of Natural History," vol. xix, p. 61. X These estimates prove to be somewhat exaggerated. Thompson gives the length of the Whitney Glacier, the longest on the mountain, as only 3,800 yards, less than two miles and a half. 18 THE ICE AGE IN NORTE AMERICA. less and less snow, and at no part of the journey were able to see a glacier. An east-and-west line divides the mountain into glacier-bearing and non-glacier-bearing halves. The as- cent was formerly always made upon the south side, where, as stated, there are no glaciers, and this is why able scientific observers like Professor Whitney and his party should have scaled the mountain without discovering their existence. . . . Upon reaching the eastern side we found in a deep canon a considerable glacier, having its origin in a broad neve which reaches to the very summit of the peak. The entire angle of this glacier can be hardly less than twenty-eight degrees. It is one series of cascades, the whole front of the ice being crevassed in the most interesting manner. Near the lower end, divided by a boss of lava, it forks into two distinct bodies, one ending in an abrupt rounded face no less than nine hundred feet in height. Below this the other branch extends down the canon for a mile and a half, covered throughout almost this entire length with loads of stones which are constantly fall- ing in showers from the canon-walls on either side. Indeed, for a full mile the ice is only visible in occasional spots, where cavities have been melted into its body and loads of stones have fallen in. From an archway under the end a consider- able stream flows out, milky, like the water of the Swiss glacier-streams, with suspended sand. Following around the eastern base of Shasta, we made our camps near the upper region of vegetation, where the forest and perpetual snow touch each other. A third glacier, of somewhat greater extent than the one Just described, was found upon the noi'theast slope of the mountain, and upon the north slope one of much greater dimensions. The exploration of this latter proved of very great interest in more ways than one. Receiving the snows of the entire north slope of the cone, it falls in a great field, covering the slope of the mountain for a breadth of about three or four miles, reaching down the caflons between four and five miles, its lower edge dividing into a number of lesser ice- streams which occupy the beds of the caflons. This mass is sufBciently large to partake of the convexity of the cone, and, judging from the depth of the caflons upon the south and southeast slopes of the mountain, the thickness can not be less OLAOIERS ON THE PAOIFIG COAST. 19 than from eighteen to twenty-five hundred feet. It is cre- Tassed in a series of immense chasms, some of them two thou- sand feet long by thirty and even fifty feet wide. In one or two places the whole surface is broken with concentric systems of fissures, and these are invaded by a set of radial breaks which shatter the ice into a confusion of immense blocks. Snow-bridges similar to those in the Swiss glaciers are the only means of crossing these chasms, and lend a spice of danger to the whole examination. The region of the terminal moraines is quite unlike that of the Alps, a larger portion of the glacier itself being covered by loads of angular debris. The whole north face of the mountain is one great body of ice, interrupted by a few sharp lava-ridges which project above its general level. The veins of blue ice, the planes of stratification, were distinctly observed, but neither moulins nor regular dirt-bands are present. N"umerous streams, how- ever, flow over the surface of the ice, but they happen to pour into crevasses which are at present quite wide.* From Mount Shasta to the Columbia Eiver the mount- ains have been but imperfectly explored. But Mount Jef- ferson, Diamond Peak, and the Three Sisters are reported by Mr. Diller and Professor Newberry as containing numer- ous glaciers, and " as affording the most interesting field for glacial studies in the United States, with the exception of Alaska." The glaciers upon Mount Hood have been more fully explored, and are of great interest, though, owing to the moderate elevation (11,000 feet) and the limited snow- fields, they are small in size. The summit of this mountain is occupied by a volcanic crater about half a mile in diameter. This serves as a fountain out of which there flow three streams of ice extending down the flanks, as glaciers, for a distance of about two miles, their subglacial streams form- ing the head-waters of the White, Sandy, and the Little Sandy Elvers. In Washington Territory, a short distance north of the Columbia River, the Cascade Mountains culminate in a clus- * "American Journal of Science," vol. ci, IS^l, pp. 158-161. rjLAO/ERS ON THE PACIFIC! COA^T. 21 ter of peaks, with Mount Eainier, or, as it is coming to be called, Tacoiua, as the center. This central peak is upward of l-±,40() feet in height, and two or tliree neighboring jieaks are upward of 10,000 feet. This great elevation, coupled with the higher latitude and fhe increasing moisture of the climate, favors the production of a most imposing series of glaciers. Even the passing traveler upon the railroad is made aware of their existence by the milky whiteness of the waters of the Cowlitz, the Nisquallj, the Puyallup, and the White Eiver, ^vhich are crossed on the way from Portland, Oregon, to Seattle in Washington Teriitory. All of these streams originate in glaciers far up on the flanks of the mountains to the east and south. Of this series, that on the north side of Mount Taco- ma, at the head of White Iliver Valley, is the largest, and Fif;. 15.— Blount Tacoraa, looking ea^^tward toward the siiinmit, from Crater Lalve. reaches down to within .5,000 feet of the sea-level. This glacier is about ten miles long, and, though comparatively narrow in its lower portioji, is in places as much as four miles wide. The extreme summit of the mountain has been ascended only by two or three parties, and the task is beset with such difHculties that it is not likely to be ascended often. Al)Ove the 9,000-foot level it is wholly enveloped in snow ; while just below tliat limit, and close up to the realm of perpetual ice and snow, flowers make the air fragrant with 22 THE ICE AGE IN NORTE AMERICA. their perfume, and the open spaces are gorgeous with their masses of brilliant color. The following is the description of the glaciers of this mountain cluster, as given by Mr. S. F. Emmons, of the United States Geological Survey, the first to ascend it : The main White Eiver glacier, the grandest of the whole, pours straight down from the rim of the crater in a northeast- erly direction, and pushes its extremity farther out into the valley than any of the others. Its greatest width on the steep slope of the mountain must be four or five miles, narrowing toward its extremity to about a mile and a half ; its length cau be scarcely less than ten miles. The great eroding power of glacial ice is strikingly illustrated in this glacier, which seems to have cut down and carried away, on the northeastern side of the mountain, fully a third of its mass. The thickness of rock cut away — as shown by the walls on either side — and the isolated peak at the head of the triangular spur . . . may be roughly estimated at somewhat over a mile. Of the thick- ness of the ice of the glacier I have no data for making esti- mates, though it may probably be reckoned in thousands of feet. It has two principal medial moraines, which, where crossed by us, formed little mountain-ridges, having peaks nearly one hundred feet high. The sources of these moraines are cliffs on the steeper mountain-slope, which seem mere black specks in the great white field above ; between these are great cas- cades, and below, immense transverse crevasses, which we had no time or means to visit. The surface water flows in rills and brooks on the lower portion of the glacier, and moulins are of frequent occurrence. "We visited one double moulin, where two brooks poured into two circular wells, each about ten feet in diameter, Joined together at the surface but sepa- rated below ; we could not approach near enough the edge to see the bottom of either, but, as stones thrown in sent back no sound, judged they must be very deep. This glacier forks near the foot of the steeper mountain- slope, and sends off a branch to the northward, which forms a large stream flowing down to join the main stream fifteen or twenty miles below. Looking down on this from a high, over- GLACIERS ON THE PAGIFIG GO AST. 23 hanging peak, we could see, as it were, under our feet, a little lake of deep, blue water, about an eighth of a mile in diameter, standing in the brown, gravel-covered ice of the end of the glacier. On the back of the rocky spur which divides these two glaciers, a secondary glacier has scooped out a basin -shaped bed, and sends down an ice-stream, having all the characteris- tics of a true glacier, but its ice disappears several miles above the mouths of the large glaciers on either side. Were nothing known of the movement of glaciers, an instance like this would seem to afEord sufficient evidence that such movement exists, and that gravity is the main motive-power. From our north- ern and southern points we could trace the beds of several large glaciers to the west of us, whose upper and lower portions only were visible, the main body of the ice lying hidden by the high intervening spurs. Ten large glaciers observed by us, and at least half as many more hidden by the mountain from our view, proceeding thus from an isolated peak, form a most remarkable system, and one worthy of a careful and detailed study.* Still farther to the north, in Washington Territory, Mount Baker, rising to an elevation of 11,000 feet, is to a limited extent a center for the dispersion of glaciers of small size. The field, hovrever, has been but imperfectly explored. Northward from Washington Territory the coast is every- where very rugged, being formed by the lofty peaks of an extension of the Cascade Range ; while the thousands of isl- ands which fringe the coast of British Columbia and Alaska are but the partially submerged peaks of an extension of the Coast Eange, from which the great glaciers of former times have scraped off nearly all the fertile soil. It is estimated that there are ten thousand islands between Washington Ter- ritory and Mount St. Elias, and all the larger of them bear snow-covered summits during the whole year. The water in the narrow channels separating these islands is ordinarily several hundred feet deep, affording, through nearly the whole distance, a protected channel for navigation. * Quoted by Clarence King, in the " American Journal of Science," vol. ci, pp. 164, 165. ^ \ liuj utiiiiki GLACIERS ON THE PACIFIC COAST. '25 Three great rivers interrupt the mountain barrier of British Columbia facing the Pacific — the Fraser, the Skeena, and the Stickeen — and the interior is penetrated for some distance by innumerable fiords. The Canadian Pacific Rail- road follows the course of the Fraser for a long distance, and passes within sight of glaciers of considerable extent, and every fiord receives the drainage of numerous decaying gla- ciers. But it is not until reaching the Stickeen Eiver, in Alaska, in latitude 5Y°, that glaciers begin to appear -which are both easily accessible and large enough to invite pro- tracted study. The water coming into the sound from the Stickeen Risrer is heavily charged with glacial mud, which spreads itself out over a great expanse. An extensive delta, forming almost the only arable land in southeastern Alaska, has been built up by the deposit at the mouth of this river. The most accurate information yet obtained concerning these glaciers is that gathered by Mr. William P. Blake in 1863. According to him, "there are four large glaciers and several smaller ones visible within a distance of sixty or seventy miles from the mouth" of the river. The second of these larger ones has attracted most attention. This "sweeps grandly out into the valley from an opening between high mountains from a source that is not visible. It ends at the level of the river in an irregular bluff of ice, a mile and a half or two miles in length, and about one hundred and fifty feet high. Two or more terminal moraines protect it from the direct action of the stream. What at first appeared as a range of ordinary hills along the river, proved on landing to be an ancient terminal moraine, crescent-shaped, and covered with a forest. It extends the full length of the front of the glacier.'" * This glacier has never been fully explored. A number of years since, a party of Russian officers attempted its ex- ploration, and were never heard from again. Mr. Blake reports that, as usual with receding glaciers, a considerable * " American Journal of Science," vol. xciv, ISBY, pp. 96-101. Fia. 17.— Map of Southeaetern Alaska. The arrow-points mark glaciers. .Simpson GLACmRS ON THE PAGIFIO GOAST. 27 portion of tlie front as it spreads out in the valley is so cov- ered with bowlders, gravel, and mud that it is difficult to tell where the glacier really ends. But from the valley to the higher land it rises in precipitous, irregular, stair-like blocks, with smooth sides, and so large that it was impossible to sur- mount them with the ordinary equipment of explorers. The glacier is estimated to be about forty miles long. Another glacier, upon the opposite side of the river, of which Mr. Blake does not speak, was reported to nie by those familiar with the country as coming down to within about two miles of the bank. The Indians are very likely correct in asserting that these two glaciers formerly met, compelling the Stickeen River to find its way to the sea through a vast tunnel. It would then have appeared simply as a subglacial stream of great magnitude. ISTorth of the Stickeen River, glaciers of great size are of increasing frequency, and can be seen to good advantage from the excursion-steamer. The Auk and Patterson gla- ciers appear first, not far north of Fort Wrangel. On approaching Holkham Bay and Taku Inlet, about latitude 58°, the summer tourist has, in the numerous icebergs en- countered, pleasing evidence of the proximity of still greater glaciers coming down to the sea-level. Indeed, the glaciers of Taku Inlet are second only in interest to those of G-lacier Bay, hereafter to be described more fully. In going from Juneau to Chilkat, at the head of Lynn Canal, a distance of about eighty miles, nineteen glaciers of large size are in full sight from the steamer's deck, but none of them come down far enough to break off into the water and give birth to icebergs. The Davidson Glacier, however, comes down just to the water's edge, and has there built up an immense terminal moraine all along its front. An illustration of the precipitous character of the south- eastern coast of Alaska is seen in the fact that it is only thirty-five miles from the head of Lynn Canal to the sources of the Yukon Eiver, which then fiows to the north and west for nearly three thousand miles before coming down to the sea- GLACIERS ON THE PACIFIG COAST. 2!) level. Lieutenant Scliwatka reports fonr glaciers of consider- able size in the coui'se of this short portage lietween Chilkat and Lake Lindemau.* The vast region through which the Yukon flows to the north of these mountains is not known to contain any extensive glaciers. But, according to the re- ports of Dall, Schwatka, and others, it is a most inhospitable country, where human life can be maintained only with the greatest difficulty ; where the thermometer sinks to 00° below Fig 19 —Davidson (ilatier, iiwir Oliilkat. Alaska, latitiiile 50° 43'. Tin- moiiiilains arc from five tliousand to se^'en thousand feet hiKh ; the Korge about three quarters of a mile wide ; the front of the glacier, three miles ; the terminal moraine, about two hundred and fifty feet high, (View from two miles distant.) zero in winter, and rises for a short period to 12(J° in the summer ; and where the ground remains perpetually frozen at a short depth below the surface. From Cross Sound, about latitude .58° and longitude 136° west from Greenwich, to the Alaskan Peninsnla, the coast is bordered by a most magnificent semicircle of mountains, opening to the south, and extending for more than a thousand miles. Throughont this wliole extent, glaciers of large size "Science," vol. iii (February 22, 1884), pp, 220-22V. 30 THi: ICE AGE IN NORTE AMERIOA. are everywhere to be seen. Elliott * estimates that, count- ing great and small, there can not be less than five thousand glaciers between Dixon's Entrance and the extremity of the Alaskan Peninsula. Little is known in detail of the glaciers of this region. But those in the neighborhood of Mount St. Elias are evi- dently the largest anywhere to be found in the northern hemisphere outside of Greenland. This mountain rises 19,500 feet above the sea ; and Lieutenant Schwatka, in his expedition of 1886, reported eleven glaciers as coming down from its southern side. One of these, which he named the Agassiz Glacier, he estimates to be twenty miles in width and fifty miles in length, and to cover an area of a thousand square miles. Another, which he named Guyot Glacier, seemed to be about the same in dimensions. These come down to the sea-level in Icy Bay, and present a solid ice wall many miles in extent, which is continually breaking off into icebergs of great size.f Vancouver's account of the glacial phenomena along this coast is still both instructive and interesting, and in places curious. Between these points [Pigot and Pakenham] a bay is formed, about a league and a half deep toward the north- northwest, in which were seen several shoals and much ice ; the termination of this bay is bounded by a continuation of the above range of lofty mountains. On this second low pro- jecting point, which Mr. Whidbey called " Point Pakenham," the latitude was observed to be 60° b^', its longitude 313° 39'. The width of the arm a.t this station was reduced to two miles, in which were several half-concealed rocks, and much floating ice, through which they pursued their examination, to a point at the distance of three miles along the western shore, which still continued to be compact, extending north 30° east ; in this direction they met such innumerable huge bodies of ice, some afloat, others lying on the ground near the shore in ten * See " Our Arctic Provinces," p. 19. \ " New Yorlf Times," November 14, 1886. GLACIERS ON THE PACIFIC COAST. 31 or twelve fathoms water, as rendered their further progress up the branch rash and highly dangerous. This was, however, very fortunately, an object of no moment, since before their return they had obtained a distinct view of its termination, about two leagues farther in the same direction, by a firm and compact body of ice reaching from side to side, and greatly above the level of the sea ; behind which extended the con- tinuation of the same range of lofty mountains, whose summits seemed to be higher than any that had yet been seen on the coast. While at dinner in this situation they frequently heard a very loud, rumbling noise, not unlike loud but distant thun- der ; similar sounds had often been heard when the party was in the neighborhood of large bodies of ice, but they had not before been able to trace the cause. They now found the noise to originate from immense ponderous fragments of ice, breaking off from the higher parts of the main body, and fall- ing from a very considerable height, which in one instance produced so violent a shock that it was sensibly felt by the whole party, although the ground on which they were was at least two leagues from the spot where the fall of ice had taken place. . . . The base of this lofty range of mountains [between Elias and Fairweather], now gradually approached the sea-side ; and to the southward of Cape Fairweather it may be said to be washed by the ocean ; the interruption in the sum- mit of these very elevated mountains, mentioned by Captain Cook, was likewise conspicuously evident to us as we sailed along the coast this day, and looked like a plain composed of a solid mass of ice or frozen snow, inclining gradually toward the low border; which, from the smoothness, uniformity, and clean appearance of its surface, conveyed the idea of ex- tensive waters having once existed beyond the then limits of our view, which had passed over this depressed part of the mountains, until their progress had been stopped by the severity of the climate, and that, by the accumulation of suc- ceeding snow, freezing on this body of ice, a barrier had become formed that had prevented such waters from flowing into the sea. This is not the only place where we had noticed the like appearance ; since passing the icy bay mentioned on the 28th 32 THE ICE AGE IN NORTH AMERICA. of June, other valleys hnd been seen strongly resembling this, but none were so extensive, nor was the surface of any of them so clean, most of them appearing to be very dirt3^ I do not, however, mean to assert that these inclined planes of ice must have been formed by the passing of inland waters thus into the ocean, as the elevation of them, which must be many hundred yards above the level of the sea, and their having been doomed for ages to perpetual frost, operate much against this reasoning ; but one is naturally led, on contemplating any phenomenon out of the ordinary course of nature, to form some conjecture, and to hazard some opinion as to its origin, which on the present occasion is rather ofEered for the purpose of describing its appearance, than accounting for the cause of its existence.* Beyond Mount St. Elias, in the neighborhood of the Copper River and Prince William Sound, glaciers are re- ported by Elliott as numerous and of great size. Mount Wrangel, in the forks of the Copper Eiver, is estimated by him to be upward of 20,000 feet in height. From the flanks of the Chugatch Alps, of which Wrangel is the eastern sum- mit, immense glaciers descend to Prince William Sound, and add greatly to the gloomy grandeur of its scenery. Glaciers also extend throughout the Kenai and Alaskan Peninsulas, as far to the westward as longitude 162°, and one even has been observed upon the island of TJnalaska. The region in the interior north of the St. Elias and Chu- gatch Alps has been but imperfectly explored; but there seems pretty general agreement that there are no glaciers there at the present time, nor is there evidence that glaciers ever existed in the country. Much of the region is now covered with tundra— thai is, with vast level areas which are so deeply frozen that they never thaw out below a few feet from the surface. These are covered with a dense growth of heath and arctic mosses, which afford food for the reindeer, but are useless for man. At Eschscholtz Bay, on Kotzebue Sound, in latitude 66° ■ "Voyage of Discovery around the World," vol. v, pp. 312-114, 358-360. GLA0IEB8 ON THE PACIFIC COAST. 33 15', Kotzebne discovered in 1818 a cliff of frozen mud and ice " capped by a few feet of soil bearing moss and grass." * Large number of bones of the " mammoth, bison (?), rein- deer, moose-deer, musk-ox, and horse, were found " at the base, where thej' had fallen down from the cliff during the summer thaws. Sir Edward Belcher and Mr. Gr. B. Seeman after- ward visited the same spot and corroborated Kotzebue's ac- count. From their report it was evident that the conditions in northern Alaska are very similar to those in northern Si- beria, where so many similar remains of extinct and other animals have been found in the frozen soil. The section de- scribed at Eschscholtz Bay seems to be simply the edge of the tundra which is so largely represented in the central portions of the Territory. In 1880 Mr. Dall visited the local- ity and gave a fuller description than had been before given. The conditions are so unique that we reproduce his account : The ice-cliffs at this point were for a considerable distance double ; that is, there was an ice-face exposed near the beach with a small talus in front of it and covered with a coating of soil two or three feet thick, on which luxuriant vegetation was growing. All this might be thirty feet in height. On climb- ing to the brow of this bank the rise from that brow proved to be broken, hummocky, and full of crevices and holes ; in fact, a second talus on a larger scale, ascending to the. foot of a sec- ond ice-face, above which was a layer of soil one to three feet thick covered with herbage. The brow of this second bluff we estimated at eighty feet or more above the sea. Thence the land rose slowly and gradu- ally to a rounded ridge, reaching the height of three or four hundred feet only at a distance of several miles from the sea, with its axis in a north-and-south direction, a low valley west from it, the shallow bay at Elephant Point east from it, and its northern end abutting in the cliffs on the southern shore of Eschscholtz Bay. There were no mountains or other high land about this ridge in any direction ; all the surface around was lower than the ridge itself. * See Prcstwich's " Geology," vol. ii, p. 463 et seq. 34 THE lOE AGE IN NORTH AMERICA. About half a mile from the sea, on the highest part of the ridge, perhaps two hundred and fifty feet above high-water mark, at a depth of a foot, we came to a solidly frozen stratum consisting chiefly of bog-moss and vegetable mold, but con- taining good-sized lumps of clear ice. There seemed no reason to doubt that an extension of the digging would have brought us to solid clear ice such as was visible at the face of the bluff below ; that is to say, it appeared that the ridge itself, two miles wide and two hundred and fifty feet high, was chiefly composed of solid ice overlaid with clay and vegetable mold. The ice in general had a semi-stratified appearance, as if it still retained the horizontal plane in which it originally con- gealed. The surface was always soiled by dirty water from the earth above. This dirt was, however, merely superficial. The outer inch or two of the ice seemed granular, like compacted hail, and was sometimes whitish. The inside was solid and transparent or slightly yellow- tinged, like peat- water, but never greenish or bluish like glacier-ice. But in many places the ice presented the aspect of immense cakes or fragments irregularly disposed, over which it appeared as if the clay, etc., had been deposited. Small pinnacles of ice ran up into the clay in some places, and, above, holes were seen in the face of the clay bank, where it looked as if a detached fragment of ice had been melted out, leaving its mold in the clay quite perfect.* After speaking of the frequency vpith which the bones of the mammoth and buffalo and other animals are found, and of portions of the earth which still has in it the odor of the decaying flesh, Dr. Dall adds : Dwarf birches, alders seven or eight feet high, with stems three inches in diameter and a luxuriant growth of herbage, including numerous very toothsome berries, grew with the roots less than a foot from perpetual solid ice. The formation of the surrounding country shows no high land or rocky hills, from which a glacier might have been derived and then covered with debris from their sides. The continuity of the mossy surface showed that the ice must be * "American Journal of Science," vol. cxxi, 1881, pp. 106-109. QLAGIEE8 ON THE PACIFIC COAST. 35 quite destitute of motion, and the circumstances appeared to point to one conclusion, that there is here a ridge of solid ice rising several hundred feet above the sea and higher than any of the land about it and older than the mammoth and fossil horse, this ice taking upon itself the functions of a regular stratified rock. The formation, though visited before, has not hitherto been intelligibly described from a geological stand- point. Though many facts may remain to be investigated, and whatever be the conclusions as to its origin and mode of preservation, it certainly remains one of the most wonderful and puzzling geological phenomena in existence. The same author elsewhere writes that the continuity of this deposit "is brohen between Kotzebue Sound and Icy Cape by rocky hills composed chiefly of carboniferous lime- stones, which bear no glaciers, and do not seem to have been glaciated. The absence of bowlders and erratics over all this area has been noted by Franklin, Beechy, and all others who have explored it." * Dnring the period of the Eussian occupancy of Alaska scarcely anything was added to our knowledge of its glaciers further than what is to be found in the notes of Vancouver's voyage. Even the existence of Glacier Bay, which is to form the subject of the next chapter, was not suspected till a comparatively recent time, and it is not noted on any map drawn previous to 1880. Muir Glacier, which is now the object of greatest interest to the host of summer tourists who crowd the steamers making the round trip from Portland, Oregon, through the waters of southeastern Alaska, was brought to the notice of the outside world by the California gentleman whose name it bears, as late as 1879, when he and Eev. Mr. Young, of the Presbyterian mission at Fort Wran- gel, made a voyage of discovery around the archipelago in a dug-out canoe. * " Bulletin of the Philosophical Society of Washington," vol. vi, p. S3 ; quoted in EusselJ, as abore, p. 354. CHAPTER III. A MONTH WITH THE MUIR GLACIEE. In, the summer of 1886 a party of three, consisting of Rev. J. L. Patton, Mr. Prentiss Baldwin, and myself, ar- ranged to visit the Mnir Glacier, at the head of Glacier Bay in Alaska, for the purpose of collecting facts concerning its ipotion, its size, its present general condition, and its probable past history and future career. The present chapter will detail with some minuteness the results of our observa- tion. On the 4:th of August, in company with two Indians for assistants, we were landed by the excursion-steamer on the east side of the inlet, directly in front of the Muir Glacier, with a dug-out canoe as our only means of escape, and two canvas tents as our only shelter. Here we remained a whole month, or until September 2d, while the steamer made a round trip to Portland, Oregon, and returned with another load of freight and tourists. The region is the most desolate imagi- nable. Indians rarely navigate its upper waters, and it is visited only by the steamer to allow tourists to behold for a few hours the wonderful spectacle of a stream of ice more than a mile in width, and four hundred feet in height, mov- ing onward with irresistible force to meet the equally irre- sistible waters of a deep tidal inlet. Those who have here, for a few hours only, witnessed the " calving " of icebergs, and heard the detonations preceding and accompanying the falling of the masses from the ice-front, can never forget the scene. Much less can we forget it, who spent a month in the majestic presence of the mighty glacier. A MONTH WITH THE MUIB GLACIER. 37 Our facilities for observations were limited by several unfavorable conditions. In the first place, though we were there in the dry season, fifteen of the twenty-nine days were so rainy that it was impossible to stir out of our tents or to see far through the mists. In the second place, the tides were so strong, and the winds at times so vio- lent, that it was hazardous to venture far away with our canoe. In the next place, the surface of the glacier is, in its central portion, so intersected by yawning crevasses that it was entirely out of the question to attempt to cross it. Plans for measurement, different from those made familiar in Professor Tyndall's book, had therefore to be devised. On the other hand, some things were favorable to obtain- ing satisfactory results. The fourteen days of fair weather were extremely clear and beautiful, and there are no trees upon the mountains to obstruct one's view or to hinder him in rambling over them. The specific results as to the movements of the ice, and as to the formation of moraines and kames, are told a little later. Here a few words will be in place concerning the general aspect of the region as we saw it in August. The mountains on each side of Muir Inlet rise immedi- ately from the water from three thousand to five thousand feet. These we often ascended, and thus were permitted repeatedly to behold one of the most marvelous views any- where to be found in the world. At that season the level places around om- feet upon these summits were carpeted with soft green grass, interspersed with large areas of flow- ers in full bloom. Here were extensive, gorgeously colored flower-beds, where bluebells, daisies, buttercups, violets, the yellow arnica-flower, and the purple epilobium, were striv- ing for mastery or for recognition. On the ^northern slopes of slight elevations great masses of snow were preserved in the very midst of these brilliant flower-gardens, and, from their melting, clear little pools of water were on every hand inviting us to drink. The track of the mount- 38 THE IGE AGE IX NORTH AMERICA. ain goat, the mountain lion, and of various smaller animals, and the songs of birds, witnessed to the abundance of ani- mal life. To the south the calm surface of the bay opened outward into Cross Sound, twenty-five miles away. The islands dot- tino- the sm-face of the smooth water below lis seemed but specks, and the grand vista of snow-clad mountains, guarding either side of Chatham Strait, seemed gradually to come to a point on the southern horizon. Westward, toward the Pacific, was the marvelous outline of the southern portion of the St. Elias Alps. The lofty peaks of Crillon (15,900 feet high) and Fairweather (15,500 feet high), about twenty- five miles away, and aboiit the same distance apart, stood as sentinels over the lesser peaks, La Perouse, Lituya, and their companions, which, anywhere else, would appear to be mountains of the first class, being more than ten thousand feet high, and rising directly from the water's edge. At one time, when on a summit overlooking Glacier Bay, it was our good fortune to see the sun go down behind this mount- ain-chain. Alternate shadows and golden rays of setting sunlight stretched across the water and climbed the peak on which we stood. The glistering summits of the western mountains were lined with the same glowing colors, while the solemn procession of glaciers on their eastern flanks was gradually fading in the growing darkness, and the more dis- tant mountain-tops in other directions were ceasing to reflect the glow of the western horizon. Tn such a setting of grandeur and beauty we gazed upon the full face of the great glacier itself lying at our feet. Below us its diminishing outlet disappeared in the waters of the bay. Distance made the rough places plain, and lent enchantment to the view. Down from the mountains in every direction from the north came the frozen torrents : Glaciers to the right of us, Glaciers to the left of us, Glaciers in front of us, Volleyed and thundered — A MONTH WITH THE MUIB GLACIER. 39 pouring into a vast amphitlieatre, and then uniting their vol- ume, preparatory to their exit through the entrance into Muir Inlet. These numerous local glaciers united to form nine main streams whose individuality could be determined all across the amphitheatre by the long lines of medial moraines which swept around in majestic curves from every quarter, like great railroad embankments in approaching some grand central depot. Such is a faint description of the scene upon which we gazed. Strength and beauty were here united as probably nowhere else in the world. But the shades of night slowly fell upon us, even in that high latitude, and we were compelled to come down closer to the thundering noises of the active glaciers and seek the prosaic quarter of our tents, and to go about the more detailed investigation of the mar- velous phenomena before us. The results I will now pro- ceed to give. The Muir Glacier enters an inlef of the same name at the head of Glacier Bay, in latitude 58° 50', longitude 136° 40', west of Greenwich (see Fig. 20). This bay is a body of water about thirty miles long, and from eight to twelve miles wide (but narrowing to about three miles at its upper end), projecting in a northwest direction from the eastern end of Cross Sound. The promontory separating it from the Pacif- ic Ocean is from thirty to forty miles wide, and contains the lofty mountain-peaks of Crillon, Fairweather, Lituya, and La Perouse, whose heights have already been indicated. To the east, between Glacier Bay and Lynn Canal, is a pen- insula, extending considerably south of the mouth of the bay, and occupied by the White Mountains, probably having no peaks exceeding ten thousand feet. Near the mouth of Glacier Bay is a cluster of low isl- ands named after Commander Beardslee, of the United States Navy. There are twenty -five or thirty of these, and they are composed of loose material— evidently glacial debris and are in striking contrast with most of the islands and shores in southeastern Alaska. These, also, like all the other land to the south, are covered with evergreen forests, though 40 THE ICE AGE IN NORTH AMERICA. FiQ. 20.— Map of Glacier Bay, Alaska, and its surroundings. Arrow-points indicate glaciated area. the trees are of moderate size ; but the islands and shores in the upper part of the bay are entirely devoid of forests. Willoughby Island, near the middle of the bay, is a bare rock, about two miles long and fifteen hundred feet high, showing glacial furrows and polishing from the bottom to the top. Several other smaller islands of similar character in this part of the bay show like signs of having been re- cently covered with glacial ice. The upper end of the bay is divided into two inlets of A MONTH WITH THE MUIR GLACIER. 41 unequal lengths, the western one being about four miles wide, and extending seven or eight miles (estimated) in the direction of the main axis of the bay to the northwest. The eastern, or Muir Inlet, is a Uttle over three miles wide at its mouth, and extends to the north about the same distance, narrowing, at the upper end, to a little over one mile, where it is interrupted by the front of the Muir Glacier. The real opening between the mountains, however, is here a little over two miles wide, the upper part on the eastern side being oc- cupied with glacial debris covering a triangular space be- tween the water and the mountain about one mile wide at the ice-front and coming to a point three miles below, be- yond which a perpendicular wall of rock one thousand feet high rises directly from the water. The mountain on the west side of Muir Inlet, between it and the other fork of the bay, is 2,900 feet high. That on the east is 3,160 feet high, rising to about 5,000 feet two or three miles back. The base of these mountains consists of metamorphic slate, whose strata are very much contorted — so much so that it is difficult to ascertain their system of folds. Upon the summits of the mountains on both sides are remnants of blue crystalline limestone preserved in synclinal axes. In the terminal mo- raine deposited in front of the glacier on its eastern side are numerous bowlders of very pure white marble brought down in medial moraines from mountain valleys several miles to the east. Granitic bowlders are also abundant. The width of the ice where the glacier breaks through between the mountains is 10,664: feet — a little over two miles. But, as before remarked, the water-front is only about one mile. This front does not form a straight line, but terminates in an angle projecting about a quarter of a mile below the northeast and northwest comers of the inlet. The depth of the water three hundred yards south of the ice- front is (according to the measurement of Captain Hunter, of the steamer Idaho) 516 feet near the middle of the chan- nel ; but it shoals rapidly toward the eastern shore. A meas- urement reported to me by Dr. Jackson, made in July, A MONTH WITH THE MUIR GLACIER. 43 1887, with the prow of the steamer within twenty feet of the ice-front, is one hundred and six fathoms (636 feet), and no bottom. Accordmg to my measurements, ialcen by level- ing up on the shore, the height of the ice at the extremity of the projecting angle in the middle of the inlet was 260 feet, and the front was perpendicular. Back a few hun- dred feet from the projecting point, and along the front nearer the shores, the perpendicular face of the ice was a lit- tle over 300 feet. A little farther back, on a line even with the shoulders of the mountains between which the glacier emerges to meet the water, the general height is 408 feet. From here the surface of the glacier rises toward the east and northeast about 100 feet to the mile. On going out in that direction on the ice seven miles (as near as I could estimate), I found myseK, by the barometer, 1,050 feet above the bay. The main body of the glacier occupies a vast amphithea- tre, with diameters ranging from thirty to forty miles. This estimate was made from various views obtained from the mountain-sumnaits near its mouth, when points whose dis- tances were known in other directions were in sight. Nine main streams of ice unite to form the grand trunk of the glacier. These branches come from every direction north of the east-and-west line across the mouth of the glacier ; and no less than seventeen sub-branches can be seen coming in to join the main streams from the mountains near the rim of the amphitheatre, making twenty-six in all. l^umerous rocky eminences also rise above the surface of the ice, like islands from the sea, corresponding to what are called nuna- taks in Greenland. The two of these visited, situated about four miles back from the front, showed that they had been recently covered with ice — their surfaces being smoothed and scored, and glacial debris being deposited everywhere upon them. Upon the side from which the ice approached these islands (the stoss side) it rose, like breakers on the sea- shore, several hundred feet higher than it was immediately on the lee side. A short distance farther down on the lee side, however, the ice closes up to its normal height at that 44 THE ICE AGE IN NORTH AMERICA. point. In both instances, also, the lee side of these islands seemed to be the beginning of important subglacial streams of water — brooks running into them as into a funnel, and causing a backward movement of ice and moraine material, as where there is an eddy in water. In both these cases, however, the lee sides of these islands were those having greatest exposure to the sunshine. The surface of the ice immediately in front was depressed from one to two hun- dred feet below the general surface on the lee side. The ice in the eastern half of the amphitheatre is mov- ing much more . slowly than that in the western half. Of this there are several indirect indications : First, the eastern surface is much smoother than the western. There is no difficulty in traversing the glacier for many miles to the east and northeast. Here and there the surface is interrupted by superficial streams of water occupying narrow, shallow chan- nels, running for a short distance and then plunging down into fissures, or, in technical language, mo\dins, to swell the larger current, which may be heard rushing along in its im- petuous course far down beneath and out of sight. The ordinary light-colored bands in the ice parallel with its line of motion are everywhere conspicuous, and can be followed on the surface for long distances. When interrupted by cre- vasses they are seen to penetrate the ice for a depth of many feet, and sometimes to continue on the other side of a cre- vasse in a difEerent line, as if having suffered a lateral fault. The color of the ice below the surface is an intense blue, and over the eastern portion this color characterized the most of the surface. Numerous holes in the ice, penetrating downward from an inch or two to several feet and filled with water, were encountered all over the eastern portion. Some- times there was a stone or a little dirt in the bottom of these, but frequently there was apparently nothing whatever in them but the purest of water. In the shallower inclosures on the surface, containing water and a little dirt, worms, about as large around as a small knitting-needle and an inch long, were abundant. A MONTH WITH THE MUIR QLAOIER. 45 The character and course of the moraines on the eastern half of the glacier also attest its slower motion. There are seven medial moraines east of the north-and-south line, four of which come in to the main stream from the mountains to the southeast (see Fig. 22). Near the rim of the glacial am- phitheatre these are long distances, in some cases miles, apart ; but, as they approach the mouth of the amphitheatre, they are crowded closer and closer together near its eastern edge, until in the throat itself they are indistinguishably mingled. The three more southern moraines unite some dis- tance above the mouth. One of these contains a large amount of pure marble. This moraine gradually approaches the others on either side until the distance between them disappears, and its marble unites with the other material to form one common medial moraine. The fifth moraine from the south is about 150 yards in width, five miles back from the mouth. It is then certainly as much as five and prob- ably eight miles from the mountains from which the debris forming it is derived. All these moraines contain many large blocks of stone, some of which stand above the general mass on pedestals of ice, with a tendency always to fall over in the direction of the sun. One such block was twenty feet square and about the same height, standing on a pedestal of ice three or four feet high. It is the combination of these moraines, after they have been crowded together near the mouth, which forms the deposit now going on at the north- east angle of the inlet just in front of the ice. Of this more win be said in connection with the question of the recedence of the glacier. Similar phenomena, though on a smaller scale, appear near the southwest angle of the amphitheatre. The dominant streams of ice in the glacier come from the north and the northwest. These unite in the lower por- tion to form a main current, about one mile in width, which is moving toward the head of the inlet with great relative rapidity. Were not the water in the inlet deep enough to float the surplus ice away, there is no knowing how much farther down the valley the glacier would extend. The 46 TSE IGE AGE IN NORTH AMERICA. streams of ice from the east and southwest have already spent the most of their force on reaching the head of the inlet ; and, were it not for this central ice-stream, a natural equilibrium of forces would be established here independent of the water, and no icebergs would be formed. The sur- face of this central current of motion is extremely rough, so that it is entirely out of the question to walk far out upon it. On approaching this portion of the glacier from the east the transverse crevasses diagonal to the line of motion increase in number and size until the whole surface is broken up into vast parallelograms, prisms, and towers of ice, sep- arated by yawning and impassable chasms scores and hun- dreds of feet in depth. Over this part of the ice the mo- raines are interrupted and drawn out into thinner lines, often appearing merely as patches of debris on separate masses of ice. This portion of the ice-current presents a lighter colored appearance than other portions, and the roughened lines of motion can be followed, as far as the eye can reach, through distant openings in the mountains to the north and the northwest. The comparative rapidity of the motion in this part of the ice is also manifest where it breaks off into the water at the head of the inlet. As already said, the perpendicular front of ice at the water's edge is from 250 to 300 feet in height. PVom this front there is a constant succession of falls of ice into the water, accompanied by loud reports. Scarcely ten minutes, either day or night, passed during the whole month without our being startled by such reports, and frequently they were like thunder-claps or the booming of cannon at the bombardment of a besieged city, and this, though our camp was two and a half miles below the ice- front. Sometimes this sound accompanied the actual fall of masses of ice from the front, while at other times it was merely from the formation of new crevasses or the enlarge- ment of old ones. Eepeatedly I have seen vast columns of ice, extending up to the full height of the front, topple over and fall into the water. How far these columns extended A MONTH WITH THE MUIR GLACIER. 47 below the water could not be told accurately, but I have seen bergs floating away which were certainly 500 feet in length. At other times masses would fall from near the summit breaking off part way down, and splashing the spray up to the very top of the ice, at least 250 feet. The total amount of ice thus falling off is enormous. Bergs several hundred feet long and nearly as broad, with a height of from twenty to sixty feet, were numerous and constantly floating out from the inlet. The steamer meets such bergs a hun- dred miles away. The smaller pieces of ice often so cover the water of the inlet two or three miles below the glacier that it is with great difficulty that a canoe can be pushed through them. One of the bergs measured, was sixty feet above water and about four hundred feet square. The por- tion above water was somewhat irregular, so that probably a symmetrical form thirty feet in height would have contained it. But even at this rate of calculation the total depth would be two hundred and forty feet. The cubical contents of the berg would then be almost 40,000,000 feet. Occa- sionally, when the tide and wind were favorable, the inlet would for a few hours be comparatively free from floating ice ; at other times it woiild seem to be full. The movements of the glacier in its lower portions are probably facilitated by the subglacial streams issuing from the front. There are four of these of considerable size. Two emerge in the inlet itself, and come boiling up, one at each comer of the ice-front, making a perceptible current in the bay. There are also two emerging from under the ice where it passes the shoulders of the mountains forming the throat of the glacier. These spout up, like fountains, two or three feet, and make their way through a channel in the sand and gravel of the terminal moraine for about a mile, and enter the inlet 250 or 300 yards south of the ice- front. These streams are perhaps three feet deep and from twenty to forty feet wide, and the current is very strong, since they fall from 150 to 250 feet in their course of a mile. It is the action of the subglacial streams near the corners of 48 THE ICE AGE IN NORTH AMERICA. the inlet which accounts for the more rapid recession of the glacier-front there than at the middle point projecting into the water soath of the line joining the east and west comers. It was also noticeable that the falls of ice were mnch more frequent near these corners, and the main motion of the ice as afterward measured was, not toward the middle point projecting into the inlet, but toward these corners where the subglacial streams emerged below the water. No small difficulty was encountered in securing direct measurements of the motion ; and, as the results may be questioned, I will give the data somewhat fully. As it was impossible to cross the main current of the glacier, we were compelled to take our measurement by triangulation. But even then it seemed at first necessary to plant flags as far out on the ice as it was safe to venture. This was done on the second day of our stay, and a base-line was established on the eastern shore, about a mile above the mouth, and the necessary angles were taken. But, on returning to repeat the observations three or four days afterward, it was found that the ice was melting from the surface so fast that the stakes had fallen, and there were no means at command to make them secure. Besides, they were not far enough out to be of much service. It appeared also that the base-line was on a lateral moraine, which was, very likely, itself in motion. But by this time it had become evident that the masses of ice uniting to compose the main stream of motion retained their features so perfectly from day to day that there was no difficulty in recognizing many of them much farther out than it was possible to plant stakes. Accordingly, another base-line was established on the east side opposite the pro- jecting angle of ice in the inlet. From this position eight recognizable points in different portions of the ice-field were triangulated — the angles being taken with a sextant. Some of the points were triangulated on five different times, at inter- vals from the 11th of August to the 2d of September. Others were chosen later and triangulated a less number of times. The base-line finally chosen (marked B on Fig. 22) was at A MONTH WITH THE MUIR GLACIER. 49 the foot of the jnountain exactly east by the compass from the projecting arjgle of ice in the inlet. The elevation of the base-liue was ^08 feet above tide— corresponding to that of the ice-front, fhe distance of this projecting point of ice (marked on Fig. 22) from the base-line was 8,534 feet, and it remained v^ry nearly stationary during the whole time— showing that the material breaking o£E from the ice- ; z z' I ' r L ^ UlR I JSTLET 1 Mtle Flo. W. — Map of Muir Inlet, showing converging moraines, and form of front. Buried foreet, A ; base of triangulation, B. 50 TEE ICE AGE IN NORTH AMEEIGA. front was equal to that pushed along by the forward move- ment. Satisfactory observations were made upon eight other points numbered and located on Fig. 22. No. 1 was a pinnacle of ice 1,476 feet north by 30° east from C. The movement from August 14th to August 24th was 1,653 feet east by 15° south. After this date the pinna- cle was no longer visible, having disappeared along the wast- ing line of front between C and the subglacial stream at the northeast corner of the inlet. This was so near the front as to be left out of the ordinary calculations. No. 2 was a conspicious pinnacle of ice 2,416 feet north by 16° east of C. Observations were continued upon this from August 11th to September 2d. The total distance moved during that time was 1,41Y feet, or about sixty-five feet per day. From August 14th to August 24th the move- ment was 715 feet, or about seventy-one feet per day. The difference is, however, perhaps due to the neglect to record the hours of the day when the observations were taken. As these observations were wholly independent of each other, their substantial concordance demonstrates that there was no serious error in the observations themselves. The direction of movement of this point of ice was very nearly the same as that of the preceding, namely, east 16° south. This also is toward the subglacial stream emerging from the northeast corner of the inlet. No. 3 was observed only from August 20th to August 24th. It was situated 3,893 feet north by 62° east of C, and moved 105 feet in a westerly direction, .about twenty-six feet per day. The westerly course of this movement probably arose from its being near where the easterly and northeast- erly currents joined the main movement. No. 4 was 5,115 feet north, 42° east of C, and moved from August 20th to August 24th 143 feet in a southeast- erly direction, or thirty-six feet per day. No. 5 was 5,580 feet north, 48° east of C, and moved 289 feet from August 20th to August 24th in a direction east by 39° south, or seventy-two feet per day. A MONTH WITH THE MUIR GLACIER. 51 No. 6 was 5,473 feet north, T0° east of C, and moved 232 feet from August 11th to September 2d in a direction south 66° east, or ten feet per day. No. T was 6,903 feet north, 59° east of C, and moved 89 feet between August 14th and August 24th, in a direction south 3° east, about nine feet per day. No. 8 was 7,507 feet north, 62° east of 0, and moved 265 feet from August 14th to August 24th, in direction south 56° east. These last three points lay in one of the moraines on the east side of the line of greatest motion and parallel with it. These moraines are much interrupted in their course by It is observable that these points are all east of the center of the main line of most rapid motion, and are tending with varying velocity toward the northeast corner of the inlet, where the powerful subglacial stream emerges from below the water-level. Doubtless, on the other side of the center of motion, and at the same relative distance from the front, the ice would be found tending toward the northwest corner of the inlet, where a similar subglacial stream emerges. From these observations it would seem to follow that a stream of ice presenting a cross-section of about 5,000,000 square feet (5,000 feet wide by about 1,000 feet deep) is entering the inlet at an average rate of forty feet per day (seventy feet in the center and ten feet near the margin of movement), making about 200,000,000 of cubic feet per day during the month of August. The preceding remarks upon the many indirect evidences of rapid motion render the cal- culation perfectly credible. "What the rate may be at other times of the year there are at present no means of knowing. The indications that the glacier is receding, and that its volume is diminishing, are indubitable and numerous. The islands of southern Alaska are ordinarily covered with forests of cedar, hemlock, and fir, up to the level of 1,500 or 2,000 feet above tide. But to this rule the shores and islands of the upper part of Glacier Bay are a striking exception. Near the mouth of the bay, forests continue to occur as in other a " Tl "l^ ^ ci p IS A MONTE WITH THE MUIR GLAOIER. 63 parts, only on a diminished scale ; but in the upper half of the bay all the shores and islands are perfectly bare of forests, and the rocks retain in the most exposed situations fresh grooves and striae of glacial origin. It would be impossible for rocks so exposed in such a climate to retain these for an indefinite length of time. Far up on the mountains, also, there are remnants of glacial debris in situations such that the material could not have resisted erosive agencies for any great length of time. The triangular-shaped terminal moraine on the eastern side, just below the ice-front, presents some interesting features bearing on the same point. This extends three miles below the glacier, and in its lower portions is thinly covered with vegetation. This covering becomes less and less abundant as the glacier is approached, until, over the last mile, scarcely any plants at all can be found. Apparently this is because there has not been time for vegetation to spread over the upper portion of the moraine since the ice withdrew, for on the mountains close by, where the exposure has been longer, there is a complete matting of grass, flower- ing plants, and shrubs. Again, in this triangular moraine- covered space there are five distinct transverse ridges, mark- ing as many stages in the recession of the ice-front (see Fig. 22). These moraines of retrocession run parallel with the ice-front on that side, and at about equal distances from each other, each one rising from the water's edge to the foot of the mountain, where they are 408 feet above tide. An inspec- tion of the upper moraine-ridge shows the manner of its for- mation. This transverse ridge is half a mile below the ice- front, and is still underlaid in some portions with masses of ice ninety feet or more in thickness, which are melting away on their sides and allowing the debris covering them to slide down about their bases. Kettle-holes are in all stages of for- mation along this ridge. The subglacial stream emerging from the southeast corner of the glacier next the mountain rushes along just in the rear of this moraine-ridge, and in front of a similar deposit in process of formation on the very edge of the ice where the medial moraines spoken of termi- 54 THE ICE AGE IN NORTH AMERICA. nate. Eventually tliis stream will break out in the rear of that deposit also, and leave another ridge similar to the one Fig. 24. — In the foreground on the ri^ht is a mass of ice. one half miie in front of the gla- cier, one hundred or Tn(H-e feet thick, covered with gravel, slowly sliding down to form the rim of a kettle-hole. The mountain back is .3,100 feet high. Near B, Fig. 33. now slowly settling down into position south of it. This first ridge south of the snbglacial stream, with its ice still melting in exposed positions under its covering of gravel, can not he many years old. Still another sign of the recent date of this whole moraine appears at various places where water-courses, coming down from the mountain, are depositing supei-ticial deltas of dehris upon the edge of the glacial deposit. These deltas are very limited in extent, though the annual deposition is by no means insignificant. At tlie southern apex of the moraine, three miles below the ice-front, and but one hundred or two hundred yards fr(.)in our camp, great quantities of dehris came tearing down in repeated avalanches during a prolonged sea- son of rain. Twenty-five years would be more than ample for the formation of the cone of dehriH at the foot of this line of avalanches. Thus there can be no reasonable doubt that A MONTH WITH THE MUIR OLAGIEB. 55 during the earlier part of this century tlie ice filled the inlet several miles farther down than now. And there can be scarcely less doubt that recently the. glacier filled the inlet, 2,500 feet above its present level near the front ; for the glacial debris and strias are very marked and fresh on both mountains flanking the upper part of the inlet up to that height, and the evidences of an ice-movement in the direction of the axis of the bay are not wanting as high as 3,700 feet on the eastern mountain, where I found fresh striae running north by south, and directly past the summit, which rises 1,000 or 1,500 feet still higher, just to the east. To this circumstantial evidence may be added what seems to be an irresistible inference from the notes of Vancouver's party in 1794. This party entered Cross Sound in small boats, and penetrated as far as the head of Lynn Oanal and Juneau. The following is the record. We should premise, how- ever, that the point referred to as seven miles from Point Dundas is probably that at the southeastern comer of Gla- cier Bay, and the " spacious inlet lying in an east-southeast " direction is probably the channel extending toward Chatham Strait. But certainly, no one looking from that point at the present time would speak, as this report does, of this inlet as seeming to be " entirely occupied by one compact sheet of ice as far back as the eye could distinguish." Nor would the observer at the present time say that, to the north and east, the two large open bays formed by the shores of the continent seemed to be " terminated by compact solid mount- ains of ice rising perpendicular from the water's edge." The ice is now full twenty-five miles away from that point, and the ice-front is not sufficiently prominent to make such an impression as this. It is hence more than probable that, at that time, the ice extended down nearly to the mouth of the bay. The morning of the 12th [July], though unpleasant, was rather more favorable to their pursuit, which was still greatly 56 TEE lOE AQE IN NORTH AMERICA. impeded by the ice. From the east point of this branch, which I have called Point Dundas, situated in latitude 58° 31', longitude 334° 1', the coast takes an irregular east-northeast direction about seven miles to a point whence this branch of the [Cross] Sound appeared to be very extensive in an east- southeast point of view, and was upward of three leagues across. The party proceeded from Point Dundas to this station, through a channel from two to three miles in width, between the continental shore and an island about seven miles long and three miles broad, lying in a northeast and southwest di- rection. This spacious inlet presented to our party an ardu- ous task, as the space between the shores on the northern and southern sides seemed to be entirely occupied by one compact sheet of ice as far as the eye could distinguish. ... To the north and east of this point the shores of the continent form two large open bays, which were terminated by compact solid mountains of ice, rising perpendicularly from the water's edge, and bounded to the north by a continuation of the united lofty frozen mountains that extend eastward from Mount Fair- weather. In these bays also were great quantities of broken ice, which, having been put in motion by the springing up of a northerly wind, was drifted to the southward, and, forcing the boats from the northern shore, obliged them to take shel- ter round the northeast point of the above island. This made Mr. Whidbey apprehensive that the still apparent connected body of ice, from side to side, would at length oblige him to abandon his researches by this route, unless he should find it possible to force a passage through this formidable obstruc- tion. In attempting this, the party succeeded far beyond their expectations, for they gained an open navigation, and by four in the afternoon arrived at a low and nearly round island about two leagues in circuit, lying from the former island north 83° east, distant three leagues. This island is moder- ately elevated, its shores pleasant and easy of access, and well stocked with timber, mostly of the pine tribe. It presented a much more inviting appearance than they had been accustomed to behold, and the weather being more favorable than for some time past, they continued along the continental shore, A MONTE WITH THE MUIB OLACIER. 5Y passing within some islets that lie about a league to the east- ward of the round island, until nine in the evening, when it became calm, and the party rested for the night at the entrance of a brook, in a bay on the northern or continental shore, which from the round island lies south 82° east, distant ten miles.* If we understand this, the bay to the north is Glacier Bay, down which the ice must then have extended south of "Willoughby Island and to within a few miles of Cross Sound. Otherwise no such description could have been given. The bay to the east is probably the extension of the sound toward the mouth of Lynn Canal, and very likely glaciers at that time came down toward the west from the White Mountains and produced the appearance described. From what has already been said of the evidence showing the present recession of the Muir Glacier, it is not at all incredible that glaciers nearly filled the whole bay a hundred years ago. All this is necessary to a comprehension of a most inter- esting problem presented by the buried forests near the south- west corner of the glacier (see A, Fig. 22). Below this corner, and extending for about a mile and a half, there is a gravel deposit, similar to that on the eastern side, except that it is not marked by transverse ridges, but is level-topped, rising gradually from about 100 feet at its southern termination to a little over 300 feet where it extends north and west of the ice-front (see Fig. 22). The subglacial stream entering the inlet just below the southwest corner of the ice emerges from the ice about a mile farther up, on the north side of the pro- jecting shoulder of the western mountain which forms that side of the gateway through which the glacier enters the in- let. This stream comes principally from the decaying western branch of the glacier before alluded to, and, after winding around the projecting shoulder of the mountain, which is 315 feet above tide, has worn a channel through the gravel * "Voyage of Discovery around the World," vol. v, pp. 420-423. 68 THE ICE A OE IN NORTH AMEIilCA. deposit lying between tlie lower mile of the glacier and the mountain a sliort distance to the southwest. About half-way down, a small brook, coming from between this latter mount- ain and that whose shoulder forms the western part of the gateway just ntjrth of it, joins the main stream issuing from the glacier on this side. Where these streams unite, at A, they are now uncovering a forest of cedar-trees in perfect preservation, standing upright in the soil in which they Fig. 25.— Shows stump of cedar-trees partly buried in glacial gi'avel. The pyramid of Rand corresponds m height to that in the cut on the opposite page. The mountain in the background is 8,000 feet high. This view is on the west side of the glacial tor- rent. See text. (From photograph, looking west.) grew, with the hvmus still about their roots. An abundance of their cones, still preserving their shape, lies aliout their roots ; and the texture of the wood is still unimpaired. One of these upright trunks measured ten feet in circumference about fifteen feet above the roots. Some of the smaller up- right trees have their liranches and twigs still intact, pre- serving the normal conical appearance of a recently dead cedar-tree. These trees are in various stages of exposure. Some of A MONTR WITH TEA' MUIR GLACIER. 59 them are uncovered to the roots ; some are washed wholly out of the soil ; while others are still buried and standing upiight, in horizontal layers of fine sand and gravel, some with tops projecting from a depth of twenty or thirty feet, others being doubtless entirely covered. The roots of these trees are in a compact, stiff clay stratum, blue in color, without grit, iutei'- sected by numerous minute rootlets, and which is, in places, twenty feet thick. There is also, occasionally, in this sub- FiG. 36.— Shows stumps of trees on east side of the glacial torrent. Note the lire of sep- aration between the envelopin;; siind and the soil in «-hicli the roots are imbedded. A stump appears on the rlL'ht. split in two, but one half standing. The gravel corre- sponds in height with that on the west side. The glacier appears m the background on the right. ' (From photograph, looking north.) stratum of clay, a small fragment of wood, as well as some smooth pebbles from an inch to two feet in diameter. The surface of this substratum is at this point 85 feet above the inlet. The deposit of sand and gravel covering the forest rises 115 feet higher, and is levehtopped at that height, but rising toward the north till it reaches the shoulder of the mountain at an elevation of 300 feet. The trees are essen- tially lihe those now growing on the Alaskan mountains. Many of them have been violently broken off fi-om five to 60 TEE ICE AOE IN NORTH AMERICA. twenty feet above their roots. This has been done by some force that has battered them from the upper side at the point of fracture. Evidently cakes of ice brought down by the streams indicated in the map, when flowing at various high- er levels than now, have accomplished this result ; for the trunks in the main stream were battered on the north side, while those in the gully worn by the lateral stream were bat- tered from the west side. From this description the explanation would seem to be evident. At some period, when the ice occupied only the upper part of the valley to the north of this point, forests grew over all the space lying southwest of the present ice- front. As the ice advanced to near its present position, the streams carrying off the surplus water from the western half of the advancing glacier were suddenly turned into the pro- tected space occupied by this forest, where they deposited their loads of sand and gravel. A cause very likely com- bining to facilitate deposition in this spot has not yet been spoken of, but is evident from a glance at the maps. A trans- verse valley passes just below this point from Muir Inlet to the western inlet into which G-lacier Bay divides. This transverse valley is at present occupied by a decaying glacier opening into both inlets, and sending a subglaeial stream through a long, narrow series of moraines, into Muir Inlet about two miles to the south. Now, when a general advance of the ice was in progress, this transverse glacier probably pushed itself down into the inlet across the path of the ice moving from the no^th, and so formed an obstruction to the water running from the southwest corner of the main glacier, thus favoring the rapid deposition which so evidently took place. When this inclosed place was filled up, and the ad- vancing ice had risen above and surmounted the projecting shoulder of the mountain just to the north, that rocky barrier protected a portion of the forest from the force of the ice- movement, causing the ice to move some distance over the top of the superincumbent gravel before exerting its full downward force. Thus sealed up on the lee side of this pro- A JJOXTH WITH THE MUIR GLACIER. 61 tecting ridge of rock, there would seem to be im limit to the leDgtli of time the forest might be preserved. I see no rea- son why this forest may not have existed before the Glacial pei'iod itself. The existence of other forests similarly' preserved in that vicinity is amply witnessed to by many facts. One upon the island near the west shore, four miles south, is now exposed in a similarly protected p(.i,sition. Furthermore, the moraine, already described on the east side of the inlet, contains much wood ground up into slivers and fragments. Indeed, our whole dependence dui'ing the month for fuel was upon such fragments lying exposed in the moraine. Occasional chunks of peat or compact masses of .sjjJufgnum formed a part of the debris of this moraine. These also occurred on some of the Fig 27 —Shows cPTitral port of the front of Muir Olacier one half mile distant. Near the lower left hand corner the ice is seen one mile distant resting for abont one half mile on travel which it had o\errnn. The ice is now retreating in the channel. The hurfed forests described lie just beyond. (From photograph.) medial moraines on the eastern side. I did not go up them far enough to learn directly their origin ; but, as no forests 62 THE ICE AOE IN NORTH AMERICA. were visible anywhere in that direction, it is presumable that they had been recently excavated from preglacial forests simi- lar in situation to that now exposed on the west below the ice-front. The capacity of the ice to move, without disturbing them, over such gravel deposits as covered the forests, is seen also in the present condition of the southwestern comer of the glacier itself. As the ice-front has retreated along that shore, large masses of ice are still to be seen lapping over upon the gravel. These are portions of the glacier still sus- tained in place by the underlying gravel, while the water of the inlet has carried the ice from the perpendicular bank clear away. This phenomenon, and that of the general per- pendicular front presented by the ice at the water's edge, accord with the well-known fact that the surface of the ice moves faster than the lower portions. Otherwise the ice- columns at the front would not fall over into the water as they do. The formation of kames, and of the knobs and kettle- holes characteristic both of kames and of terminal moraines, is illustrated in various places about the mouth of Muir Gla- cier, but especially near the southwest comer, just above the shoulder of the mountain where the last lateral branch comes in from the west. This branch is retreating, and has already begun to separate from the main glacier at its lower side, where the subglaeial stream passing the buried forest emerges. Here a vast amount of water- worn debris covers the ice, ex- tending up the glacier in the line of motion for a long dis- tance. It is evident from the situation that, when the ice- stream was a little fuller than now, and the subglaeial stream emerged considerably farther down, a great mass of debris was spread out on the ice at an elevation considerably above the bottom. Now that the front is retreating, this subglaeial stream occupies a long tunnel, twenty-five or thirty feet high, in a stratum of ice that is overlaid to a depth, in some places, of fifteen or twenty feet with water-worn glacial debris. In numerous places the roof of this tunnel has broken in, and A MONTH WITH THE MOIB 6LAGIEB. 63 the tunnel itself is now deserted for some distance by the stream, so that the dehris is caving down into the bed of the old tunnel as the edges of ice melt away, thus forming a tortuous ridge, with projecting knolls where the funnels into the tunnel are oldest and largest. At the same time, the ice on the sides at some distance from the tunnel, where the superficial dehris was thinner, has melted down much below the level of that which was protected by the thicker deposit ; and so the debris is sliding down the sides as well as into the tunnel through the center. Thus three ridges approximately parallel are simultaneously forming — one in the middle of the tunnel and one on each side. When the ice has fully melted away, this dehris will present all the comphcations of interlacing ridges, with numerous kettle-holes aiid knobs characterizing the kames ; and these will be approximately parallel with the line of glacial motion. The same condition of things exists about the head of the subglacial stream on the east side, also near the junction of the first branch glacier on the east with the main stream, as also about the mouth of the independent glacier shown on the map lower down on the west side of the inlet (see Fig. 22). The formation of kettle- holes in the terminal ridges has already been referred to. Considerable earthy material is carried out from the front by the bergs. Pebbles and dirt were frequently seen frozen into them as they were floating away. Just how many of the bergs were formed from ice that originally rested on the bot- tom of the inlet I have no means of telling. That some were so formed seems exceedingly probable, if for no other reasons because of the great amount of dehris that was sometimes seen frozen into them. It is by no means certain that the subglacial streams boiling up near the upper comers of the inlet were beneath the lowest stratum of ice. Some small streams were seen pouring out from the face of the ice half- way up from the water. It seems likely that a great amount of sediment is conveyed into cavities in the center of the glacier through the action of these subglacial streams ; and so is ready for transportation when the masses break loose. 64 THE ICE AGE IN NORTH AMERICA. My estimates concerning the amount of sediment carried out by subglacial streams are as follows : The amount of sedi- ment contained in each United States gallon (231 cubic inches) of water collected from the subglacial streams is, as deter- mined by the analysis of the late Professor H. C. Foots, of Cleveland, 708"48 grains. Estimating the total area occupied by the glacial amphitheatre to be 1,200 square miles, and the annual precipitation the same as that at Sitka (which is not far from ninety-six inches), the total amount of water which must in some form annually pass into the inlet from this area is 267,632,640,000 cubic feet. Of this amount I estimate that 77,088,000,000 cubic feet passes out as ice, or, reducing this to water, about 67,000,000,000 cubic feet of liquid water. (This part of the calculation is based on the fact approximate- ly ascertained that a section of ice one mile wide and 1,000 feet deep is moving into the inlet at a rate of 40 feet per day.) Subtracting the ice from the total amount, and esti- mating that evaporation would probably diminish the amount one eighth, the total amount of water which must issue in all the subglacial streams from this glacier is 175,000,000,000 cubic feet. Estimating the specific gravity of the sediment (which is chiefly some compound of alumina and silica) at two and a half, we have, as the total amount of sediment transported thus, 33,274,804 cubic yards. This equals not far from one third of an inch per year eroded from the total area (1,200 square miles) occupied by the glacier. This would furnish one inch of sediment per year to be spread by this single glacier over the bottom of Glacier Bay. This confirms the statements concerning the recent recession of the glacier from the lower portion of the bay, since otherwise it would now be full of sediment. Besides the Muir Glacier there are four others of large size entering the longer inlet to the west (see Fig. 20). These have their origin on the flanks of Mounts Crillon and Fair- weather. They have never been studied, but are apparently as accessible as the Muir. Professor Muir and Rev. Mr. Young are the only well-informed persons who have visited them, A MONTH WITH THE MUIB QLAGIER. 65 and their stay was brief. We went about half-way up the inlet, on its east side, and took some photographs from points where the whole outlines could be seen. We also saw them from the mountains on the east side. The general ap- pearance does not differ materially from that of the Muir Glacier. I append the record of the thermometer from August 20th to August 31st, giving the mean of three readings each day taken at 8 a. m., 2 p. m., and 8 p. m. The temperature of the water in the upper part of the inlet was uniformly 40° Fahr. "August 20, 49 •4° Fahr. August 21, 48-9° Fahr. August 22, 46-1° Fahr. August 2.3, 44-6° Fahr. August 24, 49-8° Fahr. August 25, 52-7° Fahr. August 26, 51-9° Fahr. August 27, 46-1° Fahr. August 28, 50-5° Fahr. August 29, 45° Fahr. August 30, 64-8° Fahr. August 31, 50-5° Fahr. The following is the list of plants, as identified by Pro- fessor Asa Gray, found in bloom about Muir Inlet during the month of August. Where the altitude is not given, they were found near the tide : Arabis ambigua, Brong August 26, 1,600 A. T. Arenarla peploides, L August 38. Astragalus alpinus, L August 7. Hedysarum boreale, Nutt August 28. Sanguisorba Canadensis ~ August 6. Lutkea sibbaldioides, Brong August 27. Saxifraga Lyalli, Engl August 26, 1,600 A. T. Saxifraga stellaris, L August 27, 3,000 A. T. Parnassia fimbriata, Small August 27, 3,000 A. T. Parnassia palnstris, L August 6. Epilobium latifolium, L August 6, 1,600 A. T. Epilobium origanifolium Lam. (?) August 28. Solidago multiradiata, Ait August 27. Erigeron salsuginosus, Gray, arctic form. . August 27, 3,000 A. T. Antennaria margaritacea, arctic form August 27. Achillea millefolium, L., arctic variety August 27. Arnica obtusifolia, Les August 27, 1,200 A. T. Campanula rotundifolia, L., var. Alaskana. August 28. Gentiana platy pefcala (?) August 27. Gentiana Menziesii (?) August 27. Mertensia maritima August 7. 66 THE lOE AQE IN NORTH AMERICA. Oastilleja parviflora, Brong August 28. Salix vestita, Pursh August 6. Habenaria hyperborea, K. Br August 27, 2,650 A. Luzula parviflora, Meyer. Poa alpina, variety vivipara August 26, 1,500 A. Poa alpina, L August 26, 1,600 A. Poa laxa, Ilsenlce August 26, 1,500 A. Phleum alpinum, L August 26, 1,600 A. Elymus mollis August 6. Hordeum, sp. (?) August 6. Fig. 28.— Blocks of stone supported on ice-pillars, showing how they fall toward the sun. See above, page 45. tjnited States Geological Survey (Russell). CHAPTER ly. THE GLACIERS OF GREENLAND. The continental proportions of Greenland, and the ex- tent to which its area is covered by glacial ice, make it by far the most important accessible field for glacial observa- tions. The total area of Greenland can not be less than 500,- 000 sqnare miles — equal in extent to the portion of the United States east of the Mississippi and north of the Ohio. It is now pretty evident that the whole of this area, except a narrow border about the southern end, is covered by one continuous sheet of moving ice, pressing outward on CTcry side toward the open water of the surrounding seas. For a long time it was the belief of many that a large region in the interior of Greenland was free from ice, and was perhaps inhabited. It was in part to solve this problem that Baron ISTordenskiold set out upon his expedition of 1883. Ascending the ice-sheet from Disco Bay, in latitude 69°, he proceeded eastward for eighteen days across a con- tiuuous ice-field. Eivers were flowing in channels upon the surface like those cut on land in horizontal strata of shale or sandstone, only that the pure deep blue of the ice-walls were, by comparison, infinitely more beautiful. These rivers were not, however, perfectly continuous. After flowing for a dis- tance in channels on the surface, they, one and all, plunged with deafening roar into some yawning crevasse, to find their way to the sea through subglacial channels. Numerous lakes with shores of ice were also encountered. " On bending down the ear to the ice," says this explorer, " we could hear on every side a peculiar subterranean hum, G8 THE ICE AGE IN NORTE AMEBIC A. proceeding from rivers flowing within the ice ; and occasion- ally a loud single report like that of a cannon gave notice of ''^ M ^ ^ , «^ JS /^ set 'oioi>ieN3awoN s ^^^-«%^^ ^^ the formation of a new glacier-cleft. ... In the afternoon we saw at some distance from us a well-defined pillar of mist, THE GLACIERS OF GREENLAND. 69 which, when we approached it, appeared to rise from a bot- tomless abyss, into which a mighty glacier-river fell. The vast roaring water-mass had bored for itself a vertical hole, probably down to the rock, certainly more than 2,000 feet beneath, on which the glacier rested." * At the end of the eighteen days, Nordenskiold found himself about 150 miles from his starting-point, and about 5,000 feet above the sea. Here the party rested, and sent two Eskimos forward on shidor — a kind of long wooden skate, with which they could move rapidly over the ice, not- withstanding the numerous small circular holes which every- where pitted the surface. These Eskimos were gone fifty- seven hours, having slept only four hours of the period. It is estimated that they made about 75 miles, and attained an altitude of 6,000 feet. The ice is reported as rising in distinct terraces, and as seemingly boundless beyond. If this is the case 325 miles from Disco Bay, there would seem little hope of finding in Greenland an interior freed from ice. So we may pretty confidently speak of that continental body of land as still enveloped in an ice-sheet. Up to about latitude 75°, however, the continent is fringed by a border of islands, over which there is no continuous covering of ice. In south Greenland the continuous ice-sheet is reached about thirty miles back from the shore. A summary of the results of Greenland exploration was given by Dr. Eink in 1886. From this it appears that, since 1876, one thousand miles of the coast-line have been carefully explored, by entering every fiord and attempting to reach the inland ice. According to this authority— We are now able to demonstrate that a movement of ice from the central regions of Greenland to the coast continually goes on, and must be supposed to act upon the ground over which it is pushed, so as to detach and transport fragments of it for such a distance. ... The plainest idea of the ice-forma- tion here in question is given by comparing it with an inunda- * " Geological Magazine," vol. ix, pp. 393, 399. 70 TME ICE AQE IN NORTH AMERICA. tion. . . . Only the marginal part shows irregularity ; toward the interior the surface grows more and more level, and passes into a plain very slightly rising in the same direction. It has been proved that, ascending its extreme verge, where it has spread like a lava-stream over the lower ground in front of it, the irregularities are chiefly met with up to a height of 3,000 feet, but the distance from the margin in which the height is reached varies much. While under 68^° north latitude, it took twenty-four miles before this elevation was attained ; in 63^° the same height was arrived at in half the distance. . . . A general movement of the whole mass from the central re- gions toward the sea is still continued, but it concentrates its force to comparatively few points in the most extraordinary degree. These points are represented by the ice-fiords, through which the annual surplus ice is carried ofE in the shape of bergs. ... In Danish Greenland are found five of the first, four of the second, and eight of the third (or least productive) class, besides a number of inlets which only receive insignificant fragments. Direct measurements of the velocity have now been applied on three first-rate and one second-rate fiords, all situated between 69° and 71° north latitude. The measure- ments have been repeated during the coldest and the warmest season, and connected with surveying and other investigations of the inlets and their environs. It is now proved that the glacier branches which produce the bergs proceed incessantly at a rate of thirty to fifty feet per diem ; this movement being not at all influenced by the seasons. . . . In the ice-fiord of Jakobshavn, which spreads its enormous bergs over Disco Bay, and probably far into the Atlantic, the productive part of the glacier is 4,500 metres (about 3^ miles) broad. The movement along its middle line, which is quicker than on the sides nearer the shores, can be rated at fifty feet per diem. The bulk of ice here annually forced into the sea would, if taken on the shore, make a mountain two miles long, two miles broad, and 1,000 feet high. The ice-fiord of Tor- sukatak receives four or five branches of the glacier ; the most productive of them is about 9,000 metres (five miles) broad, and moves between sixteen and thirty-two feet per diem. The large Karajak Glacier, about 7,000 metres (four miles) broad. TEE GLACIERS OF GREENLAND. YJ proceeds at a rate of from twenty -two to thirty-eight feet per diem. Finally, a glacier branch dipping into the fiord of Jtivdliarsuk, 5,800 metres (three miles) broad, moved between twenty-four and forty-six feet per diem. * Describing the " Isblink," in latitude 62^° north, Eink The whole surveyed area of the inland ice in this place is calculated at 450 square miles, and forms, by means of the tongued shape of its foremost part, in some measure a separate district, in which the principal changes of the whole margin, excepting the ice-flords, are represented. Toward the interior it is bordered by a row of nunataks,\ distant about forty miles from the seaward edge which our travelers had ascended as their starting-point. Here the origin of the ice over which they had passed was at once plainly visible ; namely, that it could not have been formed on the spot, but was brought thither from the interior of the continent. The nunataks had been an obstacle to this movement ; on the east side, fac- ing the interior, the ice was broken and piled up several hun- dred feet against the rock, like breakers of an ocean, while to the south and north, and between the nunataks, it poured down like frozen waterfalls to be embodied in and leveled with the crust over which our explorers had traveled. . . . The recent explorations, as already mentioned, have proved that what now we designate as coast-land free from ice was for- merly covered with ice like the interior. This ancient ice-cov- ering reached, in the immediate vicinity of the present inland ice, a height of 3,000 to 4,000 feet, and, farther seaward, 2,000 to 3,000 feet above the sea. All the usual traces of ancient ice-action, the erratic blocks and the ground rocks, are the same here as in northern Europe. * See "Transactions of the Edinburgh Geological Society" for February 18, 1886, vol. V, part ii, pp. 286-293. f Nunataks are simply mountain-tops projecting above the surface of the ice-fields, such as were described in the account of the Muir Glacier in Alaska. Nordenskiold was the first to describe them in Greenland, and gave them this 72 TEE ICE AGE IN NORTE AMERICA. Eink supposes the opetiing of new channels for the outlet of the ice through the fiords may have so relieved the interior as to account for this recedence of the ice.* Among the most important observations upon the rate of movement in the glaciers in Greenland are those made by the Norwegian geologist Helland, in the summer of 1875. Durinsc that season he made a series of measurements on the glacier that enters the great Jakobshavn Fiord in the north- ern part of Disco Bay, about latitude 70°. The width of this glacier near its mouth he found to be about two miles and a half. The view from the peaks in the vicinity toward the east extended to a continuous ice-field on the distant horizon. The rate of motion reported by Helland was so great, that scientific men hesitated for some time to credit it. According to his measurements, the Jakobshavn Glacier, in the central portion of its current, was moving more than sixty feet per day, as compared with the three feet per day reported for Alpine glaciers. But the subsequent measure- ments of Steenstrup, given above, and those of my own upon the Muir Glacier in Alaska (made in 1886), amply sustain the conclusions of Helland. It is proper to observe here, again, that the movement of glacial ice is affected much less by the slope of its bottom than by the size of the stream itself. The friction of the ice upon the bottom and sides of its channel is so great, that, where the stream is both shallow and narrow, the motion must be almost completely retarded. On doubling the size * The list of explorers given by Rink is worthy of being honored, and is as follows ; " Geologist K. J. V. Steenstrup (eight summers and two winters) ; Lieutenant G. Holm, of the Royal Navy (five summers and one winter) ; Lieu- tenant R. Hammer, of the Royal Navy (three summers and one winter) ; Lieuten- ant A. D. Jensen, of the Royal Navy (three summers) ; geologist Sylow (two summers) ; painter Groth (two summers) ; supernumerary officer Larsen (one summer) ; Lieutenant Garde, of the Royal Navy (two summers and one winter) ; geologist Knutsen, Norwegian (two summers and one winter) ; geologist Peter- sen (one summer) ; botanist Eberliu (two summers and one winter) ; painter Rus Carstersen (one summer). Steenstrup and Hammer did most on the fiords." THE GLACIERS OF GREENLAND. 73 of a semi-fluid stream, the relative amount of friction be- comes very much less, so that it will move more than twice Flo. 30.— Map of Frederikshaab Glacier, between 62° and 6.3°, showing conrse of Lieuten- ant Jensen in 1878 (forty-seven and a half miles). The black part, ice ; white, land ; shaded, water ; J. N., Jensen's nunataks ; D. N,, Dolager's nunataks ; white lines on the black, crevasses ; arrows, glacier-flow. Five species of plants were found on the nunataks which still survive on the White Mountains (N. H.). Dana. as fast as before. This property of a semi-fluid is made suf- ficiently evident from a homely illustration. Molasses in cold weather will scarcely run at all through a gimlet-hole, while it will run with considerable freedom through an auger-hole. Now, the glaciers of the Alps, which were the subjects of Professor Tyndall's measurements, were, in com- parison to those in Greenland and Alaska, about in the pro- portion of a small gimlet-hole to a large auger-hole, and the faster motion is really not surprising. Y4 TEE ICE AGE IN NORTH AMERICA. Helland's observations as to the amount of ice floating away from the glacier in bergs amply confirm the direct cal- culation. The depth of the Jakobshavn Fiord is about 1,200 feet, so that icebergs of vast size can float off upon its waters. The daily discharge of ice through this fiord was estimated by him to be 432,000,000 cubic feet — about three times the calculation I have made for the Muir Glacier in Alaska. In addition to the formation of large icebergs, the discharge of ice from such a glacier as that at Jakobshavn is doubtless accompanied by a continual cannonade of countless smaller fragments, keeping the heavens full of thundering sounds and the waters full of commotion. From this it fol- lows that the movement of the great glaciers must be rapid, to account for the enormous numbers of first-class icebergs wliich are encountered in the vicinity, and for the numerous and immense ice-floes composed of smaller fragments. "While the attention is fixed on the movement of the gla- ciers, we should not fail to note the uniform presence of subglacial streams of. water emerging from their fronts. Such streams are usually in proportion to the size of the gla- cier, and, as already remarked, are most powerful agencies in the transportation of earthy material. The amount of sedi- ment thus brought out by a single subglacial stream on the west coast of Greenland is estimated to be from 16,000 to 20,000 tons daily ; and the amount of water discharged in the stream is far larger than that which passes off as ice. The existence of such subglacial streams reveals much concerning the condition of the glaciers themselves. The question at once arises, Whence does the water come ? The answer is found in the facts already mentioned by Norden- skiold concerning the superficial streams of water uniformly encountered on penetrating the glaciated interior of Green- land. Doubtless, also, much water arises from the melting of the lower strata of ice through the heat produced by the friction attendant upon the motion. Mr. Whymper's descriptions add vividness to our knowl- edge of the Greenland Glacier in the latitude of Disco : THE GLACIERS OF GREENLAND. 75 In a paper commuincated to the "Alpine Journal" in 1870, I wrote in relation to this part of Greenland and the countrj' to its north and south : The great ioe-covered interior plateau of Greenland can he seen a long way off if the weather is clear. Its summit is almost a dead li m north to south. jjlu wheu one comes neaier to the coast it IS concealed b^ the hdls which are on its outskiits. Th ■nhole of the (ontei ) land on the , J \ /^ in ,(-;-'„,„,-,„ Fig. 31.— A C-!reeiiland Glacier, showing (west) Greenland coast is nlOUn- j^^ formation of icebergs. From a . . 111-1 „„„K fl,/! V,illD painting in possession of the New tainOUS, and although the hills Wk GlogvapWcal Society. scarcely ever, if ever, exceed a height'of 8,000 or 9,000 feet, they effectually conceal the inner or glacier-covered land. This latter is at a distance from the coast varying from ten to sixty or more miles, and, when it is reached, there is an end to land— all is ice, as far as the eye can see. Great as the mass of ice is which still envelops Green- land, there were times when the land was even more complete- ly covered up by it ; indeed, there is good reason to suppose that there was a time when every atom of the country was cov- ered, and that life was hardly possible for man. . . . With the exception of places where the rocks are easy of dismtegra- 76 THE ICE AGE IN NORTH AMERICA. tion, and the traces of glacier action have been to a great ex- tent destroyed, the whole country bears the marks of the grind- ing and polishing of ice ; and, judging t)y the flatness of the curves of the roches moutonnees, and by the perfection of the polish which still remains upon the rocks, after they have sus- tained many centuries of extreme variations of temperature, the Glacial period during which such effects were produced must have vastly exceeded in duration, or severity, the G-lacial period of Europe ; and the existing great interior ice-plateau of Greenland, enormous as it is, must be considered as but the remnant of a mass which was incalculably greater, and to which there is no parallel at the present time, excepting within the Antarctic Circle. And later on, in my book, "Scrambles among the Alps," 1871, pages 246, 247 : " The interior of Greenland apjDears to be absolutely covered by a glacier between 68° 30'-70° north latitude. ... On two occasions, in 1867, I saw, at a glance, at least 6,000 square miles of it from the summits of small mountains on its out- skirts. Not a single peak or ridge was to be seen rising above, nor a single rock reposing upon the ice. The country was completely covered up by glaciers ; all was ice, as far as the eye could see. . . . This vast ice-plateau, although smaller than it was in former times, is still so extensive that the whole of the glaciers of the Alps might be merged into it without its bulk being perceptibly increased. In 1872 I again traveled in northwestern Greenland, and by ascending various lofty mountains saw more of the " inland ice " ; and in the "Alpine Journal " for 1873, page 220, I wrote : From all the principal summits yon perceive the vast gla- cier-clad interior of the country, stretching from north to south in an unbroken line, with a crest as straight as a sea- horizon. There are no marks upon it which enable one to cal- culate the altitude to which it rises, or the dista,nce to which it extends. But having now seen it from several elevated and widely separated positions, as I find that its summit-line always appears lofty, even from the highest mountains which I have ascended, my impression is that its height is generally not less THE GLACIERS OF GREENLAND. 77 than 8,000 feet, and in some places, perhaps, surpasses 10,000 feet. . . . On ascending hills on the outskirts I again had extensive views to the east, finding the land, as before, absolutely cov- ered by glaciers. From the nearest parts to the farthest dis- tance that could be seen, the whole of the ice was broken up into seracs. It was almost everywhere riven and fissured in a most extreme manner, and it was obviously totally impracti- cable for sledges. . . . From the repeated views of the interior which had been seen from the coast mountains, it was clear that all this part of Greenland, except the fringe of land on the Davis Strait side, was absolutely covered by snow and ice, and that the in- terior was not broken up in those latitudes as I had conjectured it might be. . . . This vast glacier is the largest continuous mass of ice at. present known. All the glaciers of the Alps combined are as nothing to it, and the greatest of those in the Himalayas are mere dwarfs in comparison. At Jakobshavn the bergs floating away were often from 700 to 800 feet thick, and this is the only information at present possessed of its depth. The angle at which its surface rises toward the east is very slight, being seldom so much as 8°, and generally much less ; while in some places there are considerable depressions, and lakes are formed in consequence. . . . Mount Kelertingouit was 6,800 feet high, and there was a grand and most interesting view from its summit in all direc- tions. Southward it commanded the whole breadth of the Noursoak Peninsula, and extended over the Waigat Strait to the lofty island of Disco ; westward it embraced the western part of the N"oursoak Peninsula, with Davis Strait beyond ; northward it passed right over theUmenak Fiord (some thirty miles wide) to the Black Hook Peninsula ; to the northeast it was occupied by the fiord, with its many imposing islands and islets, surrounded by innumerable icebergs streaming away from the inland ice ; and in the east, extending from north- east to southeast, over well-nigh 90° of the horizon, there was the inland ice itself— presenting the characteristic features which have been mentioned in the earlier papers. The south- V8 THE 10 E AGE IN NORTE AMEBIC A. em part of the view of the inland ice, as seen from Kelertin- gouit, overlapped the northern part of it as seen on former occasions, while northward it extended to at least 71° 15' north latitude, so I had now viewed the section of the interior be- tween 68° 30' and 71° 15', equal to 190 English miles, and had everywhere found a straight, unbroken crest of snow-covered ice, concealing the land so absolutely that not a single crag appeared above its surface. The height of this straight, unbroken crest of snow was now the object of attention — the principal object for which the ascent was made. On bringing the theodolite to bear upon it, I found that it appeared to be slightly depressed below my station ; but, as it was distant more than one hundred miles, it was only lower in appearance and not in reality. On the assumption that it was no more than one hundred miles dis- tant, after making allowance for the refraction and curvature of the earth, its height was found to be considerably in excess often thousand feet.* Northward from this point we have little knowledge of the glacial conditions in addition to that obtained by the ex- peditions of Drs. Kane and Hayes, between the years 1853 and 1862. These remarkable men were associated from 1853 to 1855, in the second Grinnell Expedition in search of Sir John Franklin, which succeeded in exploring the coast on the east side of Smith Sound from Cape Alexander, in latitude 78°, to Washington Land, in latitude 80° ; while in 1861 and 1862 Dr. Hayes conducted an independent expedi- tion to Lady Franklin Bay, in latitude 82°, and resurveyed portions of his former field. In the neighborhood of Cape Dudley Digges, about lati- tude 76°, Dr. Kane's party encountered a glacier which he describes as follows : This glacier was about seven miles across at its "de- bouche" ; it sloped gradually upward for some five miles back, *" Explorations in Greenland," "Choice Literature," 1884, pp. lid, 258, 308. THE GLACIERS OF GREENLAND. T9 and then, following the irregularities of its rocky substructure, suddenly became a steep crevassed hill, ascending in abrupt terraces. Then came two intervals of less rugged ice, from which the glacier passed into the great mer de glace. On ascending a high, craggy hill to the northward, I had a sublime prospect of this great frozen ocean, which seems to form the continental axis of Greenland— a vast, undulating plain of purple-tinted ice, studded with islands, and absolutely gemming the horizon with the varied glitter of sun-tipped crystal. The discharge of water from the lower surface of the gla- cier exceeded that of any of the northern glaciers except that of Humboldt and the one near Etah. One torrent on the side nearest me overran the ice-foot from two to five feet in depth, and spread itself upon the floes for several hundred yards ; and another, finding its outlet near the summit of the glacier, broke over the rocks and poured in cataracts upon the beach below.* Between "Wolstenholme Sound and Murchison Strait, about latitude 76° 60', Tyndall Glacier comes down to the sea in a broad current ten or twelve miles in width ; while twenty or twenty-five miles to the nortli, on Northumberland Island, a curious glacier is described by Kane, which he calls a " hanging glacier," and named after his brother John. " It seemed," he says, " as if a caldron of ice inside the coast-ridge was boiling over, and throwing its crust in huge fragments from the overhanging lip into the sea below. The glacier must have been eleven hundred feet high ; but even at its summit we could see the lines of viscous movement." f Upon another point in this island a glacier was encount- ered which affords Dr. Kane opportunity to remark upon some points not often noticed. The party had encamped on a low beach at the foot of a moraine which came down be- tween precipitous cliffs of surpassing wildness. "While there, he says : * "Arctic Explorations in the Years 1853, 1854, 1855," vol. ii, pp. 270-272. f Ibid., pp. 259, 260. 80 THE ICE AGE m NORTH AMERICA. I was greatly interested by a glacier that occupied the head of the moraine. It came down abruptly from the central pla- teau of the island, with an angle of descent of more than sev- enty degrees. I have never seen one that illustrated more beautifully the viscous or semi-solid movement of these masses. Like a well-known glacier of the Alps, it had two planes of descent : the upper nearly precipitous for about four hundred feet from the summit ; the lower of about the same height, but with an angle of some fifty degrees ; the two communicating by a slightly inclined platform perhaps half a mile long. This ice was unbroken through its entire extent. It came down from the level of the upper country, a vast icicle, with the folds or waves impressed upon it by its onward motion undisturbed by any apparent fracture or crevasse. Thus it rolled onward over the rugged and contracting platform below, and thence poured its semi-solid mass down upon the plain. Where it encount- ered occasional knobs of rock it passed round them, bearing still the distinctive marks of an imperfect fluid obstructed in its descent ; and its lower fall described a dome, or, to use the more accurate simile of Forbes, a great outspread clam- shell of ice. It seemed as if an interior ice-lake was rising above the brink of the cliffs that confined it. In many places it could be seen exuding or forcing its way over the very crest of the rocks, and hanging down in huge icy stalactites seventy and one hundred feet long. These were still lengthening out by the continuous overflow, some of them breaking off as their weight became too great for their tenacity, others swelling by constant supplies from the interior, but spitting off fragment- ary masses with an unremitting clamor. The plain below these cataractine glaciers was piling up with the debris, while torrents of the melted rubbish found their way, foaming and muddy, to the sea, carrying gravel and rocks along with them. These ice-cascades, as we called them, kept up their din the whole night, sometimes startling us with a heavy booming sound, as the larger masses fell, but more generally rattling away like the random fires of a militia parade. On examining the ice of which they were made up, I found grains of neve larger than a walnut ; so large, indeed, that it was hard to re- THE GLAOTERS OF QREENLAND. 81 alize that they could be formed by the ordinary granulating processes of the winter snows. My impression is, that the sur- face of the plateau-ice, the mer de glace of the island, is made up of these agglomerated nodules, and that they are forced out and discarded by the advance of the more compact ice from higher levels.* The winter of 1853 and 1854 was spent by Dr. Kane in Van Kensselaer Harbor, in latitude 78° 60'. From this point Dr. Hayes and a small party were sent inland for the purpose of securing, if possible, some game to eke out their ship-sup- plies of food. They reported that, "after penetrating the interior about ninety miles, their progress was arrested by a glacier four hundred feet high, and extending to the north and west as far as the eye could reach." On his second ex- pedition, in 1860, Dr. Hayes penetrated this same region again, starting from Port Foulke, about twenty miles to the southwest— venturing, this time, some distance out on the surface of the glacier. The following is his own vivid de- scription of the ice-field, beginning with the narrative of his first expedition. At length we emerged upon a broad plain or valley, wider than any we had yet seen, in the heart of which reposed a lake about two miles in length by half a mile in width, over the transparent, glassy surface of which we walked. On either side of us rose rugged bluffs, that stretched off into long lines of hills, culminating in series in a broad-topped mountain- ridge, which, running away to right and left, was cut by a gap several miles wide that opened directly before us. Immediately in front was a low hill, around the base of which flowed upon either side the branches of the stream which we had followed. Leaving the bed of the river just above the lake, we ascended to the top of this hillock ; and here a sight burst upon us, grand and imposing beyond any power of mine adequately to describe. From the rocky bed, only a few miles in advance, a sloping wall of pure whiteness rose to a broad level plain of ice * "Arctic Explorations," vol. i, pp. 334-336. 82 TEE ICE AGE IN NORTH AMERICA. which, apparently boundless, stretched away toward the un- known east. It was the great mer de glace of the Arctic Con- tinent. At any subsequent period of the cruise this sight would have less impressed me ; but I had never, except in the dis- tance, seen a glacier. Here before us was, in reality, the counterpart of the river- system of other lands. From behind the granite hills the congealed drainings of the interior water- sheds, the atmospheric precipitations of ages, were moving as a solid though plastic mass, down through every gap in the mountains, swallowing up the rocks, filling the valleys, sub- merging the hills— an onward, irresistible, crystal tide, swell- ing to the ocean. Cutting the surface were many vertical crevasses, or gutters, some of great depth, which had drained ofE the melted snow. It was midnight when we made our approach. The sun was several degrees beneath the horizon, and afforded us a faint twilight. Stars of the second magnitude were dimly visi- ble in the northern heavens. When we were within about half a mile of the icy wall, a brilliant meteor fell before us, and, by its reflection upon the glassy surface beneath, greatly heightened the eSect of the scene ; while loud reports, like distant thunder or the booming of artillery, broke at intervals from the heart of the frozen sea. Upon close inspection we found the face of the glacier to ascend at an angle of from thirty to thirty-five degrees. .At its base lay a high snow-bank, up which we clambered about sixty feet ; but beyond this the ice was so smooth as to defy our efforts. The mountains, which stood like giant gate-posts on either side, were overlapped and partially submerged by the glacier. From the face of this a multitude of little rivulets ran down the gutters already mentioned, or gurgled from be- neath the ice, and formed, on the level lands below, a sort of marsh, not twenty yards from the icy wall. Here grew, in strange contrast, beds of green moss ; and in these, tufts of dwarf willows were twining their tiny arms and rootlets about the feebler flower-growths ; and there, clustered together, crouched among the grass, and sheltered by the leaves, and feeding on the bed of lichens, I found a white-blossomed draba THE OLAOIERS OF GREENLAND. 83 which would have needed only a lady's thimble for a flower- pot, and a white chickweed. Dotting the few feet of green around me were seen the yellow blossoms of the more hardy poppy, the purple potentilla, and the white, purple, and yellow saxifrages. This little oasis was literally imbedded in ice. The water which had flowed through it had frozen in the holes, and spread itself out in a crystal sheet upon the rocks and stones around. A few specimens of the tiny blossoms were laid in my note-book, a sprig of heather and a saxifrage were stuck in my button-hole, and with these souvenirs we left this garden- spot which the glacier was soon to cover forever from human eyes. . . . In the autumn of 1860 I was favored with an opportunity to make a more important exploration of this great mer de glace, having from my winter harbor at Port Foulke ascer- tained that it had broken through the mountain-chain at the head of the bay in which my harbor was situated, and was there approaching the sea. Up this glacier, which had thus forced the rocky ramparts, I made my way with a small party of men, attaining an altitude of about 5,000 feet, and extend- ing my observations seventy miles from the coast. The jour- ney possessed the more value that it was entirely novel as re- gards the interior of Greenland. I was finally driven back by a severe gale of wind, which, being accompanied by a sudden fall of temperature, placed my party, for the time, in great Jeopardy, as my tent afforded no shelter ; but I had gone far enough to determine, with some degree of accuracy, the char- acter of the interior ; and the information thus acquired, in connection with my journey with Mr. Wilson in 1853, as just related furnishes an important addition to our knowledge of the great glacier system of the Greenland Continent. Bast- ward from the position attained on both of these journeys no mountains were visible — nothing but a uniform inclined plane of whiteness, a solid sea of ice, hundreds and hundreds of feet in depth, steadily rising until lost in the distance against the sky. A full description of the journey of 1860 has been pub- lished in my " Open Polar Sea." This vast body of ice, now known as Humboldt Glacier, is 84 THE ICE A6E IN NOBTH AMERICA. the largest glacier known, being about sixty miles across, and through at least one half of that extent discharging icebergs. Like the glacier already spoken of as having broken through the mountains near Port Poulke, this Humboldt Glacier has overcome the mountain-barriers, and poured down into the sea between Greenland and Washington Land, which latter is probably an Island, lying in the expansion of Smith Sound (or Strait, as named by Dr. Kane), the water flowing to the eastward of Washington Land being now entirely replaced by the glacier. From Humboldt Glacier the face of the mer de glace sweeps around behind the mountain-chain in a curve toward Port Foulke. At the point reached by Mr. Wilson and myself, the ice was breaking through the mountains, nearly midway between these two extremes of the curve, and will, at some remote period, find its way into Smith Sound through the tortuous valley which now forms the bed of Mary Minturn Eiver. South of Port Foulke the face of the mer de glace forms a series of similar curves of greater or less extent, and through all the great valleys of the Greenland coast-range, glaciers discharge into BaflBn Bay their streams of icebergs. Several of these glaciers are from five to twenty miles across, and those of Melville Bay are doubtless much more exten- sive. * This great Humboldt Glacier enters Peabodj Bay from the east, filling the whole space from latitude 79° to 80°. There is, however, a vast movement of glacier-ice toward this point from the southeast. The face of the Humboldt Glacier is described by Dr. Kane as everywhere, for a distance of more than sixty miles, an " abrupt and threat- ening precipice, only broken by clefts and deep ravines, giving breadth and interest to its wild expression." f The party which first saw this majestic ice-front were com- pelled to traverse its entire breadth on the ice which had formed outside it in the months of September and October. A chief peril of their situation arose from the discharging * "An Arctic Boat-Journey," pp. 10-12, 877, 378. ■f "Arctic Explorations," vol. i, p. 222. THE OLAGIEES OF GREENLAND. 85 bergs of the great glacier which broke up the ice for miles around, at one time producing, directly under their tent, a fissure in the ice on which they had camped for the night. Repeatedly they were compelled to ferry themselves over the cracks in the ice on the bay by rafts of ice. * Kane gives his first impressions of this grand glacier in the following vivid description : I will not attempt to do better by florid description. Men only rhapsodize about Niagara and the ocean. My notes speak simply of the "long, ever-shining line of clifE diminished to a well-pointed wedge in the perspective " ; and, again, of " the face of glistening ice, sweeping in a long curve from the low interior, the facets in front intensely illuminated by the sun." But this line of clifE rose in a solid, glassy wall 300 feet above the water-level, with an unknown, unfathomable depth below it ; and its curved face, sixty miles in length from Cape Agas- siz to Cape Forbes, vanished into unknown space at not more than a single day's railroad-travel from the pole. The interior, with which it communicated and from which it issued, was an unsurveyed mer de glace — an ice-ocean, to the eye, of bound- less dimensions. It was in full sight — the mighty crystal bridge which con- nects the two continents of America and Greenland. I say continents ; for Greenland, however insulated it may ulti- mately prove to be, is in mass strictly continental. Its least possible axis, measured from Cape Farewell to the line of this glacier, in the neighborhood of the eightieth parallel, gives a length of more than 1,200 miles, not materially less than that of Australia from its northern to its southern cape. Imagine, now, the center of such a continent, occupied through nearly its whole extent by a deep, unbroken sea of ice that gathers perennial increase from the water-shed of vast snow-covered mountains and all the precipitations of its atmos- phere upon its own surface. Imagine this, moving onward like a great glacial river, seeking outlets at every fiord and valley, rolling icy cataracts into the Atlantic and Greenland * "Arctic Explorations," vol. i, p. 135. 86 THE ICE AGE IN NORTH AMEBIOA. seas ; and, having at last reached the northern limit of the land that has borne it up, pouring out a mighty frozen tor- rent into unknown arctic space. It is thus, and only thus, that we must form a just concep- tion of a phenomenon like this great glacier. I had looked in my own mind for such an appearance, should I ever be fortu- nate enough to reach the northern coast of Greenland. But, now that it was before me, I could hardly realize it. I had recognized, in my quiet library at home, the beautiful analo- gies which Forbes and Studer have developed between the glacier and the river. But I could not comprehend, at first, this complete substitution of ice for water. It was slowly that the conviction dawned on me that I was looking upon the counterpart of the great river-system of Arc- tic Asia and America. Yet here were no water-feeders from the south. Every particle of moisture had its origin within the polar circle, and had been converted into ice. There were no vast alluvions, no forest or animal traces borne down by liquid torrents. Here was a plastic, moving, semi-solid mass, obliterating life, swallowing rocks and islands, and plowing its way with irresistible march through the crust of an invest- ing sea.* The following summer Dr. Kane visited the scene again, and gives many additional particulars : I had not [he writes] realized fully the spectacle of this stupendous monument of frost. I had seen it for some hours banging over the ice like a white-mist cloud, but now it rose up before me clearly defined and almost precipitous. The whole horizon, so vague and shadowy before, was broken by long lines of icebergs ; and as the dogs, cheered by the cries of their wild drivers, went on, losing themselves deeper and deeper in the labyrinth, it seemed like closing around us the walls of an icy world. They stopped at last ; and I had time, while my companions rested and fed, to climb one of the high- est bergs. The atmosphere favored me : the blue tops of Washington Land [to the north] were in full view, and, 'Arctic Explorations," vol. i, pp. 225-228. THE OLAOIEBS OF GREENLAND. 87 losing itself in a dark water-cloud, the noble head-land of John Barrow. The trend of this glacier is a few degrees to the west of north. We followed its face afterward, edging in for the Greenland coast, about ■ the rocky archipelago which I have named after the Advance. From one of these rugged islets, the nearest to the glacier which could be. approached with any- thing like safety, I could see another island, larger and closer in shore, already half coyered by the encroaching face of the glacier, and great masses of ice still detaching themselves and splintering as they fell upon that portion which protruded. Kepdse was not the characteristic of this seemingly solid mass ; every feature indicated activity, energy, movement. The surface seemed to follow that of the basis-country over which it flowed. It was undulating about the horizon, but as it descended toward the sea it represented a broken plain with a general inclination of some nine degrees, still dimin- ishing toward the foreground. Crevasses, in the distance mere •wrinkles, expanded as they came nearer, and were crossed almost at right angles by long, continuous lines of fracture parallel with the face of the glacier. These lines, too, scarcely traceable in the far distance, widened as they approached the sea until they formed a gigan- tic stairway. It seemed as though the ice had lost its support below, and that the mass was let down from above in a series of steps. Such an action, owing to the heat derived from the soil, the excessive surface-drainage, and the constant abrasion of the sea, must in reality take place. My note-book may enable me at some future day to develop its details. I have referred to this as the escaladed structure of the arctic gla- cier. The indication of a great propelling agency seemed to be just commencing at the time I was observing it. These split- ofE lines of ice were evidently in motion, pressed on by those behind, but still widening their fissures, as if the impelling action was more and more energetic nearer the water, till at last they floated away in the form of icebergs. Long files of these detached masses could be traced slowly sailing off into the distance, their separation marked by dark parallel shadows 88 THE ICE AQE IK NORTH AMERICA. — broad and spacious avenues near the eye, but narrowed in the perspective to mere lines. A more impressive illustration of the forces of Nature can hardly be conceived. . . . The frozen masses before me were similar in structure to the Alpine and Norwegian ice-growths. It would be foreign to the character of this book to enter upon the discussion which the remark suggests ; but it will be seen by the sketch, imperfect as it is, that their face presented nearly all the char- acteristic features of the Swiss Alps. The "overflow," as I have called the viscous overlapping of the surface, was more clearly marked than upon any Alpine glacier with which I am acquainted. When close to the island-rocks, and looking out upon the upper table of the glacier, I was struck with the homely analogy of the batter-cake spreading itself out under the ladle of the housewife, the upper surface less afEected by friction, and rolling forward in consequence. The crevasses bore the marks of direct fracture and the more gradual action of surface-drainage. The extensive water- shed between their converging planes gave to the icy surface most of the hydrographic features of a river-system. The ice- born rivers which divided them were margined occasionally with spires of discolored ice, and generally lost themselves in the central areas of the glacier before reaching its foreground. Occasionally, too, the face of the glacier was cut by vertical lines, which, as in the Alpine growths, were evidently outlets for the surface-drainage. Everything was, of course, bound in solid ice when I looked at it ; but the evidences of torrent- action were unequivocal, and Mr. Bonsall and Mr. Morton, at their visits of the preceding year, found both cascades and water-tunnels in abundance. The height of this ice-wall at the nearest point was about three hundred feet, measured from the water's edge ; and the unbroken right line of its diminishing perspective showed that this might be regarded as its constant measurement. It seemed, in fact, a great icy table-land, abutting with a clean precipice against the sea. This is, indeed, characteristic of all those arctic glaciers which issue from central reservoirs, or m6rs de glace, upon the fiords or bays, and is strikingly in con- trast with the dependent or hanging glacier of the ravines, THE GLACIERS OF GREENLAND. 89 where every line and furrow and chasm seems to indicate the movement of descent and the mechanical disturbances which have retarded it. I have named this great glacier after Alexander von Hum- boldt, and the cape which flanks it on the Greenland coast after Professor Agassiz. The poiat at which this immense body of ice enters the land of Washington gives even to a distant view impressive indications of its plastic or semi-solid character. No one could resist the impression of fluidity conveyed by its peculiar mark- ings. I have named it Oape Forbes, after the eminent crystal- ologist whose views it so abundantly confirms. As the surface of the glacier receded to the south, its face seemed broken with piles of earth and rock-stained rubbish, till far back in the interior it was hidden from me by the slope of a hill. Still beyond this, however, the white blink or glare of the sky above showed its continued extension. It was more difficult to trace this outline to the northward, on account of the immense discharges at its base. The talus of its descent from the interior, looking far off to the east, ranged from seven to fifteen degrees, so broken by the crevasses, however, as to give the efEect of an inclined plane only in the distance. A few black knobs rose from the white snow, like islands from the sea. The general configuration of its surface showed how it adapted itself to the inequalities of the basis-country beneath. There was every modification of hill and valley, just as upon land. Thus diversified in its aspect, it stretches to the north till it bounds upon the new land of Washington, cementing into one the G-reenland of the Scandinavian Vikings and the America of Columbus.* Much less is known concerning the eastern coast of Green- land than about the western coast. For a long time it was supposed that there might be a considerable population in the lower latitudes along the eastern side. But that is now proved to be a mistake. The whole coast is very inhos- * "Arctic Explorations," vol. ii, pp. 146-153. 90 THE ICE AGE IN NORTH AMEBICA. pitable and diflBeult of approach. From latitude 65° to lati- tude 69° little or nothing is known of it. In 1822-'23 Scoresby, Cleavering, and Sabine, hastily explored the coast from latitude 69° to 76°, and reported numerous glaciers descending to the sea-level through extensive fiords, from which immense icebergs float out and render navigation dan- gerous. In 1869 and 1870 the second ISorth German Expe- dition partially explored the coast between latitude 73° and 77°. Mr. Payer^ an experienced Alpine explorer, who ac- companied the expedition, reports the country as much broken, and the glaciers as " subordinated in position to the higher peaks, and having their moraines, both lateral and terminal, like those of the Alpine ranges, and on a still grander scale." Petermann Peak, in latitude 73°, is reported as 13,000 feet high. Captain Koldewey, chief of the expe- dition, found extensive plateaus on the mainland, in latitude 75°, to be " entirely clear of snow, although only sparsely covered with vegetation." The mountains in this vicinity, also, rising to a height of more than 2,000 feet, were free from snow in the summer. Some of the fiords in this vicin- ity penetrate the continent through several degrees of longi- tude. An interesting episode of this expedition was the experience of the crew of the ship Hansa, which was caught in the ice and destroyed. The crew, however, escaped by encamping on the ice - floe which had crushed their ship. From this, as it slowly floated toward the south through sev- eral degrees of latitude, they had opportunity to make many important observations upon the continent itself. As viewed from this uniqiie position, the coast had the appearance every- where of being precipitous, with mountains of considerable height rising in the background, from which numerous small glaciers descended to the sea-level. In 1888 Dr. F. Nansen, with Lieutenant Sverdrup and four others, was left by a whaler on the ice-pack bordering the east of Greenland about latitude 65°, and in sight of the coast. For twelve days the party was on the ice-pack float- ing south, and so actually reached the coast only about lati- THE GLACIERS OP GREENLAND. 91 tude 64°. From this point they attempted to cross the inland ice in a northwesterly direction toward Ohristianshaab. They soon reached a height of Y,000 feet, and were compelled by severe northerly storms to diverge from their course, taking a direction more to the west. The greatest height attained was 9,500 feet, and the party arrived on the western coast at Ameralik Fiord, a little south of Gotthaab, about the same latitude at which they entered. The difficulties of this enter- prise proved so great, and the results are so meager, that no one is likely to cross the ice-field farther north. CHAPTEE V. GLACIERS IN OTHEE PAETS OF THE WOELD. Befoee finally concentrating our attention upon the ancient glaciated area of North America, it will be profit- able to take a glance at existing glaciers in other parts of the world. As is well known, glaciers still envelop the island of Spitzbergen and linger in the mountains of Nor- way and Sweden, of central Europe, and of southern Asia. Yast glaciers also come down to the sea-level in Patagonia, and appear higher up upon the mountains of southern Chili. The mountains of New Zealand * likewise contain numerous groups of glaciers nearly as extensive as those of the Alps. The so-called Antarctic Continent would seem to be covered with one vast sheet of ice pressing outward, and breaking off into immense icebergs. The glaciers of the Alps have been so frequently described that only a few words need be devoted to them here. It is estimated that there are as many as four hundred glaciers in the Alpine range between Mont Blanc and Tyrol, and that, all told, they cover an area of more than 1,400 square miles. In many places the ice is estimated to be 600 feet in thick- ness. The Aletsch Glacier, in the Bernese Oberland, is the longest in the Alps, being not far from twenty miles. Many others are ten miles or more in length, and are often in certain portions of their course from one mile to one mile and a half wide. The line of perpetual snow in the Alps is something * See Whitney's " Climatic Clianges," pp. 269-2'74, to which we are largely indebted for the facts presented in this chapter. GLACIERS ly OTHER PARTS OF THE WORLD. 93 more than 7,500 feet above the sea. The glaciers extend from 4,000 to 5,000 feet lower, though the^limit is by no Fig. 32. — Morteratsch Glacier, Orisons Alps. This glacier advances about seven inches per year. Centuries ago chalets stood a mile farther up the valle}'. In 1868 frag- ments of these ancient dwellings were washed out from underneath the ice. means constant from 3'ear to year. " M. Forel reports, from the data which he has collected with much care, that there have been in this century five periods in the Alpine glaciers : of enlargement, from 1800 (?) to 1815 ; of diminution, from 1815 to 1830 ; of enlargement, from 1830 to 1845 ; of dimi- nution, from 1845 to 1875 ; and of enlargement, again, from 1875 onward. He remarks further that these periods cor- respond with those deduced by Mr. C. Lang for the variations for the precipitations and temperature of the air ; and, con- sequently, that the enlargement of the glaciers has gone for- 9i THE WE AGE IN NORTH AUEUICA. ward ill the cold and rainy period, and tlie diniiimtion in the warm and the dry (' Archives 8ci. Pliys. Nat.,' May 15, 1886, p. 503)."* The _t;'laciers of Scandinavia, with their snow-tields, are estimated to cover a space of about 5,000 square miles. The Fig. 33.— The Svartisen Glacier on the west coast of Norway, juBt within the Arctic cir- cle, at the head of a fiord ten miles from the ocean. Mouth of the glacier one mile wide, and a quarter of a mile back from the water. Terminal moraine in front. (Photofiraphed by Dr. L. C. Warner.) mountains are less lofty than the Alps, the greatest altitude being about 8,500 feet. But tlie more nortlieni latitude and the moist climate are favorable to the production of glaciers. The largest single snow-field is that of Justedal, in latitude 62°, occupying a plateau about 5,000 feet above the seadevel, and an area of 580 square miles. From this plateau t\veiity- f(jur glaciers descend throuo-h the worses leading toward the kJ 0" North C4erman Sea, the largest of which is about five miles long and three quarters of a mile wide. Tlie Fondalen snow- field, in latitude ^\^° or 67°, is of nearlv the same size with * "American Journal ol' Science," vol. C-\xxii, IS.SO, p. 7'?. GLACIERS IN OTHER PARTS OF TEE WORLD. 95 the preceding, and from it glaciers descend to the ocean-level. The Folgefon snow-field, still farther north, occupies an area of about one hundred square miles, from which three glaciers of about the same rank as the preceding descend to the sea. To the north of the Scandinavian Peninsula the islands of Spitzbergen, Nova Zembla, and Franz-Josef Land, all lying above latitude T0°, and the latter north of latitude 80°, are deeply covered with glacial ice in their higher portions. Speaking of Magdalena Bay in Spitzbergen, Dr. G. Hartwig writes : Four glaciers reach down this noble inlet • one, called the Wagon-Way, is 7,000 feet across at its terminal cliff, which is 300 feet high, presenting a magniiicent wall of ice. But the whole scene is constructed on so colossal a scale that it is only on a near approach that the glaciers appear in all their impos- ing grandeur. . . . Besides the glaciers on Magdalena Bay, , Spitzbergen has many others that protrude their crystal walls down to the water's edge ; and yet but few icebergs, and the largest not to be compared with the productions of Baffin Bay, are drifted from the shores of Spitzbergen into the open sea. The reason is that the glaciers usually terminate where the sea is shallow, so that no very large mass if dislodged can float away, and they are at the same time so frequently dismem- bered by heavy swells that they can not attain any great size. * The edge of the coast of the island of Franz-Josef Land is quite generally formed by the precipitous ends of glaciers a hundred feet or more in height and of unknown depth. Iceland, too, has its glaciers in its more elevated portions, though nowhere do they come down to the sea-level. The snow-field of Vatna Jokull, with an extreme elevation of 6,000 feet, has an area of 3,000 square miles. From recent reports it would seem that the glaciers of Iceland have for some time been rapidly advancing. In Asia glaciers are found to a limited extent in the Caucasus Mountains, especially near the central portion of ' Polar World," pp. 135, 136. 96 THE ICE AGE IN NORTH AMERICA. the range, where for a distance of 120 miles the average height is 12,000 feet, while several individual peaks rise higher than 16,000 feet. The snow-line in this range is from 11,000 to 12,000 feet above the sea-level, and in no case do the glaciers descend much lower than the 6,000-foot level. The Ural Mountains — owing probably to their being so nar- row as not to afford space for large snow-fields — are entirely without glaciers. In central Asia, also, notwithstanding the high elevation (the mountains in many cases rising 20,000 feet) there would seem to be no glaciers, because of the dry- ness of the climate. But in the Himalayan range, about the head of the Indus, glaciers of great size are reported. That at the head of the Basha Biver " is over thirty miles in length, its lower part, for a distance of twenty or twenty-five miles, being about a mile and a half in width; above this — for some distance at least — it is still wider. A marked feature of this glacier seems to be its very small inclination ; along a large portion of its course it has an angle of slope of not over one and a half or two degrees. " At the head of the Braldu Valley, an easterly tributary of the Shigar, is one of the largest known glaciers — that of Baltoro. This is said by the officers of the survey to be thirty-five miles in length, ' measured along a central line from its termination up to peak K^' The Biafo Glacier, the foot of which is about ten miles west of the Baltoro, is said to be over forty miles long." * At the heads of the Sutlej and the Ganges similar glacial developments are witnessed, as well as at various other points throughout the whole length of the range. Passing to South America, we find, according to the best reports, that until reaching the southern border of ChiU gla- ciers are infrequent and relatively small. According to Mr. Whymper, no glaciers in Equador descend as low as 12,000 feet above the sea, and the glaciers in that region are largest on the eastern side. Only on Cotopaxi, Chimborazo, and lUin- * Whitney's " Climatic Changes," pp. 284, 286. GLACIERS IN OTHER PARTS OF THE WORLD. 97 issa are the glaciers comparable to those on Mont Blanc. In Chili, in the province of Colchagua, about latitude 35°, glaciers begin to appear, descending somewhat below the 6,000-foot level. " Proceeding southward from Colchagua, we pass into a region in which the climatic conditions are very different from those prevailing in the country farther north. The ranges border the sea very closely, the amount of pre- cipitation increasing and becoming more generally distributed ' throughout the year. The temperature, at the same time di- minishes, and all the conditions favorable to the formation of glaciers are found to prevail. In consequence of this, there is an extensive display of snow and ice along the southern coast of Chili, and especially at the very extremity of the conti- nent." "In Tierra del Fuego," writes Mr. Darwin, "the snow- line descends very low, and the mountain sides are abrupt ; therefore we might expect to find glaciers extending far down their flanks. Nevertheless, when on first beholding, in the middle of summer, many of the creeks on the northern side of the Beagle channel terminated by bold precipices of ice overhanging the salt water, I felt greatly astonished ; for the mountains from which they descended were far from being very lofty." * Darwin's observations upon the glaciers of South America are still standard, and are worthy of fuller reproduction. In his " Voyage of the Beagle " he says : The descent of glaciers to the sea must, I conceive, mainly depend (subject, of course, to a proper supply of snow in the upper region) on the lowness of the line of perpetual snow on steep mountains near the coast. As the snow-line is so low in Tierra del Fuego, we might have expected that many of the glaciers would have reached the sea. Nevertheless, I was as- tonished when I first saw a range, only from 3,000 to 4,000 feet in height, in the latitude of Cumberland, with every val- ley filled -with streams of ice descending to the sea-coast. * Whitney's " Climatic Changes," pp. 272, 273. 98 THE ICE AGE IN NORTH AMERICA. Almost every arm of the sea which penetrates to the interior higher chain, not only in Tierra del Fuego, but also on the coast for 660 miles northward, is terminated by "tremendous and astonishing glaciers," as described by one of the ofScers on the survey. Grreat masses of ice frequently fall from these icy cliffs, and the crash reverberates, like the broadside of a man-of-war, through the lonely channels. These falls pro- duce great waves, which break on the adjoining coasts. It is known that earthquakes frequently cause masses of earth to fall from sea-cliffs : how terrific, then, would be the effect of a severe shock (and such occur here) on a body like a glacier, already in motion, and traversed by fissures ! I can readily believe that the water would be fairly beaten back out of the deepest channel, and then, returning with an overwhelming force, would whirl about huge masses of rock like so much chaff. In Byre's Sound, in the latitude of Paris, there are immense glaciers, and yet the loftiest neighboring mountain is only 6,300 feet high. In this sound about fifty icebergs were seen at one time floating outward, and one of them must have been at least 168 feet in total height. Some of the icebergs were loaded with blocks of no inconsiderable size, of granite and other rocks, different from the clay-slate of the surround- ing mountains. The glacier farthest from the pole, surveyed during the voyages of the Adventure and Beagle, is in latitude 46° 50', in the Gulf of Penas. It is fifteen miles long, and in one part seven broad, and descends to the sea-coast. But even a few miles northward of this glacier, in the Laguna de San Rafael, some Spanish missionaries encountered "many ice- bergs, some great, some small, and others middle-sized," in a narrow arm of the sea, on the 22d of the month correspond- ing with our June, and in a latitude corresponding with that of the Lake of Geneva ! In Europe, the most southern glacier which comes down to the sea is met with, according to Von Buch, on the coast of Noj'way, in latitude 67". Now, this is more than 30° of lati- tude, or 1,330 miles, nearer the pole than the Laguna de San Eafael. The position of the glaciers at this place and in the Gulf of Penas may be put even in a more striking point of view, for they descend to the sea-coast, within 7^° of latitude, GLACIERS IN OTHER PARTS OF THE WORLD. 99 or 450 miles, of a harbor, where three species of oliva, a voluta, and a terebra are the commonest shells, within less than 9° from where palms grow, within 4|^° of a region where the jaguar and puma range over the plains, less than 2^° from arborescent grasses, and (looking to the westward in the same hemisphere) less than 2° from orchidaceous parasites, and within a single degree of tree-ferns ! * Mr. Darwin's experience was so similar to that of those who visit Alaska (see Chapter III, page 46), that another extract will prove especially instructive by way of compari- son. Speaking of the Straits of Magellan, he says : The lofty mountains on the north side compose the granitic axis, or backbone of the country, and boldly rise to a height of between 3,000 and 4,000 feet, with one peak above 6,000 feet. They are covered by a wide mantle of perpetual snow, and numerous cascades pour their waters, through the woods, into the narrow channel below. In many parts, magnificent glaciers extend from the mountain-side to the water's edge. It is scarcely possible to imagine anything more beautiful than the beryl-like blue of these glaciers, and especially as contrasted with the dead white of the upper expanse of snow. The frag- ments which had fallen from the glacier into the water were floating away, and the channel, with its icebergs, presented for the space of a mile a miniature likeness of the Polar Sea. The boats being hauled on shore at our dinner-hour, we were ad- miring from the distance of half a mile a perpendicular clifE of ice, and were wishing that some more fragments would fall. At last down came a mass with a roaring noise, and imme- diately we saw the smooth outline of a wave traveling toward us. The men ran down as quickly as they could to the boats ; for the chance of their being dashed to pieces was evident. One of the seamen just caught hold of the bows as the curling breaker reached it. He was knocked over and over, but not hurt ; and the boats, though thrice lifted on high and let fall again, received no damage. This was most fortunate for us, for we were a hundred miles distant from the ship, and we 'Voyage of the Beagle," edition of 18'72, pp. 245-247. 100 THE ICE AGE IN NORTH AMERICA. should have been left without provisions or fire-arms. I had previously observed that some large fragments of rock on the beach had been lately displaced, but, until seeing this wave, I did not understand the cause. One side of the creek was formed by a spur of mica-slate ; the head by a clifE of ice about forty feet high ; and the other side by a promontory fifty feet high, built up of huge rounded fragments of granite and mica- slate, out of which old trees were growing. This promontory was evidently a moraine, heaped up at a period when the gla- cier had greater dimensions.* Of the glaciers of New Zealand the following succinct account of Whitney must sufiBce : On the western coast of the southern island, between the parallels of 42° and 45°, rises abruptly from the sea a grand range of mountains, the culminating point of which, Mount Cook, is about 13,000 feet in elevation. Along this chain, for a length of about one hundred miles, are developed numerous groups of glaciers, some of which are not much inferior in size to the largest of those of the Alps. The Tasman Glacier is said by Haast, who first scientifically explored and described these mountains, to be ten miles in length and a mile and three quarters broad at its termination, the lower portion, for a distance of three miles, being covered with morainic de- tritus, f Of the so-called Antarctic Continent little is known ; but icebergs of great size are frequently encountered up to 58° south latitude, in the direction of Cape Horn, and as far as latitude 33° in the direction of Cape of Good Hope. Tlie number and size of these, of which more particulars will be given presently, are such as to necessitate an extensive area of glaciers about the south pole. Nearly all that is known of the Antarctic Continent was discovered by Sir J. C. Eoss during the period extending from 1839 to 1843, when, between the parallels of 70° and 78° south latitude, he en- * " Voyage of the Beagle," edition of 1872, pp. 224, 22!). t " Climatic Changes," pp. 273, 274. GLAOIERS IN OTHER PARTS OF TEE WORLD. 101 countered in his explorations a precipitous mountain-coast, rising from 7,000 to 10,000 feet above tide. Through the valleys intervening between the mountain-ranges huge gla- ciers descended, and " projected in many places several miles into the sea, and terminated in lofty perpendicular clifEs. In a few places the rocks broke through their icy covering, by which alone we could be assured that land formed the nucleus of this, to appearance, enormous iceberg." * Again, speaking of the region in the vicinity of the lofty volcanoes Terror and Erebus, between 10,000 and 12,000 feet high, the same navigator says : " We perceived a low white line extending from its extreme eastern point as far as the eye could discern to the eastward. It presented an extraordinary appearance, gradually increasing in height as we got nearer to it, and proving at length to be a perpendicular clifE of ice, between 150 and 300 feet above the level of the sea, perfectly flat and level at the top, and with- out any fissures or promontories on its even seaward face. What was beyond it we could not imagine ; for, being much higher than our mast-head, we could not see anything except the summit of a lofty range of mountains extending to the southward as far as the seventy-ninth degree of latitude. These mountains, being the southernmost land hitherto dis- covered, I felt great satisfaction in naming after Sir Edward Parry. . . . Whether Parry Mountains again take an easterly trending and form the base to which this extraordinary mass of ice is attached, must be left for future navigators to deter- mine. If there be land to the southward, it must be very remote, or of much less elevation than any other part of the coast we have seen, or it would have appeared above the bar- rier." This ice-cliff or barrier was followed by Captain Eoss as far as 198° west longitude, and found to preserve very much the same character during the whole of that distance. On the lithographic view of this great ice-sheet given in Boss's work it is described as "part of the South Polar Barrier, to 180 * Quoted by Whitney in "Climatic Changes," p. 314. 102 TSE IGE AGE IN NORTB AMERICA. feet above the sea-level, 1,000 feet thick, and 450 miles in. length." A similar vertical wall of ice was seen by D'Urville, off the coast of Adelie Land. He thus describes it : ''Its appearance was astonishing. We perceived a cliff having a uniform eleva- tion of from 100 to 150 feet, forming a long line extending off to the west. . . . Thus for more than twelve hours we had followed this wall of ice, and found its sides everywhere per- fectly vertical and its summit horizontal. Not the smallest irregularity, not the most inconsiderable elevation, broke its uniformity, for the twenty leagues of distance which we fol- lowed it during the day, although we passed it occasionally at a distance of only two or three miles, so that we could make out with ease its smallest irregularities. Some large pieces of ice were lying along the side of this frozen coast ; but, on the whole, there was open sea in the offing." * In the absence of satisfactory direct observations upon the glaciers of the Antarctic Continent, we are permitted to turn to an important source of indirect evidence furnished by the icebergs encountered in the region. Many of these are of such size as to indicate an enormous depth to the glacial ice of which they are fragments, and imply the existence of a glaciated area larger even than Greenland. In reading the accounts of icebergs we should bear in mind that the specific gravity of ice is such that where there is one cubic foot of an iceberg above the water's surface there are seven or eight cubic feet below the surface ; so that, if the form of the berg could be supposed to be symmetrical, we should multiply the height of the berg above water by eight or nine to get its perpendicular dimension. But as the forms of the ice- bergs are usually irregular, this rule can not always be ap- plied. In several of the instances to be referred to, however, the masses are so large, and the forms so regular, that we can not be far amiss in applying the rule. Some of these masses of floating ice are of almost incredible size, and their * Quoted by Whitney in " Climatic Clianges," pp. 315, 316. OLAUIERS IN OTHER PARTS OF THE WORLD. 103 origin can only have been upon large surfaces of land in every way favorably situated for the accumulation of snow and the formation of glaciers. The following are a few of the examples reported a century ago by Captain Cook : At eight o'clock saw an island of ice to the westward of us, being then in the latitude of 50° 40' south, and longi- tude 3° 0' east of the Cape of Good Hope. Soon after, the wind moderated, and we let all the reefs out of the top-sails, got the sprit-sail-yard out, and top-gallant-mast up. The weather coming hazy, I called the Adventure by signal under my stern ; which was no sooner done, than the haze increased so much, with snow and sleet, that we did not see an island of ice, which we were steering directly for, till we were less than a mile from it. I judged it to be about fifty feet high, and half a mile in circuit. It was flat at the top, and its sides rose in a perpendicular direction, against which the sea broke ex- ceedingly high. . . . At one o'clock we steered for an island of ice, thinking, if there were any loose ice round it, to take some on board, and convert it into fresh water. At four we brought to, close un- der the lee of the island ; where we did not find what we wanted, but saw upon it eighty-six penguins. This piece of ice was about half a mile in circuit and one hundred feet high and upward, for we lay for some minutes with every sail be- calmed under it. . . . At nine in the morning we bore down to an island of ice which we reached by noon. It was full half a mile in circuit, and two hundred feet high at least, though very little loose ice about it. But while we were considering whether or not we should hoist out our boats to take some up, a great quantity broke from the island. Upon this we hoisted out our boats, and went to work to get some on board. The pieces of ice, both great and small, which broke from the island, I observed, drifted fast to the westward ; that is, they left the island in that direction, and were, in a few hours, spread over a large space of sea. . . . Finding here a good quantity of loose ice, I ordered two boats out, and sent them to take some on board. While this 104 THE ICE AGE AV NORTH AMERICA. was doing, the island, which was not less than half a mile in circuit, and three or four hundred feet high above the surface Fig. 84.— Iceberg. of the sea, turned nearly bottom up. Its height, by this circumstance, was neither increased nor diminished appar- ently. . . . In the evening we had three islands of ice in sight, all of them large ; especially one, which was larger than any we had yet seen. The side opposed to us seemed to be a mile in ex- tent ; if so, it could not be less than three in circuit. As we passed it in the night, a continual cracking was heard, occa- sioned, no doubt, by pieces breaking from it. For, on the morning of the 6th, the sea, for some distance round it, was covered with large and small pieces ; and the island itself did not appear so large as it had done the evening before. It could not be less than one hundred feet high ; yet such was the impetuous force and height of the waves which were broken against it, by meeting with such a sudden resistance, that they rose considerably higher.* For a series of years the Board of Trade in England col- lected statistics from the navigators of the Southern Ocean who reported icebergs encountered in their voyages. From ■ Voyage round the World," pp. 20, 29, 48-50, 54. 0LAGIER8 IN OTHER PARTS OF TEE WORLD. 105 these reports, and from a paper of Mr. Towson upon the subject, published by the Board of Trade, Mr. Croll makes the following collection of facts concerning them, premising that, " with one or two exceptions, the heights of the bergs were accurately determined by angular measurement " : September 10, 1856. — The Lightning when in latitude 55° 33' south, longitude 140° west,, met with an iceberg 430 feet high. November, 1839.— In latitude 41° south, longitude 87° 30' east, numerous icebergs 400 feet high were met with. September, 1840.— In latitude 37° south, longitude 15° east, an iceberg 1,000 feet long and 400 feet high was met with. February, 1860. — Captain Clark, of the Lightning, when in latitude 55° 20' south, longitude 122° 45' west, found an iceberg 500 feet high and three miles long. December 1, 1859. — An iceberg, 580 feet high, and from two and a half to three miles long, was seen by Captain Smith- ers, of the Edmond, in latitude 50° 53' south, longitude 43° 58' west. So strongly did this iceberg resemble land that Captain Smithers believed it to be an island, and reported it as such, but there is little or no doubt that it was in reality an iceberg. There were pieces of drift-ice under its lee. November, 1856. — Three large icebergs, 500 feet high, were found in latitude 41° 0' south, longitude 42° 0' east. January, 1861. — Five icebergs, one 500 feet high, were met with in latitude 55° 46' south, longitude 155° 56' west. January, 1861.— In latitude 56° 10' south, longitude 160° 0' west, an iceberg 500 feet high and half a mile long was found. January, 1867. — The bark Scout, from the west coast of America, on her way to Liverpool, passed some icebergs 600 feet in height and of great length. April, 1864. — The Royal Standard came in collision with an iceberg 600 feet in height. December, 1856.— Four large icebergs, one of them 700 feet high, and another 500 feet, were met with in latitude 50° 14' south, longitude 42° 54' east. 106 THE ICE AGE IN NORTH AMERICA. December 25, 1861. — The Queen of Nations fell in with an iceberg in latitude 53° 45' south, longitude 170° 0' west, 720 feet high. December, 1856. — Captain P. Wakem, ship Ellen Eadford, found, in latitude 53° 31' south, longitude 43° 43' west, two large icebergs, one at least 800 feet high. Mr. Towson states that one of our most celebrated and talented naval surveyors informed him that he had seen ice- bergs in the southern regions 800 feet high. March 33, 1855. — The Agneta passed an iceberg in latitude 53° 14' south, longitude 14° 41' east, 960 feet in height. August 16, 1840. — The Dutch ship, General Baron von Geen, passed an iceberg 1,000 feet high in latitude 37° 33' south, longitude 14° 10' east. May 15, 1859.— The Eose worth found, in latitude 53° 40' south, longitude 133° 17' west, an iceberg as large as " Tristan d'Acunha." * Upon these facts Mr, OroU remarks : In the regions where most of these icebergs were met with, the mean density of the sea is about 1'0356. The density of ice is •93. The density of icebergs to that of the sea is there- fore as 1 to 1"115 ; consequently, every foot of ice above water indicates 8*7 feet below water. It therefore follows that those icebergs 400 feet high had 3,480 feet under water — 3,880 feet would consequently be the total thickness of the ice. The icebergs which were 500 feet high would be 4,850 feet thick, those 600 feet high would have a total thickness of 5,830 feet, and those 700 feet high would be no less than 6,790 feet thick, which is more than a mile and a quarter. The iceberg 960 feet high, sighted by the Agneta, would be actually 9,313 feet thick, which is upward of a mile and three quarters. Although the mass of an iceberg below water compared to that above may be taken to be about 8 '7 to 1, yet it would not be always safe to conclude that the thickness of the ice be- low water bears the same proportion to its height above. If the berg, for example, be much broader at its base than at its ' Climate and Time," pp. 382-885. OLACIERS IN OTHER PARIS OF THE WORLD. 107 top, the thickness of the ice below water would bear a less proportion to the height above water than as 8'7 to 1. But a berg, such as that recorded by Cap- tain Clark, 600 feet high and three miles long, must have had only 1 to 8 "7 of its total thickness above water. The same remark ap- plies also to the one seen by Cap- tain Smithers, which was 580 feet high, and so large that it was taken for an isl- and. This berg must have been 5,628 feet in thick- ness. The enor- mous berg which came in collision with the Royal Standard must have been 5,820 feet thick. It is not stated what length the icebergs 730, 960, and 1,000 feet high respectively were ; but supposing that we make considerable allowance for the possibility that the proportionate thickness of ice below water to that above may have been less than as 8-7 to 1, still we can hardly avoid the conclusion that the icebergs were considerably above' a mile in thickness. But if there are icebergs above a mile in thickness, then there must be land-ice somewhere on the southern hemisphere of that thickness. In short, the great antarctic ice-cap must in some places be over a mile in thick- ness at its edge. Fig. 35. -Floating berg, showing the proportions above and under the water. CHAPTEK VI. SIGNS OF FOEMEE GLACIATION. Befoee attempting to delineate the exact southern boundary of the ice during the height of the Glacial period in North America, it will be necessary briefly to discuss the evidence upon which the inferences concerning the Glacial period are based. The reader will ask : How is it possible to determine, with any reasonable degree of accuracy, the extent of the region formerly covered by glacial ice, but which has been free from such covering for many thousand years, and during all that time has been subjected to the disintegrating and modifying influences connected with the ceaseless operation of the ordinary forces of IsTature ? The consideration of this question will introduce us not only to some of the most interesting problems of this par- ticular subject, but to some of the fundamental principles underlying all inductive reasoning. The study of the great Ice age, like all other branches of geology, deals with the eifects of past causes. From the marks which have been left upon the surface of the earth, we endeavor by scientific processes to reproduce to our imagination the condition of things which would account for these marks. As reasonable beings, we are compelled to bring into the field of thought a past cause sufficient to produce all the results observed, both positive and negative ; and when our imagination has found ail adequate cause, true science compels us to rest with that. From observation upon living glaciers, and from the known nature of ice, we may learn to recognize the track SIG^S OF FOSiJUS OLAGIATIOX. 109 of a glacier as readily and unmistakably as we would the familiar foot-prints of an animal. Tlie indications iipon wliich glacialists have depended for their information as to the ex- tent of the glaciated region during the great Ice age are of three kinds : 1. (rrooves and scratches preserved upon the rocks in place and upon the bowlders and pebbles shoved along under the ice. 2. The extensive unstratiiied deposits called " till," which are traceable to glacial action. 3. Trans- ported material found in such positions tliat it must have been left by glacial ice rather than floating ice. In respect of the nature of ice we are compelled to ad- mit that it is capable of motion like such semi-fluids as cool- Fia. 36.-Scratched stnno from the till of Boston Xaturol size about one foot and a halt long by ten inches wide. (From photograph.) ing pitch or lava. But, though it does move, it is not capa- ble of adapting itself so perfectly as a real fluid to the SIGNS OF FORMER OLA CIA TI OK HI inequalities of the country. From this comparatively solid character of ice certain important results must follow. It is easy to see that the stones of all sizes, while being dragged along underneath the ice, would be held in a comparatively firm grasp so as to be polished and striated and scratched in a peculiar manner. On the shores of bays and lakes and in bottoms of streams we find that the stones are polished and rounded in a symmetrical manner, but are never scratched. The mobility of water is such that the edges and comers of the stones are rubbed together by a force acting successively in every possible direction. But in and under the ice the firm grasp of the stifE semi-fluid causes the stony fragments to move in a nearly uniform direction, so that they grate over the underlying rocks like a rasp, wearing down the rocks beneath and slowly grinding them to powder, and, at the same time, being worn themselves in the process. From the stability of the motion of such a substance as ice there would, from the nature of the case, result groovings and striation both on the rocks beneath and on the bowlders and pebbles which, like iron plowshares, are forced over them. Scratched surfaces of rock and scratched stones are therefore, in ordinary cases, most trustworthy indications of glacial action. The direction of the scratches upon these glaciated bowlders and pebbles is, also, worthy of notice. The scratches upon the loose pebbles are mainly in the direc- tion of their longest diameter — a result which follows from a mechanical principle, that bodies forced to move through a resisting medium must swing around so as to proceed in the line of least resistance. Hence the longest diameter of such moving bodies will tend to come in line with the direc- tion of the motion. A scratched surface is, however, not an infallible proof of the former presence of a glacier where such a surface is found, or, indeed, of glacial action at all. A stone scratched by glacial forces may float away upon an iceberg, and be deposited at a great distance from its home. Indeed, ice- berg's and shore-ice may produce, in limited degree, the phe- 112 THE IGE AGE IN NORTE AMERICA. nomena of striation whicli we have just described. One can but admire the enthusiasm with which the old defenders of the iceberg theory dwelt upon the capacity of icebergs and shore-ice to polish, groove, and scratch the surfaces over which they moved. Sir Charles Lyell tells us, in the ac- count of his first visit to America, how he stood at the foot of a cliff at Cape Blomidon, Nova Scotia, transfixed at the sight of recent furrows which were the exact counterpart of the grooves of ancient date which he had elsewhere described. So extensive were these, that they seemed for the moment to render the glacial theory unnecessary : As I was strolling along the beach at the base of these ba- saltic clifEs, collecting minerals, and occasionally recent shells at low tide, I stopped short at the sight of an unexpected phe- nomenon. The solitary inhabitant of a desert island could scarcely have been more startled by a human foot-print in the sand, than I waS on beholding some recent furrows on a ledge of sandstone under my feet, the exact counterpart of those grooves of ancient date which I have so often described in this work, and attributed to glacial action. After having searched in vain at Quebec for such indications of a modern date, I had despaired of witnessing any in this part of the world. I was now satisfied that, whatever might be their origin, those before me were quite recent. The inferior beds of soft sandstone which are exposed at low water at the base of the clifE at Cape Blomidon, form a broad ledge of bare rock, to the surface of which no sea-weed or barnacles can attach themselves, as the stone is always wear- ing away slowly by the continual passage of sand and gravel, washed over it from the talus of fallen fragments which lies at the foot of the cliff on the beach above. The slow but con- stant undermining of the perpendicular cliff forming this promontory, round which the powerful currents caused by the tide sweep backward and forward with prodigious velocity, must satisfy every geologist that the denudation by which the ledge in question has been exposed to view is of modern date. Whether the rocks forming the cliff extended so far as the water's edge, ten, fifty, or one hundred years ago, I have no SIGNS OP FORMER QLAOIATION. 113 means of estimating ; but the exact date and rate of destruc- tion are immaterial. On this recently-formed ledge I saw sev- eral straight furrows half an inch broad, some of them very nearly parallel, others diverging, the direction of the former being north 35° east, or corresponding to that of the shore at this point. After walking about a quarter of a mile, I found another set of similar furrows, having the same general direc- tion within five degrees ; and I made up my mind that, if these grooves could not be referred to the modern instrumentality of ice, it would throw no small doubt on the glacial hypothesis. When I asked my guide— a peasant of the neighborhood— whether he had ever seen much ice on the spot where we stood, the heat was so excessive (for we were in the latitude of the south of France, 45° north), that I seemed to be putting a strange question. He replied that, in the preceding winter of 1841, he had seen the ice, in spite of the tide, which ran at the rate of ten miles an hour, extending in one uninterrupted mass from the shore where we stood to the opposite coast at Parrs- borough, and that the icy blocks, heaped on each other, and frozen together or "packed" at the foot of Cape Blomidon, were often fifteen feet thick, and were pushed along, when the tide rose, over the sandstone ledges. He also stated that frag- ments of the "black stone " which fell from the summit of the cliff, a pile of which lay at its base, were often frozen into the ice, and moved along with it. I then examined these fallen blocks of amygdaloid scattered around me, and observed in them numerous geodes coated with quartz-crystals. I have no doubt that the hardness of these gravers, firmly fixed in masses of ice, which, although only fifteen feet thick, are often of considerable horizontal extent, have furnished sufficient pressure and mechanical power to groove the ledge of soft sandstone.* Stofies are also striated by other agencies than moving ice. Extensive avalanclies and land-slides furnish conditions analo- gous to those of a glacier, and might in limited and favor- able localities simulate its results. In those larger geological * " Travels in America," first series, vol. ii, pp. 144-146. §-■2 5 K'-C — d » bb O ©CO ''-'^-^ ^ SB' (ii,£! a 5 S — CO) OS 9£-C bo SIGN'S OF FORMER LACTATION-. 116 movements, also, where the crust of the earth is broken and the edges of successive strata are slioved over each other, a species of striation is produced wlaich in technical terras is called a sliokenside. Occasionally this deceives the inex- perienced or incautious observer. But by due pains all these semblances may be- detected and eliminated from the prob- lem, leaving a sufficient number of unquestionable phenom- ena due to true glacial action. A second indubitable mark of glacial action is found in the character of the deposit left after the retreat of the ice. Ice and water differ so much from each other in the extent of their fluidity, that ordinarily there is little danger of confus- ing the deposits made by them. A simple water deposit is inevitably stratified. The coarse and fine material can not be deposited by water alone, simultaneously in the same place. Along the shores of large bodies of water the deposits of solid material are arranged in successive parallel lines, the material growing finer and finer as the lines recede from the shore. The force of the waves is such in shallow water that they move pebbles of considerable size. Indeed, where the waves strike against the shore itself, vast masses of rock are oftentimes moved by the surf. But, as deeper water is reached, the force of the waves becomes less and less at the bottom, and so the transported material is correspondingly fine, until, at the depth of about seventy feet, the force of the waves is entirely lost ; and beyond that line nothing will be deposited but fine mud, the particles of which aTe for a long while held in suspension before they settle. In the deltas of rivers, also, the sifting power of water may be observed. Where a mountain-stream first debouches upon a plain, the force of its current is such as to move large pebbles, or bowlders even, two or three feet in diameter. But, as the current is checked, the particles moved by it be- come smaller and smaller until in the head of the bay, or in the broad current of the river which it enters, only the finest sediment is transported. The difference between the size of material transported by the same stream when in flood and 116 TBE ICE AGE IN NORTE AMERIOA. when at low water is very great, and is the main agent in pro- ducing the famihar phenomena of stratification. During the time of a flood vast hodies of pebbles, gravel, and sand are pushed out by the torrent over the head of the bay or delta into which it pours ; while during the lower stages of water only fine material is transported to the «ame distance ; and this is deposited as a thin film over the previous coarse deposit. Upon the repetition of the flood another layer of coarser material is spread over the surface ; and so, in successive stages, is built up in all the deltas of our great rivers a series of stratified deposits. In ordinary circumstances it is impos- sible that coarse and fine material should be intermingled in a water deposit without stratification. Water moving with various degrees of velocity is the most perfect sieve imagi- nable ; so that a water deposit is of necessity stratified. To this general principle, however, exception must be made in the case of accumulations taking place slowly in deep water containing icebergs from which bowlders and pebbles may be dropped. These, of course, will be found distributed through the fine deposit without stratification. It is thought by President Chamberlin,* however, that in cases where the bowlders dropped from the icebergs are of marked irregularity in their configuration, their position in the ooze at the bottom would betray their origin. Flattish stones, in falling through the water, would often descend edgewise, and would not uniformly lie in the mud upon their flat surface. I have myself observed numerous places in southern Ohio where the arrangement of the limestone fragments abundantly illustrates and confirms this theory. The horizontal position of such fragments in the clay of that region seems to show that they were arranged in the deposit by a moving ice-sheet, and were not dropped from floating ice. It is evident that ice is so nearly a solid that the earthy material deposited by it must be unassorted. The mud, sand, gravel, pebbles, and bowlders, dragged along underneath a * " Terminal Moraine of tlie Second Glacial Epoch," p. 297. \-''".4''' ^ ■■' o — o SO » : I -• • . -« L i jh (fi' i I i iii B m ii fi I M ii ■ ..... ■■— 118 THE ICE-AGE I?r NORTH AMERICA. moving stream of ice, must be left in an unstratified condi- tion — tlie coarse and the fine being indiscriminately mingled together. Now, this is the character of the extensive deposits of loose material which cover what we designate as the gla- ciated region. It is true that over this region there are exten- sive stratified deposits. But these invariably mark the situa- tion of abandoned lakes and water-courses. To these interest- ing formations a special chapter will be devoted. But the larger part of the region marked upon the map as glaciated is covered with an unstratified deposit, in which are mingled a variety of materials derived from rocks both of the locality and of far-distant regions. Moreover, the pebbles in this de- posit are the most of them polished and scratched after the manner of those which we know to have been subjected to glacial action. Ordinarily there can be no question of the glacial character of this formation, even when no considera- tions are taken into account except those which appear in the deposit itself. Still, in certain situations, floating ice may transport coarse material, and drop it in the midst of finer silt, so as closely to simulate a true unstratified deposit. In \ the majority of cases, however, the configuration of the coun- try is such as to exclude the agency of floating ice from tbej problem. I "We come, therefore, in the third place, to a mass of recently discovered facts which would seem to place the glacial theory above all question. When once the Kmit of these unstratified deposits, containing striated stones and transported material, had been accurately determined, it was found that the margin was exceedingly irregular in two re- spects. The southern edge of this deposit is both serrate and crenate — that is, it does not follow a straight east-and- west line, but in places withdraws to the north, and in others extends in lobe-shaped projections far to the south. This constitutes its serrate character. But it is the crenate cliar- acter of its southern border which is of most significance. The southern border, with its indentations and projections, is not determined by any natural barrier. The southern SIGN'S OF FORMER OLACIATION. 119 boundary -line rises from the level of the sea at New York to the height of the Bine Mountains in New Jersey, and descends into the valley of the Delaware, and rises again over the Blue Bidge in Pennsylvania, crossing the valley between it and Poeono Mountain, where it runs for many miles at an elevation of 2,000 feet above the sea, and descends in a nearly straight line to the East Branch of the Susque- hanna at Beach Haven, where it is not more than 500 feet above the sea. Thence it continues onward in a diagonal course across the Alleghanies, with their various subsidiary valleys, to its great turning-point in southwestern New York. Thence to the trough of the Mississippi, for a distance of 700 or 800 miles, the line winds gradually down the west- ern flanks of the Alleghanies, paying little attention in its deflections to the minor inequalities of the country. "West of the Mississippi the rise is equally gradual to northern Dakota and Montana, where the glacial border is 2,000 or 3,000, and in British America 4,000, feet above the sea. ) i Thus it is not possible, by the supposition of any conceiv- able submergence, to account for this line of demarkation. "Water which would have floated ice from the north to almost any point along that line would have floated it farther south. Consequently, the line must have been determined not by a barrier which restrained a fluid, but by the irregular losses of momentum such as would take place in a semi-fluid mov- ing in the line of least resistance from various central points of accumulation. The reader will flnd the subject of this chapter treated with great fullness by President Chamberlin in the " Third Annual Keport of the United States Geological Survey," pp. 291-402, and in the " Seventh Keport," pp. 149-248, where the subject of rock-scoring is most fully illustrated. Numer- ous illustrations in the present volume have been reserved for the latter part of Chapter X, on " Glacial Erosion and Trans- portation," which the reader will do well to consult in this connection as well as in the order in which it occurs. CHAPTER YII. BOTJNDAEY OF THE GLACIATED AREA IN NORTH AMERICA. Doubtless the Ice age both began and ended in a great number of local glaciers which became confluent and con- tinuous only during the middle of the period. One of the most interesting evidences of the independent movement of the different portions of the great North American ice-sheet is to be found in the driftless region of southwestern Wis- consin. Here is an area of several hundred square miles in extent, occupying more or less of the adjoining area in. Illi- nois, Iowa, and Minnesota, which remained as an island in the great continental expanse of ice. The ice moved past it upon both sides, and then closed together upon the south, and moved onward, a distance of about 300 miles, to the vicinity of St. Louis. When, a few years ago, attention was first directed by Mr. Clarence King,* Mr. Warren Upham,f and Professor George H. Cook;]: -to the terminal moraines of southern New England and northern New Jersey, by President T. C. Cham- berlin* to the character and connection of the kettle-mo- raine in Wisconsin, and by Dr. George M. Dawson || to the * See my paper in the " Proceedings of the Boston Society of Natural His- tory," vol. xix, pp. 60-63. f " New Hampshire Geological Report," vol. iii, pp. 300-305. t " Report upon the Geology of New Jersey for ISYS." * " On the Extent and Significance of the Wisconsin Kettle-Moraine." II " On the Superficial Geology of the Central Region of North America," from the " Quarterly Journal of the Geological Society," vol. xxxi, 1875. This is a. summary of a portion of the author's " Report on the Geology and Re- sources of the Forty-ninth Parallel," 1875. BOUNDARY OF THE GLACIATED AREA. 121 significance of the extension of the Missouri cotean in Brit- ish America, hopes were at once raised that a distinct line of terminal moraines might be traced across the continent. With this theory in mind, the late Professor H. Carvill Lewis and myself began the survey of Pennsylvania in 1881. But, upon crossing the Alleghanies and pursuing the investi- gations in the Mississippi Valley, I was compelled to aban- don this view, and to be content with finding marginal de- posits more evenly spread over the country, ending, in some cases, in an extremely attenuated border. And, upon reflec- tion, the fallacy in our original theory, that there must be a terminal moraine — that is, a noticeable ridge of glacial ac- cumulations to mark the farthest extent of the ice — is easily seen. The extent of a glacial deposit at any particular point will be determined by three factors, namely : 1. The amount of accessible loose material in the line of glacial movement which the ice can seize upon and transport. It is evident that, if the rocks over which the ice moves are hard and smooth and already denuded of loose material, there may be much motion of ice with little transportation of soil. 2. The length of time during which the ice-front remains at a given point, since time acts as a multiplier. 3. The exemption of the deposit from the action of denuding agen- cies. When a glacier melts, the torrents of water ai'ising may, in a short time, tear down and distribute as sediment to distant valleys the material accumulated by the slow movement of centuries. Indeed, it has been questioned by some whether the larger part of the grist of the glacier has not been thus ti'ansported far beyond the extreme limits reached by the ice itself. This transportation by water from the front of glaciers is certainly of immense extent. Every subglacial stream is surcharged and milky-white with sedi- ment as it emerges from the ice-front. As before stated, a traveler in Washington Territory, from Portland to Seattle, can detect the presence of glaciers in the Cascade Mount- ains, scores of miles away, simply by the milky color of 122 THE ICE AGE IN NORTH AMERICA. the streams crossed by the railroad. At Tacoma, on Puget Sound, the milk-white water, coming down from distant gla- ciers in Mount Tacoma, struggles with the dark-blue water of the sound for the occupancy of the harbor, and gives the surface of the bay the nondescript appearance of an im- mense slice of marble-cake. As one passes the mouth of the Stickeen River beyond Fort Wrangel, in Alaska, the line of demarkatioii between the clear waters of the ocean and those of the glacier-laden currents of the river is as plain as that between the water itself and the shore. One of the most interesting occupations of the leisure hours of our long encampment in Glacier Bay was to watch this struggle iov occupancy between the milk-white water of the four sub- glacial streams pouring into the inlet, and the pure blue waves urged against it by the recurring tjdes of the ocean. With a rising tide of twenty-two feet, the line of demarka- tion between the glacial water and the waters from the Pacific moved alternately backward and forward over the inlet for a distance of one or two miles, and in the shallow water, miles away, the screw of the steamer brought to the surface great quantities of the sediment which is rapidly filling up the bay. "When one considers the constancy of the operation of this cause during all glacial time, he may well be pardoned for regarding the glacial debris still remaining upon the con- tinent as but an insignificant remnant of the total amount transported and deposited by glacial action. During the whole continuance of the Glacial period in North America, subglacial streams must have sent their turbid currents down through every New England outlet, and through the Hud- son, the Delaware, the Susquehanna, and all the northern tributaries of the Mississippi. The terraces marking these glacial water-courses retain simply a part of the coarser ma- terial transported ; the fine material went constantly onward to the sea, helping to build up the immense delta of the lower Mississippi, and to line the whole coast of the Atlantic with a deposit of fine sediment ready at some day to rise above the surface as fruitful soil. BOUNDARY OF THE GLACIATED AREA. 123 With these remarks, we are prepared to come to the spe- cific subject of the present chapter, namely, the character and extent of the glacial deposits marking the southern bor- der of the glaciated area in North America. Through a por- tion of the distance these accumulations are so marked as to merit the name of terminal moraines. Through another portion tbat name is hardly applicable to anything near the glacial border. In a subsequent chapter there will be a dis- tinct discussion of the whole question of moraines. In this it is our purpose to follow somewhat minutely the boundary of the area, and detail its various aspects. Off the coast of Maine the ice, at its culminating period, extended an unknown distance into the sea, surmounting the eminences of Mount Desert and all that rock-bound coast, and leaving its terminal deposits in water so deep that there is little hope of ever determining its exact situation. But in southeastern Massachusetts the deposits emerge from the water as true moraines, and offer themselves as most interest- ing objects of study. Nantucket, Tuckernuck, Chappaquid- dick, Martha's Vineyard, No Man's Land, and Block Island are but portions of the extreme terminal moraine whose back emerges at these points from the water. Cape Cod, from Provincetown to Wood's HoU and the Elizabeth Islands, is a similar remnant of a vast moraine formed after the ice-front had withdrawn a short distance to the north. Indeed, the whole of Plymouth and Barnstable counties is " made land," as really as that of the Back Bay in Boston, only in the one case the earth was dumped, day by day, from the laborer's cart, and m the other year by year, from the melting front of the continental ice-sheet. It is an instance of misleading poetic license which per- mits us to sing of the " rock-bound " shore upon which the Pilgrim Fathers landed, for there are no rock-bound shores in southeastern Massachusetts. The hills which first greeted the eyes of the Pilgrim Fathers are the irregular morainic accumulations so frequently characteristic of glacial margins. In this case the soil composing them consists of sand, gravel. BOUNDARY OF THE GLACIATED AREA. 125 and bowlders which have been scraped off by the ice from the mountains and ledges of New Hampshire and the inter- vening portions of Massachusetts, transported to the glacial margin, and there deposited in such quantities as to consti- tute the whole southeastern portion of the latter State. The three hundred and sixty lakes of Plymouth township are nothing else than a cluster of kettle-holes. Manomet Hill, southeast of Plymouth, is not a mountain which has been thrust up by convulsive agencies, nor yet a remnant of a par- tially eroded plateau, but a glacial deposit, hundreds of feet in height and many miles in extent. From the fact of its running nearly at right angles to the backbone of Cape Cod, Manomet Hill is spoken of by some as a medial moraine. But it is doubtful if it is necessary so to regard it. The re- treat of a glacier, like the retreat of an army, is determined in part by the nature of the opposing foe. In the present instance there are abundant reasons for believing that the ice retreated by the left flank ; for it is evident that the ocean had much to do in setting bounds to the general southeastern movement of the ice-sheet in New England. From the con- tour of the coast, it can be seen at a glance that the waters of the ocean had constant opportunity to eat in upon the ice from the east as well as from the south. Nantucket marks the extreme southeastern Hmit. Here the ice maintained its position against opposing forces, until the outer line of moraines just mentioned was built up. The next line of defense taken up by the ice is that marked by the backbone of Cape Cod. The retreat had been farthest on the side most exposed to the ocean, and hence the dis- tance between the moraine of Nantucket and that at South Brewster is much greater than the distance between Martha's Yineyard and Wood's HoU. The next line of defense taken np by the ice-front along the course of Manomet Hill, left Cape Cod and the whole shore of eastern Massachusetts open to the undisputed sway of the ocean. The two lines of moraine so clearly marked in south- eastern Massachusetts can be readily traced westward through 126 THE ICE AGE IN' NORTE AMEBIOA. a considerable portion of Long Island. The exterior line, beginning at Montank Point, forms the backbone of the island as far as Brooklj'n, N". Y., the city itself being built upon it. The interior or parallel line, represented to the east by the backbone of Cape Cod and the Elizabeth Islands, disappears beneath the deeper waters of Buzzard's Bay, to emerge upon the mainland at Point Judith in Rhode Island, and give variety to the whole coast of the State westward from that point. This part of the moraine presents many features of special interest at Watch Hill, Fisher's and Plum Islands, and on the northern shore of Long Island as far as Port Jefferson. Westward from Port Jefferson only the external line of moraine hills is traceable, Staten Island in New York Harbor being a most interesting development of it. The northern and western portions of this island are covered with the peculiar combination of rounded knobs and circular depressions characteristic of moraines, while the southeastern portion of the island was just beyond the reach of the ice, and the deposits upon it are of an entirely differ- ent character. Up to this point in our investigations some doubt may attach to the inferences concerning the limits of the great ice-sheet. For, since the ocean everywhere expands to the south, it may be asked. What certainty is there that its waters do not cover a belt of glacial deposits still farther out than those now visible ? No positive answer can be made to this objection. But, on striking the coast of New Jersey opposite Staten Island, a fair field of investigation is at once offered, and doubt as to the substantial correctness of the delineation farther on need be no longer entertained. The line of moraine hills across New Jersey is a direct continua- tion of those forming the backbone of Long Island, and covering the northern half of Staten Island; Here they form a sharp line of demarkation between the glaciated region to the north and the unglaciated plains to the south. Beginning at Perth Amboy, the moraine bends northward through Baritan, Plainfield, Chatham, Morris, and Hanover, Fig. 41.— The glaciated portion is shaded. The shading on the Lehigh and Delaware Rivers indicates glacial terraces, which are absent from the Schuylkill. 128 TEE ICE AGE IN NORTH AMERICA. to Rockaway, thence a little south of west to Belvidere on the Delaware, a few miles above Easton. The innumerable throngs of passengers between New York and Philadelphia can not, after their attention has been once called to the facts, fail to notice this moraine as the southward-bound trains pass through it, and emerge into a level, sandy region free from bowlders and all irregular drift deposits. At Me- tuchen and at Plainfield the transition is almost as clearly marked as that between land and water. Before following the terminal belt farther west, where its characteristics are more or less disguised by the local topography, we will pause to consider more carefully some of the main characteristics of it as so far traced. That these hills constitute a true moraine is evident from the fact that they are composed of loose material such as, both from the nature of the case and from observation, we know is actually deposited wherever the front of a glacier rests for any great length of time. A considerable portion of them consists of material which has been transported from various localities to the north, and deposited without any stratification. Some of the bowlders are unworn and angular, as if having been carried upon the back of the gla- cier. Others are partially rounded and scratched in such a manner as to show that they have been forced along through the mass of sand and gravel which everywhere underlay the moving field of ice. Sections, however, frequently show in these hills a limited amount of stratification. But this is not at all surprising, when we consider the manner of their forma- tion ; for the ice itself to a certain extent forms barriers to, and furnishes channels for the running water which its own melting provides, and so would itself afiord the conditions necessary to a partial stratification of its own deposits. The terminal moraine where best developed may almost be said to consist of innumerable ridges, kiiolls, and kettle- holes. The kettle-holes are of all sizes, and are situated in every imaginable position with reference to the general de- posit; some of them, low down toward the base of the BOUNDARY OF THE GLACIATED AREA. 129 moraine, are filled to the rim with water ; others are but par- tially filled ; while others are for the greater portion of the year completely dry. The angle at which the earth forming their sides stands is usually as sharp as the nature of the material will allow, and bowlders are as frequently found upon the inside of the rim as upon the outside. The origin of kettle-holes has already been explained ; but it is in place here to remark upon some of the general considerations supporting the theory already advanced.* It is evident, from an inspection of the depressions themselves, that they can not be the result of erosion, since the depres- sions are too irregular and too deep to have been formed by the plunging movement of water, and the material is too coarse for a water deposit. In some respects kettle-holes resemble what are called sink-holes, frequent in limestone regions, where a great amount of material below the surface is removed in solution, leaving numerous caverns whose roofs eventually sink in, and form the depressions characteristic of such regions. But kettle-holes abound in regions where no such cavei'us could have been formed, and are distributed over the country according to a method which could not have originated by the action of underground currents of water. While the iceberg theory was in favor to account for the drift, it was not uncommon to hear these kettle-holes spoken of as places where icebergs had stranded, and in turning round and round had bored holes in the bottom of the ocean- bed over which they were floating ; but, now that the ice- berg theory is abandoned, and observations are more extended, the origin of kettle-holes is readily understood as an inevita- ble part of the glacial theory itself. Any one who will in the early spring-time take pains to observe the melting of masses of ice which have been covered by ashes and other refuse, or which have been partially buried beneath the debris of earth which some spring torrent has brought down from a neighboring hili; will find before him a very perfect * See above, p. 62 et seq. 130 TUB ICE AGE IN NORTH AMERICA. object-lesson as to the formation of kettle-holes. All the elements for their production are there, in and beneath the accumulated debris. As the heat slowly penetrates the pro- tective covering, especially upon the sides, where it is the thinnest, it melts the ice and thus undermines the earthy material, which, in due time, slides down to the base, and thus gradually leaves a cone of ice in the middle, surrounded at the edges by a continuous ridge of dirt. Eventually the ice all melts away, and a miniature kettle-hole is formed. So far as 1 know, the application of this principle to the expla- nation of the extended phenomena under consideration was first made by the late Colonel Whittlesey,* of Ohio, in study- ing the Kettle range of Wisconsin. How completely this theory was confirmed by my study of the Muir Glacier, in the summer of 1886, has already been related.-j- To prevent misapprehension, it should here be remarked that we have not intended to affirm that the whole bulk of Martha's Vineyard and Long Island consists of glacial depos- its. The nuclei of those islands certainly existed before the Glacial period, for at Gay Head, on Martha's Vineyard, and in the vicinity of Port Jefferson, on Long Island, and at some other places, there are extensive beds of Tertiary clay underlying the glacial deposits, and rising above the water- level. The glacial deposits simply form a capping of more or less thickness to these older ones. It is still true, however, that the glacial deposits have determined, in the main, the present topographical features. Eeturning now to the extreme terminal moraine, where we left it at the Delaware River, we will note its features farther westward. In following the moraine across New Jersey, we have already had opportunity to observe how in- dependent it is of the local topography. Beginning at the sea-level, its base rises to a height of more than a thousand feet, where it crosses the Blue Hills of western New Jersey ; * Report upon " The Drift Formations of the Northwest," in '' Smithsonian Contributions to Knowledge,'' 1866. \ See above, p. 53 e< seq. BOUNDARY OF THE GLACIATED AREA. 131 and everywhere its surface is characterized by the knobs and kettle-holes, whose manner of formation has just been ex- plained. Crossing the Delaware Kiver, these characteristic phe- 132 THE ICE AGE IN NORTH AMERICA. nomena arc developed in a marked degree upon tlio hills of JSTorthampton coimty, Pa., ninniiig up to the Ixise of the Kit- tatiim}' Momitaiii, some miles below the Delaware Water- Gap. West of the mouiitaiu, in Monroe county, the valley south of Stroudsburg is for several miles filled with the FiQ. 43.— A glacialcil iiflihle, iiuturnl size, from the moraine on Pocono platean, Pennsyl- vania, two thouHaiul feet above the sea, Monroe county. The striie along the longest diameter are well marked. same characteristic ridges, knobs, and kettle-holes. On going still fartlier west, and rising suddenly tifteen hundred feet to the plateau of Pocono Mountain — the southern extension of BOUNDARY OF THE GLACIATED AREA. 133 the Catskills— we find still the same characteristic deposits slightly moditied by the difterent conditions. A range of hills, sweeping around this plateau for ten oi twelve miles in a magnificent semicircle open toward the south, was discov- ered, m ISSl, l)y Professor Lewis and myself to belong to the extreme and true terminal moraine of the continental ice- sheet. South of it one may go for miles upon a level, sandy plateau (about two thonsand feet above tide) without en- countering a bowlder or any foreign material ; while the low range of hills, seventy-five or a hundred feet in height, is Fig. -W.— The previou.s pcbblt; \'iewed from the edge. The reversed side \\as free from glacial marks. ~ literally formed of bowlders ; among which may readily be recognized those of granitic origin, wrenched from ledges hundreds of miles to the north, and transported hither across the valley of the Mohawk or over the broad expanse of Lake 134 TEE ICE A OE IN NORTH AMERICA. Ontario. Nicely ensconced within their wooded shores, there are here also the numerous lakelets so often found occupying kettle-holes, and forming the sources of streams which issue from the mountain-side to water the valleys below. There are few more interesting regions in which to study an ancient terminal moraine than the plateau of Pocono Mountain, be- tween Tobyhanna and Tunkhannock townships, in Monroe county, Pa. Passing still farther westward, and descending to the North Fork of the Susquehanna, twenty miles below "Wilkes- barre, we find a remarkable terminal accumulation of glacial debris filling the valley to a great depth for some distance above Beach Haven. In the immediate vicinity of the river, which was one of the great lines of drainage during the Gla- cial epoch, the material is more or less modified by water- action ; but the unmodified moraine can be clearly traced far up on either side of the valley. A few miles to the west- ward, on the other side of Huntingdon Mountain, and but ten or twelve miles south of the southern wall of the Alleghanies, the line of terminal deposits can be traced past Asbury, through a whole township and some miles beyond, up the eastern side of Fishing Creek. Through all this distance, on the east side of the creek, there is nothing upon the surface but these irregular and confused features of a well-marked terminal deposit ; while on the west side of the creek, not a sign of glacial action can be seen. But suddenly, above the town of Benton, the belt of confused deposits of stone and gravel begins to descend the eastern side, and after crossing the valley diagonally, rises, like a broken-down Chinese wall, upon the western side. It is important to make a cautionary remark at this point. The boundary so far in eastern Pennsylvania as here delin- eated is that which was determined by Professor Lewis and myself in 1881, and it is quite possible that there is a small margin of error in our delineation. This was in the early part of our work upon the survey, and we did not at that time give the attention which we afterward gave to what we BOUNDARY OF TEE GLACIATED AREA. 135 called the " fringe "—that is, to the scattered remnants of gla- cial deposit lying for an indefinite distance in front of the more pronounced morainal deposit. While I am confident that the fringe does not in any case extend many miles in that vicinity, still I am not sure bnt the apparent definiteness of our outline has led to some serious misapprehension ; for, as we shall see later, Mr. McGee attributes certain deposits found upon the hills five hundred feet above the Susquehanna River, a short distance in front of our so-called terminal mo- raine, to the action of a glacial flood, and is led to assume a subsidence * of the region for which there does not seem to be other facts to warrant. My present impression is, that our failure to follow the extreme limit of glacial action in eastern Pennsylvania has led to a serious error on this point. The death of Professor Lewis prevents our availing ourselves of his abundant stores of information gathered on subsequent visits. But, notwithstanding the fact that he visited \ many of the places south of our line where glacial marks have been reported, and demonstrated that they were not due to glacial action, but to other causes, he makes the following significant remarks at the close of his report : In reviewing all the facts regarding the fringe presented in the foregoing pages, it will be noticed . . . that its elevation throughout is about that of the moraine back of it, although it is sometimes somewhat higher. ISTo striae have as yet been found in the narrow region cov- ered by the fringe. The hill-tops bearing the bowlders of the fringe are in all respects similar to those south of the limit of all drift, and bear no evidence of glacial erosion. * "American Journal of Science," vol. cxxxt, 1888, p. SYB et seq. See, also, the "Seventh Annual Keport of the United States Geological Survey for 1885 and 1886," pp. 53Y-646, where Mr. McGee has treated of the deposits at the head of Chesapeake Bay with great fullness. From this we learn that the Columbia formation at that point does not rise much above the twohundred- foot level ; that its mean thickness does not exceed twenty feet (p. 606) ; that the material is all from the Susquehanna Valley, and that large bowlders occur more frequently there than in the formation on the streams farther south. f "American Journal of Science," vol. cxxviii, 1884, p. 276 ei seq. 136 THE ICE AGE IN NORTH AMERICA. Nor is the material of the fringe confined to any particular elevation occurring irregularly at different heights. These facts, puzzling when first observed, and at first causing some trouble in determining the line of the accumulations of till, bowlders, and scratched stones, designated the moraine, came afterward to be commonly looked for as invariable phenomena. It is possible that traces of this fringe might be detected in eastern Pennsylvania and in New Jersey. In fact, occasional transported bowlders do occur upon several hill-tops just in front of the moraine in the vicinity of the Susquehanna and Delaware Eivers and in New Jersey, which I find it difficult to explain upon any theory of a flood, and which may be of like origin with the fringe as developed farther west. Facts observed by other geologists in more western States, and published since this report was written, confirm my im- pression that this fringe is destined to play an important part in glacial geology.* To this I will add my own testimony in contirmation. In a recent drive with Professor George H. Cook, who so early traced the moraine across New Jersey, I saw what I took to be unquestionable evidence of this fringe a few miles south of the well-defined moraine ; and in central Ohio it was a com- mon oci^urrence to find upon the hill-tops scattered remnants of glacial deposits some miles south of the more pronounced deposits of till, and yet which could not be accounted for from the action of water, being entirely above the range of the terraces. Later on, more will be said about the signifi^- - cance of the fringe, but it seems important to say this much here, in order to guard the reader against supposing that we are able to trace the southern border of the ice with strict mathematical exactness. Tine line is everywhere approximate, but I have yet to learn that there are facts showing any great margin of error in the line as we have laid it down. From Luzerne county on, the varied fortunes of the ter- minal moraine are diflScult of description. The northern-cen- * " Report on the Terminal Moraine in Pennsylvania and Western New York, Second Geological Survey of Pennsylvania, Z," p. 201. BOUNDARY OF THE GLACIATED AREA. 137 tral counties of Pennsylvania are covered with forests, and are cut up into gorges extremely difficult of access. Never- theless, in Lycoming county, upon a plateau at Eose Valley, three or four miles to the east of Lycoming Creek, and sev- eral hundred feet above it, large kettle-holes with their sur- rounding ridges of gravel, and their accompanying bowlders and striated rock-surfaces, are marked features of the land- scape at a height of 2,000 feet above the sea ; while upon the west side of the creek these features are totally absent for some miles above. Similar developments are met, at occasional intervals, all along the line of the great conti- nental divide in Potter county, whence the waters flow to the widely separated regions of the Gulf of St. Lawrence, Chesapeake Bay, and the Gulf of Mexico. The same pro- nounced features already described mark the terminus of the great ice-sheet where it crossed the valley of the various tribu- taries entering the Alleghany Hiver from the north, in Cat- taraugus county, N. Y. In that State these are specially noticeable at Ellicottville, at Little Yalley, and upon the high lands upon either side of the Eastern Branch of Cone- wango Creek near Randolph. As the glacial boundary swings back again to the Pennsylvania line, it is marked by numerous and impressive accumulations in Warren county, one of the most accessible and notable localities being where it crosses the valley of Conewango Creek at Ackley, a few miles above its junction with the Alleghany River at Warren. Here the marginal line of glacial deposits may be clearly seen as it descends from the highlands on the east diagonally to the valley, filling it to a great depth, and rising over the hills in its onward course to the southwest. At this point in the valley, as at some other places which could be mentioned, the northern side of the moraine is more abrupt than the southern. It was here that we made some of our first and most exact discoveries as to the limit of the ice west of the AUe- ghanies. Professor Lewis and myself, who were pursuing the investigations together, had already learned the charac- 138 THE ICE AGE IN NORTH AMERICA. teristies of the great lines of glacial drainage extending to the south of the glacial limit, and of which more particular mention will be made in a future chapter ; but it was here that we first detected the exact relation of these lines of gla- cial drainage to the great ice- movement. Coming up from Warren toward the glacial boundary, the valley of the Cone- wango is seen to be iilled increasingly full of gravel deposits arranged in terraces on either side of the small stream. The height of these above the stream is sixty or seventy feet. At first the gravel is rather fine, but is constituted largely of water- worn granitic fragments, which could have come with- in range of the stream only by glacial transportation from the far north. On proceeding a few miles farther north, these terrace deposits become more and more irregular, being thrown np into great ridges, and at the same time the mate- rial becomes less water-worn and much coarser. Below this point we had searched diligently in the gravel for scratched stones, but none were anywhere to be found. A mile or two from Ackley, however, we began to find pebbles upon which the glacial scratches could be dimly traced. They had been partially water-worn, but had not been rolled far enough to completely obliterate their glacial marks. Upon reaching Ackley, we found the whole valley occupied by a true gla- cial deposit, terminating abruptly to the north, and through which the stream had cut a narrow channel. The moraine ridge, or dam, as we might call it, rises from 100 to 150 feet above the present bed of the stream, and is equally well de- veloped upon both sides. Scratched stones and granitic frag- ments abound, and it is well marked upon either side of the creek by the characteristic kettle-holes. Above Ackley, for many miles, the Conewango pursues a sluggish course, and is bordered by extensive marshy land. The explanation of all this is that the ice-front remained for a somewhat pro- tracted period at Ackley, allowing the large accumulations immediately below to take place. But its retreat from Ack- ley for many miles northward was too rapid to permit of any marked terminal accumulations ; and the dam of glacial de- BOUNDARY OF TEE GLACIATED AREA. 139 iris, doubtless, maintained for a time an extensive lake in the valley of the Conewango above this point. As it would be impossible, within reasonable limits, to dwell upon everything of interest along this extreme termi- nal line, we will content ourselves with briefly glancing at a few other points. Those wishing to pursue the subject are referred to the detailed accounts already published in the va- rious State geological reports and in the scientific periodicals.* ^ Through Columbiana county, Ohio, as in the adjoining counties of Pennsylvania, there is, to the south of the heavy accumulations of till, a fringe of thinner glacial deposits from one to three miles wide. Over this margin there are scat- tered evidences of glacial action, consisting of granitic bowl- ders and patches of till here and there upon the highlands, * See G. F. Wright, " The Glacial Boundary in Ohio, Indiana, and Ken- tucky," Western Reserve Historical Society ; also " Ohio Geological Report," vol. V, pp. 750-771; T. C. Chamberlin, "Extent, etc., of the Wisconsin Ket- tle Range " (" Transactions of the Wisconsin Academy of Sciences," vol. iv) ; "Geological Survey of Wisconsin," vols, ii and iii (Madison, 1877-'80); "Pre- liminary Paper on the Terminal Moraine of the Second Glacial Epoch " (United States Report); George H. Cook, "Annual Reports for New Jersey for 1877, 1878"; G. M. Dawson, "On the Superficial Geology of British Columbia" (" Quarterly Journal of the Geological Society, 1878," vol. xxxv, p. 89) ; J. W. Dawson, "Changes of the Coast-Level in British Columbia ' ("Canadian Natu- ralist," April, 1877); G. K. Gilbert, "On Certain Glacial and Post-Glacial Phe- nomena of the Maumee Valley, Ohio" ("American Journal of Science, 1871," vol. ci, p. 339); C. H. Hitchcock, "Moraines of North America" ("Popular Sci- ence Monthly," 1881) ; Clarence King, " United States Geological Explorations of the Fortieth Parallel," vol. i ; " Systematic Geology " (Washington, 1878) ; Lewis and Wright, " Report on the Terminal Moraine in Pennsylvania and Western New York, Second Geological Survey of Pennsylvania, Z " ; Warren tJpham, " The Northern Part of the Connecticut Valley in the Champlain and Terrace Periods" ("American Journal of Science, 1877," vol. cxiv, p. 459); "The For- mation of Cape Cod" ("American Naturalist, 1879," vol. xiii, pp. 489, 552); " Geological Survey of New Hampshire," vol. iii (1878) ; " Geological and Natu- ral History Survey of Minnesota " (Report for 1879) ; " Terminal Moraines of the North American Ice-Sheet" ("American Journal of Science, 1879," vol. cxviii, pp. 81, 197); Charles Whittlesey, "Fresh- Water Drift of the Northwest- em States" ("Smithsonian Contributions, 1866," vol. xv); N. H. Winchell, "The Drift Deposits of the Northwest" ("Popular Science Monthly," vol. iii, pp. 202, 286) ; " Geology of Minnesota " (Annual Reports, 1872, etc.). 140 THE IGE AOE IN NORTH AMERICA. at an elevation of from 300 to 500 feet above the Ohio River. North of this fringe the till is continuous, and every- where of great depth. At Palestine, on the eastern edge of the county, and at New Alexandria, near the western side. Fib. 45.— Map of the glaciated region of Ohio, showing a part of its extension in Kentuclvy. wells are reported in the till fifty feet deep. This is upon the highest land in that part of the country, and the glacial deposits are marked, in a moderate degree, by the knobs and kettle-holes characteristic of the moraine upon the south shore of New England. A mile or two west of Canton, in Stark BOUNDABT OF THE GLACIATED AREA. 141 county, the accumulations of glaciated material are iipon a scale equal to anything upon Cape Cod. The northern part of Holmes county is covered with till, which is everywhere of great depth, and in numerous places near the margin dis- plays, though in a moderate degree, the familiar inequalities of the New England moraine. After the southern deflec- tion in Knox county, the glaciated region is entered near Danville, from the east, on the Columbus, Mount Yernon, and Akron Eailroad, through a cut in till a quarter of a mile long, and from thirty to forty feet in depth. At the old village of Danville, near by, upon a neighboring hill, wells are reported as descending more than a hundred feet before reaching the bottom of the till. Through Licking county, both north and south of Newark, the depth of the glacial envelope is great up to a short distance of its east- ern edge. At the old canal reservoir, in Perry county, the characteristic features of a moraine come clearly out. The hill just to the soiith of this, on which Thornville is built, is a glacial deposit in which wells descend from thirty to fifty feet without striking rock. This is upon the highest land in the vicinity. The reservoir itself seems to be simply a great kettle-hole. All through Fairfield county the glacial accumulation is of a great depth down to within a very short distance of its margin. But perhaps the most remarkable of all the portions of this line in Ohio is that running from Adelphi, in the north- east corner of Eoss county, to the Scioto Eiver. The accu- mulation at Adelphi rises more than two hundred feet above Salt Creek, and continues a marked feature in the landscape for many miles westward. Elding along on its uneven summit, one finds the surface strewn with granitic bowlders, and sees stretching off to the northwest the magnificent and fertile plains of Pickaway county, while close to the south of him, yet separated by a distinct interval, are the cliffs of Waverly sandstone, rising two hundred or three hundred feet higher, which here and onward to the south pretty closely approach the boundary of the glaciated region. 14:2 THE ICE AGE IN NORTH AMERICA. Through the southeastern corner of Highland county, and the northwestern corner of Adams, the terminal accumula- tions are less marked than in Ross county ; still, their bound- ary can be accurately and easily determined. It approaches the Ohio River, in the vicinity of Ripley and Higginsport in Brown county, and crosses it from Clermont county, entering .Kentucky half a mile north of the line between Campbell and Pendleton counties in that State. Cincinnati was cov- ered with ice during a portion of the period. There are un- doubted glacial deposits within the bounds of the city at the railroad station at Walnut Hills, and near Avondale, at a height of about four hundred feet above the river. At IS'orth Bend, twenty miles below Cincinnati, the tunnel of the rail- road leading from the Ohio to the Great Miami River is through an indubitable glacial accumulation which rises two hundred feet above the river. The northwestern part of Boone county, Ky., was also covered with the ice to a dis- tance of several miles south of the Ohio River. Through Indiana the glacial boundary, after following the Ohio River to within ten or twelve miles of LouisviUe, Ky., suddenly bends to the north, leaving a large triangular portion of the State unglaciated. The base of this ungla- ciated triangle extends from Louisville to the Illinois line, and its apex is about thirty miles south of Indianapolis. The exact course of this part of the boundary is along a line running from the neighborhood of Louisville northward through Clark, Scott, Jackson, Bartholomew, and Brown counties to Martinsville in Morgan county, where it again turns west and south nearly parallel with, and west of, the West Fork of White River, through Owen, Greene, Knox, Gibson, and Posey counties, crossing the Wabash River into Illinois, near New Harmony, the seat of Owen's cele- brated socialistic experiment. In Illinois the line continues in a southwesterly direction through White, Gallatin, Saline, and Williamson counties, where it reaches its most southern limit near the northern boundary of Johnson county, fifty or sixty miles north of BOUNDARY OF THE QLAGIATEB AREA. 143 Cairo, a little below latitude 38°. From Williamson county it starts westward upon its final course to the north, reaching the Mississippi Kiver, near Grand Tower, in Jackson county. From this poiht to the vicinity of St. Louis the Mississippi runs nearly southeast, so that the glacial boundary in its 144 THE ICE AGE IN NORTH AMERICA. northwest course is coincident with the bluffs on the north- east side of the river through Jackson, Randolph, and Mon- roe counties, 111. So far I have traced the southern boundary myself, and the information here given is nearly all at first hand. Beyond the Mississippi competent members of the United States Geological Survey have traced the course approximately to the Pacific Ocean. From these data we know that across the State of Missouri the Missouri Eiver approximates closely to the glacial limit. The Jine enters Kansas a little south of Kansas City, and runs nearly west for a hundred miles to the vicinity of Topeka, where it curves northward, crossing the State of Nebraska about one hundred miles west of the Missouri River, and reaching the southern hne of Dakota, near the junction of the Niobrara and the Missouri. In eastern Kansas and Nebraska the exact limits of the glaciated area appear, from the reports, to be somewhat difficult of determination. It would seem that the action of water and floating ice was predominant in determining the character of the glacial deposits over that region, and the theory is plausibly suggested by Professor Todd that the extension of the ice beyond the Missouri formed glacial dams across the valleys of the Kansas and Platte Rivers, so as to maintain for a short period temporary lakes of a consid- erable extent, which received and distributed the bowlder- laden fragments of ice, as well as the finer elements of the glacial deposits. The most of the glaciated portion of these States is deeply covered with fine loam, or loess, which is probably a water deposit, and, as we shall hereafter see, is on good grounds believed by Chamberhn and Sahsbury to be an " assorted variety of glacial silt directly derived from gla- cial waters." * Through Dakota the glaciated region is bounded by a line which runs northward from near the junction of the Niobrara * " Preliminary Paper on the Drif tleas Area of the Upp^r Mississippi Val- ley," by Thomas C. Chamberlin and RoUin D. Salisbury, in the " Sixth Annual Report of the United States Geological Survey," p. 304. BODNBAEY OF THE GLACIATED AREA. 145 and Missouri Rivers, and keeps pretty close to the west edge of tlie river trougli as far up as the mouth of the White River. Beyond this it breaks over the western edge of the trough for a short distance, and keeps approximately parallel with the river, from five to ten nules west of it, until reach- 146 TEE 10 E AGE IR NORTH AMERICA. ing the vicinity of Oatie, a few miles above Pierre, at the mouth of Bad River. From this point to the mouth of the Cheyenne glacial deposits do not encroach upon the plateau to the west. But, above the mouth of tlie Cheyenne, the line strikes off farther west, and crosses the Moreau River about forty-five miles back from the Missouri, and the G-rand and Cannonball Rivers at about the same distance. The Northern Pacific Railroad passes from the glaciated to tlie unglaciated region at Sims Station, about thirty miles west of Bismarck. jj In the chapter iipon " Terminal Moraines " we will speak more fully of the portion of Dakota lying east of the Mis- souri River ; but west of the Missouri the deposits belong to what we have denominated the fringe, or what President Chamberlin perhaps more appropriately calls "the attenuated border." This portion of the boundary I had the privilege of studying in the summer of 1888, driving some hundreds of miles through the Indian reservation, extending from Fort Yates south westward to the Moreau, and thence southeast- ward to the vicinity of Pierre. Here I found the border, although somewhat attenuated, to be pretty sharply defined. The glacial marks, however, consisted almost wholly of bowlders and rather coarse gravel, and was pretty evenly dis- tributed over the surface of the plateau. The formation of the region is cretaceous, so that it is easy to recognize the Laurentian bowlders. In size these sometimes attain a diame- ter of four or five feet, and frequently almost cover the ground. The elevation attained by them runs up to about six hundred feet above the river. We found, however, no scratches upon these bowlders, nor were there any exposures of till or unstratified deposit, so characteristic of the terminal moraine and of the central portion of the glaciated area. But from my own experience I have no hesitation in classify- ing these deposits witli those produced by direct action of the glacier. They are what would naturally occur on the attenu- ated margin of the ice- sheet. I found evidence, also, of a temporary line of marginal BOUNDARY OF THE OLACIATED AREA. 147 drainage, which consisted of a broad, level-topped gravel deposit from four hundred and fifty to five hundred feet above the present bed of the Cheyenne Kiver.* This old river- bottom is about two miles wide, and, where we crossed it, extended as far as the eye could reach both up and down the valley. Subsequently Mr. Eiggs found that it joined the valley sorne miles above from the northwest. The gravel is rather fine and well worn, and there is only an occasional bowlder from one to two feet in diameter to be found upon the surface. On the higher levels there are no traces of the deposit. It has, therefore, as already said, the appearance of marking a marginal line of drainage, which, north of the Moreau Kiver, was thirty or forty miles west of the Missouri, but which joined the Cheyenne just west of Fox Kidge, and followed that valley down to the vicinity of the Missouri, and ever after kept near its trough till the river passed out of the Territory at the Nebraska line. Soon after crossing the Northern Pacific Eailroad in Dakota, the glacial boundary turns abruptly to the west, crossing the Yellowstone in Montana near Glendive. We give the delineation beyond this point in the words of Presi- dent Chamberlin : " Passing north of the Judith Mountains, it again touches the Missouri in western Montana, near the mouth of the Ju- dith River, but at once swings away to the southward, to again strike and cross the river forty miles above Fort Benton, and about the same distance from the Rocky Mountains. Thence it curves rapidly to the northward, crossing the national bound- ary at the very foot-hills, and thence skirts them northward to the limits of present determination. This is the outline of the great northeastern sheet of drift. Along the Rocky Mountains, within the United States, it barely comes in contact with demonstrable glacial formations from the adjacent mount- ains, though widely intermingled with mountain ' wash.' " f * The elevations are kindly furnished me by Rer. Thomas L. Eiggs, of Oahe. \ " Proceedings of the American Association for the Advancement of Sci- ence," vol. XXXV, 1886, pp. 196, 197. 148 THE IDE AGE IN NORTH AMEBIC A. In the chapter upon " Terminal Moraines " we will speak of the extension of the Missouri coteau into British America, as determined by Dr. George M. Dawson. But, while this coteau is the limit of the heavier accumulations of Laurentian drift, it is evidently not the limit of the extent of glacial ice, for over an indefinite border to the west of it there is, accord- ing to Dr. Dawson, a large percentage of Laurentian mate- rial, amounting to nearly fifty per cent of the surface accumu- lations, mingled with about the same proportion of quartzite drift brought down from the Rocky Mountains by the numer- ous streams originating in tbem. Dr. Dawson says that these Laurentian and eastern limestone bowlders continue to occur to within twenty-five miles of the base of the Rocky Mount- ains, and up to a height of 4,200 feet above tide. The dis- tance of these traveled blocks from the nearest part of the Laurentian region is about 700 miles. Beyond this point, to the west, eastern and northern rocks are not found. The elevation of this marginal drift is about 2,000 feet above the present height of the Laurentian plateau from which it came.* " To the westward, in the' valleys of Flathead, Pend D'Oreille, and Osoyoos Lakes, and of Puget Sound, are massive deposits of drift, partly of northern and partly of local mountainous derivation. The Pend D'Oreille and Puget Sound deposits appear unquestionably to be tongues of the drift of British Columbia, which, if not constituting a continuous mantle, at least passes beyond the character of simple local mountain drift." f In the Rocky Mountain region and to the westward there were formerly extensive glaciers in Montana, Wyoming, Colorado, Utah, Nevada, and California, where now they are almost entirely absent. But the glaciation of this region was never general. According to "Whitney, there are no signs of ancient glaciers in western Nevada, though some of the mountains rise to a height of 10,000 feet. In the east Hum- * See the " Quarterly Journal of the Geologieal Society," vol. xxxi, 1878. f See Chamberlin, as above. BOUNDARY OF TEE GLACIATED AREA. 149 boldt range, local glaciers once existed in all the higher por- tions. In some of the valleys they extended for seven or eight miles. In Utah the Wahsatch Mountains were the chief center of local glaciers. The principal mountain-mass is about fifteen miles wide, and peaks above 10,000 feet high are numerous. The glaciers formerly radiating from this mass did not, however, reach a very low level. In Colorado there are evidences of former glaciers only above the 10,000- foot line. Beyond that line, such valleys as those occupied by the head-waters of the Platte and Arkansas Rivers were once filled with glaciers whose terminal moraines, in some cases, formed dams of great extent, and thus gave rise to temporary lakes. The most southern point at which signs of local glaciers in the Kocky Mountains have been noted is near the summits of the San Juan Eange in southwestern Colorado. Here a surface of about twenty-five square miles, extending from an elevation of 12,000 feet down to 8,000 feet, shows every sign of the former presence of moving ice. Il^orthward of Utah and Colorado the signs of former glacia- tion are also of the same local character — that is, glaciers everywhere radiated from the higher mountain-masses, and extended a short distance down the canons and valleys. The Upper Canon of the Yellowstone, in the famous park, was filled with glacial ice to a depth of 1,600 feet, and glacial marks were abundant down to the vicinity of Livingston. The glaciers of the Sierra Nevada and Cascade Range in California, Oregon, and "Washington Territory were on a much grander scale than those in the Rocky Mountains; but, in the one case as in the other, the glaciated areas are local, and, except in 'Washington Territory, not connected with the grand movement farther north. Upon this point Mr. Clarence King, who had most care- fully explored the region, writes : In the field of the United States Cordilleras, we have so far failed to find any evidence whatever of a southward-mov- ing coDtinental ice-mass. As far north as the upper Colum- 150 THE IGE AOE IN NORTH AMERICA. bia Eiver, and southward to the Mexican boundary, there is neither any bowlder-clay nor scorings indicative of a general southward-moving ice-mass. On the contrary, the great areas of Quaternary material are evidently subaerial, not subglacial. The rocks outside the limit of local mountain glaciers show no traces either of the rounding, scoring, or polishing which are so conspicuously preserved in the regions overridden by the northern glacier. Everything confirms the generalization of Whitney as to the absence of general glaciation. "Wherever in the fortieth parallel area a considerable mount- ain-mass reached a high altitude, especially when placed where the Pacific moisture-laden wind could bathe its heights, there are ample evidences of former glacial action, but the type is that of the true mountain glacier, which can always be traced to its local source. In extreme instances, in the Sierra Nevada and Uintah Eanges, glaciers reached forty miles in length, and, in the case of the Sierra Nevada, descended to an altitude of 3,000 or 3,500 feet above sea-level. Over the drier interior parts of the Cordilleras the ancient glaciers usually extended down to between 7,000 and 8,000 feet above the sea. In the case of the Cottonwood Glacier of the Wahsatch, a decided ex- ception, the ice came down to an altitude of 5,000 feet. . . . Not more than a thirtieth part of the entire surface of the fortieth parallel area was ever covered by glacial ice. It is characteristic of the caflons of these extinct glaciers that they give evidence of a gradual recession of the ice from its greatest extension until it is entirely melted. This retiring from its greatest bulk was not a continuous retrogression, but was marked by pauses at certain places long enough to permit the accumulation of considerable terminal moraines. In ascend- ing one of the larger caflons, as of the southern Uintah, there is observed a series of successive terminal moraines, and in passing to the upper heights of the ranges it is found that, in the great snow amphitheatres, glacial markings, rock-polish- ing, and the arrangement of morainal matter are evidently fresher than in the lower levels or points of greatest exten- sion. Whatever the greater causes may have been, the Cordilleran surface south of Washington Territory was free from an ice- BOUNDARY OF THE GLAQIATED AREA. 151 sheet, and the only ice-masses were small areas of local glaciers which did not cover two per cent of the mountain country. Supposing the arctic land configuration to be as now, and a new oscillation of climate to bring on the conditions of a glacial period, it is certain that the present ice-masses would form the nuclei of new northern ice-fields, and Greenland would probably be the point from which the glaciers would move southward to cover eastern America ; and the absolute distance from such a center would have something to do with the failure of the ice to override the Cordilleras. Dawson's suggestion of a great center of dispersion in Alaska, where an elevated and broad highland fronts the moisture-laden ocean- wind, has, it seems to me, a high degree of probability in ac- counting for the southerly-moving ice of British Columbia with- out recourse to that refage of pure imagination, a polar cap.* With this agrees the testimony of Mr. I. 0. Kussell, in his report on the "Quaternary History of Mono Valley, California," the advance sheets of which I have been kindly permitted to see. He writes : The Sierra Nevada during the Glacial epoch was covered by an immense neve field, which probably stretched continuously from a little north of latitude 36° nearly to latitude 40°. The width of this belt of perpetual snow must have been irregular, conforming to the present topography of the summit of the range, but it probably had an average width of between ten and fifteen miles. From beneath this snowy mantle trunk gla- ciers flowed both east and west down the flanks of the range. The evidence is such as abundantly to justify the conclu- sion that the ancient glacial system of the Sierra Nevada was local, and had no connection with a northern ice-sheet. The glaciers were clustered about and radiated from the higher por- tion of the range in the same manner as from the contempo- rary neve fields of the Wahsatch and Uintah Mountains, f * " Systematic Geology " in the " United States Geological Exploration of the Fortieth Parallel," 18Y8, pp. 469-461, 464. t See the " Eighth Annual Report of the United States Geological Survey," pp. 32*7, 328. 152 TEE IQE AGE IN NORTB AMERICA. The ancient area of glaciers in the Sierra Nevada Mount- ains was chiefly confined to the western slope, and was most remarkable in Tulare, Fresno, Mariposa, and Tuolumne coun- ties, California, where, as we have seen, glaciers still continue to exist. There are abundant marks of these ancient ice- streams in the upper valleys of Kings, San Joaquin, Merced, and Tuolumne Rivers. In the Tuolumne the glaciers were in some places several miles wide and twelve hundred feet deep, and extended as much as forty miles down the valley. Gla- ciers likewise iilled the Yosemite Yalley on the Merced Eiver. It is a mistake, however, according to Whitney, to suppose that the Yosemite was formed, or indeed greatly modified, by glacial action.* The vertical walls and the rect- angular recesses are such as to indicate the action of disrupt- ive rather than erosive agencies in their formation. The north-and-south valley between the Cascade Mount- ains and the Coast Range, in Washington Territory, is about one hundred miles wide. The northern half of this is pene- trated by the innumei'able channels and inlets of Puget Sound, which extends from Port Townsend south about eighty miles to the parallel of Mount Tacoma. The Olym- pian Mountains to the west rise to a height of about ten thousand feet, as does Mount Baker in the Cascade Range to the northeast. The shores and islands of Puget Sound have every appearance of being portions of a vast terminal mo- raine. They rise from fifty to two hundred feet above tide, and present a mixture of that stratified and unstratified ma- terial characteristic of the terminal accumulations of a great glacier. ~Eo rock in place appears anywhere about the sound. Bowlders of light-colored granite and of volcanic rocks are indiscriminately scattered over the surface and imbedded in the soil. One of these bowlders, near Seattle, two hundred feet above the sound, measures twenty feet in diameter, and twelve feet out of ground. The channels of the sound and * " The Yosemite Guide-Book," p. 83. See also the opinion of Mr. Russell given near the close of Chapter X. BOUSDABY OF THE GLACIATED AREA. 153 of the adjacent fresli-water lakes ha\-e a g-eneral nortli-aiid- soutli direction, parallel with the axis of the valley. This is specially noticeable near Seattle, where Lake Washington, Fia. 48. „ , ^ F t II t .fH, \\ i^hii ri 11 Tprritorv ihont t\^n hundred feet above the sound. This is on the height bi elevated sixteen feet above tide, and twenty-five miles long, is strictly parallel with the sound, and is separated from it 154 THE ICE AGE IN NORTH AMERICA. by a series of ridges showing every mark of glacial origin. Not only is tlie surface of these ridges covered with bowl- ders, but wherever the streets have cut down into the soil they show, at the depth of a few feet, an unstratiiied deposit abounding in striated stones. Superimposed upon this ridge there is a thin stratified deposit of varying depth, but in- creasing in extent down the slope toward tide-water. At Port Townsend, on the Strait of Juan de Fuca, and forty miles northwest of Seattle, the coarsely stratified de- posit is much greater in extent. A noteworthy section of this I had the privilege of studying at Point Wilson, two miles and a half northwest of Port Townsend. Here, facing ji^fi^ss^MmmB Fig. 49.— Section of the deposit at Point Wilson described in the text, showing one hun- dred and fifty feet in height, about one hundred of which is coarsely stratified, and contains layers of vegetable matter. Bowlders from till at the top have fallen down to form a talus at the water's edge. the strait toward the north, is a perpendicular bluff from one hundred and fifty to two hundred feet in height, composed BOUNDARY OF THE GLACIATED AREA. 155 of material coarsely stratified throughout its lower portion, but capped at the summit by about forty feet of coarse, un- stratified material abounding in large striated bowlders, which, as they have been washed out by the erosion of the sea, are falling down to the foot of the bluff in immense numbers. Near the bottom of the bluff there are several strata of vege- tal deposits. One of these, two feet thick, consisted almost wholly of the fragments of the bark of the fir-trees which are now so characteristic of that region. Fragments of wood project from the freshly exposed bank in great abundance. The meaning of these facts will be more readily apparent after a study of the phenomena to the north of the strait. The Strait of Juan de Fuca is from fifteen to twenty miles in width, setting in from the Pacific Ocean and run- ning east and west. Its north shore, near "Victoria, on Van- couver Island, is remarkably clear of glacial debris. The rocks, however, near Victoria, exhibit some of the most re- markable effects of glacial scoring and striation anywhere to be found. Immediately south of Victoria long parallel fur- rows rise from the shore of the inlet, and ascend the slope of the hill to the south to its summit, a hundred feet or more above the water-level. At the steamboat-landing, outside of the harbor, extensive surfaces freshly uncovered exhibit the movtonnee appearance of true glaciation, and, in addition to the finer and abundant scratching and striae, display numer- ous glacial furrows from six inches to two feet in depth, from twenty to thirty-two inches in width, and many feet in length. These grooves are finely polished and striated, resembling those with which geologists are famUiar on the islands near the the western end of Lake Erie. Like the corresponding grooves on the islands of Lake Erie, some of those near Victoria also form graceful curves, adjusting them- selves to the retreating face of the rock-wall. That the mo- tion of the ice at Victoria was to the south appears not only from the direction of the, strise, but from the fact that the stoss sides of the glaciated rocky projections are toward the north. That they are due to glacial action, and not to ice- 15G riLK lUE AGE IN NUETII AMEEIGA. y % Fig. 50. — Glacial grooviiifrs near the landing at Yietorla (^ee text), (b'roni Photuj^raph.) bergs, is evident both from tbuir cliaracter and from their analogy to numerous phenomena fartlier to the north un- questionably connected with true glaciers. BOUNDARY OF THE GLACIATED AREA. 157 Speaking of Vancouver Island and the adjacent region, the Scotch geologist, Eobert Brown, uses the following lan- guage : The chief rock in situ there is a dense, hard, feldspabhic trap, and this is plowed in many places into furrows six to eight inches deep and from six to eighteen inches wide. The ice-action is also well shown in the sharp peaks of the erupted, intruded rocks, having been broken ofE and the surface smoothed and polished, as well as grooved and furrowed, by the ice-action on a sinking land, giving to the numerous promontories and outlying islands which here stud the coast the appearance of rounded bosses, between which the soil is found to be composed of sedimentary alluvial deposits, containing the debris of ter- tiary and recent shelly beaches, which have, after a period of depression, been again elevated to form dry land, and to give the present .aspect to the physical geography of Vancouver Island. The whole surface of the country is strewn with erratic bowlders. Great masses of sixty to one hundred tons in weight — chiefly of various igneous and crystalline as well as sedi- mentary rocks, sufficiently hard to bear transportation — are found scattered everywhere over the island from north to south, and through the region lying on the western slope of the Cas- cade Mountains.* The same observer thus speaks of tlie glacial phenomena on the mainland in the same vicinity : The following section is given to show the general character of the drift at Esqtiimault Harbor : Black sandy and peaty ground, with broken shells 2 to 6 feet. Yellowish sandy clay, with oasts of shells (Cardium and Mya) and a few pebbles and bowlders ••• 6 to 8 Gravel of scratched pebbles resting on rock 2 to 3 The rocks are grooved and scratched at the junction ; the direction of the glacial markings is between north to south and north-northwest to south-southeast. In a well-sinking, at * "American Journal of Science," vol. c, 1870, pp. 320, 321. 158 THE ICE AGE IN NORTH AMERICA. Esquimault Barracks (for the boundary commission), the lower gravel was reached at forty-two feet, after going through a sandy-blue clay without shells or bowlders. The section in the cliff between Albert Head and Esquimault is as follows : Blue drift clay, with bowlders; junction with rock not seen 70 feet. Fine sand and gravel, passing upward into coarse quartzose gravel 100 to 120 feet.* I saw at Seahome (near Bellingham Bay), in the cuttings made for a tramway, the finest instances of fluting and groov- ing — evidences of glacial action — that I have ever seen on this coast. They were ninety feet in length, running north and south, according to the theory of Professor Agassiz. f Vancouver Island, which trends parallel with the shore of the continent northwest by southeast, is nearly three hun- dred miles in length and from fifty to seventy-five in breadth. In character it seems but a continuation of the Coast Range of mountains, with numerous peaks rising from four to seven thousand feet above the sea. The shore-line of the continent upon the northeastern side of the Strait of Georgia is formed by a continuation of the Cascade Range, with a general ele- vation of from three to eight thousand feet, and is penetrated in numerous places, to a distance of from twenty-five to seventy-five miles, by inlets or fiords some miles in width. Dr. George M. Dawson describes the glacial phenomena in Bute Inlet (which enters the Strait of Georgia about oppo- site the center of Vancouver Island, in latitude 50° 30') in the following language : This chasm, forty miles in length, and running into the center of the Coast Eange, is surrounded by mountains, which in some places rise from its borders in cliffs and rocky slopes to a height of from six to eight thousand feet. It must have been one of the many tributaries of the great glacier of the Strait of Georgia, and accordingly shows evidence of powerful * " Quarterly Journal of London Geological Society," 1880, p. 202. f "American Journal of Science," vol. c, pp. 322, .323. BOUNDARY OF THE GLACIATED AREA. I59 iee-action. The islands about its mouth are roches moutonnees, polished and ground wherever the original surface has been preserved. In Sutil Passage, near its entrance, grooving ap- pears to run about south 30° west. A precipitous mountain on Valdez Island, opposite Stuart Island, and directly blocking the mouth of the inlet, though 3,013 feet high, has been smoothed to its summit on the north side, while rough toward the south. The mountain-side above Arran Passage shows smooth and glistening surfaces at least two thousand feet up its face ; and, in general, all the mountains surrounding the fiord present the appearance of having been heavily glaciated, with the exception of from one to two thousand feet of the highest peaks. The high summits are rugged and pointed, and may either never have been covered by glacier-ice, or owe their different appearance to more prolonged weathering since its disappearance. In some places parallel flutings high up the mountain-sides evidence the action of the glacier, while in others it is only attested by the general form of the slopes, or detected under certain effects of light and shade. ... At the mouth of the Howathco Eiver, discharging into the head of Bute Inlet, striation shows a direction of movement south 23° east ; but in every case the motion appears to have been di- rectly down the valley, and to have conformed to its changes in course. Glacier-ice may still be seen shining bluely from some of the higher valleys at the head of the inlet, and farther up the Howathco River there are many glaciers in lateral val- leys, some of which descend almost to the river-level. Mr. James Richardson, who has had an opportunity of examining many of the inlets north of Vancouver Island, writes as follows :* "Throughout the whole of the inlets and channels which were examined, wherever the surface of the rock is exposed, the ice-grooving and scratching are very con- spicuous, from mere scratches to channels often several feet in width, and from a few inches to as much as two or three feet in depth. Often they can be distinctly seen with the naked eye from the surface of the water to upward of three thousand feet above it on the sides of the mountains. They * " Keport of Progress of the Geological Survey of Canada, 18M-"?5," p. 8. 160 THE ICE AGE IN NORTH AMERICA. run in more or less parallel lines, and are not always horizontal, but deviate slightly up or down." * Mr. Robert Brown, whom we have already quoted, gives the following additional information as to regions still far- ther north : I have not been in Alaska proper, but in 1866, in a visit to the Queen Charlotte Islands, lying some thirty or forty miles off the northern coast of British Columbia, close to the southern boundary of the former Territory, marks of the northern drift quite as marked as in Vancouver Island were found there. f As already indicated, the mountains on either side the Strait of Georgia, and northwestward to the head of Lynn Canal, about latitude 59° 20', are snow-clad throughout the whole season. The shores are everywhere rocky and pre- cipitous, retaining in many places far up their sides glacial strisB parallel with the direction of the numerous channels which thread their way through the Alexander Archipelago. I had opportunity at Loring, on the western shore of Revilla Gigedo Island, to examine minutely the striation on the shores and islands of Naha Bay. There are now no glaciers coming down from the mountains of this island, but the shores and islands abound in well-preserved glacial strise running west by 18° north, corresponding to the direction of the local val- ley, down which a glacier came in former times, entering Behm's Canal nearly at right angles to its course upon that side of the island. This is about latitude 55° 40'. Upon proceeding one degree to the north, I had oppor- tunity also to observe closely the strise at Fort Wrangel. Here, too, they show the influence of the continental eleva- tion to the east, and are moving outward in a westerly direc- tion toward the Duke of Clarence Strait. In Glacier Bay the evidence of the recent vast extension of the glaciers down the bay, and of the facility with which * " On the Superficial Geology of British Columbia," in the " Quarterly Jour- nal of the Geological Society," vol. xxxiv, February, 18T8, pp. 99, 100. f " American Journal of Science," vol. c, p. 323. BOUNDARY OF THE OLAOIATED AREA. 161 glacial ice adjusts itself to the local topography, is, as already stated, of a most explicit character.* In addition to the evi- dence already mentioned, we may add that numerous islands project from the surface of the Mnir Glacier, as from the waters of an archipelago, and that the summits of these bear every mark of having been freshly uncovered by the decreas- ing volume of ice. Also that below the mouth of the gla- cier numerous islands in the bay present exactly the same ap- pearance, except that they now project from water instead of ice. Their recent glaciation is indicated by every character- istic sign. Willoughby Island, about the middle of the bay, is as much as a thousand feet above the water. Were the ice to retreat a few miles farther back from its present front, it would doubtless uncover an extension of the bay, with numer- ous islands similar to those now dotting its surface south of the glacier. Fresh glacial dehj^'is lingers on the flanks of the mountains on either side of the inlet, to a height of 2,000 feet. The fact is also worth repeating and emphasizing that at 3,700 feet above tide striae were observed, on the east side of Muir Inlet, not pointing down the mountain, as might be expected, but parallel with the axis of the bay, showing, be- yond controversy, that the present glacier is but a remnant of an earlier ice-movement, similar in character and direction to the present, but of vastly greater dimensions, and which extended until it iilled the whole bay to its mouth in Cross Sound, a distance of twenty-five miles. At Sitka, also, the rocks of the harbor are all freshly striated— the direction of the striae being toward the west— that is, toward the open sea. Glaciers still linger in the mountains at the head of the bay to the east of Sitka. Little is known of the vast region lying to the north of the lofty mountains forming the southern border of Alaska. But from the reports of Schwatka, Dall, and others, it has long seemed improbable that there was any southerly move- ment of ice over it, thus disproving the theory, somewhat * See above, p. 52 et seq. 11 162 TEE lOE AOE IN NORTH AMERICA. prevalent, of a universal polar ice-cap. The expedition of Dr. George M. Dawson to the head-waters of the Yukon in the summer of 1887 seems to demonstrate that there was actually a movement of ice from the northern flanks of the mountains toward the interior. Thus he writes : * In the Lewes and Pelly Valleys, traces of the movement of heavy glacier-ice in northward or northwestward directions, were observed in a number of cases, the grooving and furrow- ing being equally well marked at the water-level and across the summits of hills several hundred feet higher. The facts are such as to lead to the belief that a more or less completely confluent glacier-mass moved in a general northwesterly direc- tion, from the mountainous districts south of the southern sources of the Yukon, toward the less elevated country which borders the lower river within the limits of Alaska. This ob- servation, taken in connection with the evidence of the former northward movement of glacier-ice in the arctic regions to the east of the Mackenzie, appears to have very important bear- ings on theories of general glaciation. \ From all these facts it seems evident that the supposition of a slight intensification of the present conditions so favor- able to the production of glaciers in southeastern Alaska, unravels the whole intricate web of glacial phenomena upon the western coast of North America. The present formation of glaciers on the coast of south- eastern Alaska is favored not so much by the coolness of the climate as by the elevation of the mountains and the excess- ive amount of precipitation, which, as before stated, is not far from one hundred inches annually. There is no evidence that the elevation of the coast has materially changed in recent times. But it would require only a slight change in the amount of precipitation, or a slight diminution of tem- * •' Science," vol. xi, 1888, p. 186 ; see also " Recent Observations on the Gla- ciation of British Columbia and Adjacent Regions," in the " Geological Maga- zine," vol. Ixx, August, 1888, p. 347 et seq. t "Annual Report of the Geological Survey," 1886, p. 63, R. BOUNDARY OF THE GLACIATED AREA. 163 perature, to secure all the additional force required to extend the present glaciers of southeastern Alaska, British Columbia, Fig. 51.— Generalized view of the whole glaciated region of North America. Some doubt, however, should have been expressed as to the condition of the northern part of Alaska. and of the Cascade Range of "Washington Territory and Ore- gon, until they should gorge all the channels of the Alex- 164 THE IGE AQE IN NORTH AMERICA. ander Archipelago, fill the space occupied bj the Strait of Georgia between Vancouver Island and the mainland, and cover with ice the whole valley between Mount Tacoma and the Olympian Mountains, where now we find the vast mo- raine deposits of the islands and shores of Puget Sound. A simple calculation impresses one with a sense of the unstable equihbriiim of the forces leading to the increase or diminution of a glacier. We estimated that 7T,000,000,000 cubic feet of ice annually pass through the gorge at the head of Muir Inlet, and that the area of the ice-field supplying this stream is about twelve hundred square miles. The total amount of ice entering the inlet, therefore, is only equivalent to about two feet of ice over the field of supply. If from any cause two feet more of ice should annually accumulate over this area, or two feet less should annually melt away, the amount of ice compelled to go through the gorge would be doubled, and this would doubtless fill up the whole inlet and bay to the south. When we reflect that, according to ITewcomb,* the average amount of ice which would be melted by the sun over the whole earth is something more than a hundred feet a year, and that, therefore, a change in intensity amounting to only one fiftieth of that exhibited by the present meteorological forces would produce the re- sults just mentioned, we can readily believe that oscillations in such a great glacier may be frequent in occurrence and of great magnitude. Southward, in Oregon, the Willamette Valley was filled in a similar manner by an extension of the glaciers still lin- gering on the flanks of Mounts Hood and Shasta. The absence of drift on the southern shore of Vancouver Island seems to point to a termination of the southerly movement from Alaska in the Strait of Juan de Fuca, where perhaps the confluent streams turned westward, and sent off vast drift-laden icebergs to the sea. Mount Baker, immediately to the east of this point, upward of ten thousand feet high, * " Popular Astronomy," p. 247. BOUNDARY OF THE &LAGIATED AREA. 165 and still preserving glaciers on its flanks, would have lent material aid in this seaward movement. The shores and islands about Puget Sound have the appearance of being the terminal deposits of confluent glaciers coming down from the flanks of Tacoma on the southeast, and from the lower portions of the Cascade Eange farther north, joined by smaller glaciers from the Coast Range on the west. It is clear that the earlier glacial movements on the Pacific coast were local in character, and must be studied independently of those east of the Eocky Mountains, and can be understood only by reference to the glaciers which still linger at the head of all its numerous valleys, inlets, and fiords. In these the investigator has a vera causa ever before his eyes to guide his steps and to assist his imagination. CHAPTER YIII. DEPTH OF THE ICE DURING THE GLACIAL PEEIOD. There are two sources of information concerning ■ the depth attained by the ice in North America during the Gla- cial period : First, we have direct evidence in the height of the mountains which have marks of glaciation upon their summits ; secondly, calculations can be made, with some approximation to truth, from the distance through which bowlders have been transported. Yery conveniently for the glacialist, the mountains of New England and the Middle States serve the purpose of glaciometers, preserving upon their flanks and summits in- dubitable evidence of the great depth of the ancient ice-sheet over that portion of the country. It requires but a cursory examination to see that the highest point of Mount Desert Island, on the coast of Maine, was completely covered by the glacier, showing that at the very margin of the ocean the ice must have been consider- ably more than 1,500 feet deep. Even Mount Washington, in New Hampshire, was either wholly enveloped by the ice- current, or if a pinnacle projected above the glacier it could have been no more than 300 or 400 feet higher. Professor Hitchcock having found transported bowlders to within that distance of the summit. The ice-current passed over the Green Mountains where they are from 3,000 to 5,000 feet in height in a course diagonal to that of their general direction, showing that such a mountain-chain made scarcely more of a ripple in the moving mass than a sunken log would make in a shallow river. Farther south, Mounts Monadnock, Tom, DEPTH OF THE ICE LURING THE GLACIAL PERIOD. 167 and Holyoke, the Berkshire Hills, and East and West Mount- ain, near New Haven, were also completely enveloped in ice. Between the Adirondacks and the Alleghanies the Mohawk Valley was filled nearly to the height of the Catskills, and the southern edge was pushed up in Monroe, Sullivan, Tioga, and Potter counties, Pa., to a height of 2,000 or 2,500 feet above the sea. In remarking upon the accompanying sections, Professor Lesley, who made them, says that while they do not satisfy him in several important particulars, such as the regularity of its surface, the location of possible crevasses, the descent into the plain, the distribution of bowlders, etc., they serve to give a correct generalized view — first, of the great thick- ness of the ice-sheet, by contrasting it with the sections of the solid rocks from the present surface down to the plane of sea-level ; second, to allow the reader to judge for himself of the extent of the eroding power at this point. We append Professor Lesley's reasons for constructing the scale as he has : As to the first point, I have given to the surface of the ice a gentle slope southward, by making it 600 feet thick over the mountain, and 1,800 feet thick over Cherry Creek ; which slope, if continued northward, would suffice to make the ice cover the highlands (2,000 feet above tide) farther north, as we know that it did. Thirty years ago Agassiz gave me his law of the necessary minimum thickness of a glacier for crossing a barrier. It was in a conversation immediately subsequent to his study of the strise on the top of Mount Desert, pointing from Mount Katahdin, and descending into the sea. He said that no glacier could cross a ridge unless its thickness at the summit of the ridge was at least one half the height of the ridge. By this rule he judged that the ice-sheet of Maine was 750 feet thick over the top of Mount Desert ; and this would account for the great distance south of Mount Desert of the terminal moraine. This rule was obtained by Agassiz and Desor in their long residence on the glacier of the Aar, and was based on numerous observations of local Alpine glaciers where they were crevassed s s , C^ >u o H •5^ 5^ ^ K ■*0 § ^ « s p.- 6 " s ^ 5, , ',' I ' 'is, ■ ^ ' « \ ^;;v: '.■ )" ■'. ',■ ; i:p0OO,Sl'-,',. -A '• ■■; ■>\iii, V'. ;■ ■', i''f.\ , . ' I v- i'',EPTH OF TSE ICE DURING THE OLAOIAL PERIOD. lYl miles. The difference in the elevation of the glacier could not have exceeded a thousand feet. In that direction the slope was less than on a meridian line from the Catskills southward." Professor Dana estimates that the height of the ice " above the region of New Haven, in soathern Connecticut, may have exceeded two thousand feet, and could hardly have been less than fifteen hundred." So far the evidence is direct and positive, because the glacial marks are left upon the mountain-summits mentioned. How far still above these summits it rose is not so easily determined. From this amount of direct evidence it may also reasonably be inferred that the depth of the iee over the lake and prairie region of the "West was equally great. If our interpretation of the facts implying the presence of an ice-sheet in North America is correct, we have also positive evidence of a great depth of ice over the central portion of British America between Hudson Bay and the Rocky Mount- ains. Here we find, according to Dawson, that bowlders from the Laurentian axis of the continent, which stretches from Lake Superior northward to the west of Hudson Bay, have been transported westward a distance of seven hundred miles, and left upon the flanks of the Rocky Mountains at an elevation of something over foUr thousand feet.* But nowhere does the Laurentian axis reach two thousand feet, its average elevation, according to Sir William Logan, being from fifteen to sixteen hundred feet. If these bowlders were, as we suppose, transported by glacial ice, then the ice must have accumulated over the Laurentian axis to a depth of 2,400 or 2,500 feet, and must have been several hundred feet deeper in the central part of the Red River Valley. Mr. R. G. McOonnell, of the Canadian Survey, also, from more direct evidence, estimates f that, in the plains surround- ing the Cypress Hills (in the upper valley of the South Saskatchewan, in latitude 49° 30', longitude 110°, and not * See p. 214. f See his " Report on the Cypress Hills Wood Mountain." 172 THE ICE AGE IN NORTH AMERICA. more than one hundred miles from the southwestern limit of the glaciated area in Montana), the continental glacier, or the glacial sea, according to which one of the theories of transportation is adopted, had a maximum depth of two thousand feet. This he determines by the height to which he found glacial deposits resting upon the Cypress Hills. It is the necessity of accounting for such an elevation of bowl- ders in glacier-ice which has made the Canadian geologists hesitate about accepting the glacial theory. It seemed to them at iirst more probable that there had been a depression of about four thousand feet in the Rocky Mountain region, and that these bowlders were transported from the Laurentian axis by floating ice. We think, however, that such facts as are illustrated in the diagram of Professor Lesley's on page 168, as well as other facts yet to be stated concerning the elevation of bowlders in ice, go far to remove the objections to the glacial theory urged by the Canadian geologists, and we therefore speak with considerable confidence of the great depth of the ice over the Laurentian axis. The glacial the- ory, moreover, as Dr. Dawson frankly and early admitted, relieves them of many difliculties in accounting for the noticeable absence of other indications of subsidence in the region under consideration. For example, there is, first, ac- cording to Dr. Dawson,* a complete absence of any marine animal remains in the drift over that region ; and, secondly, " the yielding, scarcely solidified " sediments over this vast region bear slight evidence of any such great change in ele- vation. The great depth of the ice over the lake-region during the Glacial period is also evident from the second mode of calculation, namely, that based upon the distance over which bowlders are known to have been transported by the direct movement of the ice. The fluidity and plasticity of ice are so slight that, where we find it moving hundreds of miles over a level country, the thielaiess at the starting-point can scarcely ' "Report on the Forty-ninth Parallel," as above, pp. 216, 244, 260, et al. DEPTH OF THE ICE DURINO THE GLACIAL PERIOD. 1Y3 have been less than that indicated by the evidence in New England. Over southern Ontario and Michigan, and over the larger part of Wisconsin, Minnesota, northern Illinois, and Iowa, the ice must have been thousands of feet in depth, or it never could have pushed southward to the latitude of Cincinnati, Louisville, and St. Louis. The uncertainties attending this mode of calculation are, however, very great, and it can be taken only for approxi- mate results. In the Alps the lowest mean slopes down which glaciers move are 2^° to 3°, or about 260 feet to the mile. But, as Professor Dana notes,* the thickness of the ice there is not over 500 feet. Mathematicians are not able to deal successfully with the problems of friction in viscous bodies. How such a body will behave in greatly increased masses can be determined only by experiment. In Green- land, where the thickness of the ice more nearly approaches that of the ice-sheet formerly covering the northern part of the United States, Jensen found the slope of the Frederik- shaab Glacier to be 0° 49', or about seventy-five feet a mile ; while Holland found the slope of the Jakobshavn Glacier to be only 0° 26', or about forty-five feet to the mile.f This latter slope of the surface of the continental glacier would, if continuous, make the thickness of the ice 10,000 feet over northern ISTew England, and about 11,000 feet over Lake Erie, while the depth of the ice in this calculation over the region north of Lake Huron and Lake Superior, from which certain bowlders in Kentucky came, would be nearly 30,000 feet, since the distance moved is 600 miles or more. Upon the supposition that the slope from the front to- ward the interior was but half a degree, Croll estimates that the depth of ice at the south pole, at the center of the Ant- arctic Continent, must be as much as twelve miles. This is on the supposition that the diameter of the continent is 2,800 miles. The same rate of calculation would, according to * " American Journal of Science," vol. cxxvi, 1883, p. 348. f " Meddelelser om Gronland," 1879, and •' American Journal of Science," vol. cxxiii, 1882, p. 364. 174 THE ICE AGE IN NORTH AMEtilCA. Hitchcock,* require tiie ice of the Glacial period to be eight miles deep over the central part of Labrador ; and, if the movement came from Greenland, the same slope of forty- five feet to the mile would reach, at that point, the astonish- ing depth of eighteen miles. It is not, however, necessary to suppose a uniform slope to the center of so vast an ice-field. The only actual transfer of ice of which we are sure is that through the space over which bowlders have been carried, which is in no case much more than 700 miles. Eut such, according to Dawson, is the distance of transport of many bowlders in central Brit- ish America. Besides, all the accumulations of snow between the center of the field and the margin would help on the peripheral motion. It is doubtful, also, if a slope of half a degree is necessary to produce movement in such vast masses of ice as existed during the Ice age. We have not been able to experiment with ice in such masses, and the law of the movement is still a mystery. Previous to the investigations of Helland on the Jakobshavn Glacier, no one had observed any motion in a glacier greater than four or five feet per dny ; but when once a glacier two miles in width was brought under observation, a movement was discovered of between sixty and seventy feet a day. Yet this motion is with a slope of the surface of less than half a degree. * " New Hampshire Geological Report," vol. iii, pp. 320, 321. GHAPTEE IX. TERMINAL MOEAINES. Since the word moraine originally designates a consider- able accumulation of glacial debris, it has been found impos- sible to apply the term to the marginal deposits along the whole boundary ; for, as was stated in the chapter treating of the subject, the glacial margin in the Mississippi Yalley is not marked by such accumulations as characterize it east of the Alleghanies. The glacial deposits south of New England are, however, truly phenomenal in their extent, and can with perfect propriety be called terminal moraines. Why glacial debris should have accumulated to such an extent along that line it is impossible to tell with certainty ; but, recurring to the principles already presented, it would seem, not only that such an extensive terminal moraine indicates an abundance of easily disintegrated rock to the north offering itself for transportation in the line of the glacial movement, but that it also indicates that the ice-front remained for a long time stationary in the latitude of New York between Nantucket and the Delaware Eiver. How much the proximity of the ocean may have had to do with the maintenance of this sta- tionary ice-front we may never fully determine ; but, both by its natural eflfect in eroding the advancing ice-column, and thus limiting its movement, and by its tendency to pro- vide moisture to the clouds which furnished the glacier of that whole region with its fresh supply of snow, the neigh- boring waters of the Atlantic would seem to be an adequate cause for the phenomenon. At any rate, in the hills of Cape Cod, of Nantucket, of Martha's Vineyard, of the Elizabeth 176 THE ICE AGE IN NORTH AMERICA. Islands, and of the soutli shore of Khode Island, and in those forming the backbone of Long Island, we have one of the most remarkable true terminal moraines anywhere to be found in the world. Throughout their whole extent these terminal accumula- tions form a marked feature in the landscape, rising for a considerable portion of the distance from one hundred and fifty to three hundred feet above the general level of the country, and being dotted over with huge bowlders transport- ed a greater or less distance from the north. Kettle-holes and the small lakelets which they inclose are also constant features in the landscape. Throughout this whole extent, also, the moraines are flanked on the south by extensive de- posits of the " over-wash " gravel carried out by the water arising from the melting ice. The line of these moraines is, of course, at right angles to the direction of the ice-movement which terminated here. It is a remarkable confirmation of the theory already pre- sented in explanation of kettle-holes, that a study of those which mark the moraines of this region reveals a strong tend- ency in them to arrange themselves with their longer diameters parallel to the general trend of the moraine. Professor B. F. Koons * has taken the exact bearings of one hundred and six of these kettle-holes upon the island of ISTaushon and upon the mainland from "Wood's Holl to Falmouth, and finds that the longer axis of eighty-two out of that number is approximately parallel to the direction of the moraine — that is, nearly at right angles to the direction of the ice-movement ; and he is doubt- less correct in his inference that " this is what we should ex- pect if the kettle-holes marked the localities where fragments of ice were broken off from the face of the glacier and buried, wholly or in part, by the earth and stones borne down by the ice-sheet." By reference to the chapter upon the Muir Glacier, with the illustration there introduced, the reader may * " American Journal of Science," vol. cxxvii, 1884, p. 260 el seg. ; vol. cxxix, 1885, p. 480 et seq. TERMINAL MORAINES. 177 easily convince him- self of the correct- ness of this sugges- tion. " Some of these kettle -holes," Professor Koons goes on to say, "are upon a truly grand scale ; for example, one which contains several smaller with- in the large depres- sion, and is like an immense amphithea- tre with the hills ris- ing upon every side of it. Its highest border is one hun- dred and fifty feet above the bottom and the outlet is forty feet above the small lake at its center ; and on the south side, near its border, but upon still higher ground, a bowlder stands projected against the southern sky like a huge sen- tinel as the observer views it from the bottom of this im- mense pit." * As * " American Journal of Science," vol. cxxvii, p. 262. 13 ITS THE ICE AOE AV NORTH AMERICA. many as eight of those examined by Professor Koons were upward of eight hundred feet long in their largest diameter, and about half as wide ; and the rims of many more were from seventy-five to a hundred (in one case one hundred and fifty) feet high. Professor Koons estimates that the twelve miles of moraine between Falmouth and the southern end of Naushon contains more than a thousand kettle-holes. The width of the moraine varies from one half of a mile to three miles. According to the same observer, these kettle-holes occa- sionally reveal themselves also to the sounding-line under- neath the water. One encountered near Wood's HoU was one hundred and twenty feet deep, with shallow water all around ; another, " in Vineyard Sound, is forty fathoms deep, with the water only ten on its borders." Professor Verrill also reports " another off Cape Ann, ninety fatlioms deep, while the water about it is from thirty to forty." * There are also distinct indications, at different places in the mo- raines, of the beds of the temporary glacial streams through which the melting waters poured themselves into the sea. "West of the Hudson Yalley, as we have already seen, it is difficult to trace a well-defined and continuous moraine along the extreme glacial boundary. Such a moraine is pretty well made out across New Jersey and a portion of the distance between the Delaware and Susquehanna Rivers in Pennsylvania ; but, beyond, the country is mountainous, and through a considerable portion of the way diiBcalt of ex- ploration. Through central New York, however, there are specially marked accumulations of glacial deh'is near the water-partings between the St. Lawrence and Mohawk Val- leys and that of the Susquehanna. President Chamberlin is inclined to correlate the accumulations just south of the " Finger Lakes " of that region f with the interior moraine * " American Journal of Science," vol. cxxix, p. 484. ■f "Preliminary Paper on the Terminal Moraine of the Second Glacial Epoch," p. 353. TERMINAL MORAINES. 179 which we have described as running through Cape Cod, Elizabeth Islands, the southern portion of Rhode Island, and the northern part of Long Island, and also with those to be described more particularly hereafter as the Kettle Range, in Wisconsin. This interior line he would designate " the ter- minal moraine of the second Glacial epoch." But, in order to avoid the assumption of a distinct second Glacial epoch before conclusive proof is presented, the happy phrase of Professor Cook, of New Jersey, seems preferable, who would call the marked accumulations in the region to the north of the gla- cial limit " moraines of retrocession." Still, there is no great impropriety in calling them simply terminal moraines, since, wherever the ice-front paused for any length of time, a spe- cial accumulation of debris would take place, and would be' terminal to the ice at that point. West of the AUeghanies, President Cliamberlin deline- ates this moraine as extending in a series of lobes pointing to the south across the States of Ohio and Indiana, making one grand loop whose axis is nearly parallel with that of Lake Erie, returning with its western arm into eastern Michigan, between Saginaw Bay and the southern end of Lake Huron. BLe discovers five minor loops in this moraine in the axes of the following river valleys : (1) the Grand and Mahoning ; (2) the Sandusky and Scioto ; (3) the Great Miand — all in Ohio ; (4) the White, in Indiana ; and (5) the Maumee and Wabash. But the accumulations called terminal in this re- gion are by no means comparable in extent with those south of New England or west of Lake Michigan, and the system is made out with some difficulty. In this portion of the territory there is another interior morainic belt of such interest that we pause to describe it more particularly. We refer to that of the Maumee "Val- ley in Ohio, and can best describe it in the words of Mr. G. K. Gilbert, its original discoverer : The Maumee Eiver occupies the axis of the broad, shallow valley which it helps to drain. This valley has no strongly 180 THE ICE AGE IX NORTH AMERICA. marked limits. Eastward it is continuous with the trough of Lake Brie, and westward with the valley of the Wabash River. At the north, or more properly the northwest, its slopes merge, ata height of five hundred to six hundred feet (above Lake Erie), with those of the valley of Lake Michigan ; and its southern slopes, reaching a height of four hundred to five hundred feet, pass into those of the Ohio valley. With these low sides and a width of 125 miles, all its inclinations are exceedingly gentle, and the title of plain can be applied to it with no less propri- ety than that of valley. North of the Maumee the general de- scent is to the southeast, and south of that river to the north- east. With slight exceptions, the smaller streams follow and indicate these slopes, but all the larger tributaries of the Mau- mee, including the St. Joseph, St. Mary's, and Auglaize Rivers, and Bean or Tiffin Creek, appear to be independent of them. The St. Joseph, for example, flows to the southwest through a country where evei-y rivulet runs to the southeast. The entire region drained by it lies on its right bank, while from its left the drainage is toward Bean Creek, the divide between the two streams being everywhere within three or four miles of the St. Joseph. In like manner, the course of the St. Mary's is west and north, and, while from its left bank the streamlets flow northeast into it, from its right they flow northeast into the Auglaize. These hydrographical peculiarities are so singular and striking as to have excited some attention and curiosity before the region was visited. Upon examination, there was found a continuous ridge, following the eastern banks of these rivers, and evidently determining their courses. Running somewhat obliquely across the slopes of the country, it turned aside all the small streams, and united them to form the St. Joseph and St. Mary's. The height of this ridge is ordinarily from twenty-five to fifty feet, and its width at base from four to eight miles. Along the St. Joseph it is not distinguished from the adjacent country by its superficial characters. In common with that, it has a gently rolling surface, with a grav- elly clay soil, supporting a heavy growth of varied timber. Farther south, where it forms the north bank of the St. Mary's River in Van Wert and Mercer counties, it is marked by such peculiarities as to divide it very sharply from the adjoining TERMINAL MORAINES. 181 plains, which are nearly level, with a soil of fine clay, and coy- ered by a heavy growth of elm, beech, ash, maple, etc. The ridge, on the contrary, presents a confused series of conical hills, chiefly of clay, but showing some pebbles and small bowlders, and clothed by a forest-growth almost exclusively of oak. Probably the only essential point in this contrast is that of hill and plain, and out of this the others have grown. There is good reason to believe that the clay deposit (Erie clay) of the plain is continuous with that on the hills. Where its surface is level, it has retained its soluble salts and accumu- lated vegetable mold, so as to form a rich soil favorable to a varied vegetation ; while from the steep hill-sides a great amount of soluble and fine material has been washed, so as to bring to the surface some of the pebbles everywhere imbedded in greater or less abundance, and the character of the vegeta- tion has been determined by that of the soil. I conceive that this ridge is the superficial representation of a terminal glacial moraine, that rests directly on the rock- bed, and is covered by a heavy sheet of Erie clay, a subsequent aqueous and iceberg deposit. Though this formation has an average depth along the upper St. Joseph of over one hundred feet, and on the upper St. Mary's of fifty feet, it has not suf- ficed to conceal a moraine of such magnitude, but has so far conformed to its contour as to leave it still visible on the face of the country — doubtless in comparatively faint relief, but still so bold as to exert a marked influence on the hydrography of the valley.* "When, a little later, we come to speak of glacial erosion, something more will be said confirmatory of this hypotheti- cal moraine of Mr. Gilbert, and of the varying movements of ice in the Lake Erie Yalley at different stages of the Glacial epoch. We shall then see abundant reason for supposing that there was, for a considerable length of time about the close of the period, an independent movement of ice in the direction of the longer axis of the lake, and that this * " On the Surface Geology of the Maumee Valley," in the " Geological Sur- vey of Ohio,'' vol. i, pp. 540-542. 12 TERMINAL MORAINES. 183 movement was directly toward tlie head of tlie Maumee Kiver. The general system of interior moraines upon Avhieh we were remarking is pretty well exhibited about the eastern shore of Lake Michigan, forming a grand loop around that lake, and connecting with two subordinate loops around the head of Grand Traverse Bay and Saginaw Bay. But it is not until reaching the country west of Lake Michigan, in Wis- consin, that these glacial accumulations become again a very prominent feature of the landscape. Here they constitute the so-called Kettle Kange, which forms a loop pointing to the southwest in the line of the longer axis of Green Bay. President Chamberlin has shown * that the ice-movement to the southward through Green Bay was in a measure inde- pendent of that through Lake Michigan ; so that the eastern arm of the Kettle Bange might more properly be called a medial moraine, to which the Lake Michigan Glacier and the Green Bay Glacier both contributed their deposits. This eastern arm runs about half-way between Fond du Lac and Sheboygan, and thence a little west of south, through "Wash- ington and Waukesha counties, between Oconomowoc and Pewaukee, and through Eagle to Milton, between Janesville and Whitewater. Thence it swings northward, passing a few miles west of Madison, and, crossing an elbow of the Wiscon- sin River, incloses in its folds Devil's Lake, near Baraboo, and thence on northward into the wilderness of northern Wisconsin, a little beyond the latitude of St. Paul, where it turns westward and with some deflection reaches the St. Croix Kiver at Hudson, a few miles ahove its junction with the Mississippi. From this point it trends southward past Min- neapolis through southeastern Minnesota, inclosing in its folds Minnetonka and many other beautiful lakes in that portion of the State. From this point on, Mr. Upham has traced the moraine in an ox-bow-shaped extension, whose * " Preliminary Paper on the Terminal Moraine of the Second Glacial Epoch," p. 315, et seq. 184 TEE ICE AGE IK NORTH AMERICA. southern extremity is near Des Moines, Iowa, and whose western limb is the Coteau des Prairies of eastern Dakota. The conclusions of Mr. Upham and President Chamber- lin, concerning the movements of the ice over the region west of the lakes, are intensely interesting and seem amply warranted by the facts. It appears that in the northwest the ice advanced in four lines of motion pointing to a center a little below Dabuque, Iowa, though the columns did not all reach the point of their apparent destination : 1. One line of advance was down the depression of Green Bay, in Wis- consin. The moraine of this lobe constitutes what is called the Kettle Kange of that State, and terminates a little west of Madison, on the eastern edge of the driftless region. 2. A second line of movement was doAvn the valley of Kewanee Bay. This movement spent its force in northern Wisconsin, reaching the vicinity of Eau Claire. 3. The third move- ment was along the line of the main axis of the western end of Lake Superior, and extended across the Mississippi past Minneapolis, as far as Lake Minnetonka, and to a line run- ning northwest from this point for a hundred miles or more. 4. The fourth movement was from the region of Lake Win- nipeg in the Hed Kiver Yalley toward the south and south- east, meeting and opposing the ice-current from Lake Supe- rior, along a line from Steams county, Minn., southeast by Lake Minnetonka to Crystal Lake in Dakota county. This is the movement which extended southward to the vicinity of Des Moines, Iowa, and whose western flank is the Coteau des Prairies. The line northwestward from Minneapolis, where these last two movements met, was an interesting battle-ground of the glacial forces. First, the Lake Superior Glacier pre- vailed, and pushed over the ground to its extreme limit, even beyond the Mississippi. This boundary-line runs from Crystal Lake through Minnetonka, Wright, and Stearns counties, Minn. A little later, the Eed River Glacier gained the ascendency, pushing the front of the Lake Supe- rior Glacier back into Wisconsin, east of the Mississippi. TERMINAL MORAINES. 185 The reality of this battle of the glaciers, and of this alter- nate advance and retreat of the opposing forces, is shown by the succession of deposits. The lower part of the ground moraine is characterized by a reddish color, and by rock- fragments from the region of Lake Superior ; while the upper portion, now upon the surface, is of a bluish color, containing bowlders and pebbles of limestone and of creta- ceous shale, and other material brought from the northwest, showing that victory was first to the Lake Superior Glacier, but finally to that of the Eed River. The evidence of the junction of these two great ice- streams appears clearly enough upon the surface when sec- tions of country a little distance apart are considered. This, Dr. G. M. Dawson had observed as early as 1875, when he wrote about it thus : A line drawn northeast and southwest, nearly parallel with the northwestern shore of Lake Superior, but lying a short dis- tance back from it, and cutting the Northern Pacific Eailway some miles west of Thomson, in this part of Minnesota, sepa- rates superficial deposits of different aspects. Northwest of this line the prevailing tint of the drift material is pale yellow- ish-gray, or drab ; southeast of it, reddish tints are almost universal, and become specially prominent on the northern part of the line of the Lake Superior and Mississippi Eailway, and continue to St. Paul. The junction of these two varieties of drift can not, of course, be exactly defined, but is interest- ing as an indication of the direction of transport of material in this region ; the reddish matter being derived from the red rocks of the lake-shore.* Some other interesting things concerning the deposits of this region can better be said when we come to treat of gla- cial dams and lakes, and of the cause of the Glacial period. The surprising thing to a glacialist, upon a first visit to * " Report on the Forty-ninth Parallel," p. 213. See also " The Fresh-Water Glacial Drift of the Northwestern States," p. 9, by Colonel Charles Whittlesey, who, it would seem, had noted the distinction as early as 1866. 186 THE ICE AGE IN NORTH AMERICA. southeastern Dakota, is the extensiveness of the apparently level areas where the till comes to the surface. This impres- sion is heightened, probably, by the absence of forests, and would very likely be the same in portions of Ohio and Indiana were it not for the timber. James Eiver Valley, in Dakota, is depressed in the center about five hundred feet below the edges, but it is, roughly speaking, seventy miles across, and therefore the slope does not strike the eye. So level is the country, that every special line of glacial accu- mulation is a prominent feature in the landscape, and the various halting-places of the ice in its retreat are readily dis- cerned. Evidently, a lobe of ice for a long time tilled the James River Valley, running parallel with that which occu- pies the upper Minnesota Valley, and extended southward into Iowa. The edges of these lobes thinned out along the north-and-south line which runs near the east margin of southern Dakota, and favored the accuujulation of the mo- rainic hills, to which we have already referred as the Coteau des Prairies. Professor Todd and others speak of this as a series of terminal moraines formed along the sides of the re- entrant angle, between the two lobes, whose apex penetrated to the vicinity of the Sisseton Agency. Perhaps, however, it would facilitate a proper understanding of the subject to speak of the Coteau des Prairies as a medial moraine, like that east of Green Bay, toward which the glacial debris, carried on the deeper portions of the ice of both the Minne- sota liiver and James River lobes, gravitated in contrary directions. But, whatever the name, certain it is that, start- ing from the Sisseton Agency, different lines of glaciiil accu- mulations stretch southward at varying angles, the later accumulations forming the more obtuse angle. Coming up the valley of the James from Yankton, one crosses the old- est of these accumulations (the Altamont Moraine) in the neighborhood of the city itself, and the second (or Gary Moraine) in the neigliborhood of Mitchell, sixty miles to the north, having run parallel with it, however, for about thirty miles. The third (or Antelope Moraine) is encountered near TERMINAL MORAINES. 187 Huron, about sixty miles nortli of Mitchell, and continues visible upon either side of the river about twenty miles dis- tant, as far up as Aberdeen. The western side of this lobe is characterized by corre- sponding lines of receding moraines, the outer of which is in the vicinity of the Missouri Eiver and on its eastern side. Together, these form the Missouri coteau. Everywhere, in coming up from the river on the west to the plateau, which is in most places from four hundred to iive hundred feet above the river, one encounters two or three bowlder-cov- ered terraces, the highest of which are at an elevation of from three hundred to four hundred feet. The moraines rise to a considerable height above the general level, and, as upon the eastern side, are everywhere marked features of the landscape. The streams entering the Missouri from the east are all of them short, none being more than forty miles in length. These streams are in all cases bordered by broad and elevated local terraces, the edges of which, where they overlook the immediate trough of the stream, are crowded with granitic bowlders. In some cases, as Professor Todd has shown, these valleys terminate abruptly in the water- parting, as if being the continuation of glacial streams from the east, which had originated upon the ice-lobe while it filled the James Valley. Professor Todd, to whom the exploration of this region was assigned, thus describes the Missouri coteau : This moraine consists of loops, convex usually toward the west and south, but in rare cases toward the northwest, as will be seen. These loops connect at re-entrant angles pointing toward the northeast and east, which are usually sharp, and sometimes are extended into elongated ridges. The moraine varies in elevation with the region on which it rests. Its rela- tive height is usually great at the head of the re-entrant angles or interlobular moraines. These frequently stand out like great promontories, rising from one hundred and fifty to four hundred feet above the plain around them. At the bot- tom of a loop the moraine is apt to be slight or wanting, if on MISSOURI COTEAU ^^''r ""'-'^ AND ITS MORAINES. By J. E, TODD, Asst. Onologist U. S. G. Si v>/-:v;:;y:->V Moraines f, 2, 3 dt4, Coiitotirs Sailroads Limit of Drift Shore of Lake Dakota FiQ 55. TERMINAL MORAINES. 189 lower land ; the flow of water from the ice probably haying carried away the debris as rapidly as it was pushed forward by the ice. On the other hand, in case the loop was pushed up an inclined plane, and the water did not find free escape, it (the loop) is well developed all around. The outer moraine in some places is very rough and stony ; at other points it is a smooth, broad ridge, with few knobs, and covered with a deep, fertile soil.* Dr. G. M. Dawson early discovered the significance of this great Missouri coteau in its extension north of the United States boundary-line, and thus describes it : On approaching its base, which is always well defined at a distance, a gradual ascent is made, amounting, in a distance of twenty-five miles, to over 150 feet. The surface at the same time becomes more markedly undulating, as, on nearing Turtle Mountain from the east, till almost before one is aware of the change, the trail is winding among a confusion of abruptly rounded and tumultuous hills. They consist entirely of drift material ; and many of them seem to be formed almost alto- gether of bowlders and gravel, the finer matter having been to a great ex,tent washed down into the hollows and basin-like valleys without outlets with which this district abounds. The ridges and valleys have in general no very determined direc- tion, but a slight tendency to arrangement in north-and-south lines was observable in some places. The bowlders and gravel of the coteau aire chiefly of Lau- rentian origin, with, however, a good deal of the usual white limestone and a slight admixture of the quartzite drift. The whole of the coteau belt is characterized by the absence of drainage valleys ; and, in consequence, its pools and lakes are often charged with salts, of which sulphates of soda and mag- nesia are the most abundant. The saline lakes frequently dry up completely toward the end of the summer, and present wide expanses of white efflorescent crystals, which contrast in * " Proceedings of the American Association for the Advancement of Sci- ence," vol. xxxiii, 1884, p. 383. TERMINAL MORAINES. 191 color with the crimson Salicornia with which they are often fringed. Taking the difference of level between the last Tertiary rocks seen near the east- ern base of the coteau and those first found on its western side, a distance of about sev- enty miles, we find a rise of six hundred feet. The slope of the surface of the under- lying rock is, therefore, assuming it to be uniform, a little less than one hundred feet per mile. On and against this gently in- clined plane the immense drift deposits of the coteau hills are piled. The average elevation of the coteau above the sea, near the forty-ninth parallel, is about 3,000 feet ; and few of the hills rise more than one hundred feet above the gen- eral level. Between the southwestern side of the coteau belt and the Tertiary plateau is a very interesting region with characters of its own. Wide and deep valleys with sys- tems of tributary coulees have been cut in the soft rocks of the northern foot of the plateau, some of which have small streams still flowing in them fed by its drainage ; but for the most part they are dry, or oc- cupied by chains of small saline lakes which dry up early in the summer. Some large and deep saline lakes also exist which do not disappear even late in the autumn. They have a winding, river-like form, and fill steep-sided valleys. These great old valleys have now no outlet ; they are evi- dently of preglacial age, and have formed a part of the former sculpture of the coun- try. The heaping of the great mass of debris of the coteau against the foot of the Tertiary plateau has blocked them up and 192 TH£ ICE AQE IN NORTH AMERICA. preyented the waters finding their way northward as before ; and, since glacial times, the rainfall of the district has never been sufficiently great, in proportion to the evaporation, to enable the streams to cut through the barrier thus formed. The existence of these old valleys, and the arrangement of the drift deposits with regard to them, throw important light on the former history of the plains. Northward, the coteau ceases to be identified with the Ter- tiary plateau, and rests on a slope of cretaceous rocks. It can be followed by Palliser's and Hector's descriptions of the coun- try to the elbow of the South Saskatchewan, and thence in a line nearly due north through the Eagle and Thickwood Hills ; beyond the North Saskatchewan, however, it appears to be- come more broken and less definite. In Dr. Hector's descrip- tion of certain great valleys without outlet in this northern region, I believe I can recognize there, too, the existence of old blocked-up river-courses similar to those just described. South of the forty-ninth parallel the continuation of the belt of drift material can also be traced. It runs southeast- ward, characterizing the high ground between the tributaries of the Missouri and the Ked River, which has already been noticed in connection with the water-shed of the continent ; but, wanting the backing of the lignite Tertiary plateau, it appears to become more diffuse, and spread more widely over the country. That the drift deposits do not form the high ground of the water-shed, but are merely piled upon it, is evi- dent, as cretaceous rocks are frequently seen in its neighbor- hood at no great depth. . . . In the coteau, then, we have a natural feature of the first magnitude— a mass of glacial debris and traveled blocks with an average breadth of perhaps thirty to forty miles, and ex- tending diagonally across the central region of the continent for a distance of about eight hundred miles.* To one familiar with the literature of the subject, it would seem that Dr. Dawson's sagacity in thus early dis- * " Quarterly Journal of the Geological Society," vol. xxxi (November, 1875), pp. 614-616. The facts are more fully stated in his governmental "Re- port on the Forty-ninth Parallel." TERMINAL MORAINES. 193 ceming the great significance of the Missouri coteau has not received from glacial writers all the recognition it fairly deserves. But here we have, as in so many other instances, fresh illustration of the fact that the minds of sagacious investigators run in the same channel. Noteworthy in- ventions and discoveries are not often due to the work of single individuals. From the references already given, it would appear that Dr. Dawson's surmise as to the signifi- cance of the Missouri coteau, President Chamberlin's theory as to the meaning of the Kettle Range, Professor Cook's delineation of the moraine across New Jerse}'^, and Clar- ence King's interpretation of the glacial accumulations on the south shore of Massachusetts were, in the minds of the authors, nearly contemporaneous and of independent origin. The subject of this chapter will not be complete without speaking of those later and more local moraines which were formed when the ice had withdrawn itself from the country in general, but still lingered everywhere in the mountains. Such moraines are numerous in all the valleys of the "White Mountains. Professor Agassiz * describes no less than fifteen terminal moraines of small size crossing the valley of the Amraonoosuc, a short distance below Bethlehem. Similar moraines exist in the valley leading down the Saco near Bartlett, and in the White Mountain branches of the Andros- coggin, as well as in the valleys leading to the vicinity of Center Harbor, on Lake Winnepesaukee. The principal local moraine of the Androscoggin is near the State line between Shelburne, N. H., and Gilead, Me. This has been described by Professor Stone and others.f Other interesting moraines described by Professor Stone in Maine are located at Read- field Village, and at Swan Island in the Kennebec Yalley, and at Sabbattusville, Machias, and "Waldoboro. Among the innumerable instances of local moraines on * " Geology of New Hampshire," vol. iii, pp. 236-238. ■f " American Journal of Science," vol. cxxxiii, 1887, p. STQ, 18 194 THE ICE AGE IN- NORTH AMERICA. the Pacific slope, the followiiig, described by Mr. I. C. Rus- sell, may serve as specimens : If one proceeds up the caflon [of Leevining Eiver, Mono county, California], he will cross five or six small terminal mo- raines which traverse from side to side the broad trench left by the ancient glacier. These are seldom more than fifteen or twenty feet high, and are separated by grassy meadows. The creek was formerly dammed by these moraines and forced to expand so as to form small lakes ; but these have long since been drained by the cutting of channels through the obstruc- tions.* Other instances have been already mentioned in describ- ing the glaciated boundary in Colorado, southern California, Oregon, and "Washington Territory. Moraines of retroces- sion characterize almost every mountain valley which was occupied by ice during the Glacial period. A brief summary of the conclusions embodied in the last two chapters will here be advantageous. The Ice age of North America doubtless began in local glaciers, which subsequently became confluent over the whole glaciated area; and it is evident that the currents of ice- movement at all times felt the direct influence of these local subcenters. An extensive region in southwestern Wisconsin was never covered by ice, though the glaciers became confluent south of it and extended onward a distance of three hundred miles. The effect of these subcenters of glacial radiation is to be seen in the lobate character of the terminal moraines, which naturally was more marked during the intermediate stages both of the advance and the retreat than at the time of maximum extension. The phrase terminal moraine is somewhat misleading, since the amount of marginal deposit is likely to vary so much in different places. Where the ice is for a long time stationary, and in front of regions where the rocks are more * " Quaternary History of Mono Valley, California," p. 334. TERMINAL MORAINES. 195 easily disintegrated or more accessible to the ice-streams than elsewhere, it is evident that unusually large terminal deposits will take place ; while, over regions where the ice both ad- vanced and retreated regularly, the deposits would be more equably spread over a wide margin of terminal frontage. This probably accounts for the lack of great ridges such as are to be found in southern New England, on the extreme mar- gin in the interior States. The total amount of deposit is as great in the interior as near the coast, but it is spread over a wider area. The " kettle moraines " of Wisconsin and the Northwest are capable of two diverse interpretations. By President Chamberlin and others they are taken as moraines of a dis- tinct second Glacial epoch, during which the ice extended no farther south in that region. Others would regard them as belonging to a stage in the recession of the ice of one great Glacial epoch in this country. Reasons for regarding the period as a unity will be ad- duced later on. Though the more accurate term moravnes of retrocession seems in itself preferable, it is not necessary to attempt to change the nomenclature in use, since a moraine of retrocession was terminal to the ice-front at the time of its deposition. On any view of the case we are compelled to regard the Glacial period as one of great vicissitudes. Evidently, as will appear in future chapters, there were great irregularities both in the advance and in the retreat of the ice. From a combination of causes which can not yet be explained there were periods of rapid advance alternating with periods of retreat, intercalated with long periods of established equilibrium. The glacialist possesses the great advantage of having a most complicated cause at hand to account for his phenomena. But while this enables him to explain everything, it as well prevents him from being over- confident in any special solution he may present of minute problems. CHAPTER X. GLACIAL EROSION AND TEANSPOJRTATION. The extent of glacial erosion is a hotly contested question. One class of writers has gone to the extreme of attributing almost all the erosion of the higher latitudes to glacial action, FiQ. 68— Canon of the Colorado. Stream-erosion in a dry climate. (Newberry.) while another class has scarcely allowed any eroding power to glacial ice, in comparison with that of running water. In considering the relative importance of these agencies GLACIAL EROSION AND TRANSPORTATION. 197 two elements enter into the problem : (1) the relative rate of action of the two forces, and (2) the relative length of time during which they have been in operation. As to time, it is evident that those have a great advantage in the argument who exalt the eroding power of running water. However slowly the drops may wear away the stone, ample amends are made in the length of the periods through which the action has continued. From the earliest ages of geological history running water has been at work counter- acting the effect of the forces which have elevated the con- tinents. River - channels are, in fact, more constant than mountain-chains. Everywhere and at all times the aceumu- .^^ Fig. 59. — Embossed floor of an ancient glacier in the valley of the upper Arkansas River. (Hayden.) lating waters on a continental area seek and find the lowest paths to the sea. The sand and gravel which these running streams push along over their beds act as the teeth of a saw upon the rising mountain-summits, so that everywhere in mountain-regions of great age we find deep, transverse val- 198 THE ICE AQE IN NORTE AMERICA. leys of erosion. Among the best-known examples in the country are those of the Mohawk and Hudson in New York ; of the Delaware and Susquehanna in Pennsylvania ; the Ohio and its tributaries on the western flanks of the AUe- ghanies ; the Mississippi and all its western branches, together with the Colorado upon the eastern flanks of the Kocky Mountains ; and the Columbia, the Fraser, and the Stickeen Elvers, which penetrate in chasms of great depth the rock- bound shore of the Pacific. At the Delaware Water-Gap the river has sawn a vertical chasm more than a thousand feet deep directly across the hard strata of the Kittatinny Mountain. For fifty miles above Lock Haven, on the West Branch of the Susquehanna, the river occupies a narrow valley of erosion more than a thousand feet in depth. For nearly twelve hundred miles, as the water runs, the Ohio River, with its extension up the Alleghany, occupies a narrow, crooked valley, one mile or more in width and several hundred feet in depth, which it has worn through nearly parallel strata of lime and sandstone rock. The trough of the Mississippi from Cairo upward is similar to that of the Ohio, except that it is two or three times as broad. The canons of the Colorado, of the Yellow- stone, and of the Columbia, are of world-wide renown. The Colorado has worn a channel with nearly perpendicular sides three hundred miles long and from three to six thousand feet deep.* Such are some of the well-recognized results produced by the long-continued mechanical action of running water. But, aside from its mechanical action, water with the acids it contains is a most eflScient chemical force, acting as a solvent upon various rocks. Every salt-spring and every spring of hard water in a limestone region is undermining the country from which it issues, and is engaged in dissolving the solid material and in transporting it in solution to lower levels. It is only a question of time when the chalk cliffs of * "Elements of Geology," by Joseph Le Conto, pp. 15-17. GLACIAL EROSION' AND TRANSPORTATION. 199 England and the extensive lime formations of the Appala- chian region in America shall all be dissolved and carried in invisible solutions to the sea. The Mammoth Cave is but the remnant of larger, longer, and more numerous caverns which have honey-combed vast regions in Kentucky and in Tennessee. Many of the extensive valleys of that region are but the depressions formed by the falling in of the roofs of innumerable caverns. The rate of chemical erosion on limestone rocks is not easy to estimate. The most elaborate attempt of which I am aware is that of Professor A. L. Ewing in the Nittany Yal- ley, of Huntingdon county, Pa.* This valley, which is known by different names, extends through a considerable portion of the Appalachian region. It consists of the remains of a great anticlinal fold, which, had it not been eroded away, would form an immense mountain-like plateau over 20,000 feet above its present height. As it is, the floor of the valley is composed of the upturned edges of the lower Silurian limestone, eroded through a thickness of 6,000 feet. The valley is flanked on either side by the overlying Medina sandstone, which forms monoclinical ridges from 600 to 1,000 feet above its floor. From data carefully collected, Professor Ewing ascer- tains that the amount of solid matter annually carried out of the valley in chemical solution is equal to a layer of ^lj,\r^^ of a metre in thickness. Hence, to lower the surface to the ex- tent of one metre by this process would require 29,173 years — that is, it would take about 9,000 years to remove one foot from the surface. It is safe to assume that had the rocks of this region been similar to those of the bordering mountains in their nature and power to resist dynamical agencies, we should have in place of Nittany Valley a mild anticlinal plateau somewhat above the * See " Proceedings of the American Association for the Advancement of Science," vol. xxxiii, 1884, p. 404. The paper was published in the " Second Geological Survey of Pennsylvania, T'," pp. 451-454. 200 TEE ICE AGE W NOliTS AMEMOA. mountains in elevation — say 1,000 feet above the present height of the valley. The erosion in the valley, then, in excess of that along the mountains has been mainly chemical, and at least a thousand feet of limestone have been thus removed. A simple further deduction shows that, accordingly, Mttany Valley has been one million years in process of formation. The limestone erosion could not begin before the latter stages of the Mesozoic era, possibly not before the Oenozoic era, as sufficient time must have elapsed subsequent to the Car- boniferous age to erode all formations of the Palaeozoic era above the Trenton limestone. One million years seems not in- consistent with other estimates of geological time. In view of sucli facts the advocate of glacial erosion can not continue to maintain that ice is the chief agency in form- ing the contour of continental areas ; but must grant that, by reason of the great length of time during which water has been about its work of corrosion and erosion, it is, with- out doubt, the most important instrument in diversifying the features of the earth's surface. Still, however short, by com- parison, have been the periods of glacial action, no one can study a glacier or a glaciated region without being deeply impressed with the eroding and transporting power of mov- ing ice. To get a full conception of its erosive power, one must either get beneath it, or be able to calculate the force of its movement from what he knows of the natm-e of ice. Neither of these plans is altogether satisfactory, but each of them is to some extent feasible. From the nature of the elements at work it has qnite generally been supposed that an advancing glacier would act like a plow or scraper, greatly disturbing and modifying the deposits over which it passed. But observation and a more careful consideration of the qualities of ice have materially modified these early impressions. From the fact that a stream of ice moves faster at the top than at the bottom, it follows that its action on iinderlying deposits is more like that of a drag than like that of a plow. The rocky frag- GLACIAL EROSION AND TRANSPORTATION. 201 ments frozen into the bottom of the ice are not held there by a perfectly firm and unyielding grasp. The same bowlder which plows a furrow in the rock beneath it, plows a longer furrow in the ice which is moving over it. This is finely illustrated by some observations of Professor Niles. In a visit to the great Aletsch Glacier, in the summer of 1878, Professor Niles had an excellent opportunity to exam- ine the under side of the ice of this glacier near its front. Here he observed jiumerous elongated ridges of rock over which the ice was flowing lengthwise, adjusting itself to all the corrugated surface. When the ice passed the lee end of the ridge it carried with it " the mold of the profile so per- fectly that for more than twenty feet the blue arch presented a series of parallel furrows, like the fiutings of a Doric column." There was there at that time another highly interesting and instructive exhibition of glacial action. Within a few feet of the down-stream end of one of these elongated rocJies mouton- nees and upon its crest, there was a bowlder fully three feet in diameter, which evidently had been slowly moving along this ridge for some distance, probably from its upper end. There were two sides of this block of stone which were not incased in ice, viz., the lower one resting upon the rock, and the one facing down the glacier. From the lower end of the ridge of rock I looked at the bowlder through a tunnel of pure, blue ice, which was continued as a deep furrow in the under sur- face of the glacier for fully thirty feet from its beginning. As this was produced by the ice moving over and beyond the bowlder, it was evident that the ice was moving more rapidly than the stone. I afterward found other examples of the same kind, but none so favorably situated for a striking exhibition of this property of ice. It will be understood that these stones were sufficiently below the upper surface of the glacier to be removed from the efEects of the ordinary changes of the tem- perature of the atmosphere. Although stones which are ex- posed to such changes may be frozen into the ice at the edges of the glacier, yet I believe these were so situated as to cor- 202 THE ICE AGE m NORTS AMERICA. rectly represent the conditions and movements of this at still greater depths. If this is correct, and I believe it is, it follows that such fragments of rock are not rigidly held in fixed posi- tions in the under surfaces of glaciers and carried irresistibly along at the same rate, but that the constantly melting ice actually flows over them, and that their motion is one of ex- treme slowness, even when compared with the motion of the glacier itself.* In a visit to the glaciers of Norway in 1886, Professor J. W. Spencer found abundant confirmation of Professor Niles's inferences concerning the low eroding power of glaciers in certain conditions. He reports that the advancing Nor- wegian glaciers " do not conform to the surfaces over which they pass, but are apt to arch over from rock to rock and point to point, especially as they are descending the ice-falls." Professor Spencer continues : Beneath the glaciers of Pondal, Tunsbergdal, and Buardal, in the northern, north-central, and south-central snow-fields of Norway, as well as under other glaciers, I observed many stones inclosed in ice resting upon the rocks, to whose surfaces — sometimes flat, sometimes sloping steeply — they adhered by friction and by the pressure of the superincumbent weight. Although held in the ice on four sides with a force pushing downward, the viscosity of the ice, or the resistance of its molecules in disengaging themselves from each other in order to flow, was less than that of the friction between the loose stones and the rock ; consequently, the ice flowed around and over the stones, leaving long grooves upon the under surfaces of the glacier. An example of the ability of the iqe to flow like a plastic body was shown in a cavern four hundred feet higher than the end of the glacier, where the temperature was 4° C, while that outside was 13° C. Upon the deiris of the floor rested a rounded bowlder, whose longer diameter measured thirty inches. A tongue of ice, in size more than a cubic yard, was hanging * "American Journal of Science," vol. cxvi, WIS, p. 366 et seq. GLACIAL EROSION AND TRANSPORTATION. 203 from the roof and pressing against the stone. In place of push- ing the stone along or flowing around it, the lower layer of ice above the tongue had yielded, and was bent backward as easily and gracefully as if it had been a thin sheet of lead, instead of one of ice a foot thick. The insufficiency of glaciers to act as great erosive agents is further shown at Pondalen, where a mass of ice thirty or forty feet thick abuts against a somewhat steep ridge of a rock ten feet or less in height. In place of a stone-shod glacier sliding up and over the barrier, the lower part of the ice appears stationary, or else is moving around the barrier, while the upper strata bend and flow over the lower layers of ice.* The whole body of facts concerning a ground moraine speaks in like manner of the limited amount of disturbance, in certain conditions, which is produced by the ice as it moves over loose material. There can be little doubt that, for a breadth of a hundred miles or more on the border of the glacial limit in North America, the ice advanced over the loose material (which is variously called " bowlder-clay," " till," and " ground moraine ") without greatly disturbing it as a body. Indeed, this great mass of firmly compacted, unassorted, and glaciated material would seem to have ac- cumulated by degrees — tlie moving ice, dragging along under it successive strata of the grist which it had ground from the surface of the rocks far to the north, where its action had been more vigorous and long continued. For another strik- ing illustration of the power of ice thus to move for a limited distance over loose material without disturbing it, one has but to refer to the description already given of the buiied forest near the southwestern comer of the Muir Glacier, Alaska-t. Here large trees in great numbers, which have been preserved for an indefinite period underneath the gla- cier, are now being uncovered, and appear standing upright * See "Glacial Erosion in Norway and in High Latitudes,'' reprinted from "Proceedings of the Royal Society of Canada," IBS'? ; extracted from " Ameri- can Naturalist," vol. xxii,1888, pp. 218, 221, 223. ■f See p. 59. 20J: THE IGE AGE IN NORTH AMERICA. with their branches intact upon them, and their roots im- bedded in the soil in which thej grew. A stratum of this soil even consists of moss and leaves and cones which origi- nally formed a carpet over the forest floor. There can be no doubt that, after the accumulation of sand burying the forest, the glacier advanced for a great distance over it, attaining a thickness at that point of two or three thousand feet. A little reflection will show that the advance of a glacier upon a new field is analogous to that of the breakers in the ocean over shallow bottoms, though the impressiveness of the scene is disguised to the physical senses by the slowness of the movement in the case of a glacier, and by the counteract- ing effect of the heat which limits the advance of the ice-front. Where the ice-movement is upon dry land, the front is ordi- narily represented by a sloping field of ice covered with debris deposited from the melting surface near its terminus. For a long time investigators were puzzled by the fact that many bowlders are found some miles south of the line mark- ing the limit of glacial scratches upon the surfaces of the rocks. But, in the light of the previous suggestions, it is easy to see that for some distance back from the southern margin there could have been no movement at all at the bottom of the ice, so that bowlders upon the surface might be transferred, as upon the summit of a breaker, from some distance back of the front to some distance beyond the far- thest limit attained by the lower strata of moving ice. This narrow belt of glacial deposits bordering the limit indicated by other glacial signs constitutes what Professor Lewis and myself * agreed to call " the fringe " of the terminal moraine. West of the Alleghanies this fringe is, as already shown, so wide as to assume commanding importance, and everywhere deserves more attention than we at first gave it. \ This characteristic of the movement of glacial ice is illus- trated by another phenomenon which has not been sufliciently * See " Second Geological Survey of Pennsylvania, Z," pp. 45, 206 et seq. f See above, p. 135. GLACIAL EROSION AND TRANSPORTATION. 205 weighed. In most of the text-books the formation of ice- bergs is represented to be by the pushing out of the ice into the water until the depth is such as to overcome its specific gravity, and lift a mass of ice up bodily and float it away. An inference from imperfect data by Dr. Kane has doubt- less done much to foster this idea. Eegarded upon a large scale, I am satisfied that the iceberg is not disengaged by debdcle, as I once supposed. So far from falling into the sea, broken by its weight from the parent- glacier, it rises from the sea. The process is at once gradual and comparatively quiet. The idea of icebergs being dis- charged, so universal among systematic writers and so recently admitted by myself, seems to me now at variance with the regulated and progressive action of Nature. Developed by such a process, the thousands of bergs which throng these seas should keep the air and water in perpetual commotion, one fearful succession of explosive detonations and propagated waves. But it is only the lesser masses falling into deep waters which could justify the popular opinion. The enor- mous masses of the great glacier are propelled, step by step and year by year, until, reaching water capable of supporting them, they are floated off, to be lost in the temperature of other regions.* Doubtless some icebergs are thus formed. But any tour- ist to Alaska may now satisfy himself that the ordinary method of the formation of an iceberg is by the breaking off of masses from the top as that portion of the ice is pushed on in advance of the lower strata. Still, as the fractures would not always reach to the bottom of a deep inlet, masses thus left below the water would eventually rise to the sur- face in case the front of the ice were retreating, so as to remove the superincumbent weight from them. Coming now to consider the direct action of glacial ice in the transportation of solid material, we speak first of that carried upon the surface. Apparently there is scarcely any * " Arctic Explorations," vol. ii, p. 148. 206 THE ICE AGE IX NORTH AMERICA. limit to the size of the fragments of rock which can be transported upon the back of a glacier. Nor would there seem to be any definable limit to the distance through which these masses of rock can thus be carried, except as there is a limit to the movement of the ice itself. In walking out on the smooth surface of the eastern part of the Muir Glacier, it was not uncommon to encounter, miles away from any mountains, cubical blocks of stone as much as twenty feet in their several dimensions, which, with countless others of smaller size, united to form a medial moraine. Slowly but surely these great bowlders have been brought to their pres- ent position, and slowly but as surely they are moving on to the front of the glacier, where, in due time, they will be deposited in the terminal moraine. Fig 60 —Vessel Rock, a glacial bowlder n & Is im, N. H. (C. H. Hitchcock.) The summary of facts published by President Hitch- cock many years ago may fitly serve as an introduction to the more detailed account to follow. In this, after remark- ing upon the great size of single bowlders, he illustrates the remark by the following examples : The block called Pierre a Bot, near NeufchAtel, contains 40,000 cubic feet. It has been transported from near Mar- tigny, more than sixty miles, across the great valley of Switz- GLACIAL EROSION ANB TRANSPORTATION. 207 erland. Professor Forbes describes another bowlder in the Alps, one hundred feet long and forty to fifty feet high ; also another, sixty-two feet in diameter, containing 244,000 cubic feet. In this country bowlders occur of equal dimensions. Thus, on Cape Ann and its vicinity, 1 have not unfrequently met with blocks of syenite not less than thirty feet in diame- ter; and in the southeast part of Bradford (Mass.) I noticed one thirty feet square, which contains 37,000 cubic feet, and weighs not less than 3,310 tons. In the west part of Sand- wich, on Cape Cod, I have seen many bowlders of granitic gneiss twenty feet in diameter, which contain 8,000 cubic feet, and weigh as much as 680 tons. Two sandstone bowlders of the same size lie a few rods distant from the meeting-house in Norton. A granite bowlder of equal dimensions lies about half a mile southeast of the meeting-house in Warwick ; and one of similar dimensions lies on the western slope of Hoosac Mountain, in the northeast part of Adams, at least one thousand feet above the valley over which it must have been transported. One of granite lies at the foot of the cliffs at Gay Head, on Martha's Vineyard, which is ninety feet in circumference, and weighs 1,447 tons. In Winchester, N. H., I recently met with a block of granite eighty-six feet in circumference. It is near the road leading to Eichmond. I noticed another in Antrim, in that State, one hundred and fifty feet in horizontal circumference. Finally, at Fall Eiver was a bowlder of conglomerate which originally weighed 5,400 tons, or 10,800,000 pounds.* A well-known bowlder in eastern Massachusetts, situated on a precipitoas cliff in the southern part of the town of Peabody, goes by the name of Ship Rock. This is a gran- ite, and measures forty-five feet in length by twenty-two in height, and twenty-five in width. Its estimated weight is 1,100 tons, and it is surrounded by many loose fragments weighing from fifty to seventy-five tons each. This rock has been purchased by the Essex Institute of Salem, and is carefully preserved from destruction.f * " Elementary Geology," pp. 242, 243. t See " Journal of Essex County Natural History Society," p. 120. 208 THE ICE AGE IN NORTH AMERICA. The largest bowlder yet described in New England is Mohegan Eock, in the town of Montville, New London county, Conn. Its dimensions, as reported to me by Mr. David A. Wells, are as follows : Length of eastern side, 54 feet ; southern side, 70 ; western side, 56 ; northern, 58 ; maximum height at least 60 feet. The weight has been esti- mated at 10,000 tons. The northern origin of much of the material composing the deposits near the margin of the glaciated area is unmis- takable, and one is as much surprised by the distance it has been moved as by the size of the masses transported. As yet, however, there has been no systematic and complete comparison of these bowlders with the rocky masses to the north, for the purpose of definitely locating the origin of the transported material ; and there is such a degree of super- ficial resemblance between the granitic formations in differ- ent localities, that no comparison will be satisfactory until the specimens have been both chemically analyzed and examined under the microscope. This almost endless task remains yet to be done. Still, a superficial examination reveals much of interest. In general, it is manifest, even to the casual observer, that everywhere east of the Mississippi River the material com- posing the till is of northern origin. In Essex county, Mass., the bowlders are readily traced to localities in New Hamp- shire, some of them but a few miles away, and others from the White Mountain region, one hundred miles and more distant. At an early stage of my own investigations, my attention was especially attracted, at Andover, to bowlders of porphy- ritic gneiss, containing very large and characteristic crystals of feldspar. The nearest ledges from which they could probably come are in the vicinity of Lake Winnepesaukee, in New Hampshire — seventy and more miles away — where this rock is abundant. So the whole mass of drift, covering Plymouth county, Mass., and that forming the hills of Cape Cod and of the islands to the south, is evidently of northern origin, though much of it is from no great distance. Plym- GLACIAL EROSION AND TRANSPORTATION. 209 outh Bock is a bowlder from the vicinity of Boston, having accomplished its pilgrimage long before the departure of the Mayflower from Holland. Professor Shaler has traced a belt of bowlders removed from a small mountain in the north of Rhode Island, and is endeavoring to estimate the amount, so as to determine the extent of the erosion taking place under a movement reaching from that point to the terminal mo- raine. The material occurs in a triangular-shaped area to the south with its apex at the mountain, and can be easily traced. The bowlders scattered over the surface of the moraine in Long Island consist largely of granite, gneiss, hornblende rock, mica-slate, and red sandstone, which are easily recog- nized as from well-known localities in Ehode Island, Con- necticut, and Massachusetts.* Professor Dana has also described f many bowlders in the vicinity of New Haven, which have been transported from well-known trap-dikes sixteen miles or more from the north- east. One of these, called Judge's Cave, on "West Eock, 365 feet above the sea, must weigh a thousand tons. The train of bowlders in Richmond, Mass., near the sum- mit of the Berkshire Hills, was long ago described by Presi- dent Hitchcock and Sir Charles Lyell, and more recently and accurately by Mr. E. E. Benton. % So much has been written about this train of bowlders, that it is worth while to give the results of this later investigation. The locality is in the towns of Lebanon, N. Y., and Richmond, Lenox, and Stockbridge, Mass., upon the western side of the Berkshire Hills. The trend of the rocky strata here is nearly northeast by southwest, and the elevation of summits of the ridges about sixteen hundred feet above tide, the valleys being about six hundred feet lower. Beginning oh the Canaan and Lebanon ridge in New York, there is a line of peculiar bowl- * " Geology of New York," Part IV, p. 165, el seq. ■f- "American Journal of Science," vol. cxxvi, 1883, p. 34'7. X Lyell's " Antiquity of Man," pp. 355-362 ; " Bulletin of Museum of Com- parative Zoology at Harvard College," Cambridge, Mass., vol. v. No. Ill, p. 41, with map. 14 210 TEE ICE AGE IN NORTH AMERICA. ders about four hundred feet wiJe, running continuously for nine miles southeast. These bowlders are composed of a chloritic schist, whose only outcrop is at Fry's Hill, on the summit of the Lebanon range. West of the ridge there are none of these bowlders, but east of the knob the train is con- tinuous. Near the knob the size of the bowlders is larger than at a distance from it. Indeed, the size gradually dimin- ishes as the distance increases. The bowlders are distributed equally over the valleys intervening and over the flanks and summits of the ridges crossed. Besides the main continuous train of bowlders there are three others more or less continu- ous for a part of the way, and originating near the same knob. The diameter of the bowlders varies from thirty feet near their origin, to an average of two feet in Stockbridge, nine miles away. In the vicinity are other bowlders from a ledge whose outcrop was from four hundred to eight hun- dred feet lower than the hills upon which they are now resting. Sir Charles Lyell's explanation of these remarkable trains of bowlders was that they were deposited when the region was depressed, so that oceanic currents carried icebergs over the summits of the intervening range, dropping their burdens along on the way. But it is surprising that the burdens of the icebergs should have been deposited so regularly, and still more surprising that icebergs should have raised bowlders and deposited them on surfaces eight hundred feet higher than the ledges from which tliey were torn. This is one of those numerous cases where the glacial hypothesis easily explains all the facts, and where it is difficult to see how any other hypothesis can do so. The only really peculiar thing about these celebrated trains of bowlders is that in their case the peak from which they are derived is isolated so that their origin can be readily traced. The prevailing rocks of this region are of such a nature that large bowlders could not readily be formed from them, whereas over most of the glaciated region the bowl- ders are so abundant and from such a variety of localities GLACIAL EROSION AND TRANSPORTATION. 211 that it is not easy to single out a particnlar train. Careful attention, however, will doubtless resolve the whole mass of till into confluent trains of bowlders and more finely com- minuted material. In New Jersey, according to Professor Cook, the bowl- ders are readily traced all along the morainic margin as be- longing to well-known outcrops of trap, blue limestone, and crystalline rocks to the northwest. Near Drakestown, in Morris county, there is a mass of blue limestone which had been worked for years as a quarry without suspecting that it was but a bowlder. " As exposed it measures thirty-six by thirty feet, and the quarrying has gone twenty feet in depth. Its vertical diameter is unknown. Around it are many gneissic bowlders and other drift materials." * This mass is about one thousand feet above the sea-level, and its native place must have been some miles to the northwest. In Pennsylvania the distance from vphich the glacial ma- terial near the border of the glaciated region has been trans- ported becomes at once more evident, because of its unlike- ness to the local rocks. West of the Kittatinny Mountain, there are no crystalline rocks within the State. Nor are there any to the north nearer than the Adirondacks in New ■York, or the highlands in Canada. Yet granitic, gneissoid, and hornblendic bowlders abound all along the glaciated border, and are an important means of determining the glacial limit. In the valley between Kittatinny and Pocono Mountains, in Monroe county, and on the summit of the Pocono plateau, 2,000 feet above the sea, granitic bowlders from one to three feet in diameter are abundant, though mingled with great piles of local fragments. The granite must have been transported a distance of 250 miles at least, and carried over the summits of the Alleghanies, intervening toward the northwest, and across the valley of the Mohawk in New York. The northern tributaries of the West Branch of the Susquehanna, likewise, bring down into that stream * Kew Jersey Keport for 1 880. 212 TEE ICE AOE IN NORTE AMEBIC A. numerous granitic pebbles, showing that glacial deposits ^ m Lycoming county contain material from the far north which has been carried bodily over the summit of the Alleghanies. On proceeding west, the granitic outcrops from which the ma- terial could come, gradually recede to the north, thus increas- ing the distance between such bowlders and their nearest known source. From Salamanca, N. Y., sonthwestward to Cincinnati the whole country is hterally covered, down to the glacial limit, with granitic, gneissoid, and hornblendic bowlders. Near Salamanca such bowlders abound at eleva- tions not far from 1,900 feet above tide, and TOO feet above the Alleghany Elver. In Beaver county they are numerous on the hills down to within six or seven miles of the Ohio Elver, and several hundred feet above it. The following are some of the more specific facts drawn from my own notes : In Columbiana county, Ohio, a granitic bowlder was found measuring thirteen by eleven feet, and eight feet out of the ground. Others near by were noted, measuring eight and five feet in diameter. In the same vicinity the till contains finely striated fragments of local sandstone, showing direct glacial action on the local rocks. In Holmes county, also, finely polished and striated pebbles of corniferous limestone occur, mingled with fragments of granite in the till. These must have been brought from the other side of the water-shed, in the vicinity of Lake Erie, 100 miles distant. JS'ear Lancaster, in Fairfield county, there is a granitic bowlder measuring eighteen feet by eleven, and six feet out of the ground. In Eoss county, near Adelphi, Chillicothe, and Bainbridge, numerous granitic bowlders were found on the hills from 400 to 600 feet above the valleys, and about 1,200 feet above tide. A hornblendic bowlder five by three by two feet was noted 550 feet above Bainbridge. In Brown, Clermont, and Hamilton counties large granitic bowl- ders abound on the hills down to the very northern edge of the trough of the Ohio. Here, also, are to be found numerous bowlders of jasper conglomerate from the region north of Lake Huron or near the lower end of Lake Superior. The OLAGIAL EROSION AND TRANSPORTATION. 213 variegated pebbles of red jasper and of darker quartzites are a striking feature in the rocks of that northern region. The bowlders of this material found in the vicinity must have been transported nearly 600 miles. Several bowlders of this description were found in Boone county, Ky., a number of miles south of the Ohio Eiver and between 500 and 600 feet above it. Bowlders of this jasper conglomerate are very abundant in Michigan, are not infrequent in northern Ohio, and occur in various localities in southern Indiana — one being observed near Nashville, Brown county, Ind., near the highest land in the State (about 1,100 feet). Granitic and hornblendic bowlders are very abundant, also, as far south as Carbondale, Jackson county. 111., below latitude 38°. The surface rocks are here distinctly striated, and the trans- portation must have been independent of any conceivable cur- rent of water. The distance from this point to the parent ledges to the north can not be less than 600 miles. All over northern Missouri, the whole of Iowa, and east- em Dakota bowlders of large size are of frequent occurrence. In some places they completely cover the ground, especially in the lines of the great moraines. Even west of the Mis- souri, for thirty miles beyond Fort Yates, granitic bowlders are so abundant as to be prominent features in the landscape. Farther north in the same Territory and in Montana they are reported as sometimes so thick that a person can walk for long distances upon them without touching the ground. As has been already remarked, the glacial movement was everywhere at right angles to the glacial boundary. We should expect, therefore, to find that the bowlders along the western border of the glaciated area beyond the Missouri River had been transported from the northeast, and such is undoubtedly the fact. In the recent excursion (in 1888) through northern Nebraska and central Dakota, already re- ferred to,* I had abundant occasion to see evidences of this transportation. On the hills in Nebraska, from 600 to 600 * See above, p. 146. 214 rSE ICE AQE in north AMERICA. feet above the Missouri River, extending in a southwest direc- tion from Yankton, Dakota, to the glacial border, a distance of about forty miles, transported bowlders of considerable size are abundant, and among them are numerous specimens of the so-called Sioux Falls quartzite, whose nearest outcrop is about forty miles to the northeast. All along upon the eastern side of the river in Dakota the glacial accumulations are on an enormous scale, and the transported bowlders without number; and on crossing the river at Fort Yates, about fifty miles south of Bismarck, granitic bowlders, in numerous instances from three to five feet in diameter, are found resting continuously over a belt of the highlands from 500 to 600 feet above the river, and extending about forty miles to the west, where they suddenly cease. There is granite in the Black Hills, 200 miles to the west, and from that source some pebbles have been brought down the Cheyenne River, which rises in that region. But with this exception, there are no granitic bowlders over the area between the Black Hills and this glacial border just mentioned. The source, therefore, of these bowlders on the west side of the Missouri River, extending from Bismarck to the Nebraska line, must lie somewhere to the northeast. Many of them might well enough have come from the vicin- ity of Lake Superior, a distance of 400 or 500 miles, though possibly some of them originated in more limited outcrops of granite in northern Minnesota. In British America, the transportation was outward from the Laurentian axis in every direction. From this axis bowl- ders in immense quantities were carried from 600 to TOO miles westward and left on the flanks of the Rocky Mount- ains, from 2,000 to 3,000 feet above their source. In Dr. G-eorge M. Dawson's report upon the extension of tlie Missouri coteau into the central region of JSTorth Amer- ica, he estimated that nearly ninety-eight per cent of this great accumulation between the Missouri and Saskatchewan Rivers was from the Laurentian axis, some hundreds of miles to the east ; and that, upon the fringe beyond tlie coteau GLACIAL EROSION AND TRANSPORTATION. 215 where there is a mingling of material brought down from the Eocky Mountains, there is still for some distance as much as forty-eight per cent of Laurentian material. Still farther north Dr. Dawson reports a movement of bowlders toward the north in the head-waters of the Yukon River, and in the northern *portion of the continent east of Mackenzie Eiver. For the arctic coast of the continent and the islands of the archipelago off it there is a considerable volume of evidence to show that the main direction of movement of erratics was northward. The most striking facts are those derived from Professor S. Haughton's appendix to McClintock's "Voyage," whei"e the occurrence is described of bowlders and pebbles from North Somerset, at localities 100 and 135 miles northeastward and northwestward from their supposed points of origin. Pro- fessor Haughton also states that the east side of King William's Land is strewn with bowlders of gneiss like that of Montreal Island, to the southward, and points out the general north- ward ice-movement thus indicated, referring the carriage of the bowlders to floating ice of the Glacial period. The copper said to be picked up in large masses by the Eskimo, near Princes Eoyal Island, in Prince of Wales Strait, as well as on Prince of Wales Island,* has likewise, in all prob- ability, been derived from the copper-bearing rocks of the Cop- permine Eiver region to the south, as this metal can scarcely be supposed to occur in place in the region of horizontal lime- stone where it is found. Dr. A. Armstrong, surgeon and naturalist to the Investi- gator, notes the occurrence of granitic and other crystalline rocks not only on the south shore of Baring Land, but also on the hills at some distance from the shore. These, from what is now known of the region, must be supposed to have come from the continental land to the southward. Dr. Bessells, again, remarks on the abundance of bowlders on the shore of Smith Sound in latitude 81° 30', which are manifestly derived from known localities on the Greenland * De Ranee, in " Nature," vol. xi, p. 492. 216 THE ICE AGE IN NORTH AMERICA. coast much farther southward, and adds, "Drawing a con- clusion from such observations, it becomes evident that the main line of the drift, indicating the direction of its motion, runs from south to north."* It may further be mentioned that Dr. K. Bell, of the Cana- dian Geological Survey, has foun^ cTidence of a northward or northeastward movement of glacier-ice in the northern part of Hudson Bay, with distinct indications of eastward glaciation in Hudson Strait, f For the northern part of the great Mac- kenzie Valley we are as yet without any very definite infor- mation, but Sir J. Kichardson notes that Laurentian bowlders are scattered westward over the nearly horizontal limestones of the district. Taken in conjunction with the facts for the more southern portion of the continent, already pretty well known, the ob- seryations here outlined would appear to indicate a general movement of ice outward, in all directions, from the great Laurentian axis or plateau which extends from Labrador round the southern extremity of Hudson Bay to the Arctic Sea ; while a second, smaller, though still very important region of dispersion — the Cordilleran glacier-mass — occupied the Rocky Mountain region on the west, with the northern and southern limits before approximately stated. J Some facts already mentioned * have prepared the way for the discussion of that most puzzling and interesting problem of the upward transportation of earthy material in moving ice. The evidence upon this point is too abundant to be ignored. Professor Charles H. Hitchcock reports finding near the very summit of Mount Washington many small bowlders which must have been elevated a considerable portion of its en- tire height. One of these bowlders, weighing ninety pounds, is now deposited in the museum of Dartmouth College ; and * " Nature," vol. ix. f "Annual Report of the Geological Survey," Canada, 1886, p. 14, D. D; and " Report of Progress," 1882-'84, p. 36, D. D. X See " Glaciation of British Columbia," " Geological Magazine," August, 1888, pp. 348, 349. * See above, pp. 169, 210. GLACIAL EROSION AND TBARSPORTATION. 217 another, of equal weight, may be found in the museum of the Boston Society of Natural History.* Professor Hitchcock writes me that while none of the very large bowlders in New Hampshire have been lifted up very much, it is safe to say that every New England mountain has bowlders on its sum- mit that have been brought there by the ice from at least as great a distance as from its immediate base. Describing a cut in till, forty feet deep, near the village of Queechee, Vt., on the Connecticut Eiver, Professor Hitchcock says it is full of small-sized glaciated stones, cemented together by thick bowlder clay. Every stone is striated. There are great numbers of the Burlington (Yt.) red sandstone, "which must have traveled from over the Green Mountains, over sixty miles, and have been raised over an acclivity of 3,000 feet altitude." f Professor A. S. Packard, Jr., reports X that, at the height of about 4,000 feet above the sea, on Mount Katahdin, Me., " is a large mass of glacial moraine matter which has escaped denudation, and this incloses frequent rounded and polished bowlders of fossils of the same species of Silurian shells, and of the same silicious slates, as are found in situ a few miles northwest, on Lakes "Webster and Telos. . . . The parent beds are but about twelve miles distant," and, according to Mr. Upham, must be 3,000 feet lower. One of the clearest instances of the elevation of bowlders in the ice is the one already alluded to, * in the vicinity of the Delaware Water-Gap, on the summit of Kittatinny Mountain. This summit consists of Medina sandstone, and is about 1,500 feet above tide. Yet Profe^^or Lewis found numerous bowlders of Helderberg limest;(5iie upon it, which he thinks must have come from Godfrey's Eidge, in Cherry Yalley, only a few miles to the north, and 1,200 feet lower. * For particulars concerning these bowlders, see Hitchcock, " New Hamp- shire Geological Report," vol. iii, pp. 204, 207, 272. f " New Hampshire Geological Report," vol. iii, p. 262. X " Memoirs of the Boston Society of Natural History," vol. i, p. 239. * See cut above, p. 168. 218 TEE ICE AGE IN NORTE AMERICA Professor Lesley adds his testimony that in all northeastern Pennsylvania there is no other source of these bowlders but the one line of outcrop mentioned by Professor Lewis ; but, as it extends a hundred miles in a northeasterly direction, he is not sure that the particular locality from which these Hel- derberg bowlders came can be determined. Still, he is cer- tain that every limestone bowlder in northern New Jersey and eastern Pennsylvania has come from some point along , this line of outcrop. Nowhere, however, does the Helder- berg Kmestone rise to an elevation of more than 1,000 feet above tide, while some of these bowlders are 1,500 feet above tide. Remarking upon this, Professor Lesley says : The only problem of prime diflBculty is, how the ice man- aged to lift the fragments from the outcrop in the valley to the crest of the Kittatinny Mountain — a problem which is re- peatedly presented for our solution at various points where the terminal moraine crosses our mountain - ridges, and where blocks from a valley to the north are left perched on a mount- ain-top to the south. And the problem is not confined to the line of the moraine, but repeats itself at points many miles back of the moraine. Twenty years ago I found Catskill red sandstone fragments which had been carried up the north flank of the Towanda Mountain, in Bradford county, and been left on the edge of a swamp upon its flat summit of coal-meas- ure sandstone ; and there is no Catskill country to the north of a higher elevation from which the ice could have brought them with a descending gradient. Professor James Hall informs me that fragments from the Mohawk Valley have been carried up over the Hel- derberg Mountain to the south of it, precisely as the Eon- dout-Walkill bowlders have been carried over the Kittatinny Mountain. So, Judging by the southeast striae, the gneiss and granite bowlders of western Pennsylvania must have been carried up from the level of Lake Brie (570 feet above tide) to elevations of 1,500 feet, along the line of the terminal moraine, 1,700 feet above tide at Lake Chautauqua, and even 3, 150 feet above tide GLACIAL EROSION AND TRANSPORTATION. 219 in Little Valley, Cattaraugus county, N. Y. ; unless we sup- pose that all the Canadian bowlders were borne upon the sur- face of the ice, which is clearly impossible. Bowlders of Alpine glaciers seem always to descend to their final resting-place, but we haye innumerable proofs that the American ice-sheet managed, in some way, to carry howlders from valleys up to mountain-tops, although the amount of elevation in many cases, if not in all cases, may be much less than we are inclined, on a first inspection of the facts, to take for granted. In the case of the Helderherg limestone bowlders men- tioned, found by Mr. Lewis on the crest of the Kittatinny Mountain, it is not necessary to suppose that it came from Godfrey's Eidge, only three miles distant (north) in the valley below, 1,000 feet beneath its present position. Indeed, the direction of the scratches on the mountain-side make such a supposition incredible. It is plain that it must have traveled down the valley of the Delaware, and may have come from the continuation of the range in the State of New York. The elevation of the surface gradually increases going east. The rise in the bed of the river for the first thirty-five miles, from the Delaware Water-Gap to Port Jervis, is about 300 feet. The rise from Port Jervis to the Eondout-Walkill divide, twenty miles, is 80 feet more. There the crest of the Helder- herg Eidge must be nearly 1,000 feet above tide. The crest of the Kittatinny Mountain where the block lies is about 1,500 feet above tide. Therefore, if the block came these sixty-five miles, it has been carried up only 500 feet above its original situation. Still, it remains a problem by what sort of internal move- ment a stone held in the ice can ascend, however gentle may be the gradient upward. That internal movements take place in all glaciers, is made visible to the spectator by their spoon- shaped stratification, and by the different rates at which their upper, lower, middle, and lateral parts move along, as well as by the fact that they press forward over rock-barriers. But so general a statement has no scientific value when evoked to explain the actual translation of a bowlder up a mountain- slope. In fact, our knowledge of how such an operation was 220 THE ICE AGE m NOETE AMERICA. performed is as vague as possible, and demands the attention of hydraulic engineers.* Elsewhere f I have briefly considered the way in which this upward movement of bowlders imbedded in the glacial current might be produced. The subject is of so much in- terest and importance that it will be well to recur to it here. As a result of the differential motion in a glacier in which each higher stratum of ice is moving slightly faster than that which is immediately below it, it follows that the inclosed bowlders are subjected to a differential strain, in which the upper portions are impelled forward with greater rapidity than the lower portions. The result of this differential strain upon the upper and lower portions of the bowlder, combined with the friction and viscosity of the ice, must produce a movement slightly upward as well as forward. The bowlder, being inelastic, or at any rate less elastic than the ice in which it is imbedded, must move as one mass, while each particle of ice moves in independence of all the others. IsTow, the portion of ice lying immediately in front of a bowlder, being protected from the differential pressure of the ice behind by the interference of the inelastic foreign object, offers more resistance in that direction than is pre- sented along a line leading diagonally upward across the lines of greater movement above. In other words, the frictional pressure of the moving ice upon the upper portion of a bowl- der is greater than that on its lower portions, and the greatest of all resistance is immediately in front. The line of least resistance is consequently always in a direction slightly up- ward. If, for example, the movement of the ice at the upper surface of a bowlder be represented by TOO, and that at the lower portion of the bowlder by 99, then one one-hundredth of the force might be supposed to be expended in producing a diagonal upward movement. Thus we can conceive that fragments of rock were picked up from beneath the glacier, * See " Second Geological Survey of Pennsylvania, Z," p. xxiii el seq. ] "Glacial Boundary in Ohio, Indiana, and Kentucky," p. 31. OLAOIAL EROSION AND TBANSPORTATIOK 221 and that, after moving a sufficient distance, they appeared upon its surface ; and we can easily believe that many bowl- ders have thus been repeatedly transferred from beneath the ice to its surface, and thence projected by the more rapid superficial motion to the front to be reincorporated into the lower strata of the mass, and re-elevated to the surface again. Upon this subject Professor Lesley offers the following ingenious suggestions : When two equal solid bodies descending opposite slopes meet, they arrest and support each other. Imagine myriads of cannon-balls rolled from both sides to meet in the middle of a symmetrical valley. Those arriving first would remain ever afterward the lowest stratum ; those which followed would arrange themselves in higher and higher layers until the valley was full or the supply exhausted. No shifting of places would take place after each had found its lowest place. But suppose opposite descending quantities of pitch, or moist clay, through which cannon-balls were scattered, to meet along the middle line of a valley ; the two advancing fronts would mash against each other, and thicken upward, the in- cluded cannon-balls rising vertically in the thickening mass, the thickening being in proportion to the height and weight and rigidity of the masses of clay pressing from the side slopes upon the middle line which had come to rest. By substituting plastic ice for moist clay, and rock-bowlders for cannon-balls, we get an idea of how the American ice-sheet may have carried up (diagonally) the masses which it tore from low-lying outcrops to higher levels, and even over mount- ain-crests. . . . A terminal moraine is often described as if it were merely the tumbled-ofE accumulation of the medial and lateral mo- raines which cover the surface and sides of a glacier. Bub the fact is, that a glacier is like a plum-pudding— full of scattered sand and stones from top to bottom and from side to side, all of which are delivered at its front end down the valley. The surface exhibition is made much stronger than it would other- wise be by the perpetual melting of the upper surface and sides 222 TEE ICE AGE I.Y NORTH AMERICA. of tlie alacier. This brings the plums in the pudding to the surface, and mixes them with tiie medial moraine blocks which <1 l! , f>- '.;' If.'! ; ' -.(11' _ ' r FJ'frXt ■ Fig. 61.— Glaciated pebble from southern Indiana. Natural size. Tliis was one of the graving tools of the glacier. The strife are parallel with the longer axis. have ridden down from the forks of the valley. A constant concentration of the debris of the whole glacier thus goes on at its surface in spite of the occasional loss of some of the stuff by dropping into crevasses, the blocks thus temporarily lost rejoining tlieir fellows at the surface, if the glacier be long GLACIAL EROSION AND TRANSPORTATION. 223 enough, lower clown the valley, or iysuing midway of its front end in the mass of the terminal moraine.* It is urged, Loth against the foregoing statement of facts and their theoretical explanation, that observations in Green- FiG. 62.— Kevtrye side of the same pebble. land are irreconcilable with them. For it seems to be ad- mitted that the surface of the great ice-cap of Greenland is * "Second Geological Sun'ey of Pennsylyania, Z," p. xxv ei seq. 224 THE ICE AGE IN NORTH AMERICA. free from bowlders, except in the lines of movement extend- ing from the nunataks. A suflacient reply to this objection is to say that a country so long subject to glacial movements as Greenland has been must already have had the looser frag- ments of rock, which can be incorporated into the ice, gath- ered up and removed to the front, so that now the floor of the whole continent is so smooth and bare of fragments that there is nothing left for the glacier to get hold of. But dur- ing the movement of the glacier over the northern part of the United States, everything favored the mode of trans- portation and elevation of bowlders as described above. As already shown,* the transportation of large bowlders, though impressive in view of the enormous masses moved, probably represents but an insignificant part of the work of erosion and transportation which took place during the Gla- cial period. Some would even regard the subglacial streams pouring out in front of every glacier, and surcharged with fine sediment, as the most important instruments of glacial transportation. How much of the glacial grist has been car- ried away by this process it is impossible to determine with accuracy. It will, however, be of some interest to learn the best attainable results. In 1864 Dolfus-Ausset measured the fine sediment carried out by the stream from below the Unter-Aar Glacier, finding 133 grammes in a cubic metre of water. This corresponds to a yearly rubbing off of about 0'6 millimetre of rock from under all parts of the glacier's basins, or an erosion of one metre in 1,666 years — about two and a half times as much as water could do in the same period. It is not determined how much of the sediment came from sand washed under the ice by side streams, but it shows that the glacier does a considerable amount of work (" Materiaux pour I'fitude des Glaciers," 1864, vol. i, p. 376). t * See above, p. 122. •j- Professor W. M. Davis, in " Proceedings of the Boston Society of Natural History," vol. xxii, 1882, p. 20. GLACIAL EROSION AND TRANSPORTATION. 225 M. B. Oollomb has made some iuteresting calculations . . . based upon the obseryations of MM. Dolfus and Desor on the Aar Glacier in 1844 and 1854. These glacialists found that the amount of water discharged from this glacier between the 30th of July and 4th of August averaged 1,378,738 cubic metres daily — the minimum being 780,000 cubic metres, and the maximum 3,100,000 cubic metres. The area occupied by the glacier is estimated at fifty-two square kilometres. Now, supposing that the old glacier of the Ehone (the area of which M. Collomb estimated at 15,000 square kilometres, but which is actually under the truth) discharged its water at the same rate, it must hare yielded a daily supply of 605,000,000 cubic metres. But if it be true, as all the facts would lead us to believe, that in the summers of the Glacial period more heat was received directly from the sun, then the daily discharge from such a glacier must have been greatly in excess of that amount. . . . MM. Dolfus and Desor found that a litre of water from the Aar Glacier contained 0'143 gramme of fine mud ; so that, according to Collomb's estimate of the area and daily dis- charge of the ancient Rhdne Glacier, the water escaping from the latter must in summer time have transported 86,000,000 kilogrammes, or about 8,500 tons (English) per diem — an esti- mate which, considering the circumstances already referred to, is probably much under the actual truth. According to Helland, the quantity of mud in the rivers that issue from the glaciers of Greenland is very variable, as may be seen from the table given by him, which is as follows : Grammes of mud in 1 cubic metre of water. River of the glacier of Jakobshavn July 9, 1875 .... 104 Alangordlek "10, " .... 2,374 Jlardlek " 17, " .... 723 " Tuaparsuit. . . . ATjgust 6, " 678 " " Umiatorflk " 20, " .... 75 " Assakak " 21, " .... 208 " " Rangerdlugssuak " 11, " .... 278 Similar observations by the same geologist on the water issuing from the snow and ice-field of Justedalsbraeen likewise 15 226 TEE ICE AOE IN NORTH AMERICA. showed that the quantity of mud varied in the different streams^ and even in the same river. The result of ten different obser- vations in the months of June and July gave a mean of 147-9 grammes of mud in 1 cubic metre of water. (See " Quarterly Journal of the Geological Society," 1877, p. 157.) * Mr. J. E. Marr gives the following facts concerning the extent to which erosion is proceeding beneath some of the Greenland glaciers : The erosive power of an ice-sheet is well seen by a glance at the observations made upon the rivers which flow into the fiords of JSTagsugtok and Isortok, and which have their origin at the ends of the tongues of ice which occupy the val- leys continuous with these fiords. The river from the first contained only 200 to 225 grammes of mud per cubic metre of water in the month of July ; while the second, in the month of June, inclosed 9,129 to 9,744 grammes. This is compared with the amount carried by the Aar where it emergies from the glacier ; it there contains only 142 grammes. The great difference presented by the rivers which fall into the two fiords is attributed to the fact that the ice moves with much greater speed toward the fiord of Isortok than toward that of Nagsug- tok. It is calculated that the quantity of fine mud carried into the former of these fiords amounts to 4,062,000,000 kilo- grammes per day. This mud is deposited in the interior of the fiord, which is filled up to such an extent in its upper por- tion that even flat-boats can not pass up it.f The amount of material carried to the sea by the subgla- cial streams during the continuance of the Glacial period in North America could not be estimated, even though we knew the rate of transportation, unless we had more definite ideas than we now have of the length of time during which glacial conditions prevailed. But if, as is probably the case, the deposits of loess in the valley of the Mississippi are of * Geikie'a " Prehistoric Europe," pp. 231, 232. f See "Geological Magazine," April, ISS^, summarized in " American Jour- nal of Science," toI. cxxxiv, ISSt, p. 313. GLACIAL EROSION AND TRAWSPOBTATIOK <2,'2n glacial origin, these may at some time give us a partial clew to the length of the period. Another method of estimating the amount of glacial ero- sion is by calculating the extent of the deposits over the glaciated region. Professor Newberry has estimated that the area south and west of the Canadian highlands, covered with glacial debris, is not much less than 1,000,000 square miles, and that the depth of the deposit over this broad marginal area can not be less, on the average, than thirty feet, and is probably twice that amount.* Professor Claypole found, upon comparing the estimates independently made in different counties of Ohio, Indiana, and Illinois, that the average depth for these three States was sixty-two feet, and Fig. 63.— Ideal section showing how the till overlies the stratified rocks. for Ohio alone fifty-six feet.f Kecent extended experi- ments in boring for gas in western Ohio and eastern Indi- ana tend greatly to increase the estimate of Professor Clay- pole. There are whole counties in southwestern Ohio so deeply covered with glacial debris, that few of the citizens have ever seen the underlying rock. In other counties, where the rock occasionally crops out at the surface, the ex- tensive spaces between are found to be valleys of immense depth, filled with the unstratified material of the ground moraine. In Professor Orton's recent report on the subject, giving the depth in fifty-three of the counties of Ohio as determined by the borings in 122 wells, the average is found to be upward of 90 feet (93 feet -|-). In some of the wells the depth was truly phenomenal, as in St. Paris, Champaign county, where, in one well, rock was not reached short of * " School of Mines Quarterly," January, 1885, p. 1. f " Glacial Erosion," by William M. Davis, " Proceedings of the Boston Society of Natural History," vol. xxii, pp. 19-88. 228 THE ICE AGE IN NORTH AMERICA. 370 feet, and in another, 530 feet of till was penetrated, and the well abandoned before rock was reached. In Dayton, Montgomery county, the glacial deposit was found to be 247 feet ; in Cridersville, Auglaize county, 300 feet ; in New- ark, Licking county, 235 feet ; in Lebanon, Warren connty, 256 feet ; in Osborn, Greene county, 207 feet ; in Hamilton, Butler county, 214 feet. These are all, perhaps, in pre- glacial valleys. But the average in various counties is cer- tainly significant. In Auglaize county, six borings give an average of 141 feet; in Butler county, four borings, 116 feet. With reference to the correctness of this representation, it should be remarked that borings prosecuted in this man- ner are more likely to give an underestimate than an over- estimate of the real facts ; for, as is well known, it is much easier and less expensive to drill through the sedimentary rocks than through deep deposits of till and looser drift ; so that the aim of the prospectors is to begin their wells at points where the rock will be reached at as small a depth as possible. But so completely have the pre-glacial lines of erosion been obliterated in many places in Ohio, that it is impossible to calculate the proximity of the rock to the sur- face. Where the deepest drift was penetrated (530 feet in Champaign county), special effort was made to locate the boring where the superficial deposits were shallow ; but, as the result proved, the surface indications were deceptive and a serious mistake was made, involving the contractor in great loss. It should, however, be observed that the deep well at St. Paris lies in the line of the so-called terminaljj.moraine of the second Glacial epoch, as indicated in the report of President Chainberlin, where the depth of glacial deposits is supposed by him to be exceptionally great. Mr. Upham estimates the mantle of drift that conceals the rocks in central Minnesota t#be between 100 and 200 feet deep. In the Red River region, to the north, and over a wide belt stretching many hundred miles along the flanks of the Rocky Mountains, in the Dominion of Canada, the GLACIAL EROSION AND TRANSPORTATION. 229 depth is equally great. In the upper valley of the South Saskatchewan, at an elevation of about 4,000 feet above the sea, and from 600 to 700 miles west of the Laurentian axis, from which much of the glaciated material came, Mr. McCon- nell reports sections of till 125 feet deep.* Professor Stone thinks the average thickness of the drift in Maine is between thirty and fifty feet. Mr. Upham's early calculations for New Hampshire were much more mod- erate, namely, ten feet-f But he now informs me that he would make a much higher estimate. Besides, in so momit^ ainous a district, we should expect a thinner deposit to re- main on the surface. The more rapid streams would trans- port a larger portion of the material to the sea than from the gentle slopes. Furthermore, the rocks of New Hampshire are better calculated to resist erosion than in some other por- tions of the country. Much of the soil of New Hampshire has been transported to the States farther south. No re- liable estimate lias been made of the average depth of the glacial accumulations over Massachusetts, Connecticut, and Rhode Island. There can be little doubt, however, that it is much greater than Mr. Upham makes for New Hamp- shire. Professor Shaler ^ sets down the total amount of drift in New England and its neighboring terminal moraines at 750 cubic miles, or more than the mass of the White Mountains. If evenly distributed, this would make a layer of about sixty- five feet. Professor Lesley says the depth of the glacial drift over the northeastern counties of Pennsylvania is not less than fifty feet. This is on the summit of the Appalachian plateau, while the old valleys, filled with glacial debris, are some of them of great depth. One on Mehoopany Creek, in Wyo- ming county, is filled with drift to a depth of more than 235 * " Report of the Cypress Hills," 1886. f " New Hampshire Geological Report," toI. iii, p. 293. % "Illustrations of the Earth's Surface: Glaciers," p. 58. 15 230 TEE ICE AGE IN NORTH AMERICA. feet. In this vicinity, Professor I. C. Wiiite * reports wells fifty feet deep which barely reach through the till, and this on elevations 1,335 feet above tide. Great as these amounts may seem, the estimation of the erosion in the Scandinavian Peninsula by Professor Helland is still larger.f Helland, after conference with several geologists familiar with the region, estimates that the aver- age depth of the drift over north Germany and northwest Russia is 150 German feet. This would indicate that the erosion from the Scandinavian Peninsula had been as nmch as 250 feet, since the material has nearly all been derived from Scandinavia, and the area of the source of supply in Scandinavia is only two fifths of that over which it was dis- tributed. This estimate of Helland for Scandinavia is not, how- ever, greater than that of Professor Levris for northern Pennsylvania. Here, on the Kittatinny Mountain, near the Delaware Water-Gap, this observer seems to have had a rare opportunity for directly measuring the eroding power of the ice at that point. :]: The summit of the mountain is crowned by compact strata of Medina sandstone, and trends northeast by southwest. The glacier surmounted the ridge on both sides of the Water-Gap, and extended twelve or fifteen miles farther south, while to the southwest the sum- mit of the mountain was outside t>i the line of ice-movement, which just here is bordered by cliffs of the crowning Medina sandstone seventy feet high, as if the ice, in moving past them, had worn down the strata underneath to that amount. Professor Lesley, at the Minneapolis meeting of the Ameri- can Association for the Advancement of Science, in 1884, took this as a measure of glacial erosion to illustrate how * See " Second Geological Survey of Pennsylvania on Wyoming, Lackawanna, Luzerne, Columbia, Montour, and Northumberland Counties G'," p. xiii. ■)■ " Ueber die Glacialeu Bildungen der nordeuropaischen Ebene, Deutsch. Geol. Gesell.," Zft. xxxi, 1879, p. OV. X " Second Geological Survey of Pennsylvania Z," pp. 70, 90. See also above, p. 168. GLACIAL EROSION AND TRANSPORTATION. 231 small it was. But Professor Newberry well replied that, if there was that amount of erosion so near the margin, what must it not have been farther back, where the stream of ice had acted for an indefinitely longer time. Probably, how- ever, Newberry is extravagant when he estimates that farther north the ice was ten times as thick, and continued to act ten times as long, making its erosive power one hundred times as great as that near the Water-Gap.* The foregoing evidence of glacial erosion drawn from the extent of marginal glacial deposits is complicated by our ignorance of the extent to which disintegration of the rocks had proceeded before the Glacial period. Professor Whit- ney f and Mr. Pumpelly X have specially pressed this point, as have Professor Sterry Hunt and the late Mr. L. S. Bur- bank,* to whom more credit is due than he has generally received for his early and sagacious suggestions upon the subject. The contrast between the glaciated and the un- FiQ. 64. — Ideal section showing result of disintegration in an unglaciated region. tChambeiiin.) glaciated region, in the extent to which the surface rocks are disintegrated by subaerial agencies, is very striking. South of the glaciated region granitic masses and strata of gneiss * " School of Mines Quarterly,'' p. 12. f " Climatic Changes," p. 1 et -leg. J " The Relation of Secular Rock-Disintegration to Loess, Glacial Drift, and Bock Basins," in the " American Journal of Science," vol. cxvli, 18'79, p. 138 et seq. ** " On the Formation of Bowlders and the Origin of Drift Material," in the " Proceedings of the Boston Society of Natural History," vol. xvi, Novem- ber 19, ISYS. 232 THE ICE AGE IN NORTH AMERICA. are often completely disintegrated to a great depth, some- times amounting to scores of feet. What seem like beds of gravel (and which can be handled with a shovel) often prove to be horizontal strata of gneiss from which the cementing material has been removed by the slow action of percolating acids brought down by the rains. North of the glacial boundary it is very rare to find any such extensive evidence of disintegrating agencies. Since the ice passed over this region there has not been sufficient time for subaerial agen- cies to produce any marked disintegrating effect. Now, it is with much plausibility contended that the action of the ice has been limited chiefly to the i/ransportor tion of this disintegrated material, and that it has had little effect as an eroding agency. The strong point in this repre- sentation is, that there is little more loose soil, over the margin of the glaciated region than would result from the simple transportation of the disintegrated material from the northern and central portions of the glaciated region to the marginal area. It certainly is clear, both from the necessities of the case and from actual observation, that the area of greatest erosion is nearest the center from which the ice radiated, and that, as the amount of deposition increased toward the margin, the erosion diminished. The advocates of the great erosive power of glacial ice appeal, also, to the general appearance of the glaciated sur- faces wherever exposed. Tlie islands near Sandusky, in the western part of Lake Erie, for example, present some of the most marked indications of glacial erosion anywhere to be found, and the facts there are justly appealed to by Professor Newberry in support of the theory that the ice was a promi- nent agent in the formation of the basins of the Great Lakes. As this is so important a region for glacial study, I will give somewhat in detail the result of my own recent obser- vations. There are twelve or fifteen islands near the west- ern end of Lake Erie, of which Kelly's, North Bass, Middle Bass, South Bass, and Pelee are the principal, each havinw GLACIAL EROSION' AND TRANHPORTATION. 233 an area of several square miles, and none of them rising 100 feet above the surface of the lake. Thej all consist of the hard limestones of the Niagara series. In every instance, as one approaches them from the eastern side, his attention is attracted by the remarkable depth and continuity of the glacial grooves running nearly east and west upon them, and whicli rise out of the water, and continue to tlie summit of the islands, or until they are covered by the ground moraine ' which has not been washed away by the waves. In some Fig. 65.— Glacial grooves on east sidu of South Bass Island, Lake Erie, running west 10" south. instances, these grooves are two or three feet deep, and ex- tend many rods in plain sight. Nor are they in all cases straight, but sometimes are extremely tortuous, winding along in their course like the channel of a sluggish stream. It is evident, in some cases, that the main features of these deep- est grooves liave been determined by preglacial or subglacial water-action, and that the ice, or the ground moraine under Missing Page Missing Page 236 TRE ICE AGE IN NORTH AMERICA. lee side of a prominence in wliich a north-and-south glaci- ated surface could be protected from the force of the eastern current, we find that it was protected, and the glaciated sur- face is as fresh on the removal of the soil as it is anywhere. In places the beveled edges of earlier grooves are perfectly distinct where the second cross-movement has obliterated a part of a groove, and left still untouched the portion of the earlier groove which was protected by a ridge. Within a half mile of each other there are grooves several inches in depth running at right angles to each other. For instance, upon the west side of Gibraltar (which is a small, rocky island at the mouth of Put-in Bay, close to South Bass Island), there are deep grooves, running north by south, made by the sec- ond movement; but they were perfectly protected by the rock, which received the brunt of the third movement as it came from the east. One half mile west, and a little south of this point on the main island, where there is a shallow valley across the island, the eastand-west grooves are equally striking. This was doubtless the movement whose record is left in the moraine of the Maumee Yalley, as already described by Gilbert.* 4. There were also indications of still a fourth movement, which set in when the ice had receded so far that the obstruc- tion presented by the elevation of the water-shed to the south would no longer compel a westerly movement. But, when the ice-front was between these islands and the south shore, the movement would again be, according to theory, to the south, at right angles to the first and third movements. This last movement, however, was probably feeble and not of long continuance. Still, there are some signs of it in the shape of shallow striae across the second set of east-and-west fur- rows. While all this is witness to the efficiency of ice as an eroding agency, it conveys the impression that the erosion accomplished by each successive movement was concentrated in special channels, and was nowhere excessive. * See p. 179. OLACIAL EROSION AND TRANSPORTATION. 23Y \ \ ' ... ■■;,'' BH^^HBj&dB^^liHM^^^B ft tfii.?^^■^^"^-'^■^•S'-;■■ ::o.... / • I \ *^^' '' ^ill:!^'!^ ^^Sfe^ ■ .. iff"*^3M&^ :;"^ ■■J./, ''■'^■■■" - ■ '?fe....-.-: ^g g^^^fe ^M^^^JBfci Fig. 07.— Section of east and west glacial furrows, on Kelly's Island. Till rests imme- diately on the rock, with washed pebbles at the surface. The following very important extracts concerning glacial erosion, from the recent report of Mr. I. C. Russell, need no introduction and no comment : That the rock-basins in the high Sierra were excavated by glaciers the writer finds no reason whatever to question. They frequently occur at the lower limit of a steep slope, which is polished and grooved, and bears every indication of having been abraded by glacial action. In such cases the slope and the direction of the furrows show that ice once descended into the basin. On examining the opposite portion of the rim of the depression, glacial markings of the same character will be found. The proof is thus positive that the ice descended into the depressions now filled with water, and emerged from them again to continue its course. As there is no other agent known capable of eroding hollows in solid rock having the character GLACIAL EROSION AND TRAMP OUT ATION. 239 of the basins observed iu the high Sierra, it seems evident that the theory of the formation of rock-basins proposed by A. C. Ramsay, from evidence obtained in Scotland and Switzerland, is substantially correct, and furnishes the true explanation of the origin of the examples before us.* The manner in which the power of moving ice is directed so as to erode depressions may be open to discussion, but the conclusion that rock-basins are a result of glacial action is now too strongly supported by facts to be questioned. . . . On examining the numerous lakes more critically, one finds that many of them occupy depressions in morainal dehris, or are confined by terminal moraines. In numerous instances, however, as in Bloody and Gibbs Cations, at the head of Eush Creek, and all about Mount Lyell and Mount Bitter, the fact that the lakes occupy depressions in solid rock is beyond all question. One may walk entirely around many of them with- out stepping off rock in place. . . . As some writers — especially those who are given to solving the mysteries of Nature from their closets — have thought that lakes filling true rock-basins are a rarity, and have even dou'bt- ed whether they exist at all, we shall be interested in examin- ing this result of glacial action, while we wait for our mule- train to, join us. The stream from above cascades over hun- dreds of feet of rock before reaching the lake ; on either hand the overshadowing cliffs tower upward for a thousand feet ; and we can walk along the lower border of the lake and find solid rock all the way across the canon. There is no doubt, therefore, that the lake occupies a basin in solid rock. The ledge confining the waters rises in places almost perpendicu- larly to the height of over a hundred feet above the lake sur- face, and indicates by its rounded contour and polished and striated sides that the ice was once forced up from the basin, now filled with water, and flowed over the ledges and down the gorge. The sounding-line tells us that the bottom of the basin is fifty-one feet below the lake surface. We thus have a rock-basin of considerable depth, in the path of a glacier, * " On the Glacial Origin of Certain Lakes," etc., " Quarterly Journal of the Geological Society of London,'' vol. xviii, 1862, pp. 185-204 240 THE ICE AGE TN NORTH AMERICA. the unmistakable markings of wliich descend into it on the u])per side and emerge again at its lower margin.* There is so much t!;encral interest in the question as to the formation of the Yosemite A^alley that we append the remarks of the same high authority concerning it : 'Tlie Quaternary History of Mono Valley, California," pp. 281, 368, S69. GLACIAL EROSION AND TRANSPORTATION. 241 It is the opinion of the writer that the excavation of many of the valleys of the Sierra Nevada began long previous to the Quaternary, and are in fact relics of a drainage system which antedates the existence of the Sierra as a prominent mountain- range. Those who seek to account for the formation of the Yo- semite and other similar valleys on the western slope of the Sierra by glacial erosion shoxild be required to point out the moraines deposited by the ice-streams that are supposed to have done the work. The glaciers of this region were so recent that all the coarse debris resulting from their action yet remains in the position in which it was left when the ice Oaelted. If the magnificent valleys referred to are the result of glacial erosion, it is evident that moraines of great magni- tude should be found about their lower extremities. Observa- tion has shown that debris piles of the magnitude and charac- ter required by this hypothesis are notably absent. It is perhaps not disgressing too far to state that the writer, while visiting the Yosemite, could not avoid adopting an hypothesis advanced some years since by J. D. Whitney, to the effect that the main characteristics of the valley are due to dislocation ; or, in other words, that the orographic block be- neath the valley has subsided. No facts were observed, how- ever, conflicting with the conclusion of Clarence King that the valley was occupied at least in part by glacial ice. The majestic domes of the Yosemite region have not been rounded by glacial action as some writers have supposed, but have been produced by the weathering of granite, in which a concentric structure on a grand scale was produced when the rocks were in a plastic condition.* Anotlier illustration of the anomalous erosive power of glacial ice is seen in the so-called cirques so abundant in gla- ciated regions containing mountains. Here, again, we are under great obligations to Mr. Bussell for his careful report upon these features of the high Sierra. Little is to be added to bis discussion of the subject. * " The Quaternary History of Mono Valley, California," pp. 350, 361. 16 242 THE ICE AGE IX N OUT 11 AMERICA. Oue of the most striking features in the sculpturing of the high Sierra is furnished by the grand amphitlieatres or cirque.^, occurring about the more elevated peaks and crests. These are deep semicircular excavations, bounded on all sides, except that through which the drainage escapes, by bold cliffs or by per- pendicular walls from a few hundred to more than a thousand Fig. 70. — Glacial furrows on Gibraltar Island, one half mile from i>roc'(.-ding, but runiiinu; nearly north and south. feet in height. The bottoms of these excavations are often depressed below that portion of the rim througii whicii the drainage escapes, and form rock-basins ; at other times the basins arc partially inclosed by (/elin'fi. and in some instances they have well-formed terminal moraines across their outlets. In these hollows there are trans])arent lakes of azure blue, whicli rcilect the grandeur of tiic sheltering walls with wonder- ful distinctness from their unruffled surfaces. A horizontal cross-section of a cirque is semicircular or horseshoe-shaped, and in certain portions of the range these arc so numerous that they give a scalloped contour to the faces of the cliffs. The GLACIAL EROSION AND TRANSPORTATION. 243 interiors of some of the amphitheatres are terraced in the same manner as are the bottoms of the eaflons leading from them, a feature which has been obseryed in the cirques of the Eocky Mountains as well. Nearly all the various branches of the ancient glaciers of the high Sierra headed in deep recesses of the character above described. In places the cirques occur on either side of a fragment of table-land, and have been eroded back until only a knife-edge of rock, so narrow and broken that the boldest mountain-climber would hesitate to traverse it, is all that divides one profound depression from another. Examples of this nature are common about Mounts Lyell and Ritter, and. find a number of typical illustrations in the cliffs of the Kuna and Koip crests. At the head of Eush Creek are a number of separate cirques, each holding a gem- like lakelet, in which the various branches of the stream drain- ing the basin have their source. Silver Creek heads in a magnificent amphitheatre formed by the union of several cirques, which during the height of the Glacial epoch was completely filled with neve. On the south side of Kuna and Koip peaks are two vast amphitheatres which rank among the finest in the Mono region. . . . No topographic delineation or word description can convey the impressive grandeur of some of these vast, shrine-like recesses that have been sculptured during the lapse of centuries from the rugged cliffs of the high Sierra. In general the cirques open northward, but many excep- tions to this rule can be found, especially about the head-waters of Eush and Silver Creeks. It is in the cirques about the higher peaks that living glaciers are still found, and those not harboring perennial ice are deeply filled with snow during a large portion of the year. The slow melting of the snow and ice in these reservoirs feeds the rills which join one with another to form the creeks flow- ing into Lake Mono. The balance between the climatic con- ditions favorable to the existence of glaciers and those which insure their disappearance is here nicely adjusted, and, should the equilibrium shift to the side of greater congelation, these ancient cirques would be the first points to exhibit the changed conditions. They were the fountains which gave birth to the 244 TEE ICE AGE IN NORTH AMERICA. ancient glaciers, and were also the last strongholds to be aban- doned when the reign of ice approached an end^ Such amphitheatres are known in all mountain-regions where glaciers have existed. It has been the good fortune of the writer to examine them on some of the higher peaks of Colorado and New Mexico, about the crests of the Wahsatch and East Humboldt Mountains, as well as in Switzerland and New Zealand. Their origin is somewhat problematic, and has occasioned much discussion, as is well known to all who have followed the growth of glacial literature. In an article ou the formation of cirques, by T. G. Bon- ney,* an attempt was made to prove that these peculiar feat- ures of mountain sculpture are the result of stream-erosion, and owe few if any of their characteristics to ice-action. The studies of B. Gastaldi,f on the efEects of glacial ero- sion in Alpine valleys, led him to reject Bonaey's hypothesis, and to conclude that they are a result of ice-erosion. The most extended as well as the most instructive essay concerning their formation that has come under the writer's notice is from the pen of Amund Helland, entitled '"'On the Ice-Fiords of North Greenland, and on the Formation of Fiords, Lakes, and Cirques in Norway and Greenland." J In this essay a clear and concise description of the cirques of Norway, Switzerland, and other regions is given, together with a brief summary of the various hypotheses that have been advanced in explanation of their origin. Strong evi- dence is also presented to show that they are a result of glacial action. In Helland's essay are included the views of Lorange, of the Norwegian Royal Engineers, who arrived at the con- clusion that they are formed principally by the effects of great changes of temperature in the vicinity of glaciers. We quote Lorange's observations and conclusions as stated by Hel- land : "Under the glaciers in cirques, where a space intervened between the bed of the cirque and the ice, he saw a great many * "Quarterly Journal of the Geological Society," London, vol. xxviii, 1872, pp. 312-324. •j- Ibid., vol. xxix, 1872, pp. 396-401. X Ibid., vol. xxxiii, 1873, pp. 142-176. GLACIAL EROSION AND TRANSPORTATION. 24-5 stones, some of which, sticking fast in the glacier, were qnite lifted np from the bed of the cirque, while others were touch- ing or resting on it ; he thinks it probable that, as the tem- perature aronnd the glacier constantly varies about the freez- Fia. 71. — Glacial groove on Middle Basg Island, running nearly east and west, with a north-and-soutti groove crossing it. Notice the beveled edge of the east-and-west groove, wliicli is here descending to the west. ing-point, the incessant freezing and thawing of the water in the cracks in the rock may si^lit it, and the glacier may do the work of transportation for the fragments thus broken loose. On examining the interior of an empty cirque, we observe that a bursting, not a scooping out, of the rocks has taken place." The writings of Penck, Lowl, and other European geolo- 246 THE ICE AGE IN NORTH AMERICA. gists might be cited here, but it is not my intention to review the entire literature of the subject. Sufficient observations have been recorded to show not only that cirques are of nearly world-wide distribution, but that they are confined to glaciated regions, and are not found in mountains where undisputed records of glacial action are absent. This in itself is good evidence that they have resulted from glacial erosion. The same conclusion is indicated by the fact that as a rule they open northward — that is, they occupy positions where glaciers first appear when a lowering of tem- perature renders their existence possible, and where they lin- ger longest when the climate ameliorates. It thus seems un- necessary to discuss in the present paper the various hypothe- ses which refer their origin to water erosion, crater elevation, etc. The descriptions presented in the essays we have cited, together with the observations of the writer, show that the cirques of the high Sierra are typical of their class, and present all the features to be seen in other similar regions. It is thus rendered evident that, if we can arrive at an acceptable expla- nation of their origin, it should explain the like phenomena in other regions as well. The writer has no mature theory to offer, but hopes to contribute something toward the desired end. It is usually diflBcult to draw a definite line between a glacial cirque and the cafion leading from it. One is a contin- uation of the other. It is evident, also, that the walls inclos- ing a cirque have many features in common with the scarps so frequent in glaciated caflons. When the cirques themselves are terraced, this analogy is rendered still more complete. The writer's studies in the high Sierra and elsewhere have led to the conclusion that such scarps and cirques result mainly but not wholly from glacial action. The initi ition of the process, at least in the high Sierra, as in the case of many glacial caflons, must have been by subaerial erosion. Lorange's observations show that when a neve fills a cirque it is capable of removing blocks of rock from the inclos- ing walls. The fact that these walls are rough and angular, instead of smoothly polished, is proof that there is but little GLACIAL EROSION AND TRANSPORTATION. 247 abrasion during the settling and consolidation of the neve in the amphitheatres in which it accumulates. At the bottom of the depressions, however, the conditions are different. Intense glaciation there takes place, as is attested by the rounded and striated surfaces, and by the occurrence of rock-basins. The ice filling a cirque impinges with great weight upon its bottom and in its motion outward tends to deepen the exca- vation. At the same time the blocks loosened from the walls of the cirque are carried away by the outward flow of the ice. There are thus at least two processes which unite in enlarging and deepening these peculiar features of glaciated mountain- tops. When a glacier leaves a cirque and flows down a canon the grade of which is uneven, the erosion of the ice-stream will also be uneven. The reason is that the ice in descending a steep slope exerts its greatest force at the base of the incline in the same manner as in the excavation of cirques. The tend- ency of a moving ice-stream in descending a steep slope is to increase the in.equalities of its bed ; this tendency, it seems probable, will lead to the formation of both scarps and cirques when the drainage of a high-grade valley is changed from a liquid to a solid form. To illustrate : The grade of mountain streams increases toward their sources, and when their gorges become occupied by ice, the irregularities of their channels — caused principally by the meandering of streams, thus leav- ing projecting bosses on either side — may cause ice-cascades in the glaciers. An ice-cascade exerts the greatest erosive power at the base of the scarp which it descends, thus aug- menting the inequality. At the same time the canon is broad- ened and the minor features resulting from stream-erosion are erased. The steeper the grade the more pronounced would be the action of the ice in remodeling and strengthening the ma- jor inequalities of its bed. The resulting scarps and terraces should therefore be most numerous and best defined near the heads of the channels in which they occur.* The prevalence of cirques is also graphically described by Dr. G. M. Dawson in his " Eeport upon the Forty-ninth * " The Quaternary History of Mono Valley, California," pp. 362-355. 248 THE ICE AGE IN NORTH AMERICA. Parallel." Speaking of the streams which rise in the Eocky Mountains, he says : The upper ends of the valleys surrounding the higher peaks and ridges are generally very abrupt and take the form of cirques, or amphitheatrical depressions of great depth, in the mountain - sides. The backs and sides of these are often nearly vertical, and they are sometimes only separated later- ally, by steep, knife-edge-like ridges, the crests of which form the most practicable paths to the summits. Each of these upper terminations of the valleys generally also shows a small lake or pond in the hollow of the surrounding cliffs, the basin of which has evidently been formed by glacier-ice — which must here have been descending almost vertically — in the moraine matter or shattered rocky floor. . . . The water of the smaller lakes in the upper ends of the valleys, as seen from the heights around, is of a beautiful semi-opalescent indigo-blue, and must be of considerable depth.* These facts confirm the theories of the leading glacialists of Europe — for instance, Dr. Albrecht Penck, who ascribes the excavations of the most important lake-basins in Bavaria, like the Ammer See and Wurni See, to glaciers, and states that " a lake-basin filled with water or sediment lies at the mouth of each of tlie Alpine valleys through which glaciers protruded in ancient times." f The Scotch lochs, and the rock-basins of Norway, would seem to be due to the same cause. It is probable also that the fiords of Norway and of British Columbia owe their greater depth near their heads to the same anomalous influ- ence of ice-erosion. Most of the arguments urged against the theory are based upon a jpriori reasons urging the impos- sibility of any such result from such a cause. Of this more will be said when speaking a little later of the irregular depo- sition of glacial debris underneath the moving ice. Not enough is known about the nature of ice to afiirm * Page 245. f Quoted by Newberry, in " School of Mines Quarterly," January, 1885, p. 10. 6LA0IAL EROSION AND TRANSPORTATION. 249 that it does not conform to the law of other moving fluids. Probably there is no reason why an ice-cascade should not produce results of erosion analogous to those of a waterfall. Summarily stated, our conclusions are that, like every- thing else connected with the action of such a complicated cause as that brought into view in the production of glacial phenomena, the exact extent of its erosive and transporting power is difficult to determine. The action of ice over the glaciated region took place after other forces had been in full operation during long ages ; and hence it is often impossible to separate the effects of the second cause from those of the first. But there can be no doubt that running water is by far the most efficient of all eroding agencies which have given shape to the contour of the continents. Most important re- sults follow from the power of water to act as a solvent. Extensive regions have been undermined and lowered through the removal by water of the soluble salts. Such has perhaps been the origin of many of the valleys of the Appa- lachian region and of some of the great lakes of the world. Running water is also a most elfective mechanical agency, continually acting along the natural lines of drainage. The sand and gravel rolled along over the bottom of a rapid stream of water act like a rasp or a saw, and have everywhere worn deep narrow channels across the slowly rising mountain- chains. Water as an eroding agency has had a great advan- tage over ice in the far greater length of time during which it has been in the field to operate. Still, it can not be doubted that ice has had no small part in transforming the appearance of the portions of the world to which it has had access. Of this the evidence is abundant in the great number and size of the bowlders scattered over the glaciated region, hundreds of miles from their native ledges, and weighing hundreds and even thousands of tons. Inasmuch as ice is frozen water, its melting furnishes the torrents to aid in the transportation. The finely comminuted material ground up underneath the ice is largely carried away 250 THE ICE AGE IN NORTH AMERICA. by the torrential subglacial streams continually pouring out from the ice-front. It is doubtful if the larger part of the glacial grist is not thus transported far beyond the limit of the glaciated region. Notwithstanding the great waste, the extent of the glacial deposits yet remaining over the southern portion of the gla- ciated region is immense. Probably not less than 1,000,000 square miles of territory in North America is covered with an average depth of fifty feet of glacial debris, forming the most permanently productive portion of the continent. It is in the extent of these glacial deposits, and in the certainly great amount of transportation by subglacial streams, that we have our most certain and impressive evidence of the enor- mous activity of erosive agencies during the Glacial age. CHAPTER XI. DEL'MLINS. " Detjmlin " is the name now used to designate the class of glacial accumulations which Professor Hitchcock origi- nally called " lenticular hills." These abound in the vicinity of Boston, and large- ly give character to the scenery of the three northeastern counties of Massachu- setts. They are not, however, evenly dis- tributed over the re- gion. Familiar ex- amples of them are Beacon Hill, Boston ; Bunker Hill, in Charlestown; Breed's Island Hill, beyond East Boston ; Green Hill, in Winthrop ; Powder - Horn Hill, in Chelsea ; Mount Revere ; Mount "Washington, in Ever- ett ; Tuft's College Hill ; Winter Hill and others, in Somer- viUe ; Bigelow Hill, in Brighton ; White's Hill, in Water- town ; Owl Hill, in Waltham ; Mount Ida, Prospect, Insti- tute, and Oak Hills, in Newton ; Corey's and Walnut Hills, FtG. 72.— Drumlins in the vicinity of Boston. (Davis.) •jr. !-5' '25-2 THE ICE AGE m NORTH AMERICA. «- M-.5«kw j in Brookline ; Parker's Hill, in ^ . Koxbury ; Bellevue and the Clarendon Hills, in West Eox- bury ; Brush Plill, in Milton; Jones's Hill, Mount Ida, and ^!F^'JS^"k ^'^P'*'' ^'^^' '° I>orchester; «?#W/'\ WoUaston Heights, Forbes, President's and Great Hills, in Quincy ; Great and King Oak I Hills, in Weymonth ; Baker's, 6 Otis, Prospect, and Turkey a Hills, in Hingham ; Scituate and Bear Hills, in Cohasset ; Strawberry and Telegraph Hills, in Hull ; and the hiUs of Deer Island in the harbor. More than a hundred others of the same character occur within this area. Mr. Warren Upham's de- '/i g scription of these interesting o features of the landscape is « most complete and satisfactory : « These hills vary in size from § a few hundred feet to a mile in " length, with usually half to two " thirds as great width. Their £ height, corresponding to their area, varies from twenty-five to two hundred feet. But, what- ever may be their size and height, they are singularly alike in out- line and form, usually having steep sides, with gently sloping, rounded tops, and presenting a very smooth and regular contour. From this resemblance in shape DRUMLINS. 253 to an elliptical convex lens, Professor Hitchcock has called them lenticular Mils, to distinguish these deposits of till from the broadly flattened or undu- lating sheets which are common through- out New England. The trend, or direction of the longer axis, of these lenticular hills is nearly the same for all of them comprised within any limited area, and is approx- imately like the course of the striae or glacial furrows marked upon the neigh- boring ledges. In eastern Massachu- setts and New Hampshire, within twen- ty-fiye miles of the coast, it is quite uniformly to the southeast, or east- southeast. Farther inland, in both of these States, it is generally from north to south, or a few degrees east of south ; while in the valley of the Connecticut Eiver it is frequently a little to the west of south. In New Hampshire, besides its accumulation in these hills, the till is frequently amassed in slopes of similar lenticular form. These have their position almost invariably upon either the south or north side of the ledgy hills against which they rest, showing a considerable deflection toward the southeast and northwest in the east part of the State. It can not be doubted that the trend of the lenticular hills, and the direction taken by these slopes, have been determined by the glacial current, which produced the striae with which they are parallel.* * " Proceedings of the Boston Society of Nat- ural History," vol. xx, pp. 224, 225. Mi I o g tn .9 1 254 TEE IDE AGE IN NORTE AMERICA. To this may be added the following interesting remarks by Professor William M. Davis : The general uniformity of outline in any single region is very noticeable ; indeed, the view from the summit of a com- manding drumlin, in the center of a group, shows as character- istic a landscape as that seen in looking from the Puy-de-D6me over the extinct volcanoes of Auvergue. ^Moreover, the con- trol that drumlins exercise over the laying out of roads and the division of property is so complete in districts where they abound, that it is the rule to find roads, fields, gardens, and even houses oriented in obedience to the march of the old ice- invasion. About Bos- ton there are hundreds of dwellings whose walls thus stand in close parallelism with the glacial scratches on bed-rock beneath them. * V Besides the groups of drumlins so prom- inent in the vicinity of Boston, there are two or three others deserving of mention, and which may per- haps be brought in- to connection with them.f One of these, about eight miles wide and twenty long, and containing thirty or forty well-marked individual hills of the character described, follows the coast from Beverly to New- buryport. Parallel to this there is a belt of country, about jl^erhill. MASSA- JCHUSETTS Fia, 75.— Drumlins in northeastern Massachusetts. (Davis.) * "American Journal of Science," vol. cxxviii, 1884, p. 409. f See map, p. 298. DRUMLINS. 255 four miles wide, over which scarcely any of these hills are found. Still farther inland, a longer range can easily be traced. Beginning in the vicinity of Portsmouth, JS . H., this interior series is well developed, in a southwest direction, through Eockingham county to Amesbury, Mass. Thence, on, it completely covers the townships in Essex county on either side of the Merrimack Eiver to Lowell, and continues, with little interruption, through Middlesex county to the vicinity of Fitchburg, Worcester county. To a limited ex- tent these same typical hills abound still farther west through the northern part of Worcester and Franklin counties to the Connecticut Kiver. Areas of them are also reported running up from Ashburnham, Mass., to Weare, IST. H. ; also in the western part of Cheshire county, N. H., and in the vicinity of Worcester, Mass., as well as about Amherst and in the northeastern part of the State of Connecticut.* The following additional facts have been collected by Professor Davis : f A fine series of drumlins stretches from about Spencer, Mass., to Pomfret, Conn., but the detailed study that it would well repay has not s ^i=^*£i2=±S^ ^rr^^ja yet been attempted. Members of this series occur near Charlton station, Boston and Albany Railroad, with their bases at an elevation of nine hundred feet above sea-level, and others stand still higher. The por- tion of the group in Connecticut is described by Percival as fol- lows : " The diitrict extending north from Hampton, through Fie. 76.- -Outltae of parallel drift-hills, in central New York. (Davis.) * Upham, in "Proceedings of the Boston Society of Natural History," vol, XX, pp. 231, 232. f "American Journal of Science," vol. cxxviii, pp. 410, 411. 256 THE ICE AGE IN NORTH AMERICA. Abington, Pomfret, and Woodstock, is characterized by a series of very smoothly rounded, detached hills, in which the rock is usually entirely concealed. These form a striking contrast with the longer and more continuous [rocky] ridges of the adjoining formation." * Professor G. H. Stone reports that drumlins of large size, like those about Boston, have not been found in Maine. Western New York, between Syracuse and Rochester, presents a surprising number of parallel north-and-south drift- hills, probably familiar to many travelers by rail. Some of them are so long, smooth, and even, that the country thereabout has been described as fluted. These were long ago described by Professor James Hall, in his " Geology of the Fourth Dis- trict of New York " (1843) ; since then they have been strangely neglected until examined by Dr. L. Johnson, who has lately published a paper, f entitled "The Parallel Drift Hills of Wesbern New York." Some of the ridges are "two or three miles long, and attain elevations of one or two hundred feet above the intervening valleys ; but the greater number are shorter and steeper. Many of them were, when first cleared of timber, very steep at their north ends, and on their east and west sides ; but, with very rare exceptions, the southern slope is gradual." These and other irregularities of form Z, I -*" Pis. 77.— Drumlins in Wisconsin. (Cliamberlin.) may require that some of the hills of this region should be separated from drumlins as here defined. In Wisconsin, the drift-hills, as described by President T. C. Chamberlin, "are arranged in lines, and their longer axes invariably lie parallel to the movement of the ice. In some localities, especially * "Geology of Connecticut," 1S42, pp. 256, 461, 479, 485. t " Tranaaotions of the New York Academy of Sciences," vol. i, 1882, p. 'Jl. BRTJMLINS. 257 in Dodge and Jefferson counties, tliese are mainly replaced by long parallel ridges, sometimes several miles in length, with corresponding linear marshes interspersed. These correspond accurately to the direction of the ice-motion." * According to Mr. Upham, drumlins are not found in the abundant drift of Minnesota. A few examples are mentioned for Pennsylvama, near its western border, by Professor Lewis, f In endeavoring to account for this class of hills, two or three facts are of sjjecial importance : 1. The material of which they are composed is very heavy and compact — almost as heavy and compact, indeed, as ice. Such masses could not have been shoved along bodily beneath the ice. In fact, Fig. vs.— Drumlins in GofEstown, S. H. (Oitchcock.) there would seem to be no reason why they might not resist the erosion of the glacier almost as well as many of the softer rocks did, especially when we remember that the pressure of the ice on the bottom need not have been uniform, but greater in some places than in others. 2. There are many indications that these hills were formed by accretion under the ice, there being, as Mr. Upham has shown, a tendency to lamination or coarse glacial stratification in the structure of * " Geology of Wisconsin," vol. i, 1883, p 283. f "Second Geological Survey of Pennsylvania, Z," pp. 29, 188. 17 258 THE ICE AGE IN NORTH AMERICA. the hills.* 3. They are not cliaracterized by kettle-holes. The surfaces are remarkably symmetrical, as if having been smoothed over by design, and all the irregular depressions are filled with homogeneous earth. 4. Up to one hundred feet above their base, the flanks of these hills in Massachusetts and New Hampshire are frequently covered with the water- worn deposits hereafter to be described and known as kames, and which are the very last work done by the ice at the points where they are found. The drumlins are, therefore, earlier than the kames. In structure these hills resemble the lower portions of till, or the ground moraine. They are only imperfectly stratified, and very compact, and filled with foreign and finely striated stones. They are, without doubt, a true glacial deposit ; but how comes the deposit to be heaped up in these localities in such vast and shapely masses? Professor Shaler surmised that they were but the remnants of a continuous ground moraine which had been eroded from the whole country', except where it was protected by pedestals or underlying rock which served to break the force of the beating waves of the ocean.f To this ingenious theory there are two fatal objections : 1. Drumlins are frequently found where there are no rocky pedestals to protect their bases. 2. They occur in the interior far above any height to which it is supposed the ocean has reached since the Glacial period. " The alti- tudes at which they occur vary from the level of the sea to fifteen hundred feet above it on the height of land between the Merrimack and Connecticut Rivers." X Mr. Clarence King would explain them as marking places in the great continental glacier where streams of water which had run for some distance in superficial channels along the surface of the glacier and collected a great amount of debris from the medial moraines, had finally plunged through a moulin into * " Proceedings of the Boston Society of Natural History," vol. xx, p. 223. \ Ibid., vol. xiii, pp. 196-283. X Ibid., vol. XX, p. 233. DBUMLINS. 259 a deeply hidden subglacial river.* Here, it is thought, vast cavities might be formed in which these accumulations vfould take place, while, by the movement of the ice, the crevasse might be transferred farther down, and so the accumulated deposit be subjected to the pressure and sculpturing power of the ice. Mr. Upham speaks on the subject as follows : " The finely pulverized detritus and glaciated stones in the bottom of the ice-sheet had a tendency to lodge on the surface of any de- posit of the same material. When such banks of the lower till became prominent obstacles to the ice-current, its level- ing force was less powerful than this tendency of adhesion, which continually gathered new material, building up these massive rounded hills." f Mr. Upham remarks upon the par- tial parallelism of these ranges of hills with the extreme ter- minal moraine, and, with his usual perspicacity, notes that both the glacial stri^ and the trend of the axes of the len- ticular hills nearest the coast in Essex and Suffolk counties, bear much more easterly than they do even a few miles in the interior. This points with much force to the effect which would be produced upon the movement of the ice near its margin when it had receded a considerable distance from the south, but especially from the east, where the waters of the Atlantic had access to the glacier along the margin of what is called the Gulf of Maine. When the waters of this gulf had eaten the ice well away into the Massachusetts shore, and thus removed the barrier to the east, the line of least resist- ance would be in that direction, and the current would natu- rally swing out toward the open sea. While recognizing the force of all Mr. TJpham says, I can not forbear repeating an additional suggestion of my own * I do not know as Mr. King haa anywhere published these views, nor, in- deed, as he would now be willing to own them, as here stated. They were given me in personal conversation, and contain so much that is worthy of con- sideration, that I venture to repeat the theory. f " Proceedings of the Boston Society of Natural History," vol, xx, p. 234. 260 TEE ICE AGE IN NORTH AMERICA. made at the same time * namely, that these hills perhaps represent an earlier moraine than that on the south shore of New England — i. e., one which was formed when, on the first advance of the ice, it had- reached the latitude of Boston, ajid where for some reason it paused until great accumulations had taken place along its front ; that afterward, upon a fresh advance, these accumulations were overrun by the ice with- out being leveled ; being merely sculptured by it, and read- justed to the changing line of general movement ; and that, finally, the retreat of the ice was so rapid over this region that there were no marked terminal accumulations ; but the superglacial debris settled gently over the whole country, constituting the more highly colored superficial blanket of debris called by Hitchcock "upper till," and furnishing the larger and more angular bowlders characteristic of the super- ficial deposits. I find also that Professor Charles Hitchcock had made the same suggestion as early as 1876.f The long discussion concerning the origin of these singu- lar hills would seem to have been brought to a close by the careful summary and discussion of facts given by Professor Davis, from which we have already quoted : The first clear reference to drumlins, as directly dependent on glacial action lor their form, was made by M. H. Close. J They are here said to be parallel to the neighboring striae, and hence, like these, dependent on the ice-sheet for their present attitude and form. The same conclusion is presented in a paper of 1866, Avhen the name drumlin was first specially pro- posed for them. Still later, when describing the physical geography of the neighborhood of Dublin, Close writes, "It is perfectly certain that it must have been the rock-scoring agent which produced the bowlder-clay ridges." Besides this, Kinnahan and Close, in a pamphlet of 1873, stated their opin- * "Proceedings of the Boston Society of Natural History," vol. xx, p. 218; also, "Prehistoric Andover," p. 4. \ Ibid., vol. xix, p. 66. Professor N. S. Shaler now favors this view ; see the "Seventh Annual Report of the United States Geological Survey," 1888, p. 321. X "Journal of the Koyal Geological Society of Ireland," vol. i, 1864, p. 3. DRUMLINS. 261 ion that/drumlins were formed in a way "similar to that by which a stream of water often makes longitudinal ridges of sand in its bed."* This is to mv mind the best suggestion yet given to account for them, j J. Geikie wrote as follows : "The remarkable linear direc- tion of certain mounds of bowl- der-clay in some districts of the Lowlands, agreeing as this does with the general bearing of gla- cial markings of the same lo- calities, induces us to believe that we have here, with certain modifications, the original con- tour of the till after the super- incumbent ice-sheet had disap- peared " ; f but he believed that these forms may be also in part dependent on marine erosion. In the "Great Ice Age," the same author briefly mentioned "the series of long, smoothly rounded banks or drums, and sow-backs, which run parallel to the direction taken by the ice," and regarded them as very Fig. 79.— Drumlins in Irelana, after Km- Tj.il j-i2 J .c ii, ■ 1 -1 nahan and Close. (Davis.) little modified from their glacial form. They are "produced by the varying direction and un- equal pressure of the ice-sheet," and are "the glacial counter- parts of those broad banks of silt and sand that form here and there upon the beds of rivers." Dr. L. Johnson says that he accepts Geikie's explanation, and applies it to the New York ridges which were "formed underneath the glaciers by alterna- tions of lateral pressure " ; but this form of statement does not commend itself so highly as the preceding. * " General Glaciation of lar-Connaught," Dublin, 18Y2. f "Transactions of the Geological Society of Glasgow," vol. iii, 1867, p. 61. 262 THE IGE AGE IN NORTH AMERICA. In this country, Professor C. H. Hitchcock and Mr. War- ren Upham, while engaged on the geological survey of New Hampshire, were the first to discover the parallelism between glacial motion and the axes of drumlins in 1875. They con- cluded that " the accumulation of these hills and slopes seems to have been by slow and long-continued addition of material to their surface, the mass remaining nearly stationary from the beginning of its deposition. Obviously this was the case with the lenticular slopes gathered behind the shelter of higher ledgy hills or upon their opposite sides."* A little later Up- ham wrote, " Although we do not discover the cause of the peculiar distribution of these hills, it seems quite certain that they were accumulated and molded in their lenticular form beneath the ice." f President Chamberlin's observations led him to a similar conclusion : " The drift presents- some pecul- iar tendencies to aggregation. ... A special tendency is ob- served over certain considerable areas lying not far from the Kettle Moraine to accumulate in mammillary or elliptical or elongated hills of smooth-flowing outline." % -A-nd again, after repeating this opinion, it is suggested that " a deeply hidden boss of rock is usually and perhaps universally the determin- ing cause of these peculiar accumulations." * In reviewing these explanations and the observations on which they are based, together with such evidence as my own studies have discovered, the conclusion that drumlins should (be compared to sand-banks in rivers appears the most satisfac- tory yet advanced. They seem to be masses of unstratified drift slowly and locally accumulated under the irregularly moving ice-sheet where more material was brought than could be carried away. The evidence for the subglacial growth of drumlins may be summarized as follows : The scratched stones in the mass of bowlder-clay show a differential motion of its several parts as they were scraped and rubbed along from a generally northern source and gradually accumulated where * " Geology of New Hampshire," vol. iii, p. 308. t "Proceedings of the Boston Society of Natural History," vol. xx, p. 223. X "Geology of Wisconsin," vol. i, 1883, p. 283. * " Third Annual Report of the United States Geological Survev," 1883, p. 306. JDEUMLINS. 263 now found. \The compactness of the mass suggests an origin under heavy pressure ; the attitude of the hills demonstrates a close dependence on the motion of the ice-sheet ; the super- position of kames on their flanks proves that their present form was essentially completed when they were uncovered by the ice-sheet, and the small change of form in the kames shows that the drumlins also can have suffered very little from post- glacial erosion ;\the faint channeling of their smooth slopes by rain measures the small amount of denudation that they have suffered since they were made. It must therefore be concluded that they were finished closely as we now see them when the ice melted away, and hence they were of subglacial construc- tion, y The supposed manner of accumulation of drumlins may be briefly sketched. It is well known that a stream of running water will at one point carry along silt and sand that must be dropped a little farther on where the current slackens, and the bank thus begun grows slowly in a form of least resistance, at- taining a maximum size when its increase of volume has so far diminished the cross-section of the stream and consequently increased the velocity that no more detritus can be dropped there ; but even then one end may be worn away while the other grows, the adjustment of velocity to channel is not per- manent. The motion of a glacial sheet has been justly com- pared to that of a broad river. The comparison may be ex- tended so as to liken the active head-waters of a stream to the presumably fa|^-moving part of the ice-sheet near its source or center of dispersion where the greatest erosion generally takes place. The delta of a river corresponds to the thinner and slower- moving marginal area of an ice-sheet, where drift brought from elsewhere is qu'etly and evenly deposited, as in Minnesota, and where erosion is relatively weak. A still fur- ther agreement is discovered in comparing the drumlins and sand-banks found in the middle course of the molten and solid streams as suggested by the several authors quoted above. In view of the irregularity of the surface on which the ice-sheet moved and of the greater weakness of some rocks than others, we must suppose an irregular velocity in the motion of the ice and an unequal distribution of the rubbish beneath it. If the 264 THE ICE AOE IN NORTH AMERICA. faster motion at one place cause an excess of erosion there, the slower motion at another place may bring about an excess of deposition. This difference of action is known to prevail be- tween the central and marginal parts of glaciated areas, and the local accumulation of drumlins in an intermediate region gives a smaller example of these two parts played by the ice. If the causes of the irregular motion of the ice lie in the gen- eral form of the country, the location of faster and slower cur- rents will be relatively permanent ; the districts of faster cur- rents would be found where the greatest volume of ice is allowed to pass, and some of the points of retardation may be the seats of long-continued drumlin-growth. The drumlins thus begun will depend less upon the immediately local form of the ground than on the topography of a more considerable district, and hence we need not suppose every drumlin to have begun its growth upon a knob of rock, although the beginning of many hills may have been thus determined. Once begun, the drumlins will go on increasing in size as long as deposition exceeds erosion, always maintaining an arched form of least resistance until a maximum size is reached or until the ice melts away ; and in their growth they will approach the form to which rough, rocky hills would be reduced by the reverse process of erosion if time enough were allowed. Under un- ending glaciation the whole surface must be rubbed down smooth.* As these theories relating to the formation of drumlins involve the general principles upon which we are to explain other evidences of the varying degrees of erosion effected by moving ice, we may as well introduce here, as anywhere, Mr. Geikie's general reply to the objections urged from tlie sup- posed nature of the case : f Our ice-sheet flowed, we can not doubt, with a differential motion : it must have moved faster in some places than in others. In steep valleys and over a hilly country its course * "American Journal of Science," vol. oixviii, pp. 413-416. f " The Preservation of Deposits of Incoherent Materials under Till or Bowlder Clay," in the " Geological Magazine," February, 18*78, pp. 2, 3, 6, *l. DRUMLIN8. 2.65 would often be comparatively rapid, but very irregular — lagging here, flowing quickly there — while in wide, open valleys that sloped gently to the sea, such, for example, as those of the Forth and the Tweed, the whole body of the ice would flow with a slower and more equable motion. As the ice-sheet approached its termination, more especially if that terminus chanced to be upon a broad and comparatively flat region, like East Anglia, the erosive power of the ice would become weaker and weaker, for two reasons : first, because of its gradual attenuation ; and, secondly, because of its con- stantly diminishing motion. These, in a few words, are the varying effects which one might a priori infer would be most likely to accompany the action of a great ice-sheet. And an examination of the glacial phenomena of this and other coun- tries shows that the actual results are just as we might have anticipated, had it been previously revealed to us that a large part of our hemisphere was, at a comparatively recent date, almost entirely smothered in ice. In places where, from the nature of the ground, we should look for traces of great glacial erosion, we find rock-basins ; in broken, hilly tracts, where the ice-flow must have been comparatively rapid but irregular, and the glaciation severe, we meet with roches moutonnees in abun- dance, but with very little till ; in the open Lowlands and in the broad valleys where the ice-sheet would advance with dimin- ished but more equable motion, we come upon wide-spread and often deep glacial deposits, and now and again with interglacial beds ; while over regions where the gradually decreasing ice- sheet crawled slowly to its termination, we discover consider- able accumulations of till, often resting upon apparently un- disturbed beds of gravel, sand, and clay. The distribution of interglacial deposits, therefore, is really in itself a proof that they have been overridden hy ice. When they occur in highly glaciated regions, it is only as mere patches which, occupying sheltered places, have been pre- served from utter destruction. In the opener, low grounds they are found in greater force, although in such places they almost invariably afEord more or less strong evidence of having been subjected to much erosion and crumpling. But the far- ther we recede from the principal centers of glaciation, and the 266 THE IGE AOE IN NORTH AMERICA. nearer we approach the extreme limits reached by the ice- sheets, the more exteDsive and the less disturbed do inter- glacial deposits become. In a word, they occur in best preser- vation where the erosive power of the ice was weakest ; they are entirely wanting where we have every reason to believe that the grinding force was strongest. . . . It is needless to refer one to the petty glaciers of the Alps and Norway to prove that glacier-ice can not both erode its bed and accumulate debris upon that bed at one and the same time. A mountain-valley glacier is one thing — a glacier ex- tending far into the low grounds beyond the mountains, and, it may be, coalescing with similar extensive ice-flows, is another and very different thing. No considerable deposit could possibly gather below Alpine glaciers like those of Switz- erland and Norway ; hut underneath glaciers of the kind that invaded the low grounds of Piedmont and Lombardy we know that thick deposits of tough bowlder-clay, crammed with scratched stones, did accumulate ; and not only so, but that tJiese glaciers flowed over incoherent deposits of sand and clay containing marine shells of late Tertiary age, without entirely oiliterating them. The deposits referred to occur now as little patches within the area bounded by the great terminal mo- raines. As physicists themselves are not yet quite agreed upon the subject of glacier motion, it is not incumbent upon the geolo- gist to explain the precise mode in which a thick mass of ice can creep over the surface of incoherent beds without entirely demolishing them. It is enough for him to show how the remarkable distribution of the interglacial beds, and the vari- ous phenomena presented by these deposits, indicate that ice has overflowed them. It is useless, therefore, to tell him that the thing is impossible. The statement has been made more than once that an ice-sheet several thousand feet thick is a physical impossibility ; but, unfortunately for this dictum, the geological facts have demonstrated that such massive ice- sheets have really existed, and there appears to be one even now covering up the Antarctic Continent. We used also to be told, not so many years ago, that the abysses of ocean must be void of life for various reasons, among which one was that the DEUMLINS. 267 pressure of the water would be too great lor any living thing to endure. Yet many delicate organisms have been dredged up from depths at which the pressure must certainly be no trifle. JSfow, there seems to be just as little difficulty in believ- ing that these organisms existed in a perfect state at the bot- tom of the ocean, as that shells imbedded in clay would remain unbroken underneath the pressure of a superincumbent ice- sheet of equal or greater weight. If the ice were in motion, the clay with its included shells might be plowed out bodily, or be merely crumpled and contorted ; or it might be ridden over with little or no disturbance ; or, on the other hand, it might become involved with subglacial debris, and be kneaded up and rolled forward — the shells in this case being broken, crushed, and striated, just as we find that the shells in certain areas of till have been. The fate of the fossiliferous beds would, in short, be determined by the rate of flow and degree of pressure exerted by the superincumbent ^'wasi- viscous body — the motion of which would be largely controlled by the physical features of the ground across which it crept. CHAPTER XII. PEEGLACIAi DRAINAGE. One of the most marked effects of the Glacial period was its influence upon drainage systems. The changes pro- duced in numerous river-courses of North America by the irregular deposits of till and modified drift, as well as by the existence of temporary barriers of ice during the con- tinuance of the continental ice-sheet, are subjects of unfailing interest to the student of physical geography, and are also of great practical significance in their relation to the economic and hygienic interests of the country. As compared with preglacial time, that which has elapsed since the close of the Ice age is admitted by all to be very short. Consequently, post-glacial erosion is much less than preglacial erosion. Before the advent of the continental ice-sheet, all the great valleys of North America had been sculptured by preglacial streams.* The effects are still to be seen even where extensive deposits of the Glacial period have partially obliterated them. The sedimentary rocks, occupying the basin of the Mississippi, and filUng it with strata thousands of feet in depth, serve as one index of the extent of preglacial erosion ; for all the material of this class of rocks has been ground up and transported by water. Coming down from the neighborhood of the "White Mount- ains, the Adirondacks. and the Archaean highlands of Canada, sediment-laden streams have, from the very earliest geologi- cal ages, been engaged in wearing away the hills, scooping * See Chapter X. PRE GLACIAL DRAINAGE. 269 out the valleys, and silting up the sea. The Alleghany Mountains were at one time the bed of the ocean upon which this sediment was deposited. The sandstones, shales, and conglomerates of the coal-measures attest the activity of the forces of that early period. The tops of the mount- ains in southern JSTew York and northern and eastern Penn- sylvania are covered with subcarboniferous conglomerates of almost incredible depth and extent, consisting largely of well-rounded quartz pebbles, of all sizes up to two or three inches in diameter. These are water-worn, and must have been rolled along by impetuous currents from far-distant re- gions. Thus the tributaries of the Mississippi are, at the present time, bringing into its valley similar deposits from the mountain plateaus on either side. But no sooner did the convulsive forces of the earth begin to lift this great, stratified ocean-bed of the Appalachian region above the water, than it too became the subject of erosion, and began to furnish material for newer deposits farther to the south and west. The Ohio River and its tributaries furnish a good exam- ple of the extent of preglacial erosion. The traveler is im- pressed with the gorge in the Niagara River below the falls, as showing the force of ranning water when concentrated in a single line of drainage. The gorge of the Niagara, how- ever, is only about seven miles long, a thousand feet wide, and three hundred feet deep. This, as will appear later,* is one of the best measures of post-glacial erosion. But the Ohio River, containing a far less volume of water, has worn a much larger and deeper trough more than a thousand miles in length. The character of the trough of the Ohio and its tributaries is readily discerned, even by the passing observer. The strata on the opposite sides are horizontal, and match each other like the ends of a plank that has been sawed asun- der. The alternate layers of conglomerate, coal, shale, and sandstone upon the one side of the river correspond to simi- * See Chapter XX. ■270 TEE ICE AGE IN NORTE AMERICA. lar layers on the other side. The width of the gap cut by the stream averages about a mile, with enlargements wher^ ever a tributary comes in from either side. The tributaries, also, occupy corresponding narrow valleys of erosion, extend- ing even to their very sources in the mountains. The nature of the cause producing these narrow troughs, and its long- continued operation in every one of these tributaries of the Ohio, are not difficult to see. They have all been formed by running water. The Tennessee, the Cumberland, the Ken- tucky, the Wabash, the Miami, the Licking, the Scioto, the Big Sandy, the Kanawha, the Hocking, the Muskingum, the Big Beaver, the Monongahela, and the Alleghany, together with their tributaries, all show the vast amount of water- erosion along these lines of preglacial drainage. But even these do not, in their present condition, reveal the whole extent of the effect of the constant and long-con- tinued erosive forces of preglacial times. The ancient bed of the Ohio River was certainly one hundred and fifty feet deeper than that over which it now flows, it having been filled with glacial debris to its present level. According to Professor Newberry — At the junction of the Anderson with the Ohio, in Indiana, a well was sunk ninety-four feet below the level of the Ohio before rock was found. In the valley of Mill Creek, in the suburbs of Cincinnati, gravel and sand were penetrated to the depth of one hundred and twenty feet below the stream before reaching rock. On the margin of the Ohio, at Cincinnati, gravel and sand have been found to extend to a depth of over one hundred feet below low-water mark, and the bottom of the trough has not been reached. The falls of the Ohio, formed by a rocky barrier across the stream, though at first sight seeming to disprove the theory of a deep continuous channel, really affords no argument against it ; for here, as in many other instances, the present river does not follow accurately the line of the old channel, but runs along one side of it. At the Louisville falls, the Ohio flows over a rocky point which projects from the north side into the old valley, while the deep PREGLAGIAL DRAINAGE. 271 channel passes on the south side, under the lowlands on which the city of Louisville is built. The tributaries of the Ohio exhibit the same phenomena. At New Philadelphia, Tuscarawas county, the borings for salt-wells show that the Tuscarawas is running one hundred and sev- enty five feet above its ancient bed. The Beaver, at the junc- tion of the Mahoning and Chenango, is flowing one hundred and fifty feet above the bottom of its old trough, as is dem- onstrated by a large number of oil-w^lls bored in the vicinity. Oil Creek is shown by the same proofs to run from seventy- five to one hundred feet above its old channel, and that chan- nel had sometimes vertical and even overhanging walls.* Additional particulars of much interest concerning buried channels in the Ohio are given by other geologists. For ex- ample, Professor Joseph F. James presents cogent reasons for believing that the northward bend of the Ohio Eiver, now culminating at Cincinnati, continued still farther north pre- vious to the Glacial period, and extended through Mill Creek up to join the valley of the G-reat Miami at Hamilton. He supposes that the main stream then ran north of the city through the valley in which Madisonville is situated. The evidence of this is, that below Cincinnati, a short distance, the present river flows " in all probability," over bedded rock between Price Hill and Ludlow, Ky. ; while borings show that up Mill Creek several miles the bed-rock lies certainly thirty-four feet below low-water mark, while at Hamilton, twenty-five miles north of Cincinnati, the preglacial valley is found to be filled up to a depth of more than two hundred feet, and the bed-rock lies ninety-one feet below low-water mark in the Ohio at Cincinnati.f Mr. M. C. Read, among other numerous references to buxied channels, describes one in Knox county, east of Gam- bier, in the valley now occupied by Owl Creek, where the * " Geological Survey of Ohio," vol. ii, pp. 13, 14. f " Journal of the Cincinnati Society of Natural History," July to October, 1888, pp. 96-101. 272 THE ICE AOE IN NORTH AMERICA. bed-rock lies eighty-two feet below the bottom of the present stream. West of Mount Liberty, in the same county, the drift conceals an old gorge two liundred and eighty-five feet deep.* Mr. P. Max Foshay, in a paper before the American Association for the Advancement of Science, in 1888, gives many reasons for supposing that Beaver Creek, which now empties into the Ohio, was connected by a buried channel with Grand Rivei-, emptying into Lake Erie at Painesville, and hence that a still laiger portion of the upper Ohio drain- age than was supposed by Mr. Carll passed into the St. Law- rence "Valley. This suggestion was first made by Professor Spencer, as appears on the accompanying map.f Every day is demonstrating that the present level appear- ance of the surface of the northwestern portion of Ohio is due to the extensive deposits of the Glacial period, whose effect has not been so much to make the hills low as to exalt the valleys. Professor Orton long ago called attention to the numerons buried channels near Springfield in Clarke county, one of which is occupied by the New York, Pennsylvania, and Ohio Kailroad.;]: The extensive explorations for stores of gas and oil now in progress in the western part of Ohio and the eastern part of Indiana are bringing to light buried channels in most unexpected places, that at St. Paris, Cham- paign county, being more than five hundred feet deep. This is an extreme case, but it illustrates what a network of pre- glacial gorges have been plastered up by the ice-movement which passed over the region. The country resembles, on a large scale, a checked and worm-eaten plank which a carpen- ter has filled with putty. One of the first effects of this filling up of the preglacial channels has been so to change the lines of superficial drain- age, in a great multitude of instances, that the streams are now made to run over rocky beds at levels far higher than * " Geological Survey of Ohio," vol. iii, pp. S25-34'7. t See Plate III, p. 278. X " Geological Survey of Ohio,'' vol. i, pp. 450-480. PREOLACIAL DRAINAGE. 2Y3 they had formerly occupied. Thus it is that the glaciated region became again a region of waterfalls. Almost eyery stream entering Lake Erie from the south exhibits waterfalls produced in this manner. In Minnesota the falls of St. Anthony, at Minneapolis, and of Minnehaha, a few miles be- low, were thus produced, and, so, are post-glacial in their origin, the ancient channels having been filled with glacial debris. The falls of Niagara are due to the same cause. The pre- glacial outlet to Lake Erie was dammed up and buried by glacial debris, so that the water was compelled to seek another channel. Before the lee age there was no Magara River, and Lake Erie is, in fact, but a glacial mill-pond. The falls of the Genesee at Eochester are also clearly due to the same cause. The preglacial valley in the Genesee is now deeply buried. "Between Mount Morris and Eochester the river follows its preglacial valley, but flows for much of this distance on an alluvial plain that closely resembles the filling of au old lake ; above and below the limits named the river has cut a new channel since glacial times, giving some of the best natural sections in the State, and its old course is choked with drift." * The falls of the Mohawk at Cohoes, also, doubtless indi- ~ cate the existence of a deeply buried channel somewhere in the vicinity, connecting, as we shall see a little later, the basin of Lake Ontario with the Hudson. The evidence that the preglacial outlet of Lake Erie was much lower than its present outlet, lies in the fact that sev- eral rivers now entering the lake from the south flow at a level two hundred feet above that formerly occupied by them, since that distance has been penetrated beneath their present bottom before reaching rock. From various borings for oil about 1860, Newberry discov- ered that the rocky floor of the Cuyahoga entering Lake Erie * See Professor W. M. Davis, in the " Proceedings of the Boston Society of Natural History," vol. xxi, p. 359. 18 274 THE ICE AGE IN NORTH AMERICA. at Cleveland is two hundred feet lower than the present bed of the stream, and that this depth extends for twenty miles IBastar. XeUtfC^ Fig, -Section across the valley of the Cuyahoga River, twenty miles above its mouth. (Claypole.) up the river. Newberry also discovered, about the same time, that Grand and Rocky Kivers formerly flowed at the same low level. In the case of Eocky River, Dr. D. T. Gould has traced the buried channel southward for a distance of nearly thirty miles, or into its upper waters in Medina county. This he has done by collecting the record of wells along the route, and by noting various places where the present stream crosses the old bed, passing out of rocky banks for a short distance, and running through clay banks and over a clay bottom to enter its new channel again between rocky walls. Professor Spencer has also shown, with great probability, what was the preglacial line of drainage through which the waters, both of Lake Huron and of Lake Erie, flowed to enter Lake Ontario on their way to the sea. The line, as he first thought, passes out of Lake Huron through the valley of the An Sable, crossing the Thomas River near London, in Canada, and entering the basin of Lake Erie a little east of Port Stan- ley. Thence, after passing around Long Point and Island, it bends northward through the valley of Grand River, and enters Lake Ontario at its extreme western point.* From later information, furnished me by letter, it appears that Pro- fessor Spencer is inclined now to make the connection be- tween Lake Huron and Lake Ontario direct, passing through Georgian Bay, and reaching Ontario in the vicinity of Toron- * " Second Geological Survey of Pennsylvania," Q", pp. 357-406. PREOLAGIAL DRAINAGE. 275 to, as shown in his accompanying map. Thence, according to Newberry, the ancient Hne of drainage passed through On- tario, and emerged in a stream occupying the valley of the Mohawk, to swell the current of the Hudson rather than that of the St. Lawrence.* The facts concerning this line of preglacial drainage had been thus succinctly stated by Professor Newberry in an ear- lier paper ; Some of the streams draining into the basin of Lake Ontario in former times cut their channels below the present ocean- level. All the salt-wells of Syracuse are sunk in one of these, which is filled with gravel and sand saturated with brine issu- ing from the salina group that forms its walls. The rock-bot- tom of this old river-bed was reached in some of these wells at a depth of fifty feet below the present level of tide-water. The valley of the Mohawk is a very deep channel of erosion, now half filled, which must have been traversed by a large stream flowing eastward at a level below that of the present ocean ; and everything indicates that this was the ancient out- let of the basin of the Great Lakes. The channel of the Hudson is apparently the only possible continuation of this long line of drainage. As has been re- marked, it is of great and yet unknown depth. The clay by which it is partially filled has been penetrated to a dejith of about one hundred feet along its margins. How deep it is in the middle portion can only be conjectured ; but Hell-Gate Channel, which has been kept comparatively free by the force of the tides, is in places known to be nearly two hundred feet deep ; and, since this is a channel of erosion formed by a stream draining into the Hudson, the ancient bed of the Hudson must be still lower, f The silting up of these preglacial outlets has enlarged Lakes Ontario and Huron far beyond their previous limits, and wholly created Lake Erie. Lakes Michigan and Superior * Paper read before the American Philosophical Society, November 4, 1881, p. 93. f "Popular Science Monthly," vol. xiii, p. 16. 276 THE ICE AGE IN NORTH AMERICA. were likewise greatly enlarged by the damming up of outlets whicli formerly conducted their waters southward into the Mississippi. These barriers also turned the surplus waters of those inland seas toward the east. The outlet of Lake Superior through the rapids of the Sault Ste. Marie is caused by the increased depths of the glacial deposits to the west across the narrow isthmus separating it from Lake Michigan, where, doubtless, a ship-canal could be cut between these two lakes without encountering any rocks at all. From the south- ern end of Lake Michigan, also, a deeply-buried preglacial channel is believed to ran southwest, through Kankakee, Livingston, and MacLean counties, toward the Mississippi. The course of preglacial drainage in the upper basin of the Alleghany River is worthy of more particular mention. Mr. Carll, of the Pennsylvania Geological Survey, has conclu- sively shown that previous to the Glacial period the drainage of the valley of the upper Alleghany north of the neighbor- hood of Tidioute, in Warren county, instead of passing south- ward, as now, was collected into one great stream flowing northward through the region of Oassadaga Lake to enter the Lake Erie basin at Dunkirk, N. Y. The proof of this is that between Tidioute and Warren the present Alleghany is shal- low, and flows over a rocky basin ; but from Warren north- ward, along the valley of the Conewango, the bottom of the old trough lies at a considerably lower level, and slopes to the north. Borings show that in thirteen miles the slope of the preglacial floor of Conewango Creek to the north is 136 feet. The actual height above tide of the old valley floor at Fentonville, where the Conewango crosses the New York line, is only 964 feet ; while that of the ancient rocky floor of the Al- leghany at Great Bend, a few miles south of Warren, was 1,1T0 feet. Again, going nearer the head- waters of the Alleghany, in the neighborhood of Salamanca, it is found that the ancient floor of the Alleghany is, at Carrollton, seventy feet lower than the ancient bed of the present strea;n at Great Bend, about sixty miles to the south ; while at Cole's Spring, in the neigh- borhood of Steamburg, Cattaraugus county, N. Y., there has PBEOLAOIAL DRAINAGE. 277 been an accumulation of 315 feet of drift in a preglaeial valley whose rocky floor is 155 feet below the ancient rocky floor at Great Bend. There must, therefore, of necessity have been some other outlet than the present for the waters collecting in the drainage-basin to the north of Great Bend. While there are numerous superficial indications of bmied channels running toward Lake Erie in this region, direct ex- ploration has not been made to confirm these theoretical con- clusions. But, if resorted to, we know, from the facts just stated, that the line of some such drainage valley must be dis- covered. In the opinion of Mr. Carll, Chautauqua Lake did not flow directly to the north, but, passing through a channel nearly coincident with that now occupied by it, joined the northerly flowing stream a few miles northeast from James- town.* It is probable, however, that Chautauqua did not then exist as a lake, since the length of preglaeial time would have permitted its outlet to wear a continuous channel of great depth corresponding to that known to have existed in the Conewango and upper Alleghany. In Professor I. C. White's report upon Pike and Monroe counties, Pennsylvania,f he gives an account of no less than twenty -three channels which have been buried by glacial debris. Among these that of the Wallenpaupack Creek is the most striking. At present this creek empties into the Lackawaxen at Paupack Falls, where it descends 260 feet in a mile. But on ascending the creek two or three miles, to the vicinity of Tafton, the course of a preglaeial valley can be easily recognized, leading into Kimball's Eun, and join- ing the Lackawaxen at Kimball's Station. This channel is now buried to a depth of 300 feet for a distance of many miles. R. W. Ells calls attention, also, to the numerous buried channels :|: in the Eastern Townships, in the province of Que- * " Second Geological Surrey of Pennsylvania," III. t Ibid., G«, pp. 52-63. % "Annual Keport of the Geological and Natural History Survey of Canada," vol. ii, 1866, p. 49, J. 18 278 THE ICE AGE IN NORTH AMERICA. bee, in the vicinity of Lake Memphremagog. These are, to a considerable extent, explored at the present time for the sake of the gold found in them. These are but a few of the innumerable facts indicating that before the great Ice age not only the Ohio, but nearly all the streams of the eastern United States, occupied deeper channels than they now do. There were then probably no Great Lakes, and few if any waterfalls, as there are now no lakes and waterfalls south of the glaciated region. All the rivers had cut their channels down so low that they drained to the bottom any lakes that may have once existed. CHAPTEE XIII. DEAESTAGE OF THE GLACIAL PEEIOD. DuBHTG the continuance of the Ice age, an extraordinary factor was in the field to modify the lines of drainage, and to give to them both a direction and a character such as they never had had at any other time. Throughout the whole ex- tent of the Glacial period the ice itself was a most important barrier, deflecting the course of the streams, and, at the same time, was a cause of irregularity in the voluifie of water such as is altogether unique in the history of the world. The vast mass of frozen water then stored up at a high level was an immense reservoir of force, ready, on proper conditions, to descend in torrents through any channel which was opened before it. As the ice of the Glacial period advanced southward from the Laurentian highlands, it reversed the currents of all the great rivers which flowed to the north. One of the first and most remarkable e£Eects of this advance must have been the damming up of Nelson Kiver, so as to cause the surplus water — ordinarily flowing through Hudson Bay into the North Atlantic — to pour over into the head-waters of the Mississippi and so into the Gulf of Mexico. Thus, from the beginning of the Glacial period to its close, the Mississippi River must have been the channel through which was carried off the waste water from the larger part of the Dominion of Canada as well as from the central portion of the United States. A little later, also, the drainage of the Great Lake region must have been obstructed toward the northeast and east; for, long before the eastern lobe of the glacier had 280 TEE ICE AGE IN NORTH AMERICA. reached the latitude of E"ew York city, the valleys of the St. Lawrence and of the Mohawk must have been closed up by ice so as to reverse their lines of drainage. The waters of Lakes Superior and Michigan must then have flowed into the Mississippi Eiver along the lines of the Fox, Wisconsin, and Illinois Eivers, while those of Lakes Huron and Erie poured into the Ohio Kiver, at first down the "Wabash, then, a little later, when the extension of the central lobe of ice cut off the western outlet of Lake Erie, over the lowest places in the water-shed into the valleys of the Scioto, the Muskingum, and Beaver Eivers ; at the same time, every northern tribu- tary of the Alleghany was a glacial flood. But the scenes to have been witnessed during the ad- vance of the ice-sheet are as nothing compared with those which must have occurred during its retreat. Even now, every spring has its freshets, when the combined action of ice and water produces floods unparalleled at other seasons of the year. If this is the case upon the melting of the small amount of snow which annually accumulates in our present winters, the floods at the breaking up of the Glacial period itself must have been inconceivably great. With every recurring spring we now look in the telegraphic summary for thrilling accounts of ice-gorges formed in the St. Lawrence, the Dela- ware, the Susquehanna, and the Missouri Eiver. By reason of these gorges, and their accompanying destructive floods. Port Jervis, on the Delaware, and Mandan, on the Missouri, have become familiar names. Eeasoning from the nature of the case, what, then, must have been the scenes during the last stages of the great Ice age, when, through the months of July, August, and September, warm southerly winds and a glowing sun were combining to dissolve, with utmost ra- pidity, the vast masses of ice which still lingered in the country ! The channels were then compelled to carry off not only the annual precipitation, and the torrents of an oc- casional cloud-burst, but the stored-up precipitation which had been accumulating as glacial ice for/ thousands of years. Nor is this altogether theoretical. Though we have no DRAINAGE OF TEE GLACIAL PERIOD. 281 telegraph to span the distances of time separating us from those events, we have come into possession of signs as intel- ligible as the lines and dots of the Morse alphabet, and even more trustworthy. These floods along the lines of glacial drainage have left their marks, and their direction and ex- tent can be traced almost as readily as in the case of the present streams. Ascending the channel of the Mississippi above its junc- tion with the Ohio, one enters a region where it is bordered on each side by rocky blufEs, and finds himself in a valley of erosion whose main features were determined in preglacial times. Above Grand Tower, in southern Illinois, and as far north as St. Louis (a distance of about one hundred and fifty miles), the extreme margin of glacial deposits rests upon the east side of the river, and an unglaciated region is upon the west ; the width of the eroded valley being from five to ten miles, and its depth several hundred feet. Above St. Louis the valley gradually narrows, though it is still from two to eight miles in width, and about the same depth as below the city. At various places, along the sides of this eroded val- ley, the observer will find gravel terraces one or two hundred feet above the present flood-plain. These terraces are, in fact, the high- water mark of the closing floods of the Ice age. But the culmination of interest is reached on coming to the present junction of the Minnesota and Mississippi Elv- ers near St. Paul. From Fort Snelling, just above St. Paul, northward, the present Mississippi Piver is a comparatively recent stream, occupying a post-glacial bed. The true exten- sion of the trough of the Mississippi follows up the Minne- sota Kiver. The gorge of the Mississippi, leading from Fort Snelling up to Minneapolis, is scarcely a quarter of a mile in width, and is about two hundred and fifty feet in depth ; while the trough of the Minnesota is from one to four miles in width, and its rocky bottom is more than one hundred and fifty feet lower than the present bed of the stream. In the bottom of this broad valley, for a distance of two hundred and fifty miles, the Minnesota River wanders about from 282 THE ICE AGE IN NORTH AMERICA. side to side as a very insignificant thing, entirely out of pro- portion to tlie valley which it occupies. Nor does the Min- nesota River have its sources in the highlands, like the Mis- sissippi. Its head is in Big Stone Lake, in the midst of this eroded trough, and but a few miles south of Lake Traverse — the head of the lied River of the North — also in the same trough and on the same absolute level. The water from Lake Traverse sometimes flows into the other lake. In short, the troughs of the Minnesota and the Eed Eiver of the North are one and continuous, and the depression joining them, known as Brown's Yalley, is to the glacialist one of the most interesting spots on the continent. The following is Mr. Upham's description : Lakes Traverse and Big Stone are from one to one and a half mile wide, mainly occupying the entire area between the bases of the bluffs, which rise about one hundred and twenty- five feet above them. Lake Traverse is fifteen miles long ; it is mostly less than ten feet deep, and its greatest depth proba- bly does not reach twenty feefc. Big Stone Lake is twenty-six miles long, and its greatest depth is reported to be from fifteen to thirty feet. The portion of the channel between these lakes is widely known as Brown's Valley. As we stand upon the blufEs here, looking down upon these long and narrow lakes in their trough-like valley, which extends across the five miles between them, where the basins of Hudson Bay and the Gulf of Mexico are now divided, we have nearly the picture that was presented when the melting ice-sheet of British America was pouring its floods along this hollow. Then the entire ex- tent of the valley was doubtless filled every summer by a river which covered all the present areas of flood-plain, in many places occupying as great width as these lakes.* Similar signs of the drainage of that period are visible in the valleys which head within the glaciated region of the * "Proceedings of the American Association for the Advancement of Sci- ence," Tol. xxxii, 1883, pp. 216, 21'7. This glacial outlet through Brown's Val- ley and the Minnesota has been fittingly named, by Mr. Upham, River Warren after General G. K. Warren, who first described it. See map in Chapter XXI. BEAINAQE OF TEE GLACIAL PERIOD. 283 basin of the Ohio River. The terraces so constantly lining the water-courses in the Middle and Western States, and which formerly were attributed to the advance into these regions of the waters of the ocean, are readily accounted for by the action of the torrents set free by the melting of tbe ice during the closing years of the great Ice age. One of the striking confirmations of the glacial theory appears in the absence of terraces in the valleys of such minor streams as have their sources south of tlie glacial limits. For example, in Ohio the small streams in the southeastern part of the State, whose sources are outside of the glacial limits, present a marked contrast to the other streams of the State flowing south, as do also the streams flowing to the north and empty- ing into Lake Erie. The troughs of the Wabash, the two Miamis, the Scioto, the Hocking, and tbe Muskingum, with their tributaries, are all lined by gravel terraces, rising from fifty to one hundred and fifty feet above the present flood- plains, showing the enormous volume of the streams whacb flowed through them at the close of the period. The coarse- ness of the materia] in these terraces also bears witness to the violence of the currents. But between the mouth of the Muskingum, at Marietta, and the mouth of the Little Beaver, the streams entering tne Ohio are devoid of terraces, the ex- planation being that their sources lie outside of the glaciated limit, so that they had access neither to the accumulations of the glacial deposits, which furnished material for the terraces, nor to the floods of water that distributed it. To the east- ward, again, upon striking the streams whose drainage-basins lie within the glaciated limit, high terraces, containing north- ern pebbles brought from beyond the water-shed, begin again to appear, both along the margins of the streams in the gla- ciated area, and also through their whole course below the boundary. In the matter of terraces, likewise, the northern tributaries of the Ohio are in striking contrast with the southern. East of the AUeghanies the same contrast appears between the streams rising within the glaciated area and those outside DRAINAGE OF THE GLACIAL PERIOD. 285 of it. The terraces on the East Branch of the Susquehanna are much more marked than those upon the West Branch, the explanation being that the East Branch lies almost wholly within the glaciated area, while only a few of the minor tribu- taries of the West Branch come down from it. But wherever they do so come, as in the case of Pine, Lycoming, and Loy- alsock Creeks, a limited amount of drift from the far north is distributed along their banks, and deposited at. their junc- tion with the main branch of the river. The Lehigh and the Delaware are likewise marked by high terraces containing pebbles from the far north, while the Schuylkill Eiver, which lies just outside of the glaciated limit, has no such terraces. Thus, both by the method of agreement and of difference, we prove the connection of these terraces of the so-called " Ter- race epoch " with the gorged and gravel-laden streams of the ■great Ice age. Before speaking of the lines of glacial drainage farther east, it will be profitable to direct our attention to another class of closely connected facts confirming the theory of the glacial origin of these terraces. The larger part of the ma- terial contained in them is derived from the glacial deposits over the region through which the several streams flow. The hard fragments of granite, quartzite, and various metamorphic rocks from the region of Lake Superior and the. Canadian highlands are eminently fitted to withstand abrasion, and can be rolled by a torrent for long distances before being ground to powder, while the softer sandstones and shales of the newer geological formations would be comminuted by the attrition of a comparatively few miles' travel. Hence it comes about that the terraces of the Middle States are composed, in a pre- dominant measure, of material brought over the water-shed by the ice from the far north, and spread broadcast over the country, and thence collected by the streams of water and rolled along as far to the south as there was force in the cur- rent to move them, or through as great a distance as the hardness of the material of which they are composed would enable them to resist complete attrition. There is no more 286 THE lOE AGE IN NORTH AMERICA. interesting verification of an hypothesis anywhere to be fouud than that furnished for the glacial theory through the study of the character of some of these terraces at and below the glacial limit. Theoretically the terraces should, for the rea- sons just stated, be more prominent and consist of coarser material, just where the streams emerge from the glacial limit ; and such, from a wide collection of facts, is proved to be the rule. I have myself examined nearly all the streams thus emerging from the glaciated area between the Atlantic Ocean and the Mississippi Eiver.* In scores of places where streams thus emerge from the glaciated region- — in Pennsyl- vania, Ohio, and Indiana — their valleys are filled with an accumulation of water-worn northern drift, which, when fol- lowed downward, becomes gradually less in amount, as well as more water-worn, and finer in its constituent elements. This is notably the case in the Delaware Yalley, at Belvi- dere, N. J. ; in the Susquehanna, at Beach Haven, Pa. ; in the Conewango (as already described), at Ackley, Warren county ; in Oil Creek, above Titusville ; in French Creek, a little above Franklin ; in Beaver Creek, at Chewtown, Law- rence county ; on the Middle Fork of Little Beaver, near New Lisbon, Ohio ; on the east branch of Sandy Creek, at East Rochester, Columbiana county ; on the Nimishillin, at Canton, Stark county ; on the Tuscarawas, at Bolivar ; on Sugar Creek, at Beech City ; on the Killbuck, at Millersburg, Holmes county ; on the Mohican, near the northeast corner of Knox county; on the Licking Hiver, at Newark; on Jonathan Creek, Perry county ; on the Hocking, at Lancas- ter ; on the Scioto, at Hopetown, just above Chillicothe ; on Paint Creek, and its various tributaries between Chillicothe and Bainbridge ; and on the "Wabash, above New Harmony, Ind. ; to which may be added the Ohio Eiver itself, at its junction with the Miami, near Lawrenceburg. Some of these instances are sufficiently interesting and * " Glaciated Area of Ohio," in the " American Journal of Science," vol, cxxvi, 1883, pp. 1-14; "American Naturalist," vol. xviii, pp. '766-'76'7. BRAINAGE OF THE GLACIAL PERIOD. 287 instructive to warrant a special description. The upper part of the Ohio River, between Pittsburg and New Kichmond, in the vicinity of Cincinnati, lies entirely outside of the gla- ciated area, while nearly all of its northern tributaries rise within that area. In the happy phraseology of Professor Dana, the Ohio River becomes, therefore, the great distribu- ter, while its northern tributaries are the principal oontribu- tors, of terrace material. Now, it is observable that, wherever a large contributor of drift material comes in from the north, there is a great increase in the extent and height of the ter- races for some distance below, and the material of which the terrace is composed is coarser at these points. For example, in the neighborhood of Cincinnati, the Ohio is joined by the Little Miami, and twenty miles below, at Lawrenceburg, by the Great Miami. Throughout their entire course these tributary streams flow through a region deeply covered with glacial deposits. As a consequence, the terraces here are of great height and width. At Cincinnati, the upper terrace upon which the original city is built is one hundred and twenty feet high ; and at Lawrenceburg the valley, from three to four miles wide, is nearly tilled to a height of one hundred and twelve feet above the flood-plain, with a ter- race deposit clearly derived from the glacial floods of the Great Miami and its tributaries. Below this point the ter- races of the Ohio gradually diminish in height, and the ma- terial becomes finer, and more and more water-worn. Above Cincinnati there is a marked development of the terraces at the mouth of the Scioto, at Portsmouth ; and again, below Marietta, at the mouth of the Muskingum, where, opposite Blennerhassett Island, the terraces are in the neighborhood of one hundred feet above the present low-water mark. It is to be noted that both of these streams were so situated as to be among the largest contributors to the Ohio, both of glacial floods and glacial debris. But the most instructive place for the observation of this class of phenomena is to be found in Pennsylvania, at the junction of Beaver Creek with the Ohio River. From the 288 TEE IGE AGE IN NORTH AMERICA. mouth of French Creek, at Franklin, Pa., to the mouth of Beaver Creek, twenty-five miles below Pittsburg, a distance of about one hundred and fifty miles, no contributors of gla- cial material enter the Alleghany or the Ohio Kiver, and the course of the river-bed lies wholly in the soft, sedimentary deposits of the coal-measures. Consequently, while this por- tion of the stream contains terraces with northern drift brought into the Alleghany above Franklin, they are of diminishing height, and contain a constantly diminishing amount of material from the glacial drift all the way down to the mouth of the Beaver. At the mouth of the Beaver there is a sudden enlarge- ment of the Ohio terrace, and it rises at once to a height of one hundred and twenty feet above the river. Upon the lower side of the Beaver, in the angle between it and the Ohio, down-stream, this terrace is very extensive, and the material very coarse, the terrace being, indeed, largely built up of pebbles and bowlders from a few inches to two feet or more in diameter, and all thoroughly rounded. On the opposite side of the Beaver, in the angle between its mouth and the upper portion of the Ohio, there is, for a limited distance, a terrace of equal height, but of entirely different . composition from that upon the lower side. The terrace u])on the upper side consists of fine material, being mostly sand and gravel derived from the coal-measures, through which the Ohio itself has cut its way. Pebbles from the northern drift are rare, and the local origin of the material is manifest at a glance. What, now, makes this difference between these terraces upon the opposite sides of this small tributary of the Ohio ? A glance at the map will show.* The Beaver River emerges from the glaciated region only a few miles to the north of its junction with the Ohio. The larger part of its drainage-basin lies in portions of northeast- ern Ohio and northwestern Pennsylvania, which are deeply covered with the terminal deposits of the continental ice- ' See map, p. 131. DRAINAGE OF THE GZAOTAL PERIOD. 289 sheet. The floods characterizing that period had access to an unlimited amount of material, which was easily swept into the current, and rolled down the torrential bed toward the Ohio Biver. Upon reaching the Ohio, the combined cur- rent of the two streams in the larger valley would have far less power of transportation than the constricted current in the channel of the glacial tributary. The bowlders would, therefore, be deposited at the mouth of the Beaver, where it joins the Ohio, and, owing to the influence of the current of the Ohio itself, would be carried below the junction of the two rivers. Hence it is that we find so many glacial bowl- ders below the junction, and so few above it. The accumu- lation of a terrace of an equal height above the junction, but consisting of fine and local material, is also what would be required by theory as well as what is found to be the case in fact. Thus we have, in this single instance, one of the best possible verifications of the glacial hypothesis. Of the glacial terraces on the Delaware Kiver, from the Water-Gap to Trenton, N. J., there will be occasion to speak more fully when treating of the subject of man's relation to the Ice age in North America. ISTor can we more than allude to those which line the Mohawk and the troughs of the Hudson and its tributaries. It is suflBcient to say that the passengers upon the New York Central Eailroad can satisfy themselves of the existence of these terraces east of Little Falls, N. Y., and between Schenectady and Albany, by merely looking out of the car- windows toward the north ; and a moment's reflection upon the topography of the coun- try will show that the terraces of the Mohawk and the up- per Hudson are much more recent than those of the Susque- hanna and the Delaware, since glacial streams could not have occupied the Mohawk until after the ice-front had retreated from northeastern Pennsylvania and the highlands of south- ern New York, in which the drainage-basins of the Susque- hanna and the Delaware River are situated. When, there- fore, the glacial floods of the Mohawk were at their height, the Delaware and the Susquehanna had been relieved of 19 290 THE ICE AGE IN NORTH AMERICA. their excessive burdens, and liad subsided to something like their present volume. As illustrating the capacity of the glacial theory to ex- plain the otherwise unaccountable facts connected with the recent changes in the drainage of the glaciated region, atten- tion is directed to two or three interesting localities east of the Alleghanies, where the dry beds of abandoned streams have been discovered. We will first consider the outlets of an interesting glacial lake which temporarily occupied the upper part of Contocook Valley in Hillsborough county, N. H., the details concerning which were furnished as early as 1878 by Mr. Upham, in the New Hampshire Greological Eeport. The Contocook River now empties into the Merrimack a little above Con- cord, and flows in a direction north-northeast. As a conse- quence, the present outlet was, toward the close of the Gla- cial period, obstructed by ice some time after it had melted off from the southeastern portion of the valley. During that period a lake was held in the portion of the valley freed from ice, at a height sufficient to turn the drainage tempo- rarily to the south and southeast. At first the drainage was over the water-shed in Rindge, through Ashburnham and Winchendon, Mass., and thence into the Connecticut. The reality of this line of drainage is evidenced by the exten- sive kames and gravel deposits extending from the Conto- cook Valley through the towns of Rindge and WinchenSon. When the ice had withdrawn a little farther north, an outlet was open to the southeast into the Souhegan River, and thence into the Merrimack. The evidence here is also con- clusive that, for a period, a stream of water eighty feet deep poured through this pass, and the lake formed in front of the ice was in its greatest extent thirty miles long, and from two hundred to two hundred and fifty feet in depth. The evidence of this remains in delta terraces at that level formed at various points where streams came into the lake. Another instance is in Grafton county, N. H., on the line of the Northern Railroad, between Grafton Centre and East DRAINAGE OF THE GLACIAL PERIOD. 291 Canaan, on the water-parting between the Merrimack and tlie Connecticut, where there is to be found the dry bed of a river which for a time flowed through a pass from the Con- necticut Valley into the Merrimack, but which is five hun- dred feet above the valleys. Here upon this mountain axis, in central New Hampshire, nine hundred feet above the sea, are numerous and large water-worn circular cavities in the rock, technically known as pot-holes, such as are formed in shallow rapids, wherever gravel and pebbles become lodged, first, in some natural slight depression, and then, through the whirling motion given them by the running water, these continue to wear a symmetrical depression so long as the supply of water continues, or until a channel has been cut through. Pot-holes may be seen in the rapids of almost any rocky stream, with the gravel and pebbles, which do the im- mediate work when set in motion, still partially filling them. Such pot-holes exist in the anomalous position mentioned in Ifew Hampshire, where no present stream could by any pos- sibility be made to flow. One of them, measured many years ago by Jackson, was eleven feet deep, four and a half feet in diameter at the top, and two feet at the bottom, and, when discovered, was tilled with earth and rounded stones.* The explanation of this phenomenon furnished by Mr. Upham, while engaged on the New Hampshire Survey, is as follows : The ice of the Connecticut Valley, being farther from the glacial front, lingered considerably longer than that in the Merrimack Valley to the southeast, so that for a con- siderable period the drainage from the ice-front in the south- eastern part of Grafton county was compelled by the ice- barrier on the west to flow over this depression into the Mer- rimack basin, thus furnishing exactly the conditions neces sary for the production of pot-holes and such other marks of water-action as have so long puzzled geologists at this point. Similar pot-holes, to be accounted for in like manner, have recently been described near Archibald, in Blakely * " New Hampshire Geological Report," vol. iii, pp. 63-li6. 292 THE ICE AGE IN NORTH AMERICA. township, Lackawanna county, Pa.* The principal one is from fourteen to seventeen feet in diameter at the top, and is forty feet deep— the sides being very smooth. The de- pression is worn through strata of sandstone, shale, and coal. The pebbles which did the wearing were still in the bottom of the hole when it was discovered, and are mostly of foreign origin, though some of them consisted of pebbles cut from the coal-bed itself. The elevation is eleven hundred and twenty-nine feet above tide, and no explanation seems possible except that which assumes that a stream of water was kept running in that position for a limited period by ice-barriers. In passing, it is interesting to remark that the study of the glacial deposits in the coal region becomes of great prac- tical interest from the relation of their buried channels to mining industries. Not only is there money to be saved by knowing the depth of the till, and the inequalities of its distribution, but the lives of the miners are seriously jeopard- ized by ignorance upon this point. On December 18, 1885, at Nanticoke, near the vicinity of the pot-hole just de- scribed, one of the most shocking mining disasters on record occurred, from miscalculating the course of a buried pre- glacial channel, which was penetrated by the miners in an unexpected place, causing a flood of quicksand, mud, and bowlders to fill the mine and immolate twenty-six miners beyond hope of rescue. Another instance of glacial drainage worthy of record is reported by Professor J. E. Todd from the Missouri coteau in Dakota. Crow Creek flows westward, and enters the Mis- souri in Buffalo county, heading well in the terminal mo- raine : f Two of its principal branches lead us into the heart of the great interlobular moraines, the Eees, and the range of which Turtle Point is the head, then by unmistakable channels through them to the inner side of the moraine and out upon * " Annual Report of the Pennsylvania Geological Survey," 1885, pp. 615-620. f Sec map, p. 188. DRAINAGE OF THE OLAOIAL PERIOD. 293 tlie great ice-sheet itself. It produces strange sensations to pass up those dry, fiat-bottomed valleys, with a steep bank on either hand fifty to one hundred and fifty feet in height, almost built of bowlders ; hug e cone s of gravel, evidently formed in ancient eddies, here and there in the valley ; simi- lar valleys joining it now and then. You press on, wondering where the beginning can be, for your map tells you that there are streams which must cut right across its course if it con- tinues as far as you might judge from its width. You press on eagerly, you note the banks rapidly subsiding, but the channel you tread still preserves its gradual rise, then suddenly you come out upon the face of the range, and a magnificent view of the plain, two hundred to three hundred feet below, bursts upon you. You look for the inclined plane which by easy steps has brought you to this altitude, and find it ending abruptly with the face of the hills. You realize, as never before, the mass of ice which once must have occupied the expanse before you. You can see that stream, scores of yards in width, leaving its icy banks, now vanished in thin air, for the stony ones which still remain.* It seems clear, also, that the disturbing effects of the great ice-sheet upon the drainage of the Northwest will account for the numerous deserted river-valleys described by Dr. G. M. Dawson in the part of British America lying between the Lake of the Woods and the Eocky Mountains. Here the conditions were somewhat peculiar. The natural drain- age is down the flanks of the Eocky Mountains eastward to the Red River Valley. The Saskatchewan River drains the northern portion of the territory, while the Assiniboin with its branches, the Qu'Appelle and Souris, drains the southern portion — the drainage-basin of the Souris joining that of the Missouri in Dakota. The Pembina River, a much smaller stream, empties into the Red River near the boundary-line, considerably south of the Assiniboin ; and the Sheyenne still farther south. * " Proceedingis of the American Association for tiie Advancement of Sci- ence," vol. xxxiii, 1884, p. 391. 294: THE ICE AGE IN NORTE AMERICA. Witli the interpretation which the present discussion has put upon the facts, the following is the order of events : Dur- ing the farthest extension of the ice to the vicinity of the Rocky Mountains, the South Saskatchewan was compelled to flow around the front of the ice-sheet to join the Milk Kiver at the boundary-line near the one hundred and tenth meridian, and thence into the Missouri. A great dry coulee, a portion of which is occupied by a large saline lake known as Peeko- pee, is a marked feature connecting these streams at the present day.* Coming eastward to the one hundred and second meridian, the Riviere des Lacs seems, without doubt, to have been the line of drainage for the Souris River for a distance of sev- enty-five or eighty miles. Characteristically enough, this ends northward, near the Souris River, "in a broad dry coulee, which shallows and dies away in a strip of bowlder- covered ground, which stretches northward toward the Souris River, five miles distant, and is somewhat lower than the general surface of the plain." At this time there was a lake-like expansion of water in the Elbow of the Souris, covering Renville, Ward, and Mc- Henry counties, Dakota, the evidence of which is still plainly seen. This lake, again, was forced to seek a southern outlet, which it found through a coulee in McHenry county into the head of the Sheyenne River, and thence followed its winding course to Lake Agassiz, near Fargo, where there is an immense delta of river gravel. Coming still farther east, the Pembina River occupies a valley very much larger than its present demands ; and, at its junction with the Red River, there is also an immense gravel delta, indicating it as a line of drainage at one time of a far larger area than now. Upon following the valley of the Pembina up, it is found to continue through the Pelican Lake to the Elbow of the Souris, near the one hundredth * " Report on the Geology and Resources of the Region in the Vicinity of the Forty-ninth Parallel," pp. 262-268. DRAINAGE OF TEE GLACIAL PERIOD. 295 meridian, and twenty-five or thirty miles south of the As- siniboin. Professor Hind finds evidence that this was the outlet temporarily not only of the Assiniboin but, through the Qu'Appelle and the Eiver that Turns, to the South Sas- katchewan. Mr. Upham writes me that he has followed this old valley for one hundred and twenty-five miles as far as Birtle, in Manitoba. According to Professor Hind, the length of the valley of the Qu'Appelle, from Birtle up to the Saskatchewan, is two hundred and sixty-eight miles in direction northwest by southeast. The valley is uniformly about one mile wide, and from one hundred and ten to three hundred and fifty feet below the general level, and eighty- five feet above the present level of the South Saskatchewan, the descent being four hundred and forty feet from the Saskatchewan to the Assiniboin. The inclosing bluffs con- sist mainly of till, and the whole trough is characterized by numerous long, shallow lakes. These lines of marginal drain- age are readily explained upon the glacial hypothesis here maintained, and are a strong proof of that hypothesis. As the ice receded, more northern outlets were opened, and these temporary channels were naturally abandoned. We have already spoken of a similar line of mai-ginal drainage entering the Cheyenne Eiver, in Dakota, a little above its mouth, probably connecting with the Moreau west of Fox Ridge. It is not unlikely also that the Missouri River occupies a post-glacial course marked out for it when it was the line of marginal drainage during the long period that the ice-front extended to the Missouri coteau, and deposited that enormous moraine. Certain it is that the valley of the James River, in the central part of eastern Dakota, is much broader and lower than that of the upper Missouri, and seems the natural northern continuation of the Missouri Yalley from below Yankton. Coming still farther southeast, we find that the Minnesota River makes a sharp turn to the north at Mankato, and so is favorably situated for having its drainage reversed while the ice rested over the counties about its junction with the Mis- 296 THE ICS AQE IN IfOBTH AMERICA. sissippi near St. Paul. The facts are found to be according to the programme. There is abundant evidence of a tempo- rary lake, covering the territory of Blue Earth and Faribault counties, which emptied through a channel known as Union Slough, about eight miles long and from one eighth to one fourth of a mile wide, with bluffs from twenty to thirty feet in height, which connect Blue Earth E-iver with the Eastern Branch of the Des Moines in Kossuth county, Iowa. This brief summary of facts concerning preglacial and glacial drainage is scarcely more than enough to excite the curiosity of the reader. The subject is boundless, and well calculated to interest local observers in all parts of the coun- try. But the interpretation of the local facts will be greatly aided by such a general summary as is here given. OHAPTEE XIY. KAMES. The word " kame " has already been defined as a local term applied to the sharp gravel ridges which abound in various parts of Scotland, and which in Ireland are called " eskers," and in Sweden " osars." As Mr. Geikie's work on " The Great Ice Age " has given currency to the Scotch name, and as the word has been adopted by those who have investigated this class of formations most fully in America, it seems best to continue its use, though either of the other names is more . euphonious. This class of ridges was first described in this country in 1842 by President Edward Hitchcock. Speaking of the gravel deposits in Ando^^er, Mass., known as Indian Ridge, he says they are " a collec- tion of tortuous ridges and rounded and even conical hills with corresponding depressions between. These depressions are not valleys which might have been produced by run- ning water, but mere holes, not unfrequently occupied by a pond." * The fuller description of their composition by Mr. James Geikie is as good for America as for Europe : The sands and gravels have a tendency to shape themselves into mounds and winding ridges, which give a hummocky and rapidly undulating outline to the ground. Indeed, so charac- teristic is this appearance, thab by it alone we are often able to mark out the boundaries of the deposits with as much precision as we could were all the vegetation and soil stripped away and * " Transactions of the American Association of Geologists and Naturalists," 1842. FAMES. 299 the various subsoils laid bare. Occasionally, ridges may be tracked continuously for several miles, running like great arti- ficial ramparts across the country. These vary in breadth and height, some of the more conspicuous ones being upward of four or five hundred feet broad at the base, and sloping upward at an angle of twenty -five or even thirty-five degrees, to a height of sixty feet and more above the general surface of the ground. It is most common, however, to find mounds and ridges con- fusedly intermingled, crossing and recrossing each other at all angles, so as to inclose deep hollows and pits between. Seen from some dominant point, such an assemblage of kames, as they are called, looks like a tumbled sea — the ground now swelling into long undulations, now rising suddenly into beau- tiful peaks and cones, and anon curving up in sharp ridges that often wheel suddenly round so as to inclose a lakelet of bright clear water.* From this description it will be seen that there are some remarkable resemblances between kames and terminal mo- raines, since both of them are characterized by confused hummocks and tortuous ridges of glacial debris, connected with numerous bowl-shaped depressions, often containing N.W, PiQ. 83. —Section of kame near Dover, New Hampshire. Length, three hundred feet ; height forty feet ; base, about forty feet above the Cochecho Eiver, or seventy-five feet above the sea. o, a, gray clay ; 6, fine sand ; c, c, coarse gravel containing pebbles from six inches to one foot and a half in diameter ; d, d, fine gravel. (TJpham.) lakelets. But in other respects there is a marked difference between them. In the first place, the material of which kames are formed is ordinarily much finer and more water- worn, and shows more abundant signs of stratification than that of which terminal moraines are composed. Secondly, while the terminal moraine forms a ridge at right angles to the motion of the glacier, and marks the limit of its exten- * "The Great Ice Age," pp. 210, 211. 300 THE ICE AGE IN NOttTH AMERICA. sion during a prolonged period, the kames approximately coincide in direction with the lines of glacial striae. A large part of New England is covered with katne deposits, arranged, in general, along the main lines of present drainage, with merely such anomalous exceptions as can readily be ex- plained by the interference which the ice itself offered to the course of the floods which characterized the last stages of the Glacial period. aoUTH SIDE NORTH BIDE. Fi. 84. — Sections of a kame at Bennington Station, New HampsWre. Scale about forty feet to tlie inch. Tlie upper figure shows a simple transverse section ; the lower figure is directly transverse on the right side, but longitudinal on the left. Counting from the top the strata are ; 1, coarse yellow gravel, with pebbles up to eight inches in diame- ter, thickness, three to five feet ; 2, fine sand, three to five feet ; 3, coarse, dark gravel, containing pebbles up to one foot in diameter, three feet ; 4, fine sand, ob- scured at the bottom by crumbling of the bank, four to eight feet ; a. a, downfall of strata with irregular, broken, steep slope, against which lies an accumulation of sand ; b, depression of two feet, similar to the foregoing ; p, fault, seen only on the south side, dislocation of strata, six inches. (Upham.) The most satisfactory conclusion with regard to the origin of kames is that they mark, in the glaciated region, lines of drainage during the closing stages of the Ice age. It is evi- dent, from a moment's reflection, that the streams of water resulting from the annual precipitation, combined with that from the wasting of the ice during these closing stages, must have been enormous, and may very likely have flowed in channels quite dift'erent from those' chosen after the ice had completely melted away. Of course, these glacial streams hiust, in the main, have followed the great valleys ; but many of the minor valleys were, at that time, so obstructed that the streams might disregard them and take a more di- rect route over the ice through the open channels and long tunnels which must then have existed. Those familiar only 302 THE IGE AGE IN NORTH AMERICA. with the contracted glaciers of the Alps are scarcely prepared to appreciate the extent to whicli currents of water flow over the larger glacial masses and rearrange and transport the su- perficial material collected upon them. The " subglacial " streams also are not always strictly subglacial, since they often flow through tunnels which are midway between the top and the bottom of the ice-mass. In the Muir Glacier, Alaska, for example, the two streams issuing from the ice- front near the sides of the glacier are several hundred feet above the level at which the two streams emerge near the center of the channel. There, also, streams of water of more or less size can occasionally be seen pouring out from the perpendicular front of the ice a hundred or more feet above the surface of the inlet. Nor is it any uncommon thing to see icebergs move off with water -worn tunnels in them which are still well filled with gravel and pebbles. In the various depressions in the surface of the glacier also, where at times extensive lakes of water are formed, there is much accumulation and assortment of earthy material far back from the terminal margin of the glacier. We will now endeavor briefly to reproduce the conditions in New England near the close of the Ice age, in order to see how the facts fit into the theory just enunciated. The main north-and-south valleys of New England are. now drained by the St. John, the St. Croix, the Penobscot, the Kennebec, the Androscoggin, the Merrimack, and the Con- necticut Rivers, with various smaller subordinate drainage- basins, such as the Machias, the Saco, and the Piscataqua. The larger valleys are also joined by various subordinate ones, tributary to them, running in various directions con- formable to the general contour of the country. But the present course of the rivers is not necessarily determined at evei'y point by barriers of any great height. For exam- jile, there are no high barriers separating the northeastern portion of the Penobscot drainage-basin from the sources of the St. Croix and Machias Eivers. South of the Rangeley Lakes, also, where Ellis River joins the Androscoggin, it is FAMES. 303 only a barrier of two or three hundred feet which causes the present deflection of the river to the east, through Lewiston to Brunswick. The great bend made by the Merrimack Eiver at Lowell, Mass., is also caused by a glacial deposit to the south of only fifty or sixty feet in height. It is easy to see that, during the period of most rapid retreat, when the waters of the wasting ice-sheet over New England were seeking their ultimate channels, the lower portion of the ice itself was an important element in deter- mining the minor deflections in these lines of drainage. An ice-barrier of a few hundred feet in the Penobscot, between Passadumkeag and Mattawamkeag, would force the drainage of the Aroostook region into the valley of the Machias, and, in the predominance of the mountains from which the west- ern branches of the Penobscot Piver descend, we have a cause favoring such an extension of the ice as would produce the results indicated. In the case of the Merrimack Eiver, the fact that, from Lowell to Newburyport, it flows in a northerly direction would also furnish a probable ice-barrier which for a time would drive the drainage of this basin di- rectly southward from Lowell and Lawrence toward Boston. It is not necessary to go into, all the details concerning the intricate network of kames which mark the lines of . drainage over New England, when ice-barriers to so great an extent directed the flow of the glacial torrents. The facts are impressive. Individual kames can be traced for long dis- tances, sometimes a hundred miles or more. The main lines in New England are shown on the accompanying map, be-, ginning on the eastern side of Maine.* A few points merit particular attention. The Connecti- cut River Valley, from its sources to the Massachusetts line, contains the remnants of what seems to be a pretty continu- ous kame, but which has been largely eroded, and in many cases covered up by subsequent deposits of river-silt. Almost * See also " Kames and Moraines of New England " in " Proceedings of the Boston Society of Natural History," vol. xx, p. 211 et seq. 304 THE ICE AGE IX NORTE AMERICA. everywhere we find illnstrations in the partial burying of kames by such river-silt that the deposition was previous SCALE OF MILES Slrittltorai t'j., Eaar%NA'. Fie. 86.— The kames of Maine and southeastern New Hampshire. The extension from New Hampshire can be seen in Fig. 83. (Stone.) to and independent of the present streams. For example, the Merrimack, between Lowell and its mouth, is crossed at right angles by two or three lines of kames, which descend into the valley from one side and come out upon the hills on the other. "While crossing the valley these are partially, EAME8. 305 and in some places completely, buried beneath the river-silt which forms the present flood-plain. In one ease, a few miles below Lowell, the end of this ridge, completely cov- ered with river-silt, may be seen where the river has cut across the old barrier. Professor Charles Hitchcock gives a similar section of a buried kame in Hanover, N. H., though . ■!- r-DT J -o ""° S S Kanover Af. Coll, Delta of 3 Tcmice ofBloody-Br., A A -i ^ ^ « " common. farm. Mink Br, S Norwich village, 525. S, §! « « M Si S4S. fioo, 564. .t3*».>Ss*^ 350ft. above sea. Pia. 87.— Section across the Connecticut Valley, from Norwich, Vermont, to Hanover, New Hampshire, distance three miles. The kame is nearly covered by later river silt. (Upham.) in this case it is parallel with the river, and not, as in the other, at right angles to it.* Inasmuch as the interpretation of the facts in the valley of the Connecticut is open to some question, and as the de- cision with respect to them will have an important bearing on our whole conception of the closing scenes of the Glacial period, it will be worth while to consider them more fully. Mr. Upham, in his survey of the Connecticut Valley,! discovered what he considered to be a line of kames extend- ing throughout nearly the whole length of the valley, though it had been much eroded in places, and in others was partially or completely buried by river-silt ; but of the character of the deposit as a trae kame he felt quite confident — that is, he con- sidered that the line of gravel ridges which he found winding from side to side down this valley were deposited as the ice retreated, after the manner we have described in channels and tunnels formed near the front. In this view these ridges in age are intermediate between the till and the regular river * "Proceedings of the American Association for the Advancement of Sci- ence," vol. xxxi, p. 328 ; also Upham, in " Geology of New Hampshire," vol. iii. f See "Geology of New Hampshire," vol. iii, pp. 3-177 ; also, "American Journal of Science," vol. cxiv, 1877, p. 459. 20 306 TEE ICE AGE IN NORTH AMERICA. terraces — being newer than the till and older than the ter- races. On going over the ground in 1881 with Mr. Uphain's notes in his hands, Professor Dana concluded that what Mr. Uphani had called kames were in reality a portion of the resular terrace formation. In Professor Dana's view, the reputed kames are merely tlie coarser part of the terrace material accumulated in excessive amount in the larger val- ley wherever tributary streams brought into it their heavier burdens from the higher land. On this theory, the height of the floods in numerous localities must have been between two hundred and three hundred feet above low water in the river ; for in various places these deposits are at that height above the river. But upon the supposition that they are Sccregated veins itLlower poitioiu 375 ft- atovc sea. I IQ. 88.— Section east from Ledyard Bridge, Hanover, New Hampshire, showing segre- ftited veins in the lower portion. Length, about seven hundred feet. Height of ame above the river, one hundred and forty feet. (Upham.) kames, deposited when the ice itself formed barriers to keep the streams in various abnormal positions, the glacial floods would not need to be more than from one hundred to one hundred and fifty feet in height, since that is all that is required for the deposition of the highest river-silt which occurs. It must be confessed that Professor Dana's estimates of the size of the Connecticut Hiver floods at that time are somewhat startling, even with all the changes of level for which he provides in his theory.* For, after reducing, by reason of the Champlain depression, the gradient of the stream during the close of the Ice period by one third, the 'American Journal of Science," vol. cxxiii, 1882, p. 19 KAMES. 307 slope of the surface of the Comiecticut would still have been more than one foot per mile. This, in a torrent 2,500 feet wide, with a depth of 140 feet, would produce a current of eight miles per hour on the surface and of six miles on the bottom. "With this size of the flood, the rate of discharge would be about four hundred cubic miles of water per an- num ; whereas, at the present time, the total discharge of a year is only about live cubic miles. To cause this enormous rate, Professor Dana supposes that, for a short period, the Connecticut glacier melted at the rate of more than a cubic mile per day. As he estimates the area of this drainage-basin to be about 8,500 square miles, this would imply that at times as much as eight inches per day melted from this surface. This rapid rate of removal in summer is not, however, sup- posed to continue for a long period — probably less than five years. Professor Dana supposes that, at that time, the long tunnels worn in the glacier by the Connecticut and its tribu- taries, wben they existed as subglacial streams, had become open channels in the ice — the tunnels having become so much enlarged that their roofs had fallen in. Thus exceptionally large areas of ice were exposed to the melting action of warm winds and sunshine. While tbore can be no doubkthat Professor Dana is, in the main, correct in his theory concerning the progress of events in the Connecticut Valley, we can not but feel tbat, if he had approached the field from a stiidy of the kames in the eastern part of New England, he would have seen less occasion to criticise Mr. Upham's interpretation of the facts. As we understand it, the Connecticut River is simply a kame-channel, modified by the remarkable size and the pecul- iar conditions of the valley. To the east, where the valleys were less pronounced and the conditions less uniform, the kames themselves are more prominent, not only because of their relative size at first, but because they have suffered less from erosion, and the ground upon which they stand has not been silted up by the temporary bodies of water which sur- round them, so much as in the Connecticut Valley. 308 THE ICE AGE IN NORTH AMERICA. The levelly stratified plains of sand and gravel which spread ont around the southern end of the kame s^'stems, and which to a greater or less extent border their margin through- out their entire length, should not be passed without notice, sand, '<>%'i^,-y'i they are Fig. 09 —Stratification of the loess in a railway cut at Plattsmoulh, Nebraska, at a deptli of eighty-four feet from the surface. (From photograph furnished by Dr. A. L. Child^ of Kansas City, Mo.) (Chamberlin.) (ITnited States Geological Survey.) always very much scattered, many bones belonging to the same individual being rarely found together, but seeming to have drifted widely apart. It is not easy to see how the cumbrous bones of the mammoth could have been widely separated in a subaerial deposit." * * "American Journal of Science," vol. cxviii, 1879, p. 110. 364 THE WE AQE IN NORTH AMERICA. It is to President Chamberlin, again,* that we are in- debted for the most careful study of this problem in the Mis- sissippi Yallej. According to him, the loess is limited pre- dominantly to the river valleys, and the belt is ordinarily not over forty miles in width. As a rule, also, it is thicker and slightly coarser in character near the banks of the great riv- ers, and there shows some signs of stratification. A comparison of loess with sand and clay reveals some interesting facts. The loess is intermediate in size and fine- ness. When the particles are suspended in water, the loess settles much more rapidly than clay — as much of the loess settling in four hours as of the clay in thirty-six hours. Out of 150,000 particles of loess examined under the microscope, 146,000, or about ninety-seven per cent, were less than -005 of a millimetre in diameter. A grain of sand one millimetre in diameter is considered fine. But this would make 200,000 particles of loess of the size mentioned, and 100,000 particles of the fineness of a large part of true clay. The size of the largest particles of loess noted by President Chamberlin was about one tenth of a millimetre in size, and consisted of scales of mica. The loess of Vicksburg is a little finer than that of Kansas City, and both a little finer than that from the Khine. The particles of loess were found to be angular and irregular. " Sharp comers and rough surfaces are the rule, and any ap- proach to regularity or smoothness is the exception." In the vicinity of the great rivers the grains are coarser than at points removed from them. In chemical composition the difference between loess and true residual clays is not very striking ; but, as compared with glacial clays, occurring in the till, there is a marked difference in several respects. The gla- cial clays have far less amount of silica and alumina, and a far larger amount of calcic and magnesian oxides and of carbonic dioxides. The glacial clays are evidently the result largely of mechanical abrasion ; but the loess would seem to consist in larger proportion of material resulting from chemical dis- * " Driftless Area of the Upper Mississippi Valley," pp. 278-307. TEE LOESS. 366 integration. But the amount of kaolinized products is nearly four times as great in the residuary clays as in glacial clays or loess. The altitude of the loess deposits in the Northwest is by no means uniform, and its variations present great theoretic difficulties. In the lower part of the Mississippi VaUey it is limited to a height of about two hundred and fifty feet. In the upper Mississippi Yalley, however, it rises to a height of seven hundred feet above the bed of the river. To account for this, the first impulse would be to suppose a general de- pression of the northern region. But this theory is excluded by various irregularities in this region itself. For example, the loess rises much higher upon the west side of the upper Mississippi than upon the east side, especially in the vicinity of the drif tless area of Wisconsin. To explain this it is necessary to resort to a rather com- plex theory. In part, perhaps, the peculiar distribution of loess is attribiitable to a period of general northerly depres- sion during the Glacial epoch. This apparent depression, however, was probably not caused wholly by an oscillation of the crust of the earth itself, since it is shown that it may be due in some degree to the attraction of the ice which had accumulated to the north. Mr. K. S. Woodward, of the United States Geological Survey, has worked out the prob- lem of attraction and its influence in producing a higher level of water at the north, on the supposition that the ice-sheet was ten thousand feet thick, and covered an area to the north 2,600 miles in diameter, and finds that the possible influence of such attraction might change the water-level at the ice- margin thirty-eight degrees from its center as much as 573 feet. But the elevation of the loess attains a height in Iowa and Wisconsin of 1,285 feet. This theory of change of the water-level by glacial attraction, therefore, though not ade- quate, very evidently comes in for a part of the credit of pro- ducing the puzzling facts relating to the deposition of loess.* * "Driftless Area of the Upper Mississippi Valley," pp. 291-301. 366 TEE IDE AGE m NORTH AMERICA. A noticeable phenomenon west of the driftless area in Wisconsin and Iowa is that the loess is thickest near the east- ern margin of the ice-lobe which passed southward through central Iowa. The distance from the driftless area to the Missouri Eiver, or the width of this lobe in the latitude of Milwaukee, is not far from three hundred and forty miles. It is urged that the presence of an ice-lobe over that region may have had influence in elevating the loess deposits. Since attraction varies inversely as the square of the distance, such a mass of ice near at hand would slightly raise the water along its margin. The objection to this is that the wider ice-lobe to the east did not produce corresponding iniluence. Among the more fruitful supplementary hypotheses brought in to aid in thfe solution of this part of the prob- lem is to be mentioned that of glacial dams similar to those already alluded to in Europe. In this country the principal field in which they may have existed is where it is most needed, namely, west of the Mississippi. An examination of our map will at once suggest that it is possible for much of the loess in northeastern Kansas and eastern Nebraska to have accumulated in temporary lakes formed by glacial dams across the mouth of the Kansas and Platte Rivers, or even of the Missouri itself. How far this cause may have operated remains yet to be determined. So complicated are the facts pertaining to the loess as they now appear, that it is not likely that, for some time to come, investigators will arrive at a perfect agreement con- cerning its manner of deposition. Those, however, who have most attentively studied glacial phenomena may be par- doned if they work the glacial hypothesis for all that it is worth. The existence of the vast body of ice which covered the glaciated area introduces a very complicated and efficient cause which it is exceedingly difficult to eliminate from the problem ; and it is as allowable for the glacialist to take refuge in supposed ice-dams, where their existence is possible and can not be disproved, as for the ordinary geologist to sup- pose vast orographic changes. The study of the loess has THE LOESS. 367 not come so much within my own field of observation as other glacial deposits have done. Still, in southern Indiana and Illinois, and in western Iowa and eastern Nebraska, I have been able to study many typical regions of this deposit ; and the more attention I have given to the subject, the more I have been led to magnify the agencies of the Ice period in producing the results both positive and negative. I have come, therefore, to set an increasingly high estimate upon the suggestions of Mr. Upham, made after he had concluded his survey of the terminal moraine in Minnesota, Iowa, and eastern Dakota : When the ice-sheet extended to the Coteau du Missouri, the Coteau des Prairies, and the Kettle-Moraine, the floods formed by its summer meltings were carried southward by the present avenues of drainage, the streams which occupied the areas between its great lobes in order from west to east being the Big Sioux, Mississippi, and Wisconsin Elvers. The vast glaciers which were gathered up on the Rocky Mountains, and the ice-fields which sloped downward to their termination at the coteaus and the moraine north and east in Minnesota and Wisconsin, supplied every summer immense floods laded with silt, sand, and gravel, that had been contained in the melting ice. Very extensive deposits of modified drift were thus spread along the course of the swollen Missouri and Mis- sissippi. The Orange sand and gravel, described by Professor E. W. Hilgard and others in the lower Mississippi Valley, ap- pear to have been deposited in this way, but during the earlier Glacial epoch, when an ice-sheet reached in Dakota beyond the Missouri River to a termination forty miles west and twenty miles southwest of Bismarck, into northeastern Kansas, half-way across the State of Missouri, and nearly to the Ohio River. In the closing stages of this epoch, and during the time succeeding, till the date of the terminal moraine of the coteaus, and especially at the final retreat of the ice-sheet of this later epoch, the deposition of the overlying, finely pulverized, are- naceous and calcareous silt, called the Bluff formation, or loess, took place. This covers considerable areas along the Mis- 368 THE ICE AGE IN NORTE AMERICA. sissippi from southeastern Minnesota to its mouth ; but its greatest thickness and extent are found in the basin of the Missouri Kirer from southern Dakota to its junction with the Mississippi, and upon the region crossed by the Platte or Nebraska Eiver, its longest tributary from the west, which takes its head-waters from a large district of the Rocky Mount- ains. The continuity of this formation from the borders of the ice-sheet and the glaciers of the Eocky Mountains to the shores of the Gulf of Mexico, the absence from it of marine shells, and the presence of land- and fresh-water shells, indicate that its deposition was by slowly descending floods, uplifted upon the surface of this sediment which was being accumu- lated during eyery summer through a long epoch, in the same manner that alluvium is now spread upon the bottom-lands of our rivers at their times of overflow. The occurrence of the loess in Guthrie, Carroll, Sac, and Buena Vista counties * in Iowa, covering the region next west of the terminal moraine, with its surface fifty feet above these drift-hills and one hundred above the undulating area of till adjoining their east side, proves that during the time of deposition of this part of the loess the ice- sheet extended to this limit, and was a barrier preventing the waters by which this sediment was brought from flowing over the lower area of till that reaches thence east to the Des Moines River. When the ice-sheet retreated beyond the water- shed of the Missouri basin, the principal source of these floods and their sediment was removed, and the subsequent work of the rivers which cross the area of the loess has been to excavate their present valleys or channels, bounded by bluffs of this formation, f A supplementary hypothesis to account for the subsidence assumed by some is, that a bodily depression of the crust of the earth was produced by the weight of the glacier. If one is inclined at first thought to reject this cause as inoperative because of its relative insignificance, he should reflect that * " The Ninth Annual Report of the Geological and Natural History Survey of Minnesota," p. SOt et seg. \ Ibid., pp. 387, 3.38. TEE LOESS. 369 thie forces maintaining tlie present contour of the earth's sur- face may be very evenly balanced, so that a slight addition at one point, and subtraction from another, might be the de- cisive influence in turning the scale. It is now pretty gener- ally believed that the long-continued and steady periods of subsidence involved in the formation of extensive sediment- ary rocks was due to the constant accumulation of the silt, out of which the rocks are made. This silt was relieving the continents of its weight, and adding to the weight along the whole line of deposition. During the Glacial period the transference of water from the ocean to accumulate as ice upon land removed an immense pressure from the ocean- beds, and added an equal amount of weight to the glaciated area. How much influence this may have had in depressing temporarily this area and its margin, we are unable to tell. But it is one of those unknown causes in the field which may be supposed to have accomplished something. Professor Alexander Winchell has, in a recent interesting paper, suggested a correlation between this pressure of the ice over the glaciated region and the enormous outflows of lava along the Rocky Mountains and the Cascade Range on the Pacific slope.* The lava-beds of that region are enormous in extent and are certainly of very recent date, and seem to have poured out from long fissures instead of craters. Under- neath these beds there are, in CaUfornia and Oregon, glacial deposits ; and it is in these lava-covered glacial deposits of southern California that human remains are supposed by Whitney to have been found. Professor Winchell's theory connecting those lava outflows with the depression prodjiced by the ice of the Glacial period in the east would relieve the subject of considerable eu)barrassment arising from the chronological difficulties that have been suggested. To the superficial objection that pressure over the Mississippi Valley would not produce volcanic eruptions at so great a distance * " Some Effect of Pressure of a Contineiital Glacier," in " The American Geologist," March, 1888, pp. 139-143. 24 370 THE WE AOE IN NORTH AMERICA. as the Pacific coast, it is readily answered : " The terrestrial globe in some of its behavior may be compared to an India- rubber ball filled with water. If indented by pressure in one place, there must be a protuberance equal in volume in another place. In a ball of uniform composition, the pro- tuberance would be spread over the entire surface beyond the region indented, and the effect in one particular spot might be insignificant. Should a small area of the caout- chouc be thinner than the rest, that part would be protruded to a greater extent than other parts of the surface. Should there be small holes or fissures through it, the water would escape and flow over the surface — that is, the protuberance resulting from local pressure would be chiefly on the outside of the shell. As we ordinarily conceive it, the water would be squeezed out like the juice of a squeezed orange."* Another supplementary theory which has been invoked to account for the loess, is, that it has, in certain sections at least, been brought to the surface by the agency of earth- worms and other animals which burrow in the ground. The remarkable facts adduced by Mr. Darwin concerning the ac- tivity of these humble agents give respectability to the theory ; and, indeed, the power of these agencies can be seen by any observer who will take the pains to notice the count- less marks of angle-worms which frequently appear upon the surface of the soil after a rain. Mr. Darwin estimated that the amount of soil brought to the surface by worms was in one case at the rate of nearly two tenths of an inch a year. It is estimated, also, by competent authority, that the number of worms to an acre is as great as fifty-four thousand and that they would weigh three hundred and fifty-six pounds. Trustworthy estimates also show that these creatures some- times raise annually to the surface fourteen tons, and again eighteen tons, to the acre.f From this it is easily seen that the predominance of fine earth upon the surface is due to the * "Some Effect of Preaaure of a Continental Glacier," p. 139. f See Darwin, on " Vegetable Mould and Earth- Worms," chaps, iii and iv. THE LOESS. 371 work of such animals as worms, crayfish, and ants. But these creatures are limited in the depth to which they can pene- trate the soil aud convey it to the surface. Still, it is not im- probable that in many places where the loess-like deposits are shallow it may be the result of these agencies. CHAPTER XVII. FLIGHT OF PLANTS AND ANIMALS DURING THE GLACIAL AGE. Among tlie most interesting incidental effects of the Glar cial period is that of its influence in distributing plants and animals over the lower latitudes. A glance at a polar pro- jection of the northern hemisphere shows to what a remark- able extent the land is clustered around the north pole, and how easy it would have been, under favorable conditions of climate, for plants and animals to spread from that vicinity along different meridians, till, in the lower latitudes, they should be on opposite sides of the earth. In conformity with these natural lines of emigration, it has long been known that both among plants and animals the species of the northern hemisphere are much more closely allied than those of the southern hemisphere, where no such land-con- nection exists ; and, as we shall presently see, the problem presented by the distribution of plants in the northern hemi- sphere is very complex and curious. For its solution we are largely indebted to the sagacity of the late Professor Asa Gray, who discovered the key in the influence of the Glacial period. In 1857, after he was already familiar, from private cor- respondence, with Darwin's theory of the origin of species. Professor Gray was called upon to study the extensive botan- ical collections brought back from Japan by the expeditions of Commodores Perry and Eodgers. Comparison of these species with those in corresponding latitudes in other por- tions of the world brought out clearly — ^what had been FLIGHT OF PLANTS AND ANIMALS. 373 before but dimly perceived — that there was a remarkable resemblance between the existing plants of eastern Asia and those of eastern North America, that more species are com- mon to Japan and Europe than to Japan and western North America, and that the resemblance is greatest of all between Japan and eastern North America. Out of three hundred species, common to the temperate regions of eastern Asia 24 374 THE ICE AGE IN NORTH AMERICA. and tte corresponding region of Nortli America, only one third is represented in western North America. The key applied b}^ Professor Gray for the solution of this problem was suggested by the investigations of Heer and others, which had brought out the fact that, during the Ter- tiary period, just before the beginning of the Ice age, a tem- perate climate, corresponding to that of latitude 35° on the Atlantic coast, extended far up toward the north pole, per- mitting Greenland and Spitzbergen to be covered with trees and plants similar, in most respects, to those found at the present lime in Virginia and North Carolina. Here, indeed, in close proximity to the north pole, were then residing, in harmony and contentment, the ancestors of nearly all the plants and animals which are now foimd in the north tem- perate zone, and here they would have continued to stay but for the cold breath of the approaching Ice age, which drove them from their homes, and compelled them to migrate to more hospitable latitudes. The picture of the flight and dispersal of these forests, and of their struggle to find, and adjust themselves to, other homes, is second in interest to that of no other migration. A single tree is helpless before such a force as an advancing glacier, since a tree alone can not migrate. But a forest of trees can. Trees can " take to the woods " when they can do nothing else, and so escape unfavorable conditions. There is a natural climatic belt to which the life of a forest is adjusted. In the present instance, as the favorable con- ditions near the poles were disturbed by the cooling influ- ences of the glacier approaching from the north, the indi- vidual trees on that side of the forest-belt gradually perished ; but, at the same time that the favorable conditions of life were contracting on the north, they were expanding on the south, so that along the southern belt the trees could gradu- ally advance into new territory, and so the whole forest-belt move southward, following the conditions favorable to its existence. It is therefore easy to conceive how, with the slow advance of the glacial conditions from the north, the FLIGHT OF PLANTS AND ANIMALS. 375 vegetation of Greenland and British America was transferred far down toward the torrid zone on both the Eastern and the "Western Continent. Being once thus transferred, the forest would be compelled to remain there until the retreat of the ice began again to modify the conditions so as to compel a corresponding retreat of plants toward their original north- ern habitat. Thus it is that these descendants of the pre- glacial plants of Greenland, arrested in their northward march, have remained the characteristic flora of the latitudes near the glacial boundary. On the other hand, the arctic species, which can not endure even a temperate climate, and which must have accompanied the advancing glacier south- ward, found their natural conditions again in two ways : 1. By following closely upon the steps of the retreating ice to extreme northern latitudes ; and, 2. To use Professor Gray's happy expression, by " taking to the mountains," and finding near their summits the necessary arctic conditions. It is thus that the mountains of New England and Labrador con- tain many species of plants nearly identical with those on the Alps in Europe. N^o better presentation of this subject can be given than that of Professor Gray himself, made in an address delivered in 1878, and which by the kindness of Mrs. Gray I am per- mitted in large part to reproduce in this connection : * The forests of the whole northern hemisphere in the tem- perate zone (those that we are concerned with) are mainly made up of the same or similar hinds. Not of the same spe- cies, for rarely do identical trees occur in any two or more widely separated regions ; but all round the world in our zone the woods contain pines and firs and larches, cypresses and junipers, oaks and birches, willows and poplars, maples and ashes, and the like. Yet, with all these family likenesses throughout, each region has some peculiar features, some trees by which the country may at once be distinguished. * " Forest Geograpby and Archaeology," a lecture delivered before the Har- yard University Natural History Society, April 18, 1878, by Asa Gray. Printed in the "American Journal of Science," vol. cxvi, pp. 86-94, 183-196. 376 THE 10 E AGE IN NORTH AMEBIOA. Beginning by a comparison of our Pacific with our Atlan- tic forests, it is to be noted that the greater part of the trees familiar on the Atlantic side of the continent are conspicu- ously absent from the Pacific forests. For example, the Pacific coast has no magnolias, no tulip- tree, no papaw, no linden or basswood, and is very poor in maples ; no locust-trees — neither flowering locust nor honey- locust — nor any leguminous tree ; no cherry large enough for a timber-tree, like our wild black ciierry ; no gum-trees {Nyssa nor Liquidambar), nor sorrel-tree, nor kalmia ; no persimmon, or bumelia ; not a holly ; only one ash that may be called a timber-tree ; no catalpa or sassafras ; not a single elm, nor hackberry ; not a mulberry, nor planer-tree, nor maclura ; not a hickory, nor a beech, nor a true chestnut, nor a hornbeam ; barely one birch-tree, and that only far north, where the differ- ences are less striking. But as to coniferous trees, the only missing type is our bald cypress — the so-called cypress of our Southern swamps — and that deficiency is made up by other things. But as to ordinary trees, if you ask what takes the place in Oregon and California of all these missing kinds, which are familiar on our side of the continent, I must answer nothing, or nearly nothing. There is the madrofla (Arbutus) instead of our kalmia (both really trees in some places) ; and there is the Califoraia laurel instead of our Southern red bay- tree. Nor in any of the genera common to the two does the Pacific forest equal the Atlantic in species. It has not half as many maples, nor ashes, nor poplars, nor walnuts, nor birches, and those it has are of smaller size and of inferior quality ; it has not half as many oaks, and these and the ashes are of so inferior economic value that (as we are told) a passable wagon- wheel can not be made of California wood, nor a really good one in Oregon. . . . Now almost all these recur, in more or less similar but not identical species, in Japan, north China, etc. Some of them are likewise European, but more are not so. Extending the comparison to shrubs and herbs, it more and more appears that the forms and types which we count as peculiar to our Atlantic region, when we compare them, as we first naturally FLIGHT OF PLANTS AND ANIMALS. 377 do, with Europe and with our West, have their close counter- parts in Japan and north China ; some in identical species (especially among the herbs), often in strikingly similar ones, not rarely as sole species of peculiar genera or in related ge- neric types. I was a very young botanist when I began to notice this, and I have from time to time made lists of such instances. Evidences of this remarkable relationship have multiplied year after year, until what was long a wonder has come to be so common that I should now not be greatly surprised if a Sarra- cenia or a BioncBa, or their like, should turn up in eastern Asia. Very few such isolated types remain without counter- parts. It is as if Nature, when she had enough species of a genus to go round, dealt them fairly, one at least to each quar- ter of our zone ; but when she had only two of some peculiar kind gave one to us and the other to Japan, Manchuria, or the Himalayas ; when she had only one, divided these between the two partners on the opposite side of the table. As to number of species generally, it can not be said that Europe and Pacific North America are at all in arrears ; but, -as to trees, either the contrasted regions have been exceptionally favored or these have been hardly dealt with. There is, as I have intimated, some reason to adopt the latter alternative. We may take it for granted that the indigenous plants of any country, particularly the trees, have been selected by cli- mate. Whatever other influences or circumstances have been brought to bear upon them, or the trees have brought to bear on each other, no tree could hold its place as a member of any forest or flora which is not adapted to endure even the extremes of the climate of the region or station. But the character of the climate will not explain the remarkable paucity of the trees which compose the indigenous European forest. That is proved by experiment, sufficiently prolonged in certain cases to justify the inference. Probably there is no tree of the north- ern temperate zone which will not flourish in some part of Europe. Great Britain alone can grow double or treble the number of trees that the Atlantic States can ; in all the latter we can grow hardly one tree of the Pacific coast. England supports all of them, and all our Atlantic trees also, and like- wise the Japanese and north Siberian species, which do thrive 378 THE ICE AGE IN NORTH AMERICA. here remarkably in some parts of the Atlantic coast, especially the cooler temperate ones. The poverty of the European sylva is attributable to the absence of our Atlantic American types, to its having no magnolia, liriodendron, asimina, negundo, no ^schulus, none of that rich assemblage of leguminous trees represented by locusts, honey-locusts, gymnocladus, and cla- drastis (even its cercis, which is hardly European, is like the Californian one mainly a shrub) ; no Nyssa, nor Liquidambar ; no EricacecB rising to a tree ; no bumelia, catalpa, sassafras, Osage orange, hickory, or walnut ; and, as to conifers, no hem- lock, spruce, arbor- vitse, taxodium, nor Torreya. As compared with northeastern Asia, Europe wants most of these same types, also the ailantus, gingko, and a goodly number of conif- erous genera. I can not point to any types tending to make up the deficiency — that is, to any not either in east North America or in northeast Asia, or in both. Cedrus, the true cedar, which comes near to it, is only north African and Asian. I need not say that Europe has no Sequoia, and shares no spe- cial type with California. Now, the capital fact is, that many and perhaps almost all of these genera of trees were well represented in Europe throughout the later Tertiary times. It had not only the same generic types, but in some cases even the same species, or what must pass as such, in the lack of recognizable distinctions be- tween fossil remains and living analogues. Probably the Eu- ropean Miocene forest was about as rich and various as is ours of the present day, and very like it. The Glacial period came and passed, and these types have not survived there, nor re- turned. Hence the comparative poverty of the existing Eu- ropean sylva, or at least the probable explanation of the ab- sence of those kinds of trees which make the characteristic difference. Why did these trees perish out of Europe, but survive in America and Asia ? Before we inquire how Europe lost them, it may be well to ask how it got them. How came these American trees to be in Europe ? And among the rest, how came Europe to have Sequoias, now represented only by our two big trees of California ? It actually possessed two species and more ; one so closely answering to the redwood of the Coast FLIGHT OF PLANTS AND ANIMALS. 379 Eanges, and another so yery like the Sequoia gigantea of . the Sierra Nevada, that, if such fossil twigs with leaves and cones had been exhumed ia California instead of Europe, it would confidently be affirmed that we had resurrected the veritable ancestors of our two giant trees. Indeed, so it may probably be. Goelum non animam mutant, etc., may be applicable even to such wide wanderings and such vast intervals of time. If the specific essence has not changed, and even if it has suffered some change, genealogical connection is to be inferred in all such cases. . . . I take it that the true explanation of the whole problem comes from a just general view, and not through piecemeal suppositions of chances. And I am clear that it is to be found by looking to the north, to the state of things at the arctic zone — first, as it now is, and then as it has been. North of our forest regions comes the zone unwooded from cold, the zone of arctic vegetation. In this, as a rule, the species are the same round the world ; as exceptions, some are restricted to a part of the circle. The polar projection of the earth down to the northern tropic shows to the eyte — as our maps do not — how all the lands come together into one region, and how natural it may be for the same species, under homogeneous conditions, to spread over it. When we know, moreover, that sea and land have varied greatly since these species existed, we may well believe that any ocean-gaps, now in the way of equable distribution, may have been bridged over. There is now only one consid- erable gap. What would happen if a cold period were to come on from the north, and were very slowly to carry the present arctic cli- mate, or something like it, down far into the temperate zone ? Why, just what has happened in the Glacial period, when the refrigeration somehow pushed all these plants before it down to southern Europe, to middle Asia, to the middle and southern part of the United States ; and, at length receding, left some part of them stranded on the Pyrenees, the Alps, the Apennines, the Caucasus, on our White and Rocky Mountains, or wherever they could escape the increasing warmth as well by ascending mountains as by receding northward at lower levels. Those 380 THE ICE AGE IN NORTH AMERICA. that kept together at a low level, and made good their retreat, form the main body of present arctic yegetation. Those that took to the mountains had their line of retreat cnt off, and hold their positions on the mountain-tops under cover of the frigid climate due to elevation. The conditions of these on different continents or different mountains are similar but not wholly alike. Some species proved better adapted to one, some to another, part of the world : where less adapted, or less adaptable, they have perished ; where better adapted, they con- tinue — with or without some change — and hence the diversifi- cation of Alpine plants as well as the general likeness through all the northern hemisphere. All this exactly applies to the temperate-zone vegetation, and to the trees that we are concerned with. The clew was seized when the fossil botany of the high ai'ctic regions came to light ; when it was demonstrated that in the times next pre- ceding the Glacial period — in the latest Tertiary — from Spitz- bergen and Iceland to Greenland and Kamchatka, a climate like that we now enjoy prevailed, and forests like those of New England and Virginia, and of California, clothed the land. We infer the climate from the trees ; and the trees give sure indications of the climate. I had divined and published the explanation long before I knew of the fossil plants. These, since made known, render the inference sure, and give us a clear idea of just what the climate was. At the time we speak of, Greenland, Spitzbergen, and our arctic sea-shore, had the climate of Pennsylvania and Virginia now. It would take too much time to enumerate the sorts of trees that have been identified by their leaves and fruits in the arctic later Tertiary deposits. I can only say, at large, that the same species have been found all round the world ; that the richest and most exten- sive finds are in Greenland ; that they comprise most of the sorts which I have spoken of, as American trees which once lived in Europe — magnolias, sassafras, hickories, gum-trees, our identical Southern cypress (for all we can see of difEerence), and especially Sequoias— x^oi only the two which obviously answer to the two big trees now peculiar to California, but sev- eral others ; that they equally comprise trees now peculiar to FLIGHT OF PLANTS AND ANIMALS. 381 Japan and China, three kinds of gingko-trees, for instance, one of them not evidently distinguishable from the Japan species which alone survives ; that we have evidence not merely of pines and maples, poplars, birches, lindens, and whatever else characterize the temperate-zone forests of our era, but also of particular species of these, so like those of our own time and country, that we may fairly reckon them as the ancestors of several of ours. Long genealogies always deal more or less in conjecture ; but we appear to be within the limits of scientific inference when we announce that our existing temperate trees came from the north, and within the bounds of high proba- bility when we claim not a few of them as the originals of present species. Remains of the same plants have been found fossil in our temperate region, as well as in Europe. Here, then, we have reached a fair answer to the question how the same or similar species of our trees came to be so dis- persed over such widely separated continents. The lands all diverge from a polar center, and their proximate portions — however different from their present configuration and extent, and however changed at different times — were once the home of those trees, when they flourished in a temperate climate. The cold period which followed, and which doubtless came on by very slow degrees during ages of time, must long before its culmination have brought down to our latitudes, with the similar climate, the forest they possess now, or rather the ancestors of it. During this long (and we may believe first) occupancy of Europe and the United. States, were deposited in pools and shallow waters the cast leaves, fruits, and occasion- ally branches, which are imbedded in what are called Miocene Tertiary, or later deposits, most abundant in Europe, from which the American character of the vegetation of the period is inferred. Geologists give the same name to these beds in Greenland and southern Europe, because they contain the remains of identical and very similar species of plants ; and they used to regard them as of the same age on account of this identity. But in fact this identity is good evidence that they can not be synchronous. The beds in the lower latitudes must be later, and were forming when Greenland probably had very nearly the climate which it has now. 382 TEE ICE AQE IN NORTH AMERICA. Wherefore the high, and not the low, latitudes must be assumed as the birthplace of our present flora, and the present arctic vegetation is best regarded as derivative of the temperate. This flora, which, when circumpolar, was as nearly homogene- ous round the high latitudes as the arctic vegetation is now, when slowly translated into lower latitudes, would preserve its homogeneousness enough to account for the actual distribution of the same and similar species round the world, and for the original endowment of Europe with what we now call Ameri- can types. It would also vary or be selected from by the in- creasing differentiation of climate in the divergent continents and on their difEerent sides in a way which might well account for the present diversification. From an early period the sys- tem of the winds, the great ocean-currents (however they may have oscillated north and south), and the general proportions and features of the continents in our latitude (at least, of the American Continent), were much the same as now, so that species of plants, ever so little adapted or predisposed to cold winters and hot summers, would abide and be developed on the eastern side of continents, therefore in the Atlantic United States and in Japan and Manchuria : those with preference for milder winters would incline to the western sides ; those disposed to tolerate dryness would tend to interiors or to re- gions lacking summer rain. So that, if the same thousand species were thrust promiscuously into these several districts, and carried slowly onward in the way supposed, they would inevitably be sifted in such a manner that the survival of the fittest for each district might explain the present diversity. Besides, there are resiftings to take into the account. The Glacial period or refrigeration from the north, which at its inception forced the temperate flora into our latitude, at its culmination must have carried much or most of it quite beyond. To what extent displaced, and how far superseded by the vege- tation which in our day borders the ice, or by ice itself, it is difficult to form more than general conjectures, so difEerent and conflicting are the views of geologists upon the Glacial period. But upon any, or almost any, of these views it is safe to conclude that temperate vegetation, such as preceded the refrigeration, and has now again succeeded it, was either thrust FLiaST OF PLANTS AND ANIMALS. 383 out of northern Europe and the northern Atlantic States or was reduced to precarious existence and diminished forms. It also appears that, on our own continent at least, a milder cli- mate than the present, and a considerable submergence of land, transiently supervened at the north, to which the Tegetation must have sensibly responded by a northward movement, from which it afterward receded. All these vicissitudes must have left their impress upon the actual vegetation, and particularly upon the trees. They furnish probable reason for the loss of American types sus- tained by Europe. I conceive that three things have conspired to this loss : ] . Europe, hardly extending south of latitude 40°, is all within the limits generally assigned to severe glacial action. 3. Its mountains trend east and west, from the Pyrenees to the Car- pathians and the Caucasus beyond, near its southern border ; and they had glaciers of their own, which must have begun their operations, and poured down the northward flanks, while the plains were still covered with forest on the retreat from the great ice-wave coming from the north. Attacked both on front and rear, much of the forest must have perished then and there. 3. Across the line of retreat of those which may have flanked the mountain-ranges, or were stationed south of them, stretched the Mediterranean, an impassahle barrier. Some hardy trees may have eked out their existence on the northern shore of the Mediterranean and the Atlantic coast. But, we doubt not, taxodium and Sequoias, magnolias, and Liquidambars, and even hickories and the like, were among the missing. Escape by the east, and rehabilitation from that quarter until a very late period, was apparently prevented by the prolongation of the Mediterranean to the Caspian, and thence to the Siberian Ocean. If we accept the supposition of Nordenskiold, that, anterior to the Glacial period, Europe was " bounded on the south by an ocean extending from the Atlantic over the present deserts of Sahara and central Asia to the Pacific," all chance of these American types having escaped from or re-entered Europe from the south and east is excluded. Europe may thus be conceived to have been for a time somewhat in the condition in which Greenland is now, 384 THE lOE AGE IN NORTH AMERICA. and, indeed, to hare been connected with Greenland in this or in earlier times. Such a junction, cutting off access of the Gulf Stream to the Polar Sea, would, as some think, other things remaining as they are, almost of itself give glaciation to Europe. Greenland may be referred to, by way of compari- son, as a country which, having undergone extreme glaciation, bears the marks of it in the extreme poverty of its flora, and in the absence of the plants to which its southern portion, extending six degrees below the Arctic Circle, might be entitled. It ought to have trees, and might support them. But, since destruction by glaciation, no way has been open for their re- turn. Europe fared much better, but suffered in its degree in a similar way. Turning for a moment to the American Continent for a contrast, we find the land unbroken and open down to the tropic, and the mountains running north and south. The trees, when touched on the north by the on-coming refrigera- tion, had only to move their southern border southward, along an open way, as far as the exigency required ; and there was no impediment to their due return. Then the more southern latitude of the United States gave great advantage over Eu- rope. On the Atlantic border, proper glaciation was felt only in the northern part, down to about latitude 40°. In the in- terior of the country, owing doubtless to greater dryness and summer heat, the limit receded greatly northward in the Mis- sissippi Valley, and gave only local glaciers to the Kocky Mountains ; and no volcanic outbreaks or violent changes of any kind have here occurred since the types of our present vegetation came to the land. So our lines have been cast in pleasant places, and the goodly heritage of forest-trees is one of the consequences. The still greater richness of northeastern Asia in arboreal vegetation may find explanation in the prevalence of particu- larly favorable conditions, both ante-glacial and recent. The trees of the Miocene circumpolar forest appear to have found there a secure home ; and the Japanese Islands, to which most of these trees belong, must be remarkably adapted to them. The situation of these islands — analogous to that of Great Britain, but with the advantage of lower latitude and greater FLIGHT OF PLANTS AND ANIMALS. 385 sunshine — their ample extent north and south, their diversified configuration, their proximity to the great Pacific Gulf Stream, by which a vast body of warm water sweeps along their accentu- ated shores, and the comparatively equable diffusion of rain throughout the year, all probably conspire to the preservation and development of an originally ample inheritance. The case of the Pacific forest is remarkable and paradoxi- cal. It is, as we know, the sole refuge of the most character- istic and wide-spread type of Miocene Coniferce, the Sequoias; it is rich in coniferous types beyond any country except Japan ; in its gold-bearing gravels are indications that it possessed, seemingly down to the very beginning of the Glacial period. Magnolias and beeches, a true chestnut, Liquidambar, elms, and other trees now wholly wanting to that side of the conti- nent, though common both to Japan and to Atlantic North America. Any attempted explanation of this extreme paucity of the usually major constituents of forest, along with a great development of the minor or coniferous element, would take us quite too far, and would bring us to mere conjectures. Much may be attributed to late glaciation ; something to the tremendous outpours of lava which, immediately before the period of refrigeration, deeply covered a very large part of the forest area ; much to the narrowness of the forest-belt, to the want of summer rain, and to the most unequal and pre- carious distribution of that of winter. Upon all these topics questions open which we are not pre- pared to discuss. I have done all I could hope to do in one lecture if I have distinctly shown that the races of trees, like the races of men, have come down to us through a prehistoric (or pre-natural-historic) period ; and that the explanation of the present condition is to be sought in the past, and traced in vestiges, and remains, and survivals ; that for the vegetable kingdom also there is a veritable archaeology. As the truth of a theory needs to be tested by the meth- od of difference as well as of agreement, we turn to inquire if it be not true that all mountains which reach above the snow-line have an Alpine flora, and whether it be not the conditions themselves which determine the peculiarities? 25 386 THE ICE AGE IN NORTH AMERICA. These questions are answered by an appeal to certain oceanic islands which were outside the influence of continental glacia- tion. According to Wallace, a striking proof of the theory presented by Professor Gray " is found on the Peak of Ten- eriffe, a mountain 12,000 feet high. In the uppermost 4,500 feet of this mountain above the limit of trees. Yon Buch found only eleven species of plants, eight of which were peculiar ; but the whole were allied to those found at lower elevations. On the Alps or Pyrenees, at this elevation, there would be a rich flora comprising hundreds of arctic plants ; and the absence of anything corresponding to them in this case, in which their ingress was cut ofE by the sea, is exactly what the theory leads us to expect." * On both continents, at the close of the Tertiary period, there occurred a remarkable extinction of animals which is doubtless connected with the advance of the continental ice- sheet. Among these we may mention two species of the cat family as large as lions ; four species of the dog family, some of them larger than wolves ; two species of bears ; a walrus, found in Virginia ; three species of dolphins, found in the Eastern States ; two species of the sea-cow, found in Florida and South Carolina; six species of the horse; the existing South American tapir ; a species of the South Amer- ican llama ; a camel ; two species of bison ; three species of sheep ; two species of elephants and two of mastodons ; a species of Megatherium, three of Megalonyx, and one of Mylodon — huge terrestial sloths as large as the rhinoceros, or even as large as elephants, which ranged over the Southern States to Pennsylvania, and the Mylodon as far as the Great Lakes and Oregon. f This wondrous assemblage of animals became extinct upon the approach of the Glacial period, as their remains are all found in post-Pliocene deposits. The intermingling of forms is remarkable. The horses, camels, and elephants which * " The Geographical Distribution of Animals," vol. i, p. 48. t Ibid., p. 129. FLIGHT OF PLANTS AND ANIMALS. 387 lived in ]S'orth America before the Glacial period were found subsequent to the Glacial period only in the Old World, while the llamas, tapirs, and gigantic Edendata are South American types. The progress of events seems to have been about as follows : In the warm period preceding the Glacial epoch, when the vegetation of the temperate zone flourished about the north pole, there was land connection between the continents, permitting the larger species of the Old "World to migrate to North America. At the same time the con- ditions in North America were favorable to the tropical spe- cies of animals which had developed and flourished in South America. The refrigeration of the climate on the approach of the Glacial period, and the advance of the ice from the north, cut off retreat to the Old World species, and gradually hemmed them in over the southern portion of the continent, where all forms of life were compelled to readjust themselves to new conditions. The struggle for existence probably re- sulted, first, in the extinction of those South American spe- cies which had invaded North America during the warmer climate of later Tertiary times ; and the more hardy emi- grants from the north would have the advantage from the similarity in climate in the southern United States during the Glacial period to that about the poles, where they had flourished immediately before. With the withdrawal of ice to the north, the struggle of these animals with the condition of existence began anew, and the mammoth and some others found themselves unable to cope with the changes to which they were compelled to adjust themselves. From the abun- dance of remains of these animals found in the peat-bogs of kettle-holes and in the glacial terraces of gravel and loess, it is evident that they followed close upon the retreating ice- front, and some of them continued the retreat to the Arctic Oirclej*where they still live and flourish ; while others, like the elephant and mastodon, perished. Few things are better calculated to impress the scientific imagination than this dispersion and final extinction in North America of so many large animals native to the Old World ; 388 THE IGE AOE IN NORTH AMERICA. wliile some of them, like the horse, were admirably adapted to the present conditions, as is shown by their rapid increase since their introduction after the discovery of America by the whites. In a succeeding chapter we shall also see that man himself participated in this struggle with the new con- ditions introduced by the Glacial period on this continent, and that, in company with the mammoth, walrus, and other arctic species, he followed up the retreating ice both upon the Atlantic coast and in the Mississippi Valley. Whether, like some of his companions, he was unsuccessful in the con- test is not certain, though there is much to be said in favor of the theory that the Eskimos of the north are the lineal descendants of the preglacial men whose implements are found in New Jersey, Ohio, and Minnesota. Much also may be said to support the theory, alluded to by Professor Clay- pole, connecting the traditions of the destniction of large portions of the human race by a flood with the extermination of species naturally brought about by the conditions accom- panying the floods which closed the Glacial period. It is interesting to observe, also, that insects as well as plants and the larger animals were compelled to reckon with the Glacial period. They, too, participated in the southern migration enforced by the advancing ice, and also shared in the vicissitudes of its final retreat, compelling them to escape from the warmer belt of climate which again advanced upon them from the south. Like the forms of arctic and Alpine vegeta- tion, a portion of the insects also took to the mountains, where they still remain, as living witnesses to the reality of the Glacial period. The summits of the White Mountains are characterized by Alpine species of insects, one of which is thus described by Mr. Samuel Scudder : But even the narrow limit of the Alpine zone of the White Mountains claims for its own a single butterfly, which probably has a more restricted range than any other in the world. One may search the season through over the comparatively vast and almost equally barren elevations within the sub-Alpine district of the White Mountains and fail to discover more than here FLIGHT OF PLANTS AND ANIMALS. 389 and there a solitary individual whirled by fierce blasts down the mountain-slopes, while, a few hundred feet above, the but- terflies swarm in great numbers. Every passage of the sun from behind a cloud brings them out in scores, and they may often be captured as fast as they can be properly secured. The contrast between the occasional and unwilling visitor in the sub-Alpine region and the swarms which flutter about the upper plateaus is most significant. Yet the Carices, the food- plant of the caterpillar, are quite as abundant in the lower regions as in the upper, even to the species G. rigida, upon which I found the larva feeding. Now, this butterfly {(Eneis semidea) belongs to a genus which is peculiar to Alpine and arctic regions ; in fact, it is the only genus of butterflies which is exclusively confined to them. It has numerous members, both in this country and in the Old World. One is confined to the Alps of Europe ; most of the European species, however, are found only in the extreme north. The genus extends across the whole continent of America, and several of its spe- cies occur on the highest elevations of the Eocky Mountains. Several species are common to Europe and America, and it is to one of these that (Eneis semidea is most closely allied. A few species descend into the Hudsonian fauna, but, as a whole, the genus has its metropolis farther north. So that, in ascend- ing Mount Washington, we pass, as it were, from New Hamp- shire to northern Labrador ; on leaving the forests, we come first upon animals recalling those of the northern shores of the Gulf of St. Lawrence and the coast of Labrador opposite New- foundland ; and, when we have attained the summit, we find insects which represent the fauna of Atlantic Labrador and the southern extremity of Greenland.* Commenting upon these and similar facts connected with other species of butterflies and vsdth several species of moths, Mr. A. R. Grote pertinently says : f The question comes up with regard to the White Mountain butterfly, as to the manner in which this species of (Eneis * "Geology of New Hampshire,'' toI. i, pp. 340, 341. f "American Journal of Science," vol. ex, ISTS, pp. 33Y, 338. 390 THE ICE AGE IN NORTH AMERICA. attained its present restricted geographical area. . . . How did the White Mountain butterfly get up the White Mount- ains ? I am disposed to answer, by the action attendant on the decline of the Glacial period. . . . The main ice-sheet had pushed them insensibly before it, and, during the continuance of the Glacial period, the geo- graphical distribution of the genus (Eneis had been changed from a high northern region to one which may well have in- cluded portions of the Southern States. And, on its decline, the ice-sheet drew them back again after itself by easy stages ; yet not all of them. Some of these butterflies strayed by the way, detained by the physical nature of the country, and destined to plant colonies apart from their companions. When the main ice-sheet left the foot of the White Mountains, on its long march back to the pole, where it now seems to rest, some of these wayward, flitting (Eneis butterflies were left behind. These had strayed up behind the local glaciers on Mount Washington, and so became separated from the main body of their companions, which latter journeyed northward, follow- ing the course of the retirement of the main ice-sheet. They had found in elevation their congenial climate, and they have followed this gradually to the top of the mountain, which they have now attained, and from which they can not now retreat. Far off in Labrador the descendants of their ancestral com- panions fly over wide stretches of country, while they appear to be in prison on the top of a mountain. I conceive that in this way the mountains may generally have secured their Alpine animals. The Glacial period can not be said strictly to have expired. It exists even now for high levels above the sea, while the Eskimos find it yet enduring in the far north. Had other conditions been favorable, we might now find arctic man living on snow-capped mountains within the temperate zone. At a height of from 5,600 to 6,200 feet above the level of the sea, and a mean temperature of about 48° during a short summer, the White Mountain butterflies {(Eneis semidea) yet enjoy a climate like that of Labrador within the limits of New Hampshire. And in the case of moths an analogous state of things exists. The species Anarta melanopa is found on Mount FLIGHT OF PLANTS AND ANIMALS. S91 Washington, the Eocky Mountains, and in Labrador. Agrotis Islandica is found in Iceland, Labrador, in the White Mount- ains, and perhaps in Colorado. As on islands in the air, these insects haTS been left by the retiring ice-flood during the open- ing of the Quaternary. On inferior elevations, as on Mount Katahdin, in Maine, where we now find no (Eneis butterflies, these may formerly have existed, succumbing to a climate gradually increasing in warmth from which they had no escape ; while the original colonization, in the several instances, must have always greatly depended upon local topography. CHAPTEK XYiri. EUROPE DUEING THE GLACIAL PERIOD. At this point it will be profitable to take a survey of the condition of some other parts of the world during the great Ice age. By the same marks which determine the extent of the glacier in America, it is evident that the existing gla- _ ciers of Switzerland and Norway are but remnants of what formerly existed in these localities. James Geikie's statement of the situation in the British Isles is sufficiently complete : During the climax of the Glacial period all Scotland was drowned in a wide-spread mer de glace, which coalesced in the north and east with a similar sheet of ice, that crept outward from Scandinavia. To the west the Scottish ice, meeting with no impediment to its course, overflowed the outer Hebrides to a height of 1,600 feet, and probably continued on its path into the Atlantic as far as the edge of the 100-fathom plateau, where the somewhat sudden deepening of the sea would allow it to break off and send adrift whole argosies of icebergs. The height reached by the upper surface of the ice that over- whelmed the outer Hebrides enables us to ascertain the angle of slope between those islands and the mainland. This was 1 in 211 — that is to say, the inclination of the surface of the ice-sheet was about twenty-five feet in the mile — an inclination which would appear to the eye almost like a dead level. . . . The ice flowed off Ireland in all directions save to northeast in Antrim, upon the coast of which it encountered the Scottish mer de glace, which forced it to turn away to northwest and southeast ; but along the whole western and southern shores, 394 THE ICE AOE JN NORTH AMERICA. where no obstacle to its passage intervened, it seems to hare swept in one broad and continuous stream out — probably as far as that of Scotland— into the Atlantic. The thickness attained by the ice that flowed into the Irish Sea from Scot- land, where it coalesced with the mer de glace coming from the eastern seaboard of Ireland, and also, as we shall presently see, with that creeping out from England and Wales, makes it quite certain that the area now occupied by that sea must at that time have been filled with glacier ice. . . . The North Sea was filled with a massive mer de glace con- tinually advancing in a general south-southwestern direction, the presence of which is distinctly traceable in the remarkable deflections of the glaciation all along the seaboard of Scotland from Stonehaven southward. It was simply owing to the superior elevation and extent of the Scottish mountains that the narrow strip of low-lying ground in the eastern maritime districts of that country was not invaded by an alien ice-stream. When we pass into England the hills become lower, and the area of low ground between the hills and the sea increases in breadth. There was thus less and less opposition ofEered to the southward advance of the North Sea mer de glace as it pressed upon the eastern shores of England, until eventually it overflowed bodily and crept southward across the midland table-land on its way to the valley of the Severn and the Bristol Channel. This remarkable glacial invasion is proved not only by the carry of local stones, and stones which have come south from the northern counties and Scotland, but by the appear- ance in the till at Cornelian Bay and Holderness of bowlders of two well-known Norwegian rocks, which were recognized by Mr. Amund Helland. . . . The ice which would thus appear to have streamed trans- versely across England eventually coalesced with that which overflowed from the basin of the Irish Sea southeast through Cheshire, together with that which streamed east from the Welsh uplands, and the united mer de glace thereafter made its way into the Bristol Channel. Here it joined the thick ice that flowed out to sea from the high grounds of South Wales, the bottom-moraine of which is conspicuous not only in the mountain-valleys of that region, but also upon the low- EUROPE DURING THE OLAGIAL PERIOD. 396 lying tracts that extend from the hills to the sea. In the south- eastern counties, so far as we know at present, the ice-sheet at the climax of the Glacial period did not extend farther than the valley of the Thames, beyond which no trace of its bottom- moraine has been met with.* Professor A. Geikie summarizes the facts conceraing the continent as follows : In Scandinavia the ice-stri£e run westward and southwest- ward on the Norwegian coasts, and eastward or southeastward across the lower grounds of Sweden. When the ice descended into the basin of the Baltic and the plains of northern Ger- many, it moved southward and southwestward, but seems to have slightly changed its direction in different areas and at difEerent times. Its movements can be made out partly from the strise on the solid rock, but more generally from the glacial drift which it has left behind. Thus it can be shown to have moved down the Baltic into the North Sea. At Berlin its movement must have been from east to west. But at Leipsic, as recently ascertained by Oredner, it came from north-north- west to south-southeast, being doubtless shed off in that direc- tion by the high grounds of the Harz Mountains. Its southern limit can be traced with tolerable clearness from Jevennaar, in Holland, eastward across the Ehine Valley, along the base of the Westphalian hills, round the projecting promontory of the Harz, and then southward through Saxony to the roots of the Erzgebirge. Passing next southeastward along the flanks of the Riesen and Sudeten chain, it sweeps across Poland into Russia, circling round by Kiev, and northward by Nijni-Nov- gorod toward the Urals. ,It has been estimated that, excluding Finland, Scandinavia, and the British Isles, the ice must have covered no less than 1,700,000 square kilometres of the present lowlands of Europe. . . . The ice is computed to have been at least between 6,000 and 7,000 feet thick in Norway, measured from the present sea-level. From the height at which its transported debris has been observed on the Harz, it is believed to have been at * "Prehistoric Europe," pp. 189, 190, 192, 193. 396 TEE IGE AGE IN NORTH AMERICA. least 1,470 feet thick there, and to have gradually risen in ele- vation as one vast plateau, like that which at the present time covers the interior of Greenland. Among the Alps it attained almost incredible dimensions. The present snow-fields and glaciers of these mountains, large though they are, form no more than the mere shrunken remnants of the great mantle of snow and ice which then overspread Switzerland. In the Bernese Oberland, for example, the valleys were filled to the brim with ice, which, moving northward, crossed the great plain, and actually overrode a part of the Jura Mountains, for huge fragments of granite and other rocks from the central chain of the Alps are found high on the slopes of that range of heights.* More recently the late Professor H. Carvill Lewis studied the field in Great Britain, and published conclusions some- what different from those which had been before accepted. He traced, according to the summary giyen by Upham, a terminal moraine "across southern Ireland from Tralee on the west to the Wicklow Mountains and Bray Head southeast of Dublin ; through the western, southern, and southeastern portions of Wales ; northward by Manchester, and along the Pennine Chain to the southeast edge of Westmoreland, thence southeast to York, and again northward to the Tees, and thence southeastward along the high coast of the North Sea to Flamborough Head and the mouth of the Humber." f Professor Lewis propounded the theory that the supposed glacial deposits in England south of this line of terminal moraine were to be explained as water- deposits in a glacial lake produced by the damming up of the Humber Eiver and a slight elevation of the earth at the Straits of Dover. It is proper to say, also, that in this theory Professor Lewis had been in part anticipated by Professor Boyd Dawkins, who had written as follows : The ice at this time was suflBciently thick to override Schihallion, in Perthshire, at a height of 3,500 feet, and the * " Text-Book of Geology," pp. 885, 886. f " The American Geologist," vol. ii, 1888, p. 375. EUROPE DURING TEE GLACIAL PERIOD. 39T hills of Galway and Mayo at 3,000 feet. Its southerii limit in Britain is uncertain. According to Professor Eamsay and Dr. James Geikie, it extended as far south as the latitude of Lon- don ; but the hypothesis upon which this southern extension is founded — that the bowlder-clays have been formed by ice melting on the land — is open to the objection that no similar clays have been proved to have been so formed, either in the arctic regions, where the ice-sheet has retreated, or in the dis- tricts forsaken by the glaciers in the Alps or Pyrenees, or in any other mountain-chain. Similar deposits, however, have been met with in Davis Strait and in the North Atlantic, which have been formed by melting icebergs ; and we may, therefore, conclude that the bowlder-clays have had a like oi-igin. . . . The English bowlder-clays, as a whole, differ from the moraine profonde in their softness and the large area which they cover. Strata of bowlder-clay at all comparable to the great clay mantle covering the lower grounds of Britain north of the Thames are conspicuous by their absence from the glaciated regions of central Europe and the Pyrenees, which were not depressed beneath the sea.* Professor Lewis's views are of such interest that our treat- ment of the subject would be incomplete without a fair pre- sentation of them, which can best be done in his own words : The great ice-sheet which once covered northern England was found to be composed of a number of glaciers, each of which was bounded by its own lateral and terminal moraines. These glaciers were studied in detail, beginning with the east of England ; and the North Sea Glacier, the Wensleydale Gla- cier, the Stainmoor Glacier, the Aire Glacier, the Irish Sea Glacier, and the separate Welsh glaciers were each found to be distinguished by characteristic bowlders, and to be defined by well-marked moraines. The terminal moraine of the North Sea Glacier, filled with Norwegian bowlders, may be seen in Holderness, extending from the mouth of the Humber to Flamborough Head, and consists of a series of conical hills inclosing meres. The moraine of the Stainmoor Glacier, char- * "Early Man in Britain," pp. 116, 117. 398 THE ICE AGE IN NORTH AMERICA. acterized by blocks of Shap granite, may be followed northward along the coast past Scarborough and Whitby ; then west along the Cleveland Hills ; then south again through Oulston to the city of York ; then west to near Allerton, where the Stainmoor Glacier is joined by the Wensleydale Glacier, a fine medial moraine marking the line of Junction. The Wensleydale Gla- cier is characterized by bowlders of carboniferous limestone and sandstone, and its lateral moraine is followed northward through Wormald Green, Markington, Fountains Abbey, and along the Permian outcrop to Masham, where it turns west to Wensleydale, passing Jervaulx Abbey, and running up the val- ley. North of Wensleydale the moraine of the Stainmoor Gla- cier is followed through Richmond to Kirby Ravensworth, and westward to the mountains, where the glacier attained an eleva- tion of two thousand feet. Thus the Stainmoor Glacier, a tongue of the great Irish Sea Glacier, had been divided into two branches by the Cleveland Hills, one branch going south to the city of York, which is built on its terminal moraine, the other branch flowing out of the Tees, and being deflected southward along the coast by the North Sea Glacier, with which it became confluent. The Irish Sea Glacier, the most important glacier of Eng- land, came down from Scotland, and being re-enforced by local ice-streams, and flowing southward until it abutted against the mountains of Wales, it was divided into two tongues, one of which flowed to Wellington and Shrewsbury, while the other went southwest across Anglesey into the Irish Sea. This great glacier and its branches are all outlined by terminal moraines. A small tongue from it, the Aire Glacier, was forced eastward at Skipton and has its own distinctive moraine. In the neigh- borhood of Manchester the great moraine of this Irish Sea Gla- cier may be followed through Bacup, Hey, Stalybridge, Stock- port, and Macclesfield, being as finely developed as the moraines of Switzerland and America. South of Manchester, it contains flint and shell fragments, brought by the glacier from the sea- bottom over which it passed." At Manchester the ice was at least fourteen hundred feet thick, beingas thick as the Eh6ne Glacier.* * Abstract by the author of a paper read at the Birmingham meeting of the British Association, September, 1886. EUROPE DURING THE GLACIAL PERIOD. 399 In a paper before the British Association in September, 1887, Professor Lewis presented his views in greater detail, and answered objections, alleging that — The hypothesis of extra-morainic fresh-water laiies, dammed up by the glaciers, is sustained by all observed facts. The most important of these lakes was one caused by the obstruc- tion of the mouth of the Humber by the N"orth Sea Glacier, whose terminal moraine crosses that river at its mouth. This large lake reached up to the 400-foot contour line, and ex- tended southward nearly to London, and westward in finger- like projections into the many valleys of the Pennine Chain. It deposited the "great chalky bowlder-clay," and erratics were floated in all directions by icebergs. It was bounded in the vale of York by the Stainmoor Glacier, and Gharnwood Forest was an island in it. At its flood period it overflowed southwestward by torrential streams into the Severn Valley and elsewhere, carrying the "Northern Drift" into the south of England. Other glaciers in England were bordered by simi- lar but smaller lakes wherever they advanced against the drain- age. Three such lakes were made by the Aire Glacier, the largest of them extending to Bradford. The Irish Sea Glacier caused many similar lakes high up on the west side of the Pennine Chain, and at its southern end north of Wolverhamp- ton. The overflow streams from the most southern of these lakes joined those issuing from Lake Humber in the Birming- ham district, characterized by a "commingling of the drift," otherwise inexplicable. An examination of the supposed evi- dences for glaciation, and for a great marine submergence in central and southern England, shows that neither theory is sustained by the facts. Thus, the supposed striae on Eowley Rag prove to be root-marks or plow-marks ; those reported at Charnwood Forest to be due to running water, or perhaps icebergs ; the supposed drift on the chalk-wolds to be a local wash of chalk-flints ; the high-level gravels on the Cotteswold Hills to be preglacial ; the shells at Macclesfield, Moel Tryfan, and Three Rock Mountain to be glacier-borne, and not a proof of submergence ; the drift on the Pennine plateau of north Derbyshire to be partly made by icebergs floating in Lake 400 TEE IGE AGE IN NORTH AMERICA. Humber, and partly a decomposed millstone grit or Bunter sandstone ; and the supposed Welsh erratics on Frankley Hill at a height of eight hundred feet to be in place and due to an outcrop of the palaBozoic floor. The conclusion that the glacial phenomena of England are due neither to a universal ice-cap nor to a marine submergence, but to a number of glaciers bordered by temporary fi-esh-water lakes, is in accordance with all the observations of the author in England and elsewhere.* It is fair to add, however, that soon after this meeting of the British Association at which this paper was read, his observations at Frankley Hill, in Worcestershire, and west- ward, led Professor Lewis to waver in his views, and he had resolved to go over all the ground again; but his untimely death prevented the accomplishment of this plan. Probably there can be little doubt of the correctness of Mr. Upham's conclusion that, if Lewis had lived, he would have accepted the opinions of the majority of the geologists of Great Brit- ain, that land-ice really extended at one time as far south as the Thames. " Still, small portions of northern England escaped glaciation ; . . . and these tracts of the high moor- lands in eastern Yorkshire and of the eastern flank of the Pennine Chain are similar to the driftless area of southwest- ern Wisconsin." It would seem from Professor Levsds's facts about a moraine in England, as well as from those presently to be stated concerning Professor Salisbury's discoveries in northern G-ermany, that tlie farthest extent of the ice-front is, in Europe as well as in America, considerably in advance of the well-defined terminal moraine, and suggests the same difference of interpretation as here — i. e., this moraine is either the rem- nant of a later glacial period or it is a moraine of retrocession. * See also "American Journal of Science," vol. cxxxii, 1886, pp. 433-438; "Proceedings of the Boston Society of Natural History," 1887; "UeberGla- cialerscheinungen bei Gommern unweit Magdeburg " (" Zeitschr. d. Deutschen geolog. Gesellschaft, Jahrg., 1883," pp. 831-848), and " Mittheilungen uebcr das Quartaer am Nordrande des Harzes" (ibid., "Jahrg, 1885," pp. 897-905), von F. Wahnschaffe, in Berlin. EUROPE BURING THE GLACIAL PERIOD. 401 The conclusion of Professor Lewis concerning the marine shells found at high elevations in glacial deposits on the mountains of Wales, and which have generally been taken to indicate a deep submergence during the Glacial epoch or at the beginning of the second Glacial epoch, are of the greatest importance, and coincide with similar discoveries recently announced by Mr. Upham concerning the shells found in the vicinity of Boston, and supposed to indicate a post-glacial sub- sidence of considerable extent in that vicinity. These shells in the British Isles, like those in the vicinity of Boston, are mostly in fragments, are very thick and compact in structure, and often water-worn and sometimes striated. Their eleva- tion in the British Isles reaches as much as thirteen hundred and fifty feet above tide. Professor Lewis thinks they were plowed up by the glacier as it passed over the trough of the Irish Sea, and were elevated to their present position by the ice in the same manner that bowlders are seen so often to have been elevated in various parts of the United States. I again adopt the words of Mr. Upham in his recent comments upon this theory : The ample descriptions of the shelly drift of these and other localities of high levels, and of the lowlands of Cheshire and Lancashire, recorded by English geologists, agree per- fectly with the explanation given by Lewis, which indeed had been before suggested, so long ago as 1874, by Belt and Good- child. This removes one of the most perplexing questions which glacialists have encountered, for nowhere else in the British Isles is there proof of any such submergence during or since the Glacial period, the maximum known being five hun- dred and ten feet, near Airdrie, in Lanarkshire, Scotland. At the same time the submergence on the southern coast of England was only from ton to sixty feet, while no traces of raised beaches or of Pleistocene marine formations above the present sea-level are found in the Orkney and Shetland Islands.* * " American Geologist," vol. ii, p. 375. 26 402 THE ICE AGE IN NORTH AMERICA. During tlie summer of 1888 Professor Salisbury, who had been for several years associated with President Cham- berlin in the glacial survey of the Northwestern States and Territories, made also a hasty survey of northern Germany for the purpose of correlating the glacial deposits of that country with those in America. The results were most im- portant and interesting.* He found a double series of termi- nal moraines back some distance from the extreme glaciated limits, as in the Northwestern States of America, and resem- bling them both in their composition and in their situation with reference to the marginal deposits. As traced by Pro- fessor Salisbury, this belt of moraines follows approximately the curve of the south shore of the Baltic Sea, and not many miles from it. Its course lies through Schleswig-Holstein, Mecklenburg, Potsdam (about forty miles north of Berlin), thence swinging more to the north, and following nearly the line between Pomerania and West Prussia, crossing the Vis- tula about twenty miles south of Dantzic, thence eastei'ly to the Spirding See, near the boundary of Poland. Among the places where this moraine can be best seen are — " 1. In Province Holstein, the region about (especially north of) Eutin ; 2. Province Mecklenburg, north of Crivitz, and between Biitow and Kropelin ; 3. Province Brandenburg, south of Keckatel, between Strassen and Barenbusch, south of Fiirstenberg and north of Everswalde, and between Pyritz and Solden ; 4. Province Posen, east of Locknitz, and at numerous points to the south, and especially about Falkenburg, and between Lompelburg and Barwalde. This is one of the best localities. 5. Province "West Preussen, east of Biitow ; 6. Province Ost Preussen, between Horn and Widikin." Comparing these with the moraines of America, Professor Salisbury remarks ; In its composition from several members, in its variety of development, in its topographic relations, in its topography, in * Professor R. D. Salisbury on " Terminal Moraines in Northern Germany," in "American Journal of Science," vol. cxxxv, 1888, pp. 405, 406. 404: THE ICE AGE IN NORTH AMERICA. its constitution, in its associated deposits, and in its wide sep- aration from the outermost drift limit, this morainic belt cor- responds to the extensive morainic belt of America, which extends from Dakota to the Atlantic Ocean. That the one formation corresponds to the other does not admit of doubt. In all essential characteristics they are identical in character. What may be their relations in time remains to be determined. CHAPTEE XIX. THE CAUSE OF THE GLACIAL PEEIOD. Fob the past few years speculation concerning the cause of the Glacial period has been largely dominated by an astronomical theory. But geologists have in general felt the impropriety, in such an important matter, of abandoning their own iield to accept the glittering results of celestial mathematics. At any rate, it would be improper for them to let the astronomical solution go unchallenged. If the geologist suffers himself to be lifted into the air, like the fabled Antseus, he labors at a disadvantage, and can be easily overcome. For this reason the glacialists of America have, of late, limited their labors chiefly to the collection of ter- restrial facts, and when asked, as they often are, " What was the cause of the Glacial period?" the first answer has fre- quently been, " That is none of our business." Still, it is by the interpretation of facts that causes are discovered, and the collection of facts concerning the glaciation of North America has advanced so rapidly during the past few years, that it is now high time to consider more fully their meaning and discuss the subject anew, if for no other reason than for the sake of finding out how little is known about it. It is easily seen that a glacier is the combined product of cold and moisture. A simple lowering of the temperature will not produce an ice age. Before an area can maintain a glacier, it must first get the clouds to drop down a sufiicient amount of snow upon it. A climate which is cold and dry mav not be so favorable to the production of glaciers as one which is temperate, but whose climatic conditions are such 406 THE ICE AOE m NORTH AMERICA. that there is a large snow-fall. For example, on the steppes of Asia, and over the Rocky Mountain plateau of our "Western States and Territories, the average temperature is low enough to permit the formation of extensive glaciers, but the snow- fall is so light that even the short summers in high latitudes cause it all to disappear ; whereas, on the southwestern coast of South America, and in southeastern Alaska, where the temperature is moderate, but the snow-fall is large, great glaciers push down to the sea even in low latitudes. The circumstances, then, pre-eminently favoring the pro- duction of glaciers, are abundance of moisture in the atmos- phere, and climatic conditions favorable to the precipitation of this moisture as snow rather than as rain. Heavy rains produce floods, which speedily transport the water to the ocean-level ; but heavy snows lock up, as it were, the capital upon dry land, where, like all other capital, it becomes con- servative, and resists with great tenacity both the action of gravity and of heat. Under the action of gravity glaciers move, indeed, but they move very slowly. Under the influ- ence of heat ice melts, but in melting it consumes an enor- mous amount of force. In order to melt one cubic foot of ice, as much heat is re- quired as would heat a cubic foot of water from the freezing- point to 176° Fahr., or two cubic feet to 88° Fahr. To melt a layer of ice a foot thick will therefore use up as much heat as would raise a layer of water two feet thick to the temperature of 88° Fahr. ; and the efEect becomes still more easily under- stood if we estimate it as applied to air, for to melt a layer of ice only one and a half inch thick would require as much heat as would raise a stratum of air eight hundred feet thick from the freezing-point to the tropical heat of 88° Fahr. We thus obtain a good idea, both of the wonderful power of snow and ice in keeping down temperature, and also the reason why it takes so long a time to melt away, and is able to go on accu- mulating to such an extent as to become permanent. These properties would, however, be of no avail if it wei'e liquid, like water ; hence it is the state of solidity and almost complete THE CAUSE OF TEE OLAGIAL PERIOD. 407 immobility of ice that enables it to produce by its accumula- tion such extraordinary effects in physical geography and in climate as we see in the glaciers of Switzerland, and the ice- capped interior of Greenland.* Theories respecting the causes of the Glacial epoch natu- rally fall into two classes, whicli may be styled respectively the "cosmical" and the "terrestrial." Altogether they number half a score. Among the terrestrial theories are in- cluded the following : 1. That the Glacial period is due to a decrease in the original heat of the planet ; 2. To the shift- ing of the polar axis ; 3. To a former period of greater moist- ure of the atmosphere ; 4. To changes in the distribution of land and water ; 5. To the elevation of the lands in northern Europe and America to a higher level than that now occu- pied. Among the cosmical theories may be mentioned : 6. Variations in the temperature of space ; 7. Variations in the amount of heat radiated by the sun ; 8. The combined effect of the precession of the equinoxes and of the changing eccen- tricity of the earth's orbit. Of these theories several may be eliminated from the problem without difficulty. If, according to the first theory, the Glacial period was due to a decrease of the original heat of the planet, the period should not have culminated in the past, but we should still be looking for its culmination in the future ; for both the earth and the sun are cooling off. "We may, therefore, drop out the first theory. If, according to the second theory, the cause had been the shifting of the earth's axis of rotation, we should not find, as we now do, evidences that the warm climate which preceded the Glacial period approached the poles along the present circles of latitude ; but, as it is, we find that the temperate flora which covered the arctic regions at the close of the Tertiary period approached the pole not only in Greenland and British America, but also in Spitzbergen and Nova * Wallace's "Island Life," pp. 12'7, 128. 408 THE ICE AOE IN NORTH AMERICA. Zembla. We may, therefore, drop this second theory out of consideration. The third theory, so ably advocated by Professor Whitney, that the Ice age was the direct result of the excessive moisture of earlier periods, and that the disappearance of glaciers is to be accounted for by a general drying up of the earth, is ruled out by the fact that there is evidence, among other things, from the vast deposits of salt existing in numerous parts of the world, that the work of desiccation has been going on in some portions of the earth from the earliest geological ages. For example, central New York is, at the present time, one of the best-watered portions of the world ; but it is underlaid by deposits of salt sixty or seventy feet in thickness, and these extend under much of the area of Upper Canada and Michi- gan. To produce this amount of salt there must have oc- curred, during the Upper Silurian age, the drying up of an inland sea over that region a mile in depth. We are com- pelled, therefore, to regard the era of the saline group of rocks, rather than the present, as the great age of desiccation. Besides, moisture in the atmosphere is efficient as a glacial cause only when it is precipitated as snow, and this must be determined by general meteorological conditions. There is probably moisture enough always in the air to produce an ice age if the conditions can be combined to precipitate it in the right form and at the right place to encourage the growth of glaciers. The fourth theory, which would attribute the growth and disappearance of glaciers entirely to changes in the distribu- tion of land and water over the surface of the globe, was, according to his general principles, ably and ardently advo- cated by Sir Charles Lyell ; and no one can read in his "Principles of G-eology" the chapters upon this subject without being greatly impressed by the possible influences of such changes. The ocean is the great equalizer of the earth's temperature. Through unimpeded ocean-currents, like the Gulf Stream of the Atlantic and the Kuro-Siwa of the Pacific, the heat of the tropics is transferred many thousands of miles THE CAUSE OF THE GLACIAL PERIOD. 409 to ameliorate the climate of even the polar regions. It is quite possible that comparatively slight changes in level in the vicinity of the West India Islands and Central America might so afEect the direction of the Gulf Stream as to produce most serious modifications of the climate in North America and Europe. Should a portion of the Gulf Stream be driven through a depression across the Isthmus of Panama into the Pacific, and an eqiial portion be diverted from the Atlantic coast of the United States by an elevation of the sea-bottom between Florida and Cuba, the consequences would neces- sarily be incalculably great, so that the mere existence of such a possible cause for great changes in the distribution of moist- ure over the northern hemisphere is sufficient to make one hesitate before committing himself unreservedly to any other theory — at any rate, to one which has not for itself independ- ent and adequate proof. It is profitable, also, in this connection, to reflect on how delicately balanced the forces of Nature now are with respect to the production of glaciers. As already noted,* the gla- ciers existing at the present time in the Alps have their peri- ods of advance and recession. A slight increase in the pres- ent snow-fall of Switzerland, if long continued, would pro- duce alarming results. From this cause alone, the glaciers would at once begin to enlarge ; and, in sympathy, the temperature would fall, and the increase of the glaciated area of Switzerland would go on until the whole country was again brought under the desolating reign of ice, or until the intervention of some coimteracting force should stay its advance. It is not without reason, therefore, that some alarm was occasioned in Switzerland a few years ago by the proposition to inundate the Desert of Sahara. Fortunately, no extensive inundation of that region is within the reach of human power. But, if it could be inundated, thus extend- ing greatly the evaporating area from which the clouds gather moisture for the Alpine heights, there is no telling * See p. 92. 410 THE IGE AOE IN NORTH AMERICA. what the result might not be. Should there be an annual increase of a foot of snow upon the Alps, a thousand addi- tional feet of snow would have to be dissolved every thou- sand years, with the enormous absorption of heat accom- panying the process. This simple calculation is sufficient to show the reality of the cause introduced by the fourth hypothesis, which would explain the Glacial period through the influence upon cHmate of changes in the distribution of land and water. This cause is so effective that it may even be conceived to be sufficient, without the introduction of any other agencies. The fifth theory, which introduces considerable change of levels in the continents, rests, without doubt, upon a true cause, which, very likely, has co-operated with others, and may in itself have been the chief agency in producing the glacial conditions which we are studying. It is very difficult to arrive at safe conclusions concerning the changes of level which accompanied the Glacial -period. There would seem to be no doubt that, in preglacial times, there was a considerable elevation of the continent east of the Mississippi Valley and on the Atlantic coast. This is shown by the fact that the trough of the lower Mississippi Eiver was eroded in preglacial times to a depth of from two hundred to five hundred feet below its present level, as was proved by the investigations of Professor Hilgard, which brought to light pebble-deposits and stump layers down to that depth. A preglacial elevation on the Atlantic coast is likewise indicated by the depth of the rock-bottom of the Hudson at New York, already alluded to, and by the sound- ings outside the harbor, which reveal the natural outlines of an old mouth and harbor just beyond the limits of the pres- ent ones. How far this preglacial elevation was participated in by the region lying between the Great Lakes and Hudson Bay there are no direct means of determining ; but it is fair to suppose that there was such a participation. The fact that the bottoms of all the Great Lakes, except Erie, are lower than the sea-level, would also seem to necessitate the TEE GAITS E OP TEE GLACIAL PERIOD. 411 supposition of a considerable elevation of the interior of the continent in preglacial times (unless we ascribe an eroding power to ice which few would be willing to grant) ; for these lakes, as we have seen, occupy valleys of erosion rather than synclinal troughs.* As to subsequent changes of level, there is much differ- ence of opinion. In the lower Mississippi the deposits of loess upon the bluffs adjoining the river have been taken by Hilgard as indicating a depression of about 450 feet. This depression probably marks the time of maximum glacial de- velopment, and, as we have elsewhere intimated, may have been partly due to the direct effect of the weight of the ac- cumulation of ice to the north. On the Atlantic coast there was also a subsidence of considerable extent, which was probably contemporaneous with that proved in the Missis- sippi Valley. The evidence of depression south of the gla- cial limit on the Atlantic coast has recently been collected and discussed by Mr. McG-ee, to whose articles reference has already been made in discussing the glacial boundary.-f- Briefly stated, it is as follows : At "Washington, on the Po- tomac, 150 feet above tide, there is an extensive gravel delta, without marine shells (upon which the city is built), which contains numerous large bowlders (some of them two or three feet through), which in all probability must have been brought down by ice from the head-waters of the Potomac when the winters were much more severe than now. Leav- ing out the extreme figures mentioned, which we have given good reason to suppose represent the height of deposits from the glacier itself, the corresponding deposits on the Susque- hanna and the Delaware are in the neighborhood of two hun- dred feet. These constitute in the Delaware what Lewis called the Philadelphia red gravel and brick-clay, and, like the Columbia deposit at Washington, contain large bowlders. Such a formation certainly imphes some depression of the region and such a refrigeration of the climate as naturally *See p. 315 e^sey. f ^ee p. 135. 412 TEE lOE AGE IN NORTH AMERICA. would accompany the Glacial period. But I am inclined to think that Mr. McGee overestimates the amount of depres- sion necessary to account for the facts. The great floods from the melting ice and from the increased snow-fall in the mountains south of the glacial limit would of themselves partly account for the higher level of the deposits. Besides, at such a situation as "Washington, where the river debouches from its narrow, rocky trough, it is possible that, in times of greater floods and more ice, the river would temporarily build up a delta-terrace of considerable height independent of any depression. It is noticeable in Mr. McGee's table that the height of these Columbian delta-deposits diminishes southward in pretty direct proportion to the diminution in the size of the bowlders inclosed. The height of the deposit, which at the mouth of the Susquehanna is set down as 245 feet, is on the Potomac liO feet ; on the Rappahannock, 125 ; on the James, 100 ; and on the Roanoke T5 feet ; while the size of the transported bowlders diminishes from fifty times the size now transported on the Susquehanna to twenty times on the Potomac, ten times on the Rappahannock, five times on the James, and only two or three times on the Roanoke. This progressive diminution in the size of the bowlders would seem to point to a gradual diminution of the size of the spring freshets and of the efficiency of ice-action, such as is naturally connected with the distance of the rivers from the ice-front, and would seem to indicate the direct agency of the streams in building up deltas independent of subsidence. Again, the evidence of any marked subsidence during the Glacial period on the southern shore of New England is not conclusive, since Mr. Upham has recently shown that the shells long ago reported from Point Shirley and Hull, near Boston, are not in place, but have been transported by the ice itself, in accordance with the theory propounded by Lewis and others to explain those found at great height on the mountains of Wales. Such shells are found in wells in the drumlins in Boston Harbor, and hence have been subject to transportation. Farther north, however, indications of TEE CAUSE OF THE GLACIAL PERIOD. 413 submergence are generally accepted along the coast — at 200 feet and upward in Maine ; 470 feet in the Gulf of St. Law- rence ; 500 feet in Labrador ; and 1,000 feet or more in the arctic regions. In the interior, also, the skeleton of a whale was found in post-glacial deposits in Yermont 150 feet above tide, and, according to Dawson, in the township of Montague, county of Lanark, in Ontario, 440 feet above tide.* Accord- ing to the same authority, the highest level at which post- Pliocene marine shells are known to occur on Montreal Mountain is at an elevation of 520 feet. According to Chal- mers, however, the depression in the lower part of the St. Lawrence was considerably less than this, no evidence appear- ing below Kiviere du Loup of any submergence exceeding 375 feet, while ranch evidence supports the view that it could not have exceeded that amount. There must, how- ever, have been at the close of the Ice age a subsidence in the upper St. Lawrence Yalley of as much as 520 feet, and this depression continued until the ice had been removed over the lower part of the valley, permitting the sea to ex- tend up into Lanark county and into the Champlain Yalley. The evidence from the lake-terraces (already alluded to) in central N^ew York points to the same concliision. The ter- races bordering the ancient glacial lake of that region increase rapidly in height from Syracuse northeastward to "Watertown. But, from other facts toward the southwest, Mr. Gilbert thinks that the axis of oscillation corresponded with the pres- ent valley of the Mohawk River, and that there was an in- creasing amount of subsidence westward from this line. Professor J. W. Spencer supposes that the depression in the upper St. Lawrence, at the close of the Glacial period, was much greater than is indicated by the evidence just presented ; in tliis agreeing with what we suppose to be the views of Sir William Dawson, both of whom would regard the glacial boundary, as we have delineated it, as the margin of a glacial sea covering the whole region of the Great * "American Journal of Science,'' vol. cxxv, 1883, p. 200. 414 THE IGE AGE IN NORTH AMERICA. Lakes.* We have, however, already, in the proper place, given the evidence precluding the theory that this boundary is a water-margin. The evidence upon which Professor Spencer relies is capable, as we suppose, of another interpretation than that which he gives to it ; for his dependence is wholly upon beach-lines in which there are no fossils, and, until his in- vestigations are published in greater detail, we may be per- mitted to regard these lines as connected in some way vdth local bodies of water held at varying levels by obstructions of ice. One familiar with glaciers of a larger sort may readily believe that beaches may be formed on the margin of small glacial lalces, at various altitudes, upon the surface of the glacier itself. It would be no unlikely thing for the higher summits of land, as they emerged from the ice as it was melting away, to be surrounded on one or more sides by an extensive lake, held in a hollow of the ice. Such are re- ported on some of the glaciers in India at the present day, and the Merjelen See and others like it in the Alps furnish analogous instances. Our conclusion, therefore, is, that the only changes of level connected with the Ice age in North America of which there is satisfactory evidence are the following : 1. A pre- glacial elevation of a few hundred feet in the valley of the Mississippi, on the New England coast, and presumably in the Canadian highlands. 2. A depression of the same regions during the maximum extent of the ice, amounting in the lower Mississippi Valley to between three hundred and four hundred feet below the present level, and on the Atlantic coast to about one hundred and fifty feet in the vicinity of Philadelphia, five hundred and twenty feet at Montreal, and about one thousand feet in arctic regions. 3. Before the close of the Glacial period a re-elevation of the region about Lake Superior, if indeed this were ever much depressed. The subsidence in the St. Lawrence Valley continued until * See "Science," January 27 and August 31, 1888. THE CAUSE OF THE GLACIAL PERIOD. 415 tlie Ice age was nearly past, when that portion gradually as- sumed its present level. These would seem to be all the changes of level which are absolutely required to explain the facts in hand. But even these changes of level can not be proved to be amoug the causes of the Glacial period. They are more likely to be effects than causes. The weight of the ice and its power of attraction have already been spoken of as real causes tending to explain the subsidence indicated during and following the maximum glaciation. To this Mr. McGee would add the effect of refrigeration underneath the ice, tend- ing to a contraction of the strata occupying the glaciated area. This would naturally not operate to any great extent in pro- ducing depression until about the time of maximum develop- ment of ice, and naturally would continue for some time after it. But, while we can not point to positive evidences of such an elevation of country to the north as would produce a glacial period, we are led, from the instability of the earth indicated in the preceding statements, to conclude that this can not be arbitrarily ruled out of the case. Probably, how- ever, it did not act alone. The sixth theory and the seventh naturally go together. That there may be variations in the temperature of space is entirely within the realm of possibility, and that the sun maybe a variable star is a statement which can not be proved absolutely false. Indeed, the hypothesis that the heat of the sun is kept up by a bombardment of meteoroids is defended by eminent astronomers. In case this were true, a known natural cause for the production of variability is certainly in the field, since the cometary bodies, which are circulating irregularly through space, are probably themselves but ganglia of meteoroids which may readily get entangled within the predominant sphere of the sun's attraction, and become a means of increasing for a long period the amount of the sun's heat. This theory can not be positively affirmed to be true ; but, as long as it can not be disproved by astronomical con- siderations, it remains in the field to diminish the confidence with which we support other hypotheses. 416 THE ICE AGE IN NORTH AMERICA. The eighth and last theory mentioned is the one whicli has for the past few years been most popular, having had in Mr. Croll and in Mr. James Geikie most able and convincing advocates. Briefly stated, the theory is this. As is well known, the earth's orbit is not a circle, but an ellipse, whose longer diameter at the present time exceeds its shorter by about 3,000,000 miles. The sun is not in the center, but is in one of the foci of the ellipse, which at the present is about 1,500,000 miles outside of the center. As matters are now situated, therefore, the earth on the 21st of June (when it is said to be in aphelion) is 3,000,000 miles farther from the sun than it is on the 21st of December (when it is said to be in perihelion) ; that is, during the pres- ent winters of the northern hemisphere we are 3,000,000 miles nearer the sun than we are during the summers. Pig. 103.— Exaggerated view of the earth's orbit, showing the effects of precession of the equinoxes. A, condition of things now ; B, as it will be 10,500 years from now. Eut, if a line be drawn through the earth's orbit joining the equinoxes — that is, the points passed through on the 20th of March and the 22d of September— we shall iind that the winter is shorter than the summer. The period from the 20th of March to the 22d of September, which constitutes the summer, is one hundred and eighty-six days, while that THE CAUSE OF THE GLACIAL PERIOD. 417 from September 22d to March 20th is only one hundred and seventy-nine days, or seven days less. So that, while the earth is farther away from the sun during the summer, and receiving daily less heat, the additional seven days occupied by that part of the journey makes ample compensation, and the absolute amount of heat received during the longer time exactly equals that received by the earth during the shorter half of its journey. It will be observed, also, that when the summer in the northern hemisphere occurs at aphelion, the summer in the southern hemisphere occurs in perihelion, thus exactly reversing the conditions. Now it is claimed by Mr. Croll and others that when, in either hemisphere, the winter is short and occurs in perihelion, the climate of that hemisphere will be less favorable to the production of glaciers than in the hemisphere where the winter is long and in aphelion. Consequently, according to their theory, the situation of the earth is now favorable to the production of glaciers in the southern hemisphere, and un- favorable to their production in the northern. This theory is not based, however, on the idea that the hemisphere whose winter is in aphelion receives less heat from the sun than the other hemisphere during that season, but upon the supposition that the greater period occupied by the sun in passing through aphelion when the winter nights are long, gives more oppor- tunity for the loss of heat during winter in that hemisphere by radiation. Now, if it be correct that a winter in aphelion is favorable to glaciation, and a winter in perihelion unfavorable, then, from the astronomical changes which transfer this condition periodically from one hemisphere to another, we can reason that the northern and southern hemispheres are alternately subjected to conditions favorable to glaciation. For, through what is called the •' precession of the equinoxes," the periods of aphelion and perihelion are exactly reversed in their rela- tions to the two hemispheres every 10,500 years. The points at which the equinoxes occur are slowly slipping around, making a revolution once in about 21,000 years; so that, 27 418 THE lOE AGE IN NORTE AMERICA. 10,500 years from the present time, the winter in the north- ern hemisphere will occur in aphelion instead of perihelion ; and, if the supposition be correct concerning the influence of this increased length of the winter and distance of the earth from the sun, the conditions would favor the return of a glacial period 10,000 or 11,000 years hence, and would imply that similar favorable conditions existed 10,000 or 11,000 years ago. According to this theory, also, there should have been a succession of glacial periods every 21,000 years during long ages past. But there is still another periodicity in the movements of the earth about the sun with which to combine the preced- ing. The shape of the earth's orbit is not permanent, but through the influence of the attraction of the planets upon it is subject to periodic changes. In astronomical terms, the " eccentricity " of the earth's orbit is subject to variations ; that is, there are sometimes very much greater differences than at present between the longer and the shorter diameters of the earth's orbit. When this difference is greatest it amounts to no less than 7,000,000 miles ; so that at certain times the earth is 14,000,000 miles farther from the sun in winter than in summer, and vice versa. At the time of greatest eccentricity, also, the difference in length between the summers and winters would amount to thirty-six days, instead of seven or eight as now. These periods of greatest eccentricity in the earth's orbit during which, on Mr. CroU's theory, the conditions were ex- tremely favorable for the production of glacial epochs, are somewhat unevenly distributed. One of them culminated 200,000 years ago ; another, 750,000 ; another, 850,000 ; another, 2,500,000 ; and another, 2,600,000. In the future they will occur 500,000, 800,000, 900,000 hence. In the present condition of the earth's orbit this supposed cause is at its minimum. Of the astronomical changes affecting the eccentricity of the earth's orbit, we are certain. But the value of Croll's theory depends upon the correctness of the original assump- THE CAUSE OF THE GLACIAL PERIOD. 419 tion, that when the winters occur in aphehon there will be a great increase of snow-fall and a marked lowering of tem- perature during the winter ; and that, during the summers, heat would have less than its average influence in removing the snows of the previous winter. As already remarked, however, it should be noted that Mr. CroU's calculations upon these two points do not rest upon any difference in the estimate of the absolute amount of heat received during these periods. But he assumes that an excessive amount of heat would be lost from the earth by radiation during the long winters in aphelion ; so that the effect from that cause, when accumulated during a number of centuries, would be marked by a noticeable increase in the glacial fields of the hemisphere whose winter was in aphelion, and this would be connected with the decrease of glaciers in the other hemisphere, whose winters were correspondingly short and in perihelion. Having got glaciation started in one hemisphere during periods when the winters were in aphelion, Mr. Croll ad- duces an additional cause to help on the refrigeration, in the effect which this cause itself would have upon the winds and the ocean-currents. He estimates that the heat conveyed by the Gulf Stream into the Atlantic Ocean is equal to one fifth of all the heat possessed by the waters of the North Atlantic ; or to the heat received from the sun upon a mill- ion and a half square miles at the equator, or two million square miles in the temperate zone. " The stoppage of the Gulf Stream would deprive the Atlantic of 77,479,650,000,- 000,000,000 foot-pounds of energy in the form of heat per day." The cause of the Gulf Stream, therefore, becomes a most important element in the problem. What is the force driving onward this immense body of warm water, which he estimates " to be equal to that of a stream fifty miles broad, a thousand feet deep, flowing at the rate of four miles an hour," and whose temperature as it emerges from the Straits of Florida averages as high as 65° Fahr., twenty-five degrees of which is eventually parted with to ameliorate the climate of the North Atlantic ? 420 THE ICE AOE IN NORTH AMERICA. With great cogency of reasoning, Mr. Croll shows that the trade-winds are the predominant cause of the present course of the Gulf Stream. After attempting to show the failure of all other theories to account for ocean-currents, and for the direction of the Gulf Streatn in particular, Mr. CroU calls attention to the general correspondence between the direction of the winds and that of the great currents of the ocean, and shows how powerful this agency must be in giv- ing motion to the surface of the water, and by constancy of action, finally, to the lower strata of water. Now, from some cause or other, at the present time the southeast trade- winds are considerably stronger than the northeast. As a result, the southeast trades sometimes extend as far as latitude 10° or 15° north of the equator ; while the northeast trades rarely extend even as far south as the equator.* The geo- FiG. 104.— Map showing couree of currents in the Atlantic Oeoan. * and b' are currents set in motion by opposite trade winds ; meeting they produce the equatorial current wliich divides into c and c' continuing on as a and a' and e. * CroU's " Climate and Time," p. 70. 27 423 THE ICE AGE IN NORTH AMERICA. graphical position and contour of South America give special significance to this cause in its relation to the Gulf Stream and to all the regions dependent on it for warmth of climate. Cape St. Eoque, the easternmost point of South America, is only five degrees south of the equator, and is fifty degrees, or three thousand miles, east of the longitude of Florida. With the present relation of the trade-winds to each other, this situation of the continent of South America is favorable to the production of the Gulf Stream. For it is evident at a glance that the movement of water caused in the South Atlantic by the southeast trades will be at its maximum over the tropical belt in the vicinity of Cape St. Koque, and that the movement there will be in a general northwest direction. Hence there is great significance in the present contour of the continent, it being such that there is nothing to impede the movement of water once begun by the trade-winds in a northwest direction (at least, so mucli of it as is north of Cape St. Eoque) ; but the current thus started must keep on its way until it reaches the cul-de-sac formed by the Gulf of Mexico, and from this there is no escape except through the passage between the West India Islands and Florida. Here we have a deep, strong current, produced by the same hydrostatic law which propels the comparatively small stream in the hydraulic ram. Or, to draw an illustration from a grander spectacle, the movement is like that of the tides when passing up through gradually restricted channels. In such cases the thin onward movement of the tidal wave over a wide space is translated into a narrow but deeper and more powerful current up the gradually restricted channel into which it is forced ; so that, whereas the general height of the tidal wave is but two or three feet, it sometimes in re- stricted channels, like the Bay of Fundy, rises to the height of seventy feet, and the so-called "bore," characteiistic of many tidal rivers, like the Orinoco and the Amazon, becomes the terror of navigators. Through such a translation of the gentle but steady pres&ure of the southeast trades over the wide area of the South Atlantic Ocean, the waters of the THE CAUSE OF THE GLACIAL PERIOD. 423 Gulf Stream are projected through the Straits of Florida with a force sufficient to carry the in across the Atlantic and to the shores of Iceland and Norway. So far Mr. CroU's explica- tion is certainly very plausible, and seems to proceed from well-known physical principles, and is pretty generally ac- cepted by scientific men. This, however, is only the first step in his argument. Another glance at the map will show that if from any cause the relations of the trade-winds should be reversed, so that the northeast trades should predominate over the southeast, and extend some degrees south of the equator, then Cape St. Roque would intercept the movement caused by the south- east trades, and the warm water from the South Atlantic, which is now forced into the Caribbean Sea, would all of it do what part of it now does, namely, turn to the south, and, after following for a while the southwestern trend of the South xlmerican coast, would join the slow-moving whirlpool of the South Atlantic, whose center is on the parallel joining Montevideo and Cape Colony. It will thus appear that, in searching for the cause of the Gulf Stream, we are ultimately compelled to search for the cause of the present preponder- ance of the southeast trades in the Atlantic Ocean. This sends us backward upon a receding series of causes. The southeast trades preponderate because the southern hemisphere is cooler than the northern hemisphere, for wind is but the movement of the cooler and therefore heavier at- mosphere of one region toward a partial vacuum produced by a superior degree of heat in another. This conclusion pushes us back one step further to find the cause for the present lower relative temperature of the southern hemi- sphere, and here we strike what is probably a coincidence, but which Mr. Croll and his followers have too readily ac- cepted as a cause. Mr. Croll thinks he finds the cause of the low temperature of the southern hemisphere in the pres- ent prevalence there, in moderate degree, of the astronomical conditions to which he has attributed the production of gla- cial periods. The winters of the southern hemisphere now THE CAUSE OF THE GLACIAL PERIOD. 425 occur in aphelion — that is, when the earth is farthest away from the sun, and are seven days longer than the summers. While admitting that there is not as yet enough known about the laws governing the absorption and retention of the sun's heat upon the earth's surface to permit us to say with confi- dence that the acknowledged glacial condition of the southern hemisphere is not produced by the astronomical cause under consideration, it must be added that we are also unable to prove the inadequacy of other causes to produce the same results. In assuming the reality of Mr. Croll's cause, we are in danger of resting on a theoretical may ie rather than on well-established premises. At any rate, Woeikoff, in the ablest review that has yet appeared upon the subject, thinks the glaciation of the southern hemisphere may readily enough be accounted for without the aid of Croll's theory, and sums up the case thus : The extent and depth of the oceans of the southern hemi- sphere give a greater steadiness and force to the winds of that hemisphere, and the difference is even more marked if we com- pare the westerly winds of middle latitudes rather than the trades, though also well seen in the latter. Now, land acts in two ways on the trade-winds' : it weakens them, first, by the increase of friction. But this is not all. The trades, few ocean regions excepted, are not strong winds ; they are impor- tant on account of their extent and steadiness. The gradient which causes them is small. Now, in such cases, land, even if it is not a continent but only a cluster of small islands, has a great influence on trade-winds in causing local gradients which may have even an opposite direction to the general gradients, thus causing different and even opposing winds. The land- and sea-breezes and the monsoons are cases in point. Even where the disturbances of the normal ocean gradients are not . large enough to cause monsoons, we see generally the trades oftener interrupted in summer, when they are weaker and when local thunder-storms and rains are more frequent on land. For the two reasons given, the trades of the southern hemi- sphere must be more extensive and stronger than those of the northern. 426 TEE ICE AGE IN NORTE AMERICA. The relatively small extent of sea in middle latitudes of the northern hemisphere, in comparison with the southern, must tend to warm the seas of the former, even if the quantity of warm water from the tropical seas reaching them be equal. Thus, generally in the middle latitudes the evaporation goes on at a higher temperature from the seas of the northern than the southern hemisphere. Now, this has a very great influ- ence on the resulting precipitations ; when the evaporation goes on at or near 33°, there is much more probability that the resulting precipitation will be snow and not rain, even on low lands ; the higher the temperature at which the evaporation takes place, the greater must be the height at which snow can fall, on account of cooling by expansion. Not all cold seas are favorable to glaciation. If they are surrounded by land on which the winters have a temperature considerably below 32°, they will be covered with ice, and thus evaporation will be checked just at the time when it is most favorable to snow-fall. The ice of the seas will be covered with snow, the temperature of the air over it may be very low, but the snow-fall will not be great, and thus the conditions not favorable to glaciation. Such is the condition of many seas of the northern hemisphere, as the Arctic Ocean north of Siberia, the Kara Sea, the bays and inlets north of the North American Continent, the Sea of Okhotsk, etc., which are covered with ice during many months. These conditions are favorable to cold of many months' duration, but not to a large snow-fall and the resulting glaciation. The observations made at many points off the coasts of Siberia and the North American Archi- pelago have shown that the snow-fall is exceedingly light. The seas between 45° and 70° of southern latitude are deep and not surrounded by land, and thus by far not so ice-bound, both on account of the absence of very low temperatures favorable to the formation of ice, and of the rupture of ice, when formed, by winds and currents.* Thus it is shown that the depth and relative extent of the southern ocean furnish a sufficient cause for its present glacial conditions. * "American Journal of Science," vol. cxxxi, 1886, pp. 175, 176. TEE CAUSE OF THE GLACIAL PERIOD. 427 As already intimated, the weak point in Mr. Croll's the- ory is the general state of uncertainty as to the laws regulat- ing the absorption, retention, and distribution of the sun's heat upon the earth. It is evident that the heat upon which the earth is dependent is that of the sun ; since, as Professor Newcomb has shown, the total amount of heat received from the stars is probably not one-millionth part of that received from the sun.* Now, as all admit that the annual amount of heat received from the sun is not affected by changes either in the eccentricity of the earth's orbit, or in the rela- tion of the poles to that eccentricity, it is only the question of the retention and distribution of heat with which we have to do. And here we come to a most obscure realm of sci- entific investigation, where ignorance is still profound. The reason why the summit of a mountain is cold is not because of lack of heat from the sun, but it arises rather from the facility with which the heat is dissipated by radiation. On the contrary, the reason why the atmosphere of a greenhouse is warmer than that upon the outside is not because it reoewes more heat, but because it retains more. The intenser heat- rays of the sun readily penetrate the glass cover, while the less intense rays of radiated heat from the earth are unable to do so in return. It is well known, also, that clonds pre- vent a frost by checking the radiation from the surface of the earth. The laws regulating the influence of the atmosphere and the floating particles contained in it, over the retention of the sun's heat in its lower strata, are as yet but little un- derstood. There is here an almost unlimited field for inves- tigation and discovery. And this, as just remarked, is the weak point of Mr. Croll's theory. Everything here depends upon the forces which distribute the heat and moisture over the land-sur- faces. It is by no means certain that, when the winters of the northern hemisphere occur in aphelion, they will be colder than now. Whether they would be so or not depends * "American Journal of Science," vol. cxi, ISTB, p. 264; vol. cxxvii, 1884, p. 22. 428 THE IGE AGE I.Y NORTE AMERICA. upon the action of forces wliose laws can not now be accurately calculated. As Woeikoff goes on to show, there are some very singular facts in the distribution of heat over the earth's sur- face — proving that the equator is not so hot as theoretically it ought to be, and that the arctic regions are not so cold ; and this in places which could not be affected by oceanic currents. For example, at Iquitos, on the Amazon, only three hundred feet above tide, three degrees and a half south of the equator, and more than a thousand miles from the Atlantic (so that ocean-currents can not abstract the heat from its vicinity), the mean yearly temperature is but Y8° Fahr., while at Verkhojansk, in northeast Siberia, which is 67° north of the equator, and is situated where it is out of the reach of ocean- currents, and where the conditions for the radiation of heat are most favorable, and where, indeed, the winter is the coldest on the globe (January averaging —56° Fahr.) the mean yearly temperature is two degrees and a half above zero ; so that the difference between the temperature upon the equator and that at the coldest point on the sixty-seventh parallel is only about 75° Fahr. ; whereas, if temperature were in proportion to heat received from the sun, the difference ought to be 172°. Again, the difference between the actual January temperature on the fiftieth parallel and that upon the sixtieth is but 20° Fahr., whereas, the quantity of solar heat received on the fiftieth parallel during the month of January is three times that received upon the sixtieth, and the difference in temperature ought to be about 170° Fahr. upon any known law in the case. But to be quite sure to get beyond the influence of ocean- currents, I will take the mean January temperature in the strictly continental climate of eastern Siberia, under 130° east. According to Ferrel's tables : Under 50° north we have 0° Fahr. Under 60° north we have —30° Fahr. If the January temperature decreased from 50° to 60° north, according to the hypothesis of Dr. Oroll, it should be on the 60° north -155° Fahr. THE CAUSE OF THE GLACIAL PERIOD. 429 But to be quite sure of taking the most fayorable case for the hypothesis of Dr. Oroll, I take the highest January tem- perature on the 50° north in Ferrel's tables, that is, that on 20° east =44° Pahr., and the coldest January temperature on the 60° north, that is, that of 130° and 130° east,= -30° Fahr. Yet in proportion to the quantity of heat received, the mean temperature of January on 60° north should be —140° Fahr. The following table gives the results of the three cases con- sidered : Mean tem- perature 50° N. MEAN TBMPEEATUBE, 60° H. Difference. On the hypothe- sis of JJr. Croll. Actual, Mean January temperature of all meridians 21-3 0° 440 - 147-9 - 15o-3 - 140-0 -30 -30 149-6 Mean January temperature in 120° E. (east Siberia) Mean January temperature of warmest meridian 50° N., and coldest meridian on 60° N 125-3 110-0* These facts, and many others like them, make it evident that we understand very little about the laws governing the distribution of heat over the surface of the earth. Other things besides ocean-currents are active in the matter, and some of them must be far more potent than any cause which we now clearly discern. "We quote again the words of the same high authority : How can we judge of the change of temperature resulting from this or that distance from the sun, even if we knew accu- rately the temperature of space, when we do not know the dia- thermancy of the atmosphere under different conditions ? We know only that it is exceedingly different, according to the different quantities of carbonic acid and aqueous vapor con- tained in it, and in a far higher degree, according to the ab- sence or presence in different quantities of suspended liquid and solid particles (clouds, dust, smoke, etc.). Thus, when * Woeikoff in " American Journal of Science," vol. cxxxi, p. 166. 430 THE ICE AOE IN NORTH AMERICA. we do not know in how far the loss of heat is impeded, even an accurate knowledge of the temperature of space would be of small use in this matter. I will illustrate this by a homely example. Take a room where the fire is extinguished and the hearth or stove cold in the evening, and try to guess at the temperature the room will have in the morning. If we follow the method of Dr. Oroll, we should inquire only about the out- side temperature, and not about the thickness of the walls, the windows, etc. I think that, taking the average construction of Eussian, English, and Italian houses, if the inside tempera- ture was in all three cases 60° in the evening, and the outside temperature 30° in Eussia, 33° in England, and 45° in Italy, the morning temperature in the room would not be very differ- ent, and probably even higher in the Eussian room, owing to its thick walls, double windows, etc. . . . Thus it is easy to see that the question how great will be the temperature of the air at a given place, say in midwinter, when the distance of the sun is greater or less than at present, can not be answered, even approximatively, especially in the exceedingly crude way it is put by Br. CroU — that is, without distinguishing high and low latitudes, continent and ocean, etc. One thing is certain, that such a change will certainly have a greater influence on the temperatures in the interior of continents than on the oceans and their borders. The caloric capacity of water is so great, and the mobility of its particles so effectual in resisting a lowering of the surface temperature, by the convection currents it causes, that I doubt very much if, during a great eccentricity and winter in aphelion, the sur- face temperature of the oceans can be lower in winter than now ; the difference in the quantity of sun-heat is too small and too short-continued to give an appreciable difference in winter ; and, as in the year there is no difference in the quan- tity of heat received by the waters, I think there will be no difference in the temperature of the waters, and thus no influ- ence of great eccentricity with winter in aphelion on the ocean temperatures, and also no greater snow-fall than now. As to the continents, I admit that, though we are unable to calculate the rate of decrease of temperature of the winter months in these conditions, there is no doubt that it will be appreciable, and be THE CAUSE OF TEE OLACIAL PERIOD. 431 the greater the less a given place is under the influence of the seas.* We may test the theory still further by an appeal to geo- logical facts. According to Mr. Croll, there must have been a succession of glacial periods in the past, and it would seem that numerous indications of such epochs, if they occurred, must exist in the successive geological strata. If siich indi- cations are not found in requisite amount, the adyoeates of Mr. CroU's theory are bound to give a satisfactory explana- tion of the failure. To a consideration of this evidence Mr. Croll devotesjthe seventeenth and eighteenth chapters of his book on " Climate and Time," and at the outset confesses that " the facts which have been recorded as evidence in favor of the action of ice in former geological epochs are very scanty indeed." To ac- count for this deficiency of evidence, he adduces, first, " the imperfection of the geological records themselves ; and, sec- ond, the little attention hitherto paid toward researches of this kind." Mr. Croll's presentation of the reasons, from the nature of the case, why the evidence of glaciation in the earlier geo- logical periods should be in large degree obliterated, is prob- ably as strong as can be made. He argues that the present land-surfaces in nearly all cases represent former ocean-beds, hence sedimentary strata deposited during the Glacial age must consist of the water-worn material which had been car- ried out from glacial streams into the bordering seas and oceans, so that the most distinct signs of glacial action which we could expect to find in sedimentary strata woiild be de- posits of pebbles, forming conglomerate rocks, and the occur- rence in these conglomerates of occasional angular fragments, such as could only be transported on ice. Mr. Croll, also, very naturally, dwells upon the extent to which the land-surfaces exposed between two geological epochs must have suffered from denudation. Erosive agen- * WoeikofE in '' American Journal of Science," vol. cxxxi, pp. 169, 172. 432 THE ICE AGE IN NORTH AMEBIC A. cies would operate in the ordinary way during the whole period of elevation, the streams carrying down to the sea a large amount of material every season. But when the period of depression had proceeded so far as to bring the surface below the level of the sea, Mr. CroU believes the action of the waves would greatly hasten the operation, and would thoroughly sort out and roll the pebbles, washing the finer particles into deeper water. A careful consideration of the forces in operation, how- ever, does not seem wholly to justify this reasoning of Mr. Croll. In the first place, there must have been at various geological epochs, over the area now most studied, extensive land exposures, continuing through a long period of time. The Tertiary deposits contain many vegetable remains as well as animal, showing the existence of land areas of no small extent. The Carboniferous period reveals whole con- tinents maintaining, over a large portion of their extent, an elevation near the sea-level, in which there were continual but slight oscillations, tending, however, on the whole to subsidence. So that land-plants accumulated in sufficient quantity to form the coal-beds — the periods of depression being marked by sedimentary rocks formed by the consolida- tion of the wash that was spread over the whole region dur- ing the times of depression. Now, it does not seem possible that a glaciated area so ex- tensive as is that of North America, and so deeply covered with glacial debris, could be so completely removed by or- dinary denuding agencies that no more signs of it should appear than are found of such phenomena in the earlier geo- logical epochs ; for the till, or ground-moraine, is not readily removed by the action of water, even where subjected to the shore-waves of the ocean. The bowlders which are washed out of it form a protecting barricade around the base of the deposit, so that islands like those in Boston Harbor, com- posed wholly of till, are as nearly proof against the waves as are those of ordinary rocks. If there were in progress a sub- sidence of the glaciated area of North America, instead bf THE UAUSE OF THE OLAOIAL PERIOD. 433 having the waves wash the glaciated surfaces away gradually from the edges inward, we should find merely an encroach- ment made here and there upon the border during a portion of the subsidence, until, finally, when the waters covered the whole, all but a very thin stratum of the upper portion would be protected from further disturbance. Especially must the till remain in the innumerable buried channels of the glaci- ated region, and over the extensive protected northern slopes. It is thus difficult to conceive how there should ever be any such complete removal of the ground-moraine from the im- mense glaciated area of North America as Mr. Croll sup- poses to have occurred several times over in preceding gla- cial epochs. The facts supposed to prove, by direct evidence, the ex- istence of glacial periods in the various successive geological epochs, can be briefiy stated.* Beginning with some of the oldest sedimentary strata, Professor Archibald Geikie has discovered what he believes to be unmistakable signs of glacial action in the north of Scotland, in Sutherlandshire, on rocks of Cambrian age — that is, just below the base of the Silurian system. Here he reports extensive surfaces of gneiss rock worn into the char- acteristic " rounded bossy surface " of glaciated regions, and this evidently runs under an extensive deposit of breccia of glacial origin, made up of fragments eroded by ice at that early period of glaciation.f Some of the fragments of this overlying breccia are said to be from five to six feet long. A second instance of early glaciation, mentioned by Pro- fessor Eamsay, occurs in the south of Scotland, in Ayrshire and Wigtonshire, in the Lower Silurian formations of that region. Here are extensive sedimentary rocks, containing * On the whole, the best summary of the eridence upon this subject, and the one to which we are mainly indebted for the facts here presented, was given by Sir A. C. Ramsay, Director-General of the Geological Survey of the United King- dom, in his Presidential Address before the British Association of Swansea in 1880. (See "Nature," vol. xxii, p. 388 et seq.) f See communication to "Nature," vol. xxii, pp. 400-403, .28 434 THE IGE AOE IN NORTH AMEBIGA. characteristic Lower Silurian fossils, in which are numerous erratic blocks of gneiss and granite, some of them as many as nine feet in length. Both Dr. Ramsay and Mr. James Geikie believe that the nearest source from which these fragments could come is one hundred miles or more to the north. Their theory is that, in the early Silurian times, the region occu- pied by the Hebrides and the adjoining coast of northern Scotland consisted of an immense granitic mountain uplift, down which glaciers descended to the sea, sending off bowlder- laden icebergs, which wandered to the vicinity of Ayrshire and Wigtonshire, and there dropped their burdens. In India, also, according to Dr. Eamsay, Medlicott and Blanford describe " old slates supposed to be Silurian, con- taining bowlders in great numbers," which these experienced authorities believe to be of glacial origin. They also de- scribe other very ancient transition beds which overlie rocks " marked by distinct glacial striatious." Again, Dr. Eamsay describes bowlder-beds in the south of Scotland, on the Lara- mermoor Hills, south of Dunbar, which "contain what seem to be indistinctly ice-scratched stones." These beds lie " un- conformably on Lower Silurian strata," and are now gener- ally believed by the members of the Geological Survey of Scotland to be of glacial origin. Dr. Ramsay goes on to say : I know of no bowlder formations in the Carboniferous series, but they are well known as occurring on a large scale in the Permian brecciated conglomerates, where they consist of peb- bles and large blocks of stone, generally angular, imbedded in a marly paste ; . . . the fragments have mostly traveled from a distance, apparently from the borders of Wales, and some of them are three feet in diameter. Some of the stones are as well scratched as those found in modern moraines or in the ordinary bowlder-clay of what is commonly called the Glacial epoch. In 1855 the old idea was still not unprevalent that during the Permian epoch, and for long after, the globe had not yet cooled sufficiently to allow of the climates of the exter- nal world being universally affected by the constant radiation of heat from its interior. For a long time, however, this idea THE CAUSE OF THE GLACIAL PERIOD. 435 has almost entirely vanished, and now, in Britain at all events, it is little if at all attended to, and other glacial episodes in the history of the world have continued to be brought forward and are no longer looked upon as mere ill-judged conjectures. The same kind of brecciated bowlder-beds that are found in our Permian strata occur in the Eotheliegende of Germany, which I have visited in several places, and I believe them to have had a like glacial origin. Mr. G. W. Stow, of the Orange Free State, has of late years given most elaborate accounts of similar Permian bowlder-beds in South Africa. There great masses of moraine matter not only contain ice-scratched stones, but on the banks of rivers where the Permian rock has been removed by aqueous denuda- tion the underlying rocks, well rounded and mammillated, are covered iy deeply incised glacier grooves pointing in a direction which at length leads the observer to the pre-Permian mount- ains whence the stones were derived that formed these ancient moraines. Messrs. Blanford and Medlicott have also given, in " The Geology of India," an account of bowlder-beds in what they believe to be Permian strata, and which they compare with those described by me in England many years before. There the Talchir strata of the Gondwana group contain numerous bowlders, many of them six feet in diameter, and in one in- stance some of the Hocks were found to be polished and striated, and the underlying VindJiyan rochs were similarly marked. The authors also correlate these glacial phenomena with those found in similar deposits in South Africa, discovered and de- scribed by Mr. Stow. In the Olive group of the Salt range, described by the same authors, there is a curious resemblance between a certain con- glomerate " and that of the Talchir group of the Gondwana system." This "Olive conglomerate" belongs to the Creta- ceous series, and contains ice-transported erratic bowlders de- rived from unknown rocks, one of which, a red granite, "is polished and striated on three faces in so characteristic a man- ner that very little doubt can exist of its having been trans- ported by ice." One block of red granite at the Mayo salt- mines of Khewra "is seven feet high and nineteen feet in cir- 436 THE ICE AGE IN NORTH AMERICA. cumference." In the "transition beds" of the same authors, which are supposed to be of Upper Cretaceous age, there also are bowlder-beds with erratic blocks of great size. I know of no evidence of glacial phenomena in Eocene strata excepting the occurrence of huge masses of included gneiss in the strata known as riysch in Switzerland. On this question, however, Swiss geologists are by no means agreed, and I attach little or no importance to it as affording evidence of glacier ice. Neither do I know of any Miocene glacier deposits except- ing those in the north of Italy, near Turin, described by the late eminent geologist, Gastaldi, and which I saw under bis guidance. These contained many large erratic bowlders de- rived from the distant Alps, which, in my opinion, were then at least as lofty as or even higher than they are now, especially if we consider the immense amount of denudation which they underwent during Miocene, later Tertiary, and post- Tertiary times.* In North America Professor Shaler would attribute the conglomerates of Jurassic age in the valley of the Con- necticut, in a part of which lie the celebrated bird-tracks, to glacial origin. This he infers, from the great thickness of the beds, the absence of life from the accompanying sand- stones, the subangular forms of many of the pebbles, and from the similarity in composition of the pebbles of that conglomerate with that of those found in the modern di-ift of the region.f Upon this conclusion, howeyer, it is proper to remark that the drift in the lower Connecticut Yalley would, to a great extent, come from the same region, whether brought by ice or water, and the extent to which the pebbles would have been reduced to uniformity and smoothness by attrition depends upon the distance to which they have been rolled, or the length of time to which they have been sub- jected to wave-action. From what appears, the evidence is not clear that the fragments from which the pebbles are * " Nature," vol. xxii, p. 389. f See " Illustrations of the Earth's Surface : Glaciers," by N. S. Shaler and W. M. Davis. Boston: James E. Osgood & Co., 1881, p. 95, TEE CAUSE OF TEE GLACIAL PERIOD. 437 made may not have originated in the near vicinity, and so their subangular condition need not imply glacial agency in transportation. Professor Shaler also is inclined to attribute the exten- sive conglomerate deposits of the Carboniferous age in the Appalachian district of North America to glacial action ; and certainly the extent of these conglomerate deposits un- derlying the coal-beds is surprising. " In Pennsylvania they are about one thousand feet ; in eastern Kentucky and east Tennessee their thickness rises to about two thousand feet." Similar conglomerate deposits everywhere underlie the Car- boniferous system. According to Professor Shaler, " we find it from southern France to Scotland, from Alabama to New Brunswick, in India, and elsewhere." For the most part, however, the pebbles of this conglomerate consist of quartz or quartzite, well rounded, and seldom of larger size than can readily be transported by water ; though Professor Newberry is reported to have " found a bowlder of quartz- ite seventeen inches by twelve inches, imbedded in a seam of coal." Altogether it seems more likely that we have in these conglomerates underlying the Appalachian coal-fields of America the wash brought down by large rivers heading in the mountain plateau toward the north and east, of which the Archaean range on the Atlantic border, together with the hills of New England and the Adirondacks of New York, are but the remnants. That floating ice may have played some part in the streams coming down from these mountain-heights is not improbable ; but it is doubtful whether the facts war- rant us in inferring anything more. Professor Shaler would also attribute a still lower series of conglomerates whose typical development is in eastern Tennessee and western North Carolina, and which rests un- conformably upon Laurentian rocks, to glacial action ; though he confesses that no scratched bowlders have yet been dis- covered in these deposits, but he writes: "Recollecting that we know of no force that is competent to bring together such masses of pebbles derived from a wide-spread surface save 438 THE ICE AGE IN NORTE AMESIGA. glacial action, we are justified in believing that this deposit is a product of ice-action, though the waste has evidently been worked over by water since its production." The thickness of the deposits he estimates to be in some places nearly twenty thousand feet. These deposits correspond in age to the Koxbury conglomerates in Boston, which are about five hundred feet in thickness, and " are composed of materials derived from various points in eastern Massachusetts and southern New Hampshire. The pebbles are rarely over a foot in diameter." But Professor Shaler thinks " their frequently subangular forms and the vride range of sub- stances associated together make it pretty clear that they have a glacial origin." Upon this the same remark is applicable which was made in a preceding section, namely, that along this whole Appala- chian border there were formerly Archaean highlands of in- definite height, of which the stumps are all that now remain in the present hills and mountains. The erosion of these mountains on their western flanks has furnished the material of the vast sedimentary deposits of the eastern part of the Mississippi basin. For all we know, the material spread out over this area of sedimentary rocks was all within reach of rivers coming down from Archaean heights, and so there is no necessity of supposing extensive glacial transportation from more northern water-sheds such as we are compelled to suppose in the glacial age of recent date. The same remark may be extended to. all the evidence adduced in the preced- ing sections concerning a succession of glacial periods.* In all cases they are of such limited character that local glaciers coming down from isolated mountain-masses, such as now come down from the mountains of Alaska, Patagonia, and at no very distant date from those of New Zealand, are sufficient to account for the facts. Eeturning to the point under discussion, it is proper to remark that the conclusions here presented with reference to * See Lyell, " Principles of Geology," vol. i, pp. 203-210. THE CAUSE OF THE GLACIAL PERIOD. 439 Mr. CroU's theory are those pretty generally adopted at the present time by the American geologists best qualified to in- terpret the facts. Thus, among the more eminent American geologists, Mr. G. K. Gilbert wrote, in 1883 : It deals with a series of physical laws and physical condi- tions which interact upon each other in an exceedingly complex way — in so complex a way that meteorologists, who have to deal with only a portion of them, do not claim and scarcely hope for a complete analysis of their combinations. The op- portunities for arguing in a circle are most seductive, and the a priori probability that important considerations have been overlooked is not small. The only manner in which so comprehensive and intricate an hypothesis can be established is by stimulating inquiry which shall lead to corroborative evidence, and this is precisely what CroU's hypothesis, after eight years of wide publicity, has failed to do. If it is true, then epochs of cold must have occurred with considerable frequency through the entire period repre- sented by the stratified rocks ; and iceberg drift, if no other traces, should have been entombed at numerous horizons. It has not been found, however, and of the eight horizons claimed by OroU to show evidence of glacial action, the treatise under consideration [A. Geikie's "Text-Book of Geology"] mentions only two with confidence, and two others with doubt. In the two instances to which queries are not attached, the phenomena appear to indicate local and not general glaciation. If the hypothesis is true, the cold of the Glacial epoch must have been many times interrupted by intervals of exceptional warmth, but little has been added to the evidence adduced by Croll for such an interruption, and in America, where there is now great activity in the investigation of glacial phenomena, the evidence of a single inter-glacial period is cumulative and overwhelming, while there is no indication whatever of more than one.* With this agrees the opinion of President Chamberlin : The various astronomical hypotheses seem to be the worse for increased knowledge of the distribution of the ancient ice- * " Nature," vol. xxvii, p. 262. iiO THE ICE AOE IN NORTH AMERICA. sheet. I think I speak the growing conviction of active work- ers in the Amei-ican field, that even the ingenious theory of Croll becomes increasingly unsatisfactory as the phenomena are developed into fuller appreciation. I think I may say this without prejudice, as one who, at a certain stage of study, was greatly drawn toward that fascinating hypothesis. But the more we know and ponder upon the enormous de- velopment of ice upon the plains of northeastern America, and contrast it with the relatively feeble development and dispersion from the mountainous regions of Alaska, which now bear the greatest glaciers outside of the ai'ctic regions, and the relative absence of such accumulations in northeastern Asia — in short, the more we consider the asymmetry of the ice distribution in latitude and longitude, and its disparity in elevation, the more difllcult it becomes to explain the phenomena upon any astro- nomical basis, correlated though it be with oceanic and aerial currents and geographical features, by whatsoever of inge- nuity.* Professor Le Conte remarks, in similar strain : Of the recurrence of many glacial epochs in the history of the earth there is as yet no reliable evidence, but much evidence to the contrary. It is true that what seem to be glacial drifts, with scored bowlders, etc., have been found on several geologi- cal horizons, but these are usually in the vicinity of lofty mountains, and are probably, therefore, evidence of local glacia- tion, not of a glacial epoch. On the other hand, all the evi- dence derived from fossils plainly indicates warm climates even in polar regions during all geological periods until the Quater- nary. The evidence at present, therefore, is overwhelmingly in favor of the uniqueness of the Glacial epoch. This fact is the great objection to Croll's theory, f The sum of the whole matter, so far as theory is con- cerned, seems to be that as yet we do not know what was the ultimate cause of the Glacial period. But even this is, in * " Proceedings of the American Association -for tlie Advancement of Sci- ence," vol. XXXV, p. 211. f " Elements of Geology," p. S*??. THE CAUSE OF THE GLACIAL PERIOD. 441 the present case, a conclusion of no small importance, since it clears the field for an impartial survey of the terrestrial facts bearing upon the date of the Glacial epoch, and of some other allied questions of great interest. Our inability to assign a definite cause to the Glacial period is, however, no bar to our believing in such a period. We are, indeed, in the matter of glacial theory, troubled with an embarrassment of riches. A great many things may have produced the Glacial period. There is no lack of seemingly adequate causes to be invoked. There is, for all we know to the contrary, always moisture enough in the air, and instabil- ity enough in the crust of the earth, to make a glacial period imminent. A little change in the direction of the aerial cur- rents may readily be conceived to start a period of glaciation, and when it is once started it is somewhat difficult to see how it can stop. Leaving, then, the discussion of these the- ories of ultimate causation where it belongs — to the more enlightened meteorologists of the future — let us take a closer look at the proximate causes which have been at work in determining the limits of the glaciated region in North America. The questions are sure to be asked : "Why is the boundary of the glaciated region in the United States so irregular? "What was the cause of its withdrawing so far north in western New York, and of its sudden bend to the south in eastern Ohio, and of its lobe-like projections in southeastern Indiana and southern Illinois ? And what was the cause, at a later stage, of the lobate contour of the moraines west of Lake Michigan in "Wisconsin, Iowa, Minnesota, and Dakota ? These questions we can only answer by saying that the distance to which the great American ice-sheet penetrated the southern latitudes was evidently not dependent, to any great extent, upon the elevation of the land over which it was compelled to move. The ice did not uniformly move farther south where there was a depression of the land, and the boundary does not ordinarily retreat to the north on account of the higher elevations opposing its progress. South of New England the 442 THE lOE AGE IN NORTH AMEBIC A. glacial front was at the sea-level, and the ocean itself may have kept it from advancing farther. In Pennsylvania the boundary-line crosses the Alleghany Mountains diagonally, being nearly as high on Pocono Mountain, in the eastern part of the State (about two thousand feet), as in the southwestern part of New York, where it is sixty or seventy miles farther north. In Ohio the highest portion of the State is in Logan county, almost directly north of Ripley, in Brown county, the most southern point reached in that State. The unglaci- ated portion in southern Indiana, projecting about seventy miles northward into the glaciated region, is indeed some- what higher than the land on either side, but nowhere is its elevation greater than that of the larger portion of Ohio. The farthest extension of the ice in Illinois is closely coinci- dent with the trough of the Mississippi Valley, and westward of the Mississippi River the edge of the ice withdrew farther and farther north pretty nearly in proportion to the increas- ing elevation of the country, until, at Sim's Station, in the vicinity of Bismarck, Dakota, it is nineteen hundred and sixty feet above tide, and continues thence to ascend northward to a height of three or four thousand feet in the upper valley of the Saskatchewan. There is, thus, a general conformity in its southern exten- sion to the valley of the Mississippi. The ice of the Glacial period as a whole did, indeed, move down that great valley, and its most southern point is in the middle of it, where it it is not more than five hundred feet above the sea ; but it is evident that the total width of the southern portion of this ice-sheet is so great, and the slope itself so slight, that this depression could not have been the main cause of the great extension to the south in Ilhnois. The width of the glaci- ated area from southern Ohio to eastern Kansas, on the thirty- ninth parallel, not far from the extreme limit of glaciation, is nearly a thousand miles. The cause, therefore, of the lo- bate character of the southern boundary must, in all proba- bility, be sought in the irregular areas of excessive snow-fall existing to the north during the advance and continuance of THE CAUSE OF THE GLACIAL PERIOD. 443 the Glacial period. From a glance at the map it would seem, therefore, that the greatest area of snow-fall was somewhere in the vicinity of Lake Superior, and that a secondary area of large snow-fall was in the vicinity of Labrador ; for the south- ern boundary of the glaciated region consists essentially of the arcs of two circles whose centers would fall within the areas indicated. In speaking of these two areas as centers of radiation for the glaciers of the great Ice age in North America, it is not affirmed that the movement received no impulse from still farther north. It is not improbable that the upper portion of Baffin Bay was filled and crossed by the glaciers still lingering over the continental area of Greenland, and that this Greenland ice aided in the movement which covered the northern part of the United States with its icy mantle. But it was probably by reaction rather than by direct action that aid came from that quarter. The accumulations to the north prevented an outflow in that direction, and so compelled a southerly movement from the vicinity of the Laurentian highlands. It is not, however, probable that any Greenland ice ever reached the United States. None of the bowlders so common in the United States are, so far as known, more than a few hundred miles distant from their parent ledges. There was doubtless all the while an ice-stream down Baf- fin Bay toward the Atlantic Ocean, with a movement into it from both sides. But even if this were not the case, the areas south of Hudson Bay and in Labrador would still be the predominant influence in determining the southern out- line of the glacial boundary. The snow that piled up from year to year over those centers would be compelled to move off in the lines of least resistance. ISTow, ice can be an ob- struction to other ice as well as to water ; and what the Greenland glacier probably did was to close up the upper portion of Baffin Bay, so that the excess of snow-fall over the subcenter referred to as north of Lake Superior could not move off to the northeast, but was compelled to spend all its force in a southerly movement. It is evident also that 44i TEE ICE AOE m NORTE AMEBIOA. every subcenter where the snow-fall was larger than the average would, to some extent, make its influence felt upon the shape of the margin. Here is a tield for the mathematician. When the prop- erties of ice are more fully understood from experimental investigations, and the laws of its fluidity brought under mathematical formulae, it will doubtless be possible, from a study of the contour of the glacial boundary, to calculate the position of all the principal areas of largest precipitation during the Glacial period. Those remarkable lobe-like projections in southern Ohio and Indiana, for example, indicate subcenters of accumu- lated ice not far back from the margin. The still more remarkable prolongation of the loops in the kettle-moraine in Wisconsin, and its extension through the States farther west, point, as President Chamberlin sagaciously and cor- rectly supposes, not merely to greater snow-fall over the regions from which the ice-movement radiated, but to the conservative influence of the deeper valleys and depressions to the north, which were filled with ice. These loops of the kettle-moraine sustain a remarkable relation to the valley of Green Bay, and to the northeast and southwest axis of Lake Superior, while the ice-lobe which occupied the valley of the Minnesota and extended to the center of Iowa is evidently related to the great valley of the Red River of the North. It is not improbable that the depth of ice in such a depression as Lake Superior would, by its very thickness, tend for a long time to increase the snow-fall over its own area, and in other ways to resist the antagonistic agencies which were gradu- ally driving the ice-front back to the north. The driftless area of southwestern Wisconsin is situated just where it es- capes these several ice-movements dominated by the depths of Lake Michigan and Lake Superior, and it is to this day — as Professor Dana has pointed out — a region of light precipi- tation. If this discussion of the cause of the Glacial period seems unsatisfactory, the justification is that the present THE CAUSE OF THE GLACIAL PEBIOD. 445 knowledge of the whole subject is in an extremely unsatis- factory condition ; and in this, as in other things, the first requisite of progress is to squarely face the extent of our ignorance upon the question. The causes with which the glacialist deals are extremely complicated, and yet they are of such a nature as to invite investigation, and to hold out the hope of increasing success in mastering the problem. There is opportimity yet for some Newton or Darwin to come into the field and discover a clew with which success- fully to solve the complicated problem which has so far baf- fled us. To the genuine investigator it is a source of inspi- ration rather than of depression to have such an untrodden field before him. Conclusion. — Geology is pre-eminently a terrestrial sci- ence, and there is danger of a misdirection of effort when the geologist forms an alliance with the astronomer. Astro- nomical data are so largely theoretical, and the quantities which the astronomer multiplies are often so nearly infinitesi- mal, that quantitative error is in peculiar danger of becom- ing enormous in large calculations. Hence, we can not count it altogether an advantage that astronomical speculation has been so rife during the past few years in determining the causes and the chronology of the Glacial period. Of the various cosmical theories to account for the Gla- cial period, that of Mr. CroU is by far the most plausible and interesting. It must be admitted that his data concerning the various distances at which the earth is fomid from the sun during the winters of different periods, and concerning the periodical variations in the length of the winters, rest upon well-ascertained facts. It is no doubt true that about one hundred thousand years ago the winters were at times several days longer than now, and the northern hemisphere was receiving daily considerably less heat than now, since it was several millions of miles farther away from the sun. But the distribution of the earth's heat by winds and oceanic currents is a subject concerning which much less is known. The phenomena presented in a hot-house are puz- 446 THE ICE AGE IN NORTH AMERICA. zling. The heat of the sun goes through the glass, but can not readily get out again. It is well known, also, that a shght increase of moisture in the atmosphere, or a slight film of cloud over the sky, prevents a frost. The real prob- lem lies, therefore, in the meteorological iield. Now, during Mr. Croll's " aphelion " winters, the summers are in " perihel- ion," and the summer heat in this hemisphere while in peri- helion is more intense than at other times. In fact, the earth receives at all times the same absolute amount of heat from year to year. Thus, we can not avoid the conclusion that the predominant influence in climate may consist in the power of moisture-laden atmosphere to retain and transport the heat, thus determining its distribution. As a matter of fact, we find that the equator is not so hot as theoretically it should be, and the arctic regions are by no means so cold as, on Croll's theory, they ought to be. The difference between the mean temperature on the equator and that at the coldest point on the sixty-seventh parallel is really only about 76° Fahr. ; whereas, if the temperature at these points were pro- portionate to the amount of heat received from the sun, the difference would be 172°. Such facts as these lead meteor- ologists to regard Mr. Croll's theory with much less favor than formerly. But the glacialist is not so much concerned to know the ultimate cause of the Glacial period as he is to collect the facts which characterize the period. The truth is, that the meteorological forces of Nature are so powerful and complex that there is an embarrassment of riches in the field of gla- cial theory. It is easy to see that a slight increase of snow- fall over the Alps would cause a permanent enlargement of all the glaciers of Switzerland, and threaten every interest of that republic, and perhaps of central Europe ; for the ulti- mate effects of a climatic disturbance in one such center can not well be estimated. Much light upon the condition of things during the Gla- cial period in America must yet come from a careful study of the lobate contour of the terminal moraines. The shapes THE CAUSE OF THE GLACIAL PERIOD. 447 of these moraines, coupled with what may yet be learned concerning the nature of ice and concerning the shifting course of the atmospheric currents, will, in all probability, eventually furnish the data for the solution of the question of the true cause of the Glacial period. A fair field here invites the active and prolonged attention of some future meteorological Darwin or Newton, and promises immortality such as they have attained. CHAPTER XX. THE DATE OF THE GLACIAL PEKIOD. Two causes have combined in recent years to favor erro- neous calculations concerning glacial chronology. Of these, the first has been the almost unquestioned acceptance of the astronomical theory subjected to examination in the preced- ing chapter. If Mr. Croll's theory of the cause of the Gla- cial epoch is correct, then we should no longer speak of an ice age, but of a succession of such ages, whose dates could be readily determined from a table showing the periods of high- est eccentricity in the earth's orbit. According to this table, the modem period most favorable to the production of a glacial epoch began about two hundred and forty thousand years ago and ended about seventy thousand years ago. The whole intervening time was one of high eccentricity, when, during the recurring intervals in which the winters occurred at aphelion, the excess of winter over summer ranged from fourteen to twenty-six days, and the intensity of the heat, received from the sun during those aphelion winters was ten per cent less than at the present time. During the time in- tervening between seventy thousand and two hundred and forty thousand years ago, there occurred, therefore, according to this theory, a succession of glacial and interglacial periods in which geologists and archaeologists are invited to distrib- ute their remarkable discoveries concerning glacial man. Undue confidence in this theory has had no small infiuence in diverting attention from the more legitimate lines of in- vestigation. A second source of error has been an incorrect interpre- THE DATE OP THE GLACIAL PERIOD. 449 tation of Lyell's principle of uniformity in Nature's opera- tions. This has led to an exaggerated estimate of everything pertaining to geological time. There is a prevalent popular impression that all geological events happened a great while ago. This impression arises largely from the imperfect ap- prehension of the extent to which changes are now going on in the world. In reality, however, Lyell's greatest service consisted in quickening our conception of the instability of the present condition of things, and of the intensity of present natural forces. He riveted the attention of his readers upon the cnmulative effect of such earthquakes as are now of daily occurrence, and occasionally of enormous influ- ence, and continually reminded them of the frequency and intensity with which new volcanoes are now from time to time bursting forth. Continuity, therefore, rather than uni- formity, is the word which most properly expresses the prin- ciple underlying the theories of this great geologist. A pe- rusal of his works makes it evident that evolution, and not dull repetition, characterizes the processes of Nature. There is therefore really nothing in Lyell's working principle to raise any antecedent presumption in favor of an extreme antiquity for the Glacial period. On the contrary, the present tendency, both among as- tronomers and geologists, is to diminish estimates of geologi- cal time in almost every period. The himdreds of millions of years claimed, not long ago, as necessary for the deposition and metamorphism of the geological strata, and for the ele- vating and eroding forces to produce the present contour of the earth's surface, have on geological evidence been reduced to much more moderate limits. Thirty million years is now shown to be ample for the deposition, by forces still in op- eration, of all the sedimentary strata of which we have knowledge. At the same time the astronomers affirm that life can not have existed on the earth earlier than twelve million or fifteen million years ago.* Before that period * See Newcomb's "Popular Astronomy," pp. 51S-519. 29 4:50 THE IGEAOE m NORTH AMEBICA. the radiation of the sun's heat was so intense that life must have been impossible upon our globe. But he who pauses to reflect upon how long a period one million years is, and takes the pains to multiply the annual changes of the present time by that number, will not feel cramped by the limits which astronomers are now setting to geological time. In this general shortening of our concep- tion of geological periods, the evolutionists also find no small relief in the more moderate estimates made concerning the date of the close of the Glacial period ; for it is very clear that the changes in species since the great Ice age are trifling. The flora and fauna of the world during the Glacial period were essentially the same as those of the present time. Even man is believed to have been an inhabitant of America, as well as of Europe, before the ice had withdrawn from the head-waters of the Delaware Eiver, and from the mountains of Scotland and the north of England. If these changes in the organic world have been so slight since the Glacial epoch, it follows that, the further back that period is placed in time, the greater are the difiSculties of the evolutionists. The more the evolutionists are limited in time by the astrono- mers, the more do they need a rapid rate of change as the basis of their calculations. If, therefore, the Glacial period should prove to have ended only ten thousand years ago in- stead of seventy thousand, the Darwinian would be relieved from no small embarrassment. Thus, so far as there is likely to be any odium theologiGwm in the case, the desire to sup- port a short biblical chronology and the counter-desire to dis- credit Darwinism, and viae versa, may be left to counteract each other. In view of the doubt expressed in the preceding chapter concerning Mr. OroU's theory, it does not seem proper for geologists to rest satisfied with mere astronomical calcula- tions respecting glacial chronology. We may, therefore, be permitted to turn to the more congenial task of considering the direct geological evidence bearing on the question. In this field there are three classes of facts to which we can 4:52 THE lOE AGE IN NORTH AMERICA. confidently look for light: 1. The amount of erosion and disintegration which has occurred since the Glacial period. 3. The extent to which lakes and kettle-holes have been filled with sediment. 3. The apparent freshness of organic re- mains in glacial deposits. Beginning with the extent of erosion which has taken place since the withdrawal of the ice, our attention is, natu- rally enough, directed first to the gorge below the Falls of Niao-ara. How this comes to be a glacial chronometer has already been seen.* The old outlet of Lake Erie, which must have existed as the result of preglacial erosion, was filled up during the great Ice age, and the Niagara Kiver is the outlet of the pond thus created. Originally the water plunged over the escarpment at Queenston, about seven miles below the present cataract. This escarpment is formed by an outcrop of a thick deposit of Magara limestone, the summit of which is about three hundred feet above the level of Lake Ontario. This is underlaid by a softer rock, which is more rapidly disintegrated than the upper strata ; hence the upper strata always project beyond the lower, and every- thing favors the continuance of the cataract at the head of the gorge as it wears back. The problem is to de- termine how long the Niagara Eiver has been in wearing out the gorge between Queenston and the present cata- ract. The solution of that problem will furnish an answer to the other question, How long has it been since the ice- barrier across the valley of the Mohawk was removed, so that the damraed-up waters of Lake Ontario and Lake Erie would subside sufficiently to permit the formation of the cataract at Queenston and the commencement of erosion in the present gorge ? The problem is comparatively simple, and its bearing upon the date of the Glacial period is clear. The Falls of Niagara have receded about seven miles. The conditions have been from the first so nearly uniform that the present rate of erosion can not differ largely from * See chapter xii, p. 273 et mq. THE DATE OF THE QLAGIAL PERIOD. 453 the average rate. There is some evidence that a part of the work above the Whirlpool had been done by a local stream FiQ. 108.— Section of strata along the Niagara gorge from the falls to the lake. Ij 3, strata of hard rock ; 2, 4, of soft rocli. which formerly passed from the Wliirlpool westward to St. Davids, since there is no doubt of the existence of a filled-up preglaeial channel running from the Whirlpool to St. Dav- ids. But the evidence that this channel extended above the Whirlpool toward the present cataract is so imperfect that we must leave it out of the question, and take the whole length of the gorge from Queenston to Niagara as our dividend. The problem remaining is to find the rate of recession, and this will serve as a divisor. The comparative youth of the Niagara gorge is evident from the present condition of its mouth at Queenston. This is narrow, and its walls abrupt ; but it is well known that, by the inevitable action of natural forces, the mouth of a river- gorge must become, in process of time, very much enlarged, since from the beginning its sides have been exposed to the eroding action of the elements and to the undermin- ing action of the river. In the unglaciated region the mouths of such gorges are universally wide and Y-shaped, and the banks much obscured at the bottom by the accumu- lation of debris. In this respect the mouth of the present gorge at Queenston is in striking contrast with that of the old gorge which opened at St. Davids. As will be seen by reference to the map, this, though narrower where it left the Whirlpool than the present main gorge below the Whirl- Fig. 109. — Map of the Niagara River below the falls, showinsf the buried channel from the whirlpool to St. Davids. Small streams a. b, c, fall into the main gorge over a rocky escarpment. No rock appears in the channel at d, but the rocky escarpment reappears at e. THE DATE OF THE GLACIAL PERIOD. 455 pool and not over half as long, is still at its mouth by St. Davids several times as wide as that of the present Niagara gorge at Queenston. Coming to the main question, ai)d taking the whole of the gorge from Queenston to the present cataract as the work done by the Niagara River since the ice-barrier in the valley of the Mohawk gave way, the problem is to find the rate of recession. Until very recently the estimates of this rate have been scarcely more than mere guesses. The emi- nent French glacialist Desor thought it could not have been greater than one foot in a century, which would place the beginning back 3,500,000 years. In 1841 Sir Charles Lyell and Professor James Hall examined the gorge together ; and Sir Charles, in his lectures in Boston before the Lowell Insti- tute soon after, estimated that the maximum rate of reces- sion could not be greater than one foot a year, which would fix the minimum date of its beginning at about thirty-five thousand years ago. On the contrary, all the guides of that period who had observed the falls for many years, were con- fident that the rate of recession was as much as two feet a year ; * while Mr. Bakewell, an eminent English geologist, who had given much personal study to the question, estimated that, for the forty years previous to 1830, the rate of recession had been about three feet a year. Mr. Bakewell's son care- fully reviewed the phenomena again in 1846, in 1851, and in 1856, and found no occasion to revise his father's estimate.f To furnish the basis for more accurate calculations, Pro- fessor James Hall had a map of the falls made in 1842, from a trigonometrical survey, so that there should be a fixed standard for future comparison. Within the past few years, accurate surveys have again been made, both by.the geolo- gists of the State of New York, and by members of the United States Coast Survey. In 1886 the American Associa- tion for the Advancement of Science held its annual meeting * Lyell's "Travels in America" (first series), vol. i, p. 27. I "American Journal of Science," vol. Ixxiii, 1857, pp. 87, 93. 456 THE ICE AGE IN NORTH AMERICA. at Buffalo and great interest was naturally centered upon this question of the rate of the recession of the falls. Mr. Gr. K. Gilbert, of the United States Geological Survey, whose au- thority is unsurpassed on such subjects, gave it as his conclu- sion that the " maximum length of time since the birth of the falls, by the separation of the lakes, is only seven thousand years, and that even this small measure may need significant reduction." At the same time Mr. R. S. Woodward,* of Washington made a new survey, and gives the results definite- ly as follows : The length of the front of the Horseshoe Fall is twenty-three hundred feet. Between 1842 and 1875 four and a quarter acres of rock were worn away by the recession of the falls. Between 18Y5 and 1886 a little over one acre and a third disappeared in a similar manner, making in all, from 1842 to 1886, about five and a half acres removed, and giving an annual rate of recession of about two feet and a half per year for the last forty-five years. But in the central parts of the curve, where the water is deepest, the Horseshoe Fall retreated between two hundred and two hundred and seventy-five feet in the eleven years between 1875 and 1886. The foregoing estimates concerning the recession of the Niagara gorge assume a uniform rate, and that all the work has been done since the Glacial period. As to the first of these assumptions. Dr. Julius Pohlman,f of Buffalo, adduces some evidence to show that the present course of the Niagara from the Whirlpool to Queenston follows an old line of drain- age, in which a small stream had eroded a shallow valley previous to the Ice period, and thus, by reducing the thick- ness of the upper layer of hard limestone along its course, had greatly facilitated the work of recession, when the whole torrent of Niagara began to pour over the escarpment. Dr. Pohlman has also greatly increased our conception of the work already done before the Glacial period by the stream which had its exit from the Whirlpool to St. Davids. This * Report in " Science," September 3, 1886. f " Transactions of the American Institute of Mining Engineers," 1888. 458 THE ICE AGE IN NORTH AMERICA. stream, composed of the waters of the Tonawanda and Chip- pewa Creeks, was of considerable extent, and by its action had doubtless predetermined the course of the present river above the Whirlpool, and may actually have worn a consid- erable part of the present gorge above the Whirlpool. Tak- ing all things into consideration, it seems, to both Mr. Gil- bert and Mr. Woodward, not improbable that the average rate of recession is as much as five feet per annum ; so that Mr. Woodward and Mr. Gilbert substantially agree in their results, and seven thousand years may, with a good deal of confidence, be taken as the age of the lower part of the Ni- agara gorge. This, of course, does not take us back to the period when the front of the glacier lay in the head-waters of the Delaware and the Little Miami River, and when gla- cial floods were depositing the gravel at Trenton, New Jer- sey, and at Loveland and Madisonville, Ohio, where Drs. Abbott and Metz have found palaeolithic implements ,*• but it does bring us back to within a comparatively short distance of that period, the difference being merely the time necessary for the melting back of the ice from the summit of the Cats- kills to the southern flanks of the Adirondacks, and from the water-partings of the Ohio to the north shore of Lake Erie. A second typical place for the study of the recession of post-glacial waterfalls is presented in the gorge of the Missis- sippi River below the Falls of St. Anthony at Minneapolis. The problem here presented has been carefully studied by Professor N. H. Winchell, the State Geologist of Minnesota, who thinks he can pretty closely approximate to the truth concerning its antiquity.* Prom the Falls of St. Anthony to Port Snelling the gorge between the rock-bluffs is somewhat less than a quarter of a * Professor Winchell's estimates were first published in the " Quarterly Jour- nal of the Geological Society of London for 1876," pp. 886-901 ; afterward in the "Fifth Annual Report of the Minnesota Geological Survey," pp. 156-189. The whole discussion is restated and reviewed, with abundant illustrations, in the "Geology of Minnesota," 1882-'85, vol. ii of the "Final Report," pp. 813- 316, 340, 341. The extracts and summary here given are from the latter work. TEE DATE OF THE GLACIAL PERIOD. 459 mile in width, and the rock has a freshly broken appearance, the large fragments thrown down by the action of the water on the easily crumbled sand-rock, as the falls have receded, still existing in the talus along the bluffs. Throughout this distance (about eight miles) the strata are horizontal, the thick- ness of the drift-sheet overlying them nearly uniform, and all other conditions, so far as they can be seen, that would affect the rate of recession, seem to have exerted an unvarying in- fluence. The inference is inevitable that the rate of recession has been practically uniform between the two points named. There is an aspect of age, and long weathering, presented by the rock in the bluffs of the Mississippi below Port Snelling. It has a deeply changed color, a light- yellow, oxidized exterior, which marks all old bluffs. The blue color is found at greater depths from the surface than it is in the rock of the bluffs above Port Snelling. This stained condition also pervades the lime-rock at the mouth of Bassett's Creek and at the quarries in the ancient river-bluffs near the mouth of Shingle Greek, on both sides of the river. Another notable difference between the bluffs above Port Snelling and those below consists in the absence of caves, and subterranean streams entering the river, above Port Snelling. Although the Trenton limestone exists in full force about St. Paul, in the bluffs east and north of the city, yet it had been cut through by some means prior to the drift so as to allow the entrance and exit of streams of water at levels below its horizon through the sandstone. None such are found above Port Snelling. The surface drainage is shed by the limestone, and is precipitated over the brink of the gorge, forming several beautiful cascades. When such streams enter the river below Port Snelling, they either enter some subterranean passage and appear at the mouths of caverns in the sandstone, or as springs in the drift along the talus, or they find an ancient ravine down which they plunge, by a se- ries of rapids over bowlders, to the river-level, rarely striking either the lime-rock or the underlying sand-rock. Again, the rock-bluffs at St. Paul, and everywhere below Port Snelling, are buried under the drift-sheet. Their angles are sometimes seen Jutting out from some wind-beaten corner, but nearly everywhere they are smoothed over by a mantle of drift and 460 TEE ICE AGE IN NORTH AMERICA. loam. Even the immediate river-bank, where the lime-rock should be intact, shows that it has been extensively disrupted and its debris, often coarse and water-worn, in pieces from four to ten feet long, is mixed with the coarse bowlders, gravel and the drift, at the height of fifty to seventy-five feet above the water-level, the heterogeneous mass lying on the worn upper surface of the St. Peter sandstone. But above Fort Snelling the upper edge of the lime-rock is intact all the way to the falls, and shows afresh-cut section. It is surmounted by a continuous sheet of drift, which rises from the water-level in one blufE coincident with the rock-cut. Its individual strata show that they were cut by the recession of the falls in the same manner as the strata of the rock. They do not conform in their undulations to the outline of the rock, as if the gorge were present when they were formed, as at St. Paul. There is no spreading of loam over these cut edges, except such as has fallen down from above at the time of their removal or subsequent to it. At Fort Snelling, the direction of the Mis- sissippi changes abruptly at a right angle. The change is caused by entering the wide gorge which runs in that direc- tion. This gorge is that in which the Minnesota runs, and is out of proportion with the amount of water which it carries. This valley continues in the same direction, and with the same width, beyond the confluence of the Mississippi, but takes the name of the latter stream. At one mile below the mouth of the Minnesota it is a mile and a half wide. These features of greater age, pertaining to the bluffs of the Mississippi below Fort Snelling, are seen in the old rock- bluffs of the river above the mouth of Bassett's Creek as far as to Shingle Creek. The rock there is deeply changed in color, and is hid by the drift, and the bluffs, as left by the more an- cient river, are far apart, the old gorge being three or four times as wide as that between the falls and Fort Snelling. These rock-bluffs, consisting of the same limestone as that which at the falls is below the water, here rise from thirty to forty feet above the river, and are buried under loam, or under drift and loam. This part of the old valley continues south- wardly, by way of Bassett's Creek (below its last turn), across the western suburbs of Minneapolis, through the valleys occu- THE DATE OF THE GLACIAL PERIOD. 461 pied by Lakes Oalhouu and Harriet, and joins the Minnesota at some point above Fort Snelling, the precise locality being Fio. 111.— Map of Mississippi Eiver from Port Snelling to Minneapolis and the vicinitj', showing the extent of the recession of the Falls of St. Anthony since the great Ice age. Notice the greater breadth of the valley of the Minnesota Eiver as described in the text. hid by a subsequent deposit of drift. It was cut down into the St. Peter sandstone over one hundred feet at least, as shown by the well at the Sumner school-house, and about two 462 THE lOE AGE IN NORTH AMERICA. hundred and seyenty-five feet, as shown by the deep well at the Lakewood Cemetery. This would show that probably the an- cient valley of the Minnesota where it passes Fort Snelling, and all the way through Ramsey county and below, has been filled more than two hundred feet by drift that originated since the excavation of the gorge. This supposition is borne out by all borings that have been made between the rock-bluffs at lower points, as at West St. Paul and at Lake City. Such excavation is not found in the river-gorge between Fort Snel- ling and the Falls of St. Anthony ; but, below the water, are found, first, some large fragments of limestone, and some bowl- ders of foreign origin, the whole being generally less than twenty-five feet in thickness, and below that the undisturbed St. Peter sand-rock is found, suitable for the foundation of piers for bridges. These facts warrant the conclusion that that part of the Mississippi gorge above Fort Snelling has been excavated by the recession of the falls since the last general drift movement, and that prior to that event there was a gorge which passed from the present channel of the Mississippi at the mouth of Bassett's Creek southward to the great gorge of the Minnesota at some place above Fort Snelling. It is probable that this gorge was then occupied by waters that drained from the northern part of the State, and had existed through many ages, dating back to pre-Cretaceous times. It seems to have been filled first by a blue till, or partly filled, and to have remained free for the passage of the Mississippi during the on-coming of the Glacial epoch, till the advent of the ice of the last Glacial epoch, when morainic accumulations so choked it that the water of the river was driven out and compelled to seek an- other passage to the Minnesota. When this last event took place, the Falls of St. Anthony probably began at Fort Snel- ling, the water being precipitated over the rock-bluff of the pre-existing old gorge, unless the whole valley was too deeply buried under water. Whether this was at the beginning or at the acme of cold, or at the recession of the ice, is a question which may well be considered, but at this time the only point that is claimed is that it was not earlier than the beginning of the last Glacial epoch, and was probably near the acme of cold. THE BATE OF TEE GLACIAL PERIOD. 463 Having thus established the post-glacial origin of the gorge below the FaUs of St. Anthony, the next point was to determine the rate at which the recession has been proceed- ing. Fortunately, upon this point an abundance of evidence is available. The falls were first visited and described as early as 1680 by the Jesuit missionary Hennepin. His de- scription is found in tlie Amsterdam edition of his works, printed in 1704. The falls were again visited in 1766, eigh- ty-six years later, by Carver, another Jesuit missionary. In addition to his description this traveler made a sketch of the falls, which was engraved to accompany his travels, pubKshed in London in 1778. Subsequent travelers who describe it are Major Z. M. Pike, in 1805 ; Major Stephen S. Long, in 1817 ; Schoolcraft, in 1820 ; Professor William Keating and Mr. Beltrami, Eev. "W. T. Boutwell and Schoolcraft, in 1832 ; and Mr. G. W. Featherstonhaugh, in 1835. In addition vari- ous artists have gathered descriptions of the falls as they ap- peared in 1842, 1848, 1853, and in 1857, and daguerreotypes were taken in 1851 ; while in 1853, before the erection of saw- mills, Mr. J. "W. Bond gave a careful description of the falls as they then existed, and numerous living witnesses fix their position in 1856, when artificial changes were introduced, which so modified the rate of recession as to disturb further calculation. The period, then, during which evidence is avail- able for calculation is that between Hennepin's visit in 1680 and the year 1856 — one hundred and seventy-six years. The descriptions are so minute that Professor "Winchell is able to fix beyond doubt the various stages of recession between these dates. In 1680 the falls were near the south end of Hennepin and Spirit Islands, not far above the present Tenth Ave- nue Bridge. In 1766, at the time of Carver's visit, the falls had receded about four hundred and twelve feet, and were at Carver's Island. In 1 856 the west falls were about five hun- dred feet below their present position, which is now made sta- tionary by artificial means. According to Professor Winchell, the recession from 1680 to 1766, between - Hennepin and 464 TEE ICE AGE IN NORTH AMERICA. Carver, was four hundred and twelve feet ; and between 1766 and 1856, six hundred feet, making a total between 1680 and 1856, of one thousand and eighteen feet. " These give re- spectively the rates 4-79, 6-73, and 5-08 feet per year, and for the corresponding periods necessary for the recession of the falls from Fort Snelling (a distance of a trifle over eight miles) 8,819 years, 6,276 years, and 8,315 years. The aver- age of these three results is 7,803 years." Professor "Winchell then proceeds to discuss the possible elements of error in this calculation : 1. That arising from difference in the volume of the river. The terraces already described in the chapter on " Preglacial Drainage," as characterizing both the Minnesota Eiver and the upper Mississippi, reveal the existence of enormous floods during the closing stages of the Glacial pe- riod. Indeed, these floods in the Minnesota River were so high as to fill it up to the level of the lime-rock at Fort Snel- ling, about one hundred feet. During the existence of this high water, therefore, there could have been no cataract at Fort Snelling or farther up the Mississippi. The Falls of St. Anthony could have begun only after the floods of the Minnesota began to shrink so as to uncover the lime-rock at Fort Snelling. 2. Difference in the height of the falls at various points from Fort Snelling up to its present position. This is shown to be comparatively insignificant, so that it can be left out of the account. 3. The stage of the Glacial period when the recession began. Upon this we quote again at length : This point has already been considered in the possible va- riations in the volume of the river. It is probable that the Mississippi, in diminutive form, began to flow in its new channel at the acme of the cold,* since the moraine of the sec- ond Glacial epoch runs across the country, approximately through this region, and since it would have remained in its ■ See map of Minnesota in next chapter (Fig 1S8). THE DATE OF THE GLACIAL PERIOD. 465 preglacial channel till it was driven out by the encroaching moraine. It was the easier removed from its old channel by reason of its reduction in volume. When it began its course in its new channel, it flowed over a broad plain of gravel and sand, the then latest accumulations of glacial torrents. This plain of gravel and sand extended throughout the adjoining space now occupied by such drift deposits. The same kind of deposits filled the whole Minnesota Valley, from side to side, and rose as high as the plains back of Fort Snelling. The river, being comparatively small, had but little efEect on these deposits. If it excavated any channel, the torrents from the ever-present glacier-ice filled them at once — indeed, it exca- vated, it refilled, as it was glacier-born. It was on the retire- ment of ice, bringing a greater drainage area into contribution to swell the main streams at this latitude, that these rivers began to deposit the fine loam-sand which covers the coarse gravel and sand of these terraces. It was still later, when the rivers were shrunk, by the partial or complete with- drawal of the glaciers from their remote sources, that they began to excavate through the loam and the gravel and sand and finally entered on the slow erosion of rock-gorges. Thus it appears that the date from which the recession of the falls must be reckoned was after the outlet of Lake Agassiz had been opened toward the north, one of the last acts of the Ice age. . . . Finally, if all the supposed irregularities be allowed their full force, and all the elements of doubt be admitted, their combined efEect would not, at the most, more than slightly modify the result. And even if it should double the first result, or should reduce it to one half, the chief value of the calcula- tion is not impaired. That consists in showing the lateness of the last G-lacial epoch compared with the enormous time that has sometimes been supposed to have elapsed since its de- parture. If the occurrence of our winter in aphelion, caused by the precession of the equinoxes and the revolution of the line of the apsides, about eleven thousand three hundred years ago, was the cause of our last Glacial period, it follows that it re- quired about thirty-five hundred years for the withdrawal 30 466 THE ICE AGE IN NORTH AMERICA. of the ice-margin from the vicinity of Fort Snelling to that place where the discharge of Lake Agassiz was opened to- ward the north, reducing the Minnesota to nearly its present size. This change must have given prominence and erosive effect to the waterfall at Fort Snelling, if it did not give it birth. These calculations concerning the age of Niagara and the Falls of St. Anthony are amply sustained by the study of various minor waterfalls and gorges in Ohio to which I have myself given special attention. For example, at Elyria, twenty-five miles west of Cleveland, Black Eiver plunges over the outcropping Waverly sandstone, and flows onward to the lake through a wide valley in the Erie shale, which was doubtless preglacial, though no buried channel above has yet been discovered. The gorge below the falls, which has been eroded since glacial times, and which approximately represents the work done by Black River during that time, is only a trifle over two thousand feet long. The water flowing over the falls represents the drainage of about four hundred square miles, and the sandstone which forms the precipice over which the water plunges is underlaid by soft shale very favorable to rapid erosion. A few years ago a mass of rock fell which was so large that the concussion shook the whole town and produced the semblance of an earthquake. With the present forces in operation at this point, it would seem incredible that the average rate of re- cession should not be considerably more than one foot in fifty years. Yet thus infinitesimal would be the rate if one hundred thousand years must be allowed for the time separat- ing us from the birth of the present waterfall at Elyria. The shortness of this and other similar gorges in that region points to a great reduction of the prevalent estimates of glacial chronology. Another interesting confirmation of this moderate esti- mate is to be found in Paint Creek Valley, in the southern part of Ohio, to which attention was directed in a previous chapter. As was discovered by Professor Orton several years THE DATE OF TEE GLACIAL PERIOD. 467 ago, this stream, a few miles above its junction with the Scioto, at Chillicothe, abandoned its preglacial valley in a most singular manner.* The preglacial valley of Paint Creek for about twenty miles above its junction with the Scioto runs in a northeast direction from the town of Bain- bridge. The valley is nearly a mile wide at the bottom, and about five hundred feet below the general level. But the pres- ent stream, after it has abandoned this old valley, occupies for two or three miles a narrow gorge not over five hundred feet wide, cutting directly through the table-land, and re- entering the old valley considerably lower down in its course. The only satisfactory explanation of this is found in a study of the local glacial phenomena. The lower or northeastern part of this preglacial valley is exactly on the line of the gla- cial boundary, and was for a certain period obstructed by the most advanced portions of the glacier, which dammed up the water and raised it to a level at which it would be forced in front of the ice across a tongue of the table-land, thus eroding the present channel. f This portion of the channel, as already indicated, is about three miles long, from three hundred to five hundred feet deep, five hundred feet wide at the top, and two hun- dred at the bottom. The walls near the top consist of fifty or sixty feet of Waverly sandstone, while all below is a soft shale crumbling very readily. The question in glacial chro- nology is to find the age of this gorge, which is clearly post- glacial. The true solution of the problem comes from a study of one of the lateral gorges formed by a small tributary entering the main gorge midway from the south. This tributary, though dry a portion of the year, is at other times a raging torrent, and drains an area of two or three square miles. Yet in the soft shale, so favorable for rapid erosion, it has worn a gorge less than six hundred feet long, but hav- ing a mouth of nearly the same width where it joins the * " Geological Survey of Ohio," vol. ii, p. 653. \ See map, p. 333. 4f!S THE ICE AGE IN NORTH AMERICA. main channel. It can scarcely be possible that these forces have been in operation iu their pi'esent position for many thousand years ; for, ac- cording to the testimony of Mr. Long, who has been a resident upon the ground for fifty years, and has definite data for calculation, this tributary creek has worn back sev- eral feet since his re- membrance. If the rate of recession for this trib- utary gorge were as little as one foot in twenty years, only twelve thoii- sand years would be re- quired for the accf)ni- ])lishment of the work done. If we should go liaek to the period as- signed by ilr. Croll's the- ory to find the Glacial period, the rate of recession would be incredibly slow, and far below what is pretty certainly the rate at the present time. Another means of measuring the amount of erosion since the Glacial period is found in post-glacial river-valleys by estimating the amount of material which has been car- ried out by the present streams fi-oni the glacial deposit itself. Professor Ilicks, of Granville, Oliio, reported in 18S4,"" some iin])ortant results of such an investigation in the valley of Kaccoon Creek, Licking couuty, neai- the glacial border. The present flood plain of this creek is now bordered on either side by graxel terraces aboiit fifty feet high, which are I (, 11 —II I t 11)1 t 1 tr^ h 1 tl 1 III sum I 1 \ the fi i ncnceswli n r s r n his j roceedcd to a ^,rcat e\ftiit In t 1 stitcs teol(i(»icil Sumv ) ( hail I rl •- " Baptist Quarterly " for July, I8S4, See also Fig. 81, p. 284. THE DATE OF THE GLACIAL PERIOD. 469 evidently the remains of a modified glacial deposit formerly filling the wliole valley to that height. Since the Glacial period the present stream has been occupied with the task of slowly removing this material. The number of cubic yards ■\\diich it has already carried away can be approximately esti- mated. The rate of removal is more difBcult to determine. Assuming the rate to be the same per cubic foot of water as that which is transported by the Mississippi River past l^ew Orleans, which doubtless is far too small, the time recpiired would be, according to the calculation of Professor Hicks, less than fifteen thousand years. A similar estimate I have made for Plum Creek, a small stream ^^'hich passes through Oberlin, Lorain county, Ohio. This creek is about twelve miles long, and drains a basin whose area is about twenty-five square miles, and its course is all the way through a bed of till, with no rocky ob- structions. The channel which it has worn aver- ages about twenty feet in depth above its iiood- plain, and five hundred feet in width. Calcula- ting from these data, we have the following re- sults : The total amount of material eroded from the trough of the creek is 633,600,000 cubic feet. Estimating the rainfall at forty inches per year, and that one Fig. 11.3.— a country, in contrast witb tluit on tlie opposite page, in wtiicli tlie draina^^e has been disturbed by glacial deposits and the streams ore beginning to wear new channels, (Cham- berlin.) quarter finds its way through the channel, the total annual discharge is .580,800,000 cubic feet of water. Accepting for this, also, the estimate of Humphreys and Abbott, for the 470 THE lOE AGE IN NORTH AMEBIOA. Mississippi liiver, that one foot of silt is carried away by every 2,610 cubic feet of flowing water, we find that this in- significant stream carries away 222,528 cubic feet of silt every year. Estimating that one half of this comes from the surface and tributary streams, leaving one half to come from the valley itself, it would require at present rates only 5,694 years for the erosion of this valley. But allowance must be made for the slower rate of erosion when the country was cov- ered with forests. This rate, however, can scarcely be re- duced for the valley to more than one half, which would make the period not far from eleven thousand years. In view of the effects produced, it is incredible that this stream could have occupied its bed for a much longer period than that, and the time may be considerably shorter. Another class of fncts which seems to set moderate limits to glacial chronology relates to the amount of superficial ero- sion of glacial deposits of various sorts, and the extent to which the rocks have been disintegrated since that period. President T. C. Chamberlin, when State geologist of Wisconsin, remarked that no sensible denudation had taken place there since glacial times.* Even Mr. CroU expresses surprise at the small amount of erosion which has taken place since the kames of Scotland were deposited. Both in Europe and in America these peculiar relics of the Glacial period retain a sharpness of outline which it is difficult to be- lieve could have survived the protracted period of one hun- dred thousand or even of forty thousand years, according to Hitchcock's reckoning. When, also, one considers the chemi- cal agencies at work to decompose the rocks everywhere protected by a covering of till, the freshness of the glaciated surfaces never ceases to be a cause of astonishment. Ma- terial is accumulating for more trustworthy estimates at some time from such data ; but it is as yet too early to formulate the conclusions more definitely. Closely connected with the preceding class of facts are * " Geology of Wisconsin," vol. ii, p. 632. THE DATE OF THE GLACIAL PERIOD. 471 tlie observations made upon the extent to which the lakes, dating from the Glacial period, have been filled with sedi- ment. Little reflection is required to make it evident that our present lake-basins could not always have existed ; for, except where counteracting agencies are at work, the " wash " of the hills will, in due time, fill to the brim all inclosed areas of depression. Mr. Upham, of the Minnesota Geo- logical Survey, expresses surprise at the small extent to which the numerous lakes of that State have been filled with the sediment continually washing into them. " The lapse of time since the Ice age has been insuificient for rains and streams to fill these basins with sediment, or to cut out- lets low enough to drain them, though in many instances we can see such changes slowly going forward." * Dr. E. Andrews, of Chicago, has made calculations, de- serving of more attention than they have had, concerning the rate at which the waters of Lake Michigan are eating into the shores, and washing the sediment into deeper water or toward the southern end of the lake.-j- The United States Coast Survey have carefully sounded the lake in all its parts, and have ascertained the width of the area of shallow water extending inward from the shores. It is well known that waves are limited in their downward action, so that there will be a surrounding shelf, or shoulder of shallow water, in cases where the waves of a deep lake are eroding its banks. This fringe of shallow water encircling Lake Michigan is only a few miles wide ; and from such data as have been gath- ered, the average rate of erosion is found to be as much as five or six feet per annum ; which would indicate that the lake-basins had not been in existence more than seventy-five hundred years. Leaving these more indefinite and in many respects un- satisfactory efforts to estimate the age of lake-basins, we may get some assistance in approximating to a correct chro- * " Minnesota Geological Report" for 1879, p. 73. f " American Journal of Science," toI. xcviii, 1869, pp. 172 et seq. 472 THE ICE AGE IN NORTH AMERICA. nology of the Glacial age by studying the smaller kettle-holes which constitute so marked a feature in the kames and moraines of the glaciated region. As already shown, the most satisfactory explanation of these curious depressions is, that they mark places where masses of ice were buried in the debris of sand and gravel brought down by the streams of the decaying glacier ; and where, upon the melting of the buried ice, a cone-shaped depression was left with sides as steeply inclined as the nature of the soil would permit. At any rate, there can be no question that the kettle-holes were formed during the closing stages of the Glacial period. As typical of numberless others we present the facts concerning a kettle-hole near Pomp's Pond in Andover, Mass.* Pomp's Pond is itself a moraine basin about a quarter of a mile in diameter, and but slightly above the level of the Shaw- shin Eiver, into which it empties. Upon its north side is an accumulation of gravel and sand, with pebbles intermingled, in which there are several of the smaller characteristic bowl- shaped depressions of which we have spoken. Their appear- ance is much like that of volcanic craters. You ascend a sharp acclivity from every side to a rim of gravel, and then descend as rapidly into the bowl-shaped or crater-like depres- sion. A section carried across will present the idea. Fig. 114.— Section of kettle-hole near Pomp^p Pond, Andover, Massachusetts. (See text.) (For general view of the situation, see Fig. 88, p. 298.) Prom the level of the pond, and two or three rods from the edge, you begin to ascend at an average rate of about one foot in three, till the south side of the rim is reached, at a height of fifty-two and five tenths feet above the pond, (a) (This rim is not, however, of a uniform height. On the east side it rises * I here transfer a few paragraphs from my " Studies in Science and Re- ligion." THE DATE OF THE GLAGIAL PERIOD. 473 into a pyramid seTenty-seven feet high.) {b) Then, descending fifty and five tenths feet vertically, you are carried one hundred and thirty-eight feet horizontally, reaching at that point the edge of a circular mass of peat which is ninety-six feet in di- ameter, (c) From the opposite side the ascent of the northern rim begins, and you descend from its top to the valley, repeat- ing almost exactly the first descent from the pond. The dis- tance from rim to rim, or the diameter is three hundred and eighty feet. It is evident that since the first formation of this crater- shaped depression no material can have reached the bottom, except from three sources : 1. The wash from the side ; 3. The decay of vegetation which grew within the circumference of the rim ; 3. The material brought by the winds. It is equally evident that what is once in can not get out. Dust, leaves, and twigs carried by the winds inevitably lodge in such depressions more thickly than in other places, since the atmosphere in such hollows is comparatively quiet. For the same reason the surrounding trees as they are blown down are more likely to fall toward the center of the kettle- hole ; and the ashy material which their roots abstract from the sides of the depression is no insignificant factor in the problem. N'ow, from the angle of the declivity, the original depth of the depression can be approximately estimated. If the angle be still the same as at first, the first three terms of the propor- tion would be 138 : 50*5 : : 48 : 17^f, making the original depth below the present surface of the peat a trifle over 17 '5 feet. If, however, we suppose the original slant to have been steeper and the rim higher, we can still see that there must have been a limit to the depth. Suppose the rim to have been one third higher and the slant one third steeper, we then should have in round numbers the proportion 138 : 68 : : 48 : 33|4, making the original depth of the depression nearly twenty-four feet below the present surface of the peat. From the nature of the material it is impossible that the depth could originally have much exceeded that amount. Accepting this conclusion, the problem is, to determine the time it would require the agencies mentioned above to fill the 474 THE lOE AGE IN NORTH AMERICA. bottom of this bowl to a depth of twenty-four feet — a cone ninety-six feet in diameter at the base and twenty-four feet to the apex — which would be equal to a deposit of only eight feet over the present surface of the bottom. The question is, Could this hare stood with so little change for eighty thousand years ; or even for forty thousand years, if we were to accept Professor Charles H. Hitchcock's estimate of the prolongation of the effects of CroU's period ? * Is not the supposition of ten thousand years sufficiently extravagant ? If the close of the great Glacial period be so far back as Mr. Croll estimates, we must believe that sediment would accumulate, in the situ- ation above described, over the sui-face of the present peat-bog, at the rate of only one inch in a thousand years ; while, if we put the close of this period back ten thousand years, the rate of accumulation would seem to be as slow as our imagina- tion can well comprehend. One hundred inches, which is little more than eight feet, divided into one hundred thousand parts, would be only -001 of an inch ; that is, if this depression has been in existence one hundred thousand years, we must believe that with all the dust there is in the air, and all the soil that would wash down the steep incline of all the sides, and all the vegetable matter growing in and falling into the depression, one thousand years would be required for one inch of sediment to accumulate J If we reduce this supposed period to 50,000, 35,000, and 13,500 years successively, the time re- quired for the accumulation of an inch of sediment would be proportionally 500, 350, and 135 years. If any one will be at the trouble of dividing an inch into 125 equal parts, he will probably be surprised at the insignificance of the quantity. The slowest rate at which Boucher de Perthes calculates for the accumulation of peat over Roman pottery in the valley of the Somme is three centimetres, or a little over an inch, in a century. We do not bring railing accusation against those who, from astronomical considerations, confidently speak of the close of the Glacial period as an event which occurred scores of thou- sands of years ago ; but it is important to know what other ' Geology of New Hampshire," vol. iii, p. 327. THE DATE OF TEE OLAOIAL PERIOD. 475 beliefs that long chronology carries with it. If any one chooses to believe that kettle-holes can stand one hundred thousand years, and fill up only twenty-four feet from the apex of the inverted cone, he must run the risk of being con- sidered credulous. In rejecting the theory of Mr. OroU concerning an in- definite succession of glacial periods, we did not mean to foreclose the discussion connecting the question whether there have not been two glacial epochs in this country. This question must, therefore, now be considered with more particular reference to its bearing upon matters of chro- nology. As the reader doubtless observed in the remarks upon Croll's theory, quoted from Mr. Grilbert and President Ohamberlin, in the preceding chapter, each of them spoke of an " Interglacial Period " as clearly indicated in North American geology. The calculations just made relate to the chronology of what President Chamberlin calls the " second glacial epoch." Niagara Falls, the Falls of St. Anthony, the kettle-holes of Massachusetts, and the valley of Plum Creek, are none of them upon the extreme border of the glaciated region. Kaccoon Creek is nearer the margin. Cal- culations respecting those interior points, therefore, do not give the date of the extreme marginal deposits. Hence it becomes a matter of prime importance to consider to what extent the Ice age in North America was bipartite. I pre- sume the majority of authorities of most weight would prefer to speak of two glacial epochs in America, believing that the ice-front withdrew to the Laurentian highlands during the so-called interglacial period, and thence readvanced to the outer line of moraines which we described in the chapter upon that subject. In view of some important questions bearing upon the antiquity of man, to be discussed in the next chapter, it is necessary to consider this question here more fully than we have yet done. The most obvious evidence adduced in favor of an inter- glacial epoch in America consists of the so-called " inter- 476 THE lOE AGE IN NORTE AMERICA. glacial " forest-beds.* These forest-beds and vegetal de- posits occur over a wide area, and in places have glacial de- posits both under them and over them. The first supposi- tion with regard to them was that these various forest-beds were contemporaneous, and indicated a general retreat of the ice after its iirst invasion of North America until it had entirely disappeared or lingered only in the Canadian high- lands; whereupon there was a readvance of the ice, over- whelming the forests and other vegetal deposits which had collected in kettle-holes and other depressions, and burying them beneath a second sheet of ground-moraine, where they are opened to present inspection whenever wells penetrate them or eroding streams expose them on their banks. But it is not clear that these interglacial forest-beds might not originate in front of the margin of the slowly retreating ice if only there were comparatively brief periods of readvance along successive lines of latitude. Thus they may belong to various times of oscillation, both during the general advance and during the general retreat of the glacier. If, for ex- ample, at any time during the period of advance there had been a retrocession of the ice-front for a short distance, for- ests and vegetable growth would soon have spread over the marginal belt from which the ice had retreated, and, upon a readvance, these would be overwhelmed and covered with a new stratum of glacial deposition. In case of some of the peat-beds, it is probably necessary to suppose that they were formed where they are, and are really interglacial ; but, in case of many of the fragments and logs of wood found in the glacial deposit, we are not compelled to suppose an interglacial origin. Wood will stand transportation in the ground-moraine almost as well as bowlders, and it is by no means certain that much of the timber found in the till may * See Charaberlin, " Geology of Wisconsin," vol. i, chap, xv, especially pp. 271-291 ; "Driftless Area," pp. 211-216; N. H. Winchell in "Proceedings of the American Association for the Advancement of Science," vol. xxiv, 1875, pp. B, 43-66 ; " Geology of Minnesota," vol. i of the " Final Report," pp. 363 etseq.; 3. S. Newberry, "Geological Survey of Ohio," vol. ii, pp. 90-83. Fig. 115. — Perpendicular section of till at Oxford. Ohio, showing; a ])iect' of wood three inches in diameter projecting from the face. This has evidently been transported in the till like a bowlder. The section is about fifty feet ; portion shown, about fifteen feet, near the middle. (United States Geological Survey.) (Wright.) 478 THE ICE AGE IN NORTH AMERICA. not have belonged to the original forests whicli covered tlie country in front of the first sheet of advancing ice. These logs may have been picked up like the bowlders, and trans- ferred to the south a long time after their original deposi- tion. Thus, it may be that the " forest-beds " near the mar- gin of the glaciated area are of more recent origin than those some distance back, since the ice in its final retreat may have proceeded with few a^nd slight oscillations. As Presi- dent Charaberlin suggests, also, " certain subaqueous deposits so closely resembled true till that they have been mistaken for it, and there is perhaps no case of superposition of beds supposed to represent two glacial periods that is not still open to these doubts." * President Chamberlin, whose knowledge of the facts bearing on this subject is wider than that of any one else, therefore does not rely so much upon the existence of in- closed forest-beds and a supposed superposition of distinct beds of glacial debris, in proof of distinct glacial epochs, as upon certain other considerations of a more general nature, such as the following : The earlier drift is characterized, in the interior basin, by a wide but relatively uniform distribution, manifesting only occasional and feeble tendencies to aggregation in mo- rainic ridges. It is not bordered, except in rare instances, by a definite terminal moraine, but ends in an attenuated border. It is not characterized by the prevalence of prominent drum- lins or other similarly ridged aggregations. The phenomena of glacial erosion connected with it are generally feeble. Glacial striae are indeed present, even in the peripheral portions, but the surface of the rock is not usually extensively planed. The whole aspect of the deposit indicates an agency which spread the drift over the surface smoothly, and relatively gently, with little forceful action. The drainage phenomena are also of the gentle order. We have yet failed to find evidence of very vigorous drainage connected with the older drift of the in- * See " Geology of Wisconsin," vol. i, p. 272. THE DATE OF THE GLACIAL PERIOD. 4Y9 terior basin except in osars and kames, whose conditions of formation were exceptional, but, on the contrary, abundant . proof of slow-moving waters and imperfect drainage, indicat- ing low slope of the surface. The later Glacial epoch, on the contrary, was character- ized by strong glacial action, planing the rock-surface vigor- ously, even up to the very limit of its advance. The glaciers plowed up immense moraines about their edges, except on smooth plains whose slope was away from the ice-movement. The drainage was usually vigorous, and immense trains of glacial gravel stretch away from the margin of the ice-sheet, reaching great distances down the valleys and frequently filling them to great depths with well-assorted material. The vigorous action of the glaciers of the second epoch and the rapid drain- age, in general, stand in marked contrast with the gentle action and imperfect drainage of the earlier epoch. One of the conditions that determined the distinction was probably the difference in elevation that characterized the two epochs. The interval between these two leading epochs we regard as the chief interglacial epoch, representing a greater lapse of time and a greater change in the dynamic agencies of the age than the several other interglacial intervals, or episodes of deglaciation, which mark the complicated history of the Ice age. As belonging to the earlier Glacial epoch, we recognize two drift-sheets that have been described by the geologists of the re- spective States as occurring in southwestern Ohio, southern In- diana, central and southern Illinois, eastern and southern Iowa, northern Missouri, eastern Nebraska, and southeastern Min- nesota. Between these occur, at numerous points, vegetal and fer- ruginous accumulations and other evidences of a non-glacial interval. To this horizon belong the larger number of de- posits described under the term "old forest-bed," but very many vegetal deposits so referred do not, in our judgment, belong there, but are referable to several distinct horizons. * Others adduce as evidence of two distinct Glacial epochs in North America the greater oxidization and general de- * " Driftless Area," pp. 214, 215. 480 TEE IGE AGE IN NORTH AMERICA. composition of the material upon tlie extreme border of the glaciated region as compared with that of the kettle-moraine in Wisconsin, and what is considered to be a moraine of corresponding age in the regions both east and west. It is with diffidence that I venture to present any inter- pretation of facts difEering from that of such eminent authori- ties as have defended the foregoing theory. But, even if my interpretation should not be altogether satisfactory, its presentation may help to a better understanding of the sub- ject. With reference to these evidences, therefore, it is to be noted : 1. That the more complete oxidization of the glacial debris along the southern border and the greater decomposi- tion of the granitic bowlders and pebbles distributed over this border, are naturally accounted for by the obvious fact that for the most part the material along the southern border, and for some distance back from it, was that which was first picked up by the advancing ice, and was probably already oxidized and partially decomposed by the long-continued action of preglacial agencies when the ice began its removal. Its oxidization, therefore, may not be any true indication of the remoteness of its transportation and deposition. It is evi- dent that every successive period of movement from the north would operate upon lower strata of rock and upon the masses which had been less affected by secular agencies of decomposition. Thus it is natural that the more northern moraines and glacial deposits, of various kinds, should appear fresher than the southern. 2. The uniformity in the distribution of the till over the southern portion of the glaciated area in the Mississippi Valley is partly an illusion, due to the fact that the great amount of loess covering the region, especially in southern Indiana and Illinois and in eastern Nebraska, prevents, to a considerable extent, observations upon the original surface, and this loess, as has already been shown, is doubtless the finer part of the glacia#c?e5m carried soutliward by the gla- cial streams — so that, upon any theory, we should expect a TEE DATE OF TEE GLACIAL PERIOD. 481 much larger accumulation of loess over the southern portion of the area. 3. The theory of a general depression of the glaciated area with reference to the sea-level may apply to a certain portion of a single period as well as to one of two distinct .periods. We may suppose a low slope of a surface and the consequent imperfect drainage and slow-moving waters dur- ing the maximum extent of a single glacial epoch as well as during the lirst of two epochs. The theory that the weight and attraction of the ice were tangible factors in pro- ducing the relative depressi(m of land which characterized a portion of the Ice age would lead us to expect the greatest depression during the period of maximum extension. When the ice-front had retreated from Carbondale, 111., to Mad- ison, Wis., the intervening area had been relieved from an enormous amount of pressure. 4. With reference to the comparative absence of glacial strise and of planing and grooving over the southern area, it should be noted, first, that fresh exposures of rock in that region are very infrequent, owing to the great depth of till and loess ; and, secondly, that upon any theory the gla- cial grooving and striation would necessarily grow fainter as the boundary was approached, because the movement of ice over that portion was so much less than over the central and northern portions; and, thirdly, the absence of planation is not relatively so great as is sometimes represented. The grooves and strise in Highland and Butler counties, Ohio, very near the margin, and in southwestern Indiana and southern Illinois, still nearer the margin, are as clear and distinct as can anywhere be found. Also, upon the surface of the limestone rocks, within the limits of the city of St. Louis, where the glacial covering was thin, and disintegrat- ing agencies had had special opportunities to work, I found very clear evidences of a powerful ice-movement; and at Du Quoin, 111., only forty or fifty miles back from the ex- treme limit of glaciation, I was greatly impressed with the extent to which the surface rock had been planed, by ex- 31 482 THE IGE AGE IN NORTH AMERICA. amining the fragments brought up from a shaft which had recently been sunk tirst through fifty or sixty feet of surface soil, and then for some distance into the rock. The small fragments from the surface of the rock thrown up were most beautifully planed and striated. A thorough study of the condition and distribution of the buried forest-beds bears strongly, as I can not but think, against the complete separation of glacial epochs in North America. In addition to the facts about to be enumerated, it is a significant circumstance that the buried vegetable de- posits under consideration do not mark a warm climate, but a climate much colder than the present — such a vegetation, in fact, as would naturally flourish near the ice-margin. The buried forests of southern Ohio have a striking resemblance to those we described in Glacier Bay, Alaska. Peat and hardy coniferous trees are predominant. One of the most instructive localities in which to study organic remains embodied in glacial deposits is in the region included in the southern part of Montgomery and the north- ern part of Butler county, Ohio. The glacial deposits con- taining organic remains in that vicinity were first described by Professor Orton, of the Ohio Survey, in 1870.* Near Germantown, on Twin Creek, in Montgomery county, about thirty miles north of Cincinnati, there is exposed, at a sharp angle of the stream, a perpendicular bank of drift ninety-five feet in height. Underneath this is a deposit of peat as much as fourteen feet thick. The upper portion of the peat " con- tains much undecomposed sphagnous mosses, grasses, and sedges." Both the stratum of peat and the clayey till above "contain many fragments of coniferous wood, some of which can be identified as red cedar {Jvniperus Vlrginia- nus)." Immediately above the peat-bed there is from fifteen to twenty-five feet of what seems to be true till. This shows no sign of stratification, and abounds in striated stones. Next above occurs a band about ten feet thick of stratified * " American Journal of Science," vol. c, 1S70. THE DATE OE THE OLACIAL rEETOD. 483 material containing cuarse gravel and a gund deal t— - Z Fig. 134. — ], a, convex surface of a chert implement found at the mouth of Little Elk Eiver, Morrison County, Minnesota, supposed to be a scraper. 1, 6. profile view of the same.t 2, a, convex surface of a chert implement found at Little Falls, Minnesota. 3, 6. profile vievp of the same. The figures do not perfectly represent the evident- ly chipped edges. (Winchell.) Another locality especially worthy of attention, in which palseoliths have been found, is at Little Falls, Morrison county, Minnesota, the situation of which can readily be seen by reference to the map on page 546. The first discover- ies at this point were made as long ago as 1877, and an ac- count of them was given in the " Sixth Annual Geological Eeport of Minnesota." % These implements were made from chert and quartz, and were recognized by Professor IST. H. * See Mr. Cresson's report on the subject, in the " Proceedings of the Boston Society of Natural History," toI. xxiv, p. 150 et seq. f This specimen is regarded a finished implement by Putnam. \ Pp. 83-58. 538 THE ICE AGE IN NORTH AMERICA. Winchell as belonging to the age of tlie glacial deposits which here line the trough of the Mississippi. A little later, Miss Franc E. Babbitt examined the locality more carefully, and found a large number of additional implements. Her discoveries were first reported in a paper read before the Minnesota Historical Society in February, 1880. A fuller account was presented at the meeting of the Ameri- can Association for the Advancement of Science at Minne- apolis in August, 1883. At that time also the subject was thoroughly canvassed by the numerous geologists present, and a paper was read upon the subject by Mr. Warren Upham, to whose work upon the surface geology of the Northwest we have so often had occasion to refer. To get the whole subject before our readers we can do no better than to append the principal portion of an elaborate paper read by Mr. Upham before the Boston Society of Natural History, on December 31, 1887, which will be the more readily understood by reason of the previous chapters of the present volume detailing the general results of Mr. Upham's work in that region : The recession of the ice-sheet of the last Grlacial epoch in Minnesota seems to be clearly marked by 'as many as ten stages of halt or readvance, in which distinct marginal mo- raines were accumulated, besides the moraine on the limits of its farthest extent. Six summers of geologic field-work in that State have been spent by the writer chiefly in the examination of its glacial and modified drift, of these mo- raines, and of the beaches and deltas of the glacial Lake Agassiz, which was formed in the valley of the Eed River of the North and of Lake Winnipeg by the barrier of the dejDarting ice-sheet. In their bearings upon this subject, my observation and study of that region convince me that the rude implements and fragments of quartz discovered at Little Falls were overspread by the glacial flood-plain of the Mis- sissippi River, while most of the northern half of Minnesota was still covered by the ice, contemporaneously with its for- mation of the massive moraines of the Leal Hills and with MAN AND TEE GLACIAL PERIOD. 539 the expansion of Lake Agassiz on their west side, respectively sixty and eighty-five miles west of Little Falls. This was during the highest stage of Lake Agassiz, and previous to its extension beyond the north line of Minnesota and Dakota, More than twenty lower beaches of this glacial lake have been traced, belonging to later stages in the recession of the ice- sheet, before it was melted so far as to reduce Lake Agassiz to its present representative, Lake Winnipeg. Estimated by comparison with the series of moraines and beaches formed during the glacial recession, the date of the gravel plain at Little Falls appears to be about midway between the time of maximum extent of the last • ice-sheet and the time of its melting on the district across which the Nelson Eiver flows to Hudson Bay. The town of Little Falls is on the east bank of the Mis- sissippi River, in Morrison county, near the geographic center of Minnesota. It is about a hundred miles northwest from St. Paul and Minneapolis, and nearly an equal distance southeast from Itasca Lake. The elevation of Itasca Lake is about 1,450 feet above the sea ; of the Mississippi, at the head of the rapids or Little Falls, from which the town derives its name, 1,090 feet ; and at the head of St. Anthony's Falls in Minneapolis, 796 feet. Following the general course of the river, without regarding its minor bends, its descent from Lake Itasca by Little Falls to Minneapolis' averages about two feet per mile, and is approximately uniform along the entire distance. Considered in a broad view, this central part of the State may be described as a low plateau, elevated a few hundred feet above Lake Superior on the east and the Eed River Valley on the west. In most portions it is only slightly undulating or rolling, but these smooth tracts alternate with belts of knolly and hilly drift, the recessional moraines of the ice- sheet, which commonly rise fifty to one hundred feet, and in the Leaf Hills one hundred to three hundred and fifty feet above the adjoining country. The bed-rocks are nearly every- where concealed by the drift-deposits, into which the streams have not eroded deep valleys, their work of this kind being mostly limited to the removal of part of their glacial flood- plains. The upper portions of the Mississippi and of its 540 THE ICE AGE IN NORTH AMERICA. chief tributaries, and all the smaller streams throughout this region, flow in many places through lakes which they have not yet filled with silt nor drained by cutting down their outlets. At Little Falls the glacial flood-plain of the Mis- sissippi is about three miles wide, reaching two miles east, and one mile west from the river. Its elevation is twenty- five to thirty feet above the river at the head of the rapids, which have a descent of seven feet. The Mississippi here flows over an outcrop of Huronian slate, and the same forma- tion is also exposed by the Little Elk River near its mouth, on the west side of the Mississippi three miles north of Little Falls. Veins of white quartz occur in the slate at both these localities, and were doubtless the source of that used by man here in the Glacial period for the manufacture of his quartz implements. The locality and section of the modified drift, where these worked fragments of quartz were found by Miss Babbitt, and the account of their discovery, are best told in her own words from her paper read before the Anthropological Section of the American Association for the Advancement of Science at its Minneapolis meeting in 1883. I quote as follows : " Rudely worked quartzes had previously been discovered here by the State Geologist of Minnesota, Professor N". H. Winchell, by whom they had been described and figured in the State Geological Report for 1877. . . . The find reported by Professor Winchell consists of chipped objects of a class generally ascribed to what is called the rude stone age. Of these many appear to be mere refuse, while others are regarded as finished and unfinished implements. The Winchell speci- mens have been assigned, upon geological ground, to a pre- historic era antedating that of the mound-building races, and reaching back to a time when the drift material of the terrace- plain was just receiving its final superficial deposit. It is found that, at intervals, the surface soil of the terrace con- tains these quartzes to a depth of not unfrequently three or four feet. "The lowest and newest formation at this place is the present flood-plain of the river. It is still in process of depo- sition, being yet subject to partial overflows at periods of MAW AND TEE GLAOIAL PERIOD. 541 exceptionally high water. In that portion of the town of Little Falls situated east of the Mississippi, this bottom-land is limited on the east by a high, ancient river-terrace, which has here an average elevation of about twenty-five feet above the river. . . . This older terrace, like the present flood-plain, has been spread out by the immediate action of water. . . . Fig. 135.— Quartz imijlement, found by Miss F. E. Babbitt, 1878, at Little Falls, Minneso- ta, in modified drift, fifteen feet below surface, a, face view ; 6, profile view. The black represented on the cut is the matrix of the quartz vein. (No. 31,323.) (Put- nam.) While occupied in examining the river bank at Little Falls in quest of wrought quartzes, one day during the season of 1879, 1 had occasion to ascend a slope lying between the new flood-plain and the older terrace, by a path leading through a sort of gap or notch in the latter (three hundred and ten rods, very neaiiy, or almost one mile north of the east-west road by Vasaly's Hotel ; ten rods west of the road to Belle Prairie ; and thirty- eight rods from the river). ... It seemed that at some past period a cut had been effected here by drainage, and that the wash-out thus formed had afterward been deepened by being 542 THE ICE AGE IN NORTH AMERICA. used, now and then, as a vagon-track. In this notch I dis- coyered the soil to be thickly strewed with pieces of sharp, opaque quartz. These were commonly of a white color, and ranged in size from minute fragments to bits as large as a man's hand, and in some instances even larger. There were many hundreds of these chips visible, scattered over an area Fig. 136.— Quartz implement, found by Miss F. E. Babbitt, 1878, at Little Falls, Minne- sota, in modified drift, fifteen feet below surface, a, face view ; b, side view. (No. 31,316.) (Putnam.) the width of the wagon road, and ten or fifteen yards in length. They were conspicuously unwaterworn, and likewise mostly unweathered, though occasionally a bit was picked up having some one of its surfaces weathered, while fractured or wrought faces appearing upon other parts of it, looked as fresh as if the work of yesterday. On the other hand, the mass of stone rubbish upon and among which the quartzes were strewed is much water-worn, many of the pieces being well rounded, while none of them are wholly angular. " By continued observations at this locality, I found that many of these quartz chips were brought to light at every suc- ceeding freshet of the season, being washed out of the saud by MAN AND TEE GLACIAL PERIOD. 543 descending drainage. Their immense and continually increas- ing numbers seemed to warrant the belief that they had re- sulted from systematic operations of some sort, once conducted, for unknown purposes, upon this particular spot. A portion of the studied specimens subsequently yielded evidence of having received shape from human hands, and therefore it was assumed provisionally that the site of exposure represented a prehistoric workshop. " Prolonged investigation ensued ; and investigation estab- lished the hitherto unsuspected fact that no quartz chips nor fragments were inclosed in the upper part of the gravel and sand terrace at the notch, nor within a considerable distance at either hand, though they were sought with careful scrutiny. . . . Ultimately it was ascertained that the notch quartzes had dropped to the level at which they were seen from a thin layer of them once lying from ten inches to two feet above it, and subsequently broken up through the wearing away of the sand underneath by drainage. This layer or stratum was still intact on the north and south and partially so on the east, in which direction it had, however, at certain points, suffered some dis- placement by wagoning. It extended in a nearly horizontal plane into the terrace, in the sloping edge of which the notch, opening into its west bank and truncated at its edge, is cut. . . . The qiiartz-bearing layer averaged a few inches only in thickness, varying a little as the included pieces happened to be of smaller or larger size. The contents were commonly closely compacted, so much so that one might sometimes ex- tract hundreds of fragments, many of them very small ones, of course, from an area of considerably less than a square yard. " The quartz bed, so far as examined, rested upon a few inches of sandy soil, which passed downward into a coarse water-worn gravel, immediately overlying till. Above the quartz chips, stratified gravel and sand extended up to the sur- face of the terrace. The pebbles of the gravel lying directly on the quartz-bearing stratum were small and well rounded, and were noticeably less angular than those of the gravel below. The stratum of quartz chips lay at a level some twelve or fif- teen feet lower than the plane of the terrace-top. " These observations show that the quartz chips were spread 544 THE lOE AOE IN NORTH AMEBIC A. originally upon an ancient surface that has been since covered deeply by the modified drift which forms the terrace. It will be remembered that the quartz chips and implements discovered by Professor Winchell in this vicinity are contained in the up- per stratum of the terrace-plain ; but the notch quartzes do not occur at the terrace-top, and can not have been derived from it, but are confined strictly to a single stratum of the lower gravels closely overlying the till. Hence the two sets of objects can not be synchronous, though they may have been produced by the same race at different stages of its existence. The notch quartzes must, of course, be older than those de- scribed by Professor Winchell, by at least the lapse of time required for the deposition of the twelve or fifteen feet of modi- fied drift forming the upper part of the terrace-plain, above the quartz-bearing stratum." This description by Miss Babbitt shows that these imple- ments and fragments of chipped quartz occurred in a well-de- fined thin layer in the modified drift forming the glacial flood- plain of the Mississippi River, as shown in the section which I have drawn (see the following figure). I have examined the Fig. 137.— Section across the Mississippi Valley at Little Falls, Minnesota, showing the stratum in which chipped quartz fragments were found by Miss F. E. Babbitt, as de- scribed m the text. (Upham.) terraces and plains of this valley drift from St. Paul and Min- neapolis to Brainerd, some twenty-five miles north of Little Falls, and find them similar in material and origin with the modified drift terraces in the valleys of the Merrimack, Con- necticut, and other rivers in New England. These water-, courses extending southward from the region that was covered by the ice-sheet became the avenues of drainage from it during its retreat. A part of the drift which had been contained in the lower portion of the ice was then washed away by the streams formed on the ice in its rapid melting and was depos- ited as modified drift, forming layers of gravel, sand, and fine silt, in the valleys along which the floods supplied by this melting descended toward the ocean. Along the Mississippi MAN AND THE GLACIAL PERIOD. 545 the flood-plain of modified drift at Brainerd has a height of about 60 feet above the river ; at Little Falls, as before noted, its height is 25 to 30 feet ; at St. Cloud, 60 feet ; at Clear- water and Monticello, 70 to 80 feet ; at Dayton, 45 feeb ; and at Minneapolis, 35 to 30 feet above the river at the head of St. Anthony's Falls. The modified drift at Little Falls lies on the till or direct deposit of the ice-sheet, and forms a surface over which the ice never readvanced. It lies far within the area that was ice- covered in the second and latest principal epoch of glaciation, and by reviewing the steps in the recession of the ice of that epoch we shall be able to ascertain approximately what were the outlines of its receding margin when the gravel and sand plain of Little Falls was deposited, inclosing these evidences of man's pi-esence. The ice-sheet, supplying both this modi- fied drift and the floods by which it was brought, still covered much of the upper part of the Mississippi basin, which only reaches about a hundred miles north of Little Falls ; and the courses of massive morainic belts show the continuation of the glacial boundary northwestward across Dakota and with less clearness eastward across the Lauren tiaa lakes. When the latest North American ice-sheet attained its greatest area, its southern portion from Lake Brie to Dakota consisted of vast lobes, one of which reached from central and western Minnesota south to central Iowa. This lobe in its maximum extent ended near Des Moines, and its margin was marked by the Altamont moraine, the first and outermost in the series of eleven distinct marginal moraines of this epoch which are recognizable in Minnesota. When the second or Gary moraine was formed, it terminated on the south at Min- eral Eidge in Boone county, Iowa. At the time of the third or Antelope moraine, it had farther retreated to Forest City and Pilot Mound in Hancock county, Iowa. The fourth or Kiester moraine was formed when the southern extremity of the ice-lobe had retreated across the south line of Minnesota and halted a few miles from it in Freeborn and Faribault counties. The fifth or Elysian moraine, crossing southern Le Sueur county, Minnesota, marks the next halting-place of the ice. At the time of formation of the fifth moraine, the south 35 546 THE ICE AGE IN NORTH AUERIGA. end of the ice-lobe had been melted back a hundred and eighty miles from its farthest extent, and its southwest side, which at first rested on the crest of the Ooteau des Prairies, had retired thirty to fifty miles to the east side of Big Stone Lake and the east part of Yellow Medicine county. During its next stage of retreat this ice-lobe was melted away from the whole of Le Suenr county, and its southeast extremity was withdrawn to Waconia, in Carver county, where it again halted forming its Fio. 138.— Map showing the ot recession of the ice in Minnesota as described in the teit. (Upham.) sixth or Waconia moraine. The seventh or Dovre moraine marks a pause in its recession when its southeast end rested on MAN ANB THE GLACIAL PERIOD. 547 Kandiyohi county. Probably nearly all of the southern half of Minnesota was at this time divested of its ice-mantle, while nearly all of the northern half was still ice-coyered, the glacial boundary across the State passing in an approximately east to west course not far from Little Palls. By its next recessions the ice-border was withdrawn to the eighth or Pergus Palls moraine, and the ninth or Leaf Hills moraine. These are merged together in the prominent accu- mulations of the Leaf Hills, which reach in a semicircle from Pergus Palls to the southeast, east, and northeast, a distance of fifty miles, marking the southern limits of this ice-lobe when it terminated nearly due west of Little Palls and half-way be- tween the south and north borders of Minnesota. Conspicuous morainic hills a few miles east of Little Palls, and others in the north part of Morrison county and along its west side, seem to be correlated with the Pergus Palls moraine. Much of the modified drift of the Mississippi Valley at Little Palls was probably deposited when the ice-sheet terminated at these hills five to fifteen miles distant on the east, north, and west. Eastward from Morrison county, this moraine continues north- east to the north side of Mille Lacs, thence probably through the south edge of Aitkin county and the north part of Pine county, and onward northeasterly to the west end of Lake Superior. The Leaf Hills moraine extends from the northeast part of the Leaf Hills, near the Leaf Lakes, east across northern Todd county and northwestern Morrison county and then north-northeast by Gull, Pelican, White Pish, and Crooked Lakes. Next it probably takes an eastward course, crossing the Mississippi several miles north of Sandy Lake and the St. Louis Eiver near the mouth of the Cloquet, and thence an east-northeast course passing not far south of the Cloquet Eiver and reaching the north shore of Lake Superior about half-way between Duluth and Pigeon Point. The upper portion of the modified drift at Little Falls, probably including the stratum of chipped fragments of quartz, is referable to the time of the recession of the ice-sheet north from the Pergus Palls moraine to the Leaf Hills moraine. At the west end of the Leaf Hills and thence through a distance of fifty miles northward, this stage of recession carried the ice-border 548 THE ICE AGE IN NORTH AMERICA. back only five to ten miles ; and in the main Leaf Hills, as before noted, the two moraines are united. Across the Mis- sissippi basin the glacial recession between them uncovered an area mainly twenty to forty miles wide. The portion of the ice-sheet nearest to Little Falls at the time of the I^eaf Hills moraine was in the vicinity of Fish-Trap Lake and Lake Alexander, in northwestern Morrison county, only twenty miles distant. There, as in the Leaf Hills, this moraine and that of Fergus Falls come together. Ascending the Mississippi, a distance of eighty miles intervened between Little Falls and the ice-border at the time of the Leaf Hills moraine, which extends approximately parallel with the river and ten to twenty miles from it on its northwest side in passing north-northeast- ward from Morrison county. During the formation of the tenth or Itasca moraine, and of the eleventh or Mesabi moraine, crossing the lake region at the head of the Mississippi, the gravel and sand of the modi- fied drift were probably wholly deposited north of Little Falls. Later moraines, formed at times of halt or readvance, inter- rupting the recession of the ice-sheet between northern Minne- sota and Hudson Bay, have not been determined, but I believe that they exist and await discovery when the glacial drift of that wooded and very scantily inhabited region shall be fully explored. The many beaches of Lake Agassiz, all showing an ascent northward when compared with the level of the present time, but with this ascent gradually decreased during the suc- cessive stages of the lake, probably find their explanation in the manner of retreat of the ice in Canada, interrupted there, as farther south, by pauses and the formation of moraines. Contemporaneously with the deposition of the glacial flood- plain at Little Falls and the accumulation of the Leaf Hills, the ice-front forming the north shore of Lake Agassiz crossed the Red River Valley between Fargo and Grand Forks, and ex- tending northwesterly across northern Dakota, as shown by its moraines remarkably developed along the south side of Devil's Lake and onward to Turtle Mountain. Toward the east, the ice-sheet at this time had receded from the south- west part of Lake Superior, which was held about five hundred feet higher than now and overflowed to the St. Croix and MAN AND THE OLA CIAL PERIOD. 549 Mississippi Eivers by the way of the Bois Brule River and Upper St. Croix Lake. It seems nearly certain also that the ice- border continued across Green Bay and the north part of Lake Michigan ; and farther east I think that it probably crossed southwestern Ontario and the central or northern portions of New York, Vermont, New Hampshire, and Maine. The Lau- rentian lakes were dammed by the retreating glacial barrier and overflowed at the lowest points on their southern water- shed. The time when the Little Falls stone implements and fragments from their manufacture were covered by the modi- fied drift seems therefore somewhat later than that of the im- plements found in southern Ohio and in New Jersey ; for, if this was the course of the ice-boundary east from the Leaf Hills of Minnesota, it had already receded beyond the region where the glacial floods could be discharged by the Little Miami and Delaware Rivers. If the question be asked. How many thousand years ago was this ? a reply is furnished by the computation of Professor N. H. Winchell, that approximately eight thousand years have elapsed during the erosion of the post-glacial gorge of the Mississippi from Fort Snelling to the Falls of St. Anthony ; of Dr. Andrews, that the erosion of the shores of Lake Michi- gan, and the resulting accumulation of dune-sand drifted to the southern end of that lake, can not have occupied more than seventy-five hundred years ; of Professor Wright, that streams tributary to Lake Erie have taken a similar length of time to cut their valleys and the gorges below their waterfalls ; and of Mr. Gilbert, that the gorge below Niagara Falls has required only seven thousand years or less. These measures of time carry us back to the date of the Little Falls quartz-workers, when the ice-sheet of the last Glacial epoch was melting away from the basins of the upper Mississippi and of the Laurentian lakes. Plants and animals doubtless followed close upon the retir- ing ice-border, and men living in the region southward would make journeys of exploration to that limit, but probably they would not take up their abode for all the year so near to the ice as Little Falls at the time of the Fergus Falls and Leaf Hills moraines. It may be that the chief cause leading men 550 TEE ICE AGE IN NORTH AMERICA. to occupy this locality, so soon after it was uncoyered from the ice, was their discovery of the quartz-veins in the slate there and on the Little Elk River, affording suitable material for making sharp-edged stone implements of the best quality. Quartz-veins are absent or very rare and unsuited for this use in all the rock-outcrops of the south half of Minnesota that had become uncovered from the ice, as well as of the whole Mississippi basin southward, and this was the first spot acces- sible whence quartz for implement-making could be obtained. While the deposition of the valley-drift at Little Falls was still going forward, men resorted there, and left, as the remnants of their manufacture of stone implements, multitudes of quartz fragments. By the continued deposition of the modified drift, lifting the river upon the surface of its glacial flood-plain, these quartz-chips were deeply buried in that formation. The date of this valley-drift must be that of the retreat of the ice of the last Glacial epoch, from whose melting were supplied both this sediment and the floods by which it was brought. The glacial flood-plain, beneath whose surface the quartz frag- ments occur, was deposited in the same manner as additions are now made to the surface of the bottom-land ; and the flooded condition of the river, by which this was done, was doubtless maintained through all the warm portion of the year, while the ice-sheet was being melted away upon the region of its head-waters. But in spring, autumn, and winter, or, in exceptional years, through much of the summer, it seems prob- able that the river was confined to a channel, being of insuffi- cient volume to cover its flood-plain. At such time this plain was the site of human habitations and industry. After the complete disappearance of the ice from the basin of the upper Mississippi, the supply of both water and sediment was so diminished that the river, from that time till now, has been occupied more in erosion than in deposition, and has cut its channel far below the level at which it then flowed, excavating and carrying to the Gulf of Mexico a great part of its glacial flood-plain, the remnants of which are seen as high terraces or plains upon each side of the river. CHAPTEE XXII. MAN AND THE GLACIAL PEEIOD {continued). The preceding instances all belong, without question, to the later stages of the Ice age in America. Those geologists who speak of two glacial periods would classify the gi'avels at Trenton, N. J., Madisonville and Loyeland, O., Medora, Ind., and Little Falls Minn., as belonging to the later stages of the second Glacial epoch — the deposits at Little Falls being, as Mr. Upham has already stated, somewhat subse- quent to any of the others. We come now to a still more startling discovery, made by Mr. Oresson, near his summer residence, on the Delaware Eiver, at Claymont, Del. For a general idea of the situation the reader is referred to the map of New Jersey in the general chapter upon the glaciated area.* The discovery was made while an extensive excava- tion was in progress connected with the building of the Balti- more and Ohio Railway. It was my privilege, in November 1888, to examine the locality in company with Mr. Cresson, and I here give the results of the observations : The point is located about one mile and a half west of the river bank, and about one hundred and fifty feet above tide-water. The ascent from the river at Claymont is by three or four well-marked benches. These probably are not ter- races, in the strict sense of the word, but shelves marking dif- ferent periods of erosion, when the land stood at these sev- eral levels, but now thinly covered with old river deposits. The cut where the discovery was made is well shown in our * See p. 127. MAN AND THE GLACIAL PEBIOD. 553 accompanying illustration, reproduced from a photograph. The lower part of this cut consists of decomposed schist rocks in place and of deposits which are preglacial. These extend in the illustration to the top of the light band run- ning through the picture. The portion above this light band belongs to what was described in the preceding chapter as pertaining to the formation denominated by Professor Lewis the Philadelphia red gravel and brick-clay, being identical with that at Philadelphia both in its composition and in its stratigraphical relations, and extending continuously down the river from that city (nineteen miles). By Mr. McGee this would be denominated the Columbia formation, since he cor- relates the deposits in the Delaware Valley with those in the District of Cohimbia in the valley of the Potomac. The age of this deposit we have already discussed,* and we need here only repeat that it is without doubt a glacial formation of a much earlier period than that farther up the valley at Tren- ton, N. J., where Dr. Abbott made his important discoveries. While that at Trenton belongs to the later stages of the Ice age, this at Claymont is to be connected with the ice when at its maximum extension, and when the level of the region was depressed one hundred feet or more. In a preceding chapter I have given my reasons for questioning the theory of Mr. McGee, who would connect this deposit with a glacial age previous to, and entirely distinct from, that which was concerned with the deposits at Trenton, and which he would make from three to ten times as remote. But, whichever view upon this point prevails, whether that of two distinct glacial epochs, or of one prolonged epoch with various halts in the retreat of the ice, the Philadelphia red gravel and brick-clay must be regarded on the least calculation as some thousands of years older than the deposits at Trenton, N. J., Loveland and Madison ville, O., Medora, Ind., and Little Falls, Minn. The circumstances of the discovery are thus reported by Mr. Cresson : * See pp. 501, 523 a seq. Fia. 140.— N'riuer view i>( llir Siimc, witli the rni),'iT pointiiij; tii tllu prctif'i' pl^lt't' '" ^ic bank where the implement was I'ounU. (From phntuKraph by Crcssou.) MAN AND TEE GLACIAL PERIOD. 555 Toward midday of July 13, 1887, while lying upon the edge of the railroad cut, sketching the bowlder line, my eye chanced to notice a piece of steel-gray substance strongly re- lieved in the sunlight against the red-colored gravel just above where it joined the lower grayish-red portion. It seemed to me like argillite, and, being firmly imbedded in the gravel, was decidedly interesting. Descending the steep bank as rapidly as possible, the specimen was secured. . . . Upon examining my specimen, I found that it was unquestionably a chipped implement. There is no doubt about its being firmly imbed- ded in the gravel, for the delay I made in extricating it with my pocket-knife nearly caused me the unpleasant position of being covered by sevei-al tons of gravel. . . . Having duly reported my find to Professor Putnam, I began at his request a thorough examination of the locality, and on May 35, 1888, the year following, discovered another implement, four feet below the surface, at a place about one eighth of a mile from the first discovery. . . . The geological formation at which the implement was found seems to be a reddish gravel mixed with schist. The implements thus discovered by Mr. Cresson in this early deposit of the Glacial period must be connected with others near by, found by him several years before in a shel- ter-cave, since destroyed by the railroad excavation. This was situated near the small building that appears at the right of our picture.* Interested as a youth in the reports of cave explorations in Europe, he carefully excavated this rock shelter in 1866, making notes of and preserving everything he found. As recorded at the time, the lower part of this cave was filled to a depth of about six feet with a deposit ap- parently identical with the Philadelphia red gravel and brick- clay. This contained only palaeolithic implements of argil- lite similar to those (figured on the following page) from the railroad cut near by. Argillite implements, mingled with others of jasper, quartzite, and bone, with fragments of pot- * See Fig. 139, p. 552; also "Proceedings of the Boston Society of Natural History," vol. xxir, p. 145. 556 THE ICE AGE IN NORTH AMEBIOA tery atid human bones and charcoal, were found nearer the surface. The total depth of the deposit was about fifteen feet. Pig. 141.— Argillite implement, found by H. T, Creeson, 1887, in Baltimore and Ohio Rail- road cut, one mile from Claymont, Delaware, in Columbia gravel, eight to nine feet below the overlying clay bed. a, face view ; b, Bide view. (No. 46,726.) (Putnam.) The progress of the race from the Palteolithic to the Neo- lithic age here suggested corresponds in part to that indicated by Dr. Abbott's discoveries at Trenton, where the transition from the palsEolithic type of implement to the moi'e modern types, though sudden at the top of the gravel itself, is gradual from the top of the gravel to the surface of the soil. For, according to him,* argillite implements occur in greatest * " Proceedings of the American Association for the Advancement of Sci- ence," vol. xxxvi). MAN' AND THE GLACIAL PEBIOD. 557 abundance at the base of the deposit of " black soil" which overlies the gravel to an average depth of about one foot. " The flint implements known as Indian relics belong to this superficial or black soil," and they are found abundantly on the surface, more sparingly near the surface, and "more sparingly still the deeper we go," until, on reaching the gravel proper, they disappear entirely. In this connection it is interesting to note that at the mouth of Naaman's Creek, the nearest point on the river from the shelter-cave just described, Mr. Cresson has also discovered remains of prehistoric wooden structures below the level of low tide. These consist of the ends of rude piles which had evidently been fashioned by stone implements, but for what purpose intended it is not evident. In dredging here, he found numerous rude argillite implements of the palaeolithic type, which, in the vicinity of two of the structures, were mingled with those of a modern type. Thus the valley of the Delaware would seem to contain a record of the passage of the race on the Atlantic coast from the Palseolithic to the Neolithic age. Here, about as far below the ice-front at that time as the shore of Greenland now is, the hardy hunters who had been driven before the advancing cold of the great Ice age found ample space for their pursuits, and excellent shelter in the dense forests which everywhere bordered the southern front of the great snow- fields. The proximity of the ocean furnished, doubtless, a supply of fish, while numerous animals, long since extinct in this region, were for a time fellow-fugitives with man from the advancing northern foe. Among these companions of man we may pretty certainly include the mastodon (one of whose tusks, as already remarked, was found by Professor Cook in the Trenton gravel itself), the walrus, the Greenland reindeer, the caribou, the bison, the moose, and the musk-ox, for the remains of all these animals are found either in the superficial gravel deposits of southern New Jersey, or in the adjoining region of country to the south and west. The picture of human life during that period in the valley of the 558 THE ICE AGE IIT NORTH AMERICA. Delaware is substantially the same as that presented by the archaeologists of Europe for southern England and northern France in the declining years of the Glacial period. With the exception of some implements reported in 1874 by Professor Aughey * from the loess deposits of Nebraska, but whose connection with the Glacial age does not seem to be well established, the foregoing are all the discoveries east of the Eocky Mountains which definitely connect man with the great Ice age. Mr. McGeef has, however, reported the discovery of a finely wrought obsidian spear-head, of modern type, in the lacustral clays of the ancient bed of Lake Lahontan, whose relation to the Glacial period was described in a previous chapter. But the uncertainty as to the exact age of this lake, and the doubt which even Mr. McGee expresses of the implement being in place, together with its evident modern type, make it unnecessary to say more about it at present. The discussion of man's rela- tion to the Glacial period in North America, however, can not be complete without considering the reported discover- ies both of human implements and human remains in the gold-bearing gravel deposits in California. The first evidence upon the point to which we will turn attention is that produced by Professor J. D. Whitney,;]: of Harvard University, concerning human remains beheved by him to have been found in strata which mark the closing period of the Tertiary epoch in California. The following clear description of the age of the deep placer deposits in which these remains have been found is given by Le Conte, and is necessary for an appreciation of the value of the evidence : * " Annual Report of the United States Geological Survey of the Territories," by F. V. Hayden, 1874, pp. 247, 254. f " Geological History of Lake Lahontan, a Quaternary Lake of Northwest- era Nevada," by I. C. Russell, being Monograph No. XI, " United States Geo- logical Survey," p. 247. X " Report on the Auriferous Gravels of the Sierra Nevada," ]879,p.258«< seq. MAN AND TEE OLAGIAL PERIOD. 559 There are in many parts of California two systems of river- beds, an old and a new. The old belongs. to the Tertiary ; the new, to the Quaternary and present. The change took place during the oscillations of the Quaternary. The old river-system is substantially parallel to the present river-system, though in some places the one cuts across the other. It is probable, therefore, that there was but little change in the general direc- tion of the slope, produced by the oscillations of this epoch. These old river-beds are filled with drift-gravel, and often cov- ered with lava-streams. These drift-gravels probably repre- sent the beginning of the Glacial epoch, though Whitney thinks an earlier or Pliocene epoch. The present river-system sometimes cuts across, sometimes runs parallel to, the lava- filled beds of the old river-system, and the beds of the former have in their turn been eroded two thousand to three thousand feet in solid rock. In these also have been accumulated im- mense quantities of gravel and bowlder drift, evidently brought Fie. 142.— Lava-Btream cut through by rivers ; a, a, basalt ; b, b, volcanic ashes ; e, c, ter- tiary ; d, d, cretaceous rocks ; H, S, direction of the old river-bed ; iS', B', sections of the present river-beds. (Le Conte from Whitney.) down from the glacial moraines by the swollen rivers of the Champlain and early Terrace epochs. These facts are illus- trated by Figs. 143 and 143, in which R' represents the present Pig. 143. — Section across Table Mountain, Tuolumne County, California, b, lava ; G, grav- el ; «, slate ; B, old river-bed ; B', present river-bed. (Le Conte.) river-system, in Pig. 142, cutting across, and in Fig. 143 run- ning parallel to, the old system E. Although it is impossible to synchronize with certainty 560 THE IGE AGE IN NORTE AMERICA. these events with the changes in the eastern portion of the continent, yet the order of sequence is evident ; and that the greater part, if not all, occurred in the Quaternary, is also evi- dent. . . . The history of changes shown in these sections is suffi- ciently obvious. In the time of the old river-system, R was a river-bed, doubtless with a ridge on either side represented by the dotted lines. In this bed accumulated gravel, containing gold. Then came the lava-flow, which of course ran down the valley, displacing the river and covering up the gravels. The displaced rivers now ran on either side of the resistant lava, and cut out new valleys, two thousand feet deep, in the solid slate, leaving the old lava-covered river-beds and their aurifer- ous gravels high up on a ridge. In other cases the convulsion which ejected the lava also changed greatly the general slope of the country, and therefore the direction of the streams. In such cases of course the present river-system cuts across the old river-beds and gravels, and their covering lavas, as shown in Fig. 14a. The age of the old river-gravels is still doubtful ; that of the newer river-gravels is undoubtedly Champlain or early Ter- race. Below we give a list, taken from Whitney, of the re- mains found in these gravels : Neiver Placers. — Great mastodon, mammoth, bison, tapir (modern), horse (modern), man's works. Deep Placers. — Great mastodon,* mammoth, tapir (mod- ern), rhinoceros (ally), hippopotamus (ally), camel (ally), horse, extinct species. It will be seen that the fauna of the deep placers unite Pli- ocene and Quaternary characters. The great mastodon, the mammoth, and the tapir, are distinctively Quaternary, while the others are Pliocene. The plants, according to Lesquereux, are decidedly Pliocene. Therefore Whitney has not only placed the deep placers in the Pliocene, but made them the repre- sentative of the whole Pliocene, and probably Miocene, and the lava-flow as the dividing-line between the Tertiary and the * Whitney states that the mastodon is not found here, but it has been since found. MAN- AND THE GLACIAL PERIOD. 561 Quaternary. But, all the facts considered, it seems most prob- able that both the filling of the old riyer-beds, and their pro- tection by lava, took place comparatiyely rapidly, and were together the closing scene of the Tertiary drama. The deep gravels, therefore, may be placed indifferently in the latest Pliocene or earliest Quaternary. The newer gravels are un- doubtedly Quaternary and recent. Certain it is that the deep placer-gravels are similar in all respects to the Quaternary gravels all over the world, except that, by percolating alkaline waters containing silica, they have been cemented in some cases into grits and conglomerates. This is because they are covered with lava which yields both the alkali and the soluble silica. In any case, we have here an admirable illustration of the immensity of geological times. The whole work of cutting the hard slate-rock two thousand feet or more has been done since the lava-flow, and therefore certainly since the beginning of the Quaternary.* It will readily be seen that these upper gravels, whether we call them Tertiary or Quaternary, are with reference to the historical period very ancient, though recent if spoken of from a geological point of view. The question of man's antiquity does not turn on the name of the formation ; but upon the reality of the existence of his remains in the upper gravels. Indeed, there does not seem to be any hai'd-and- fast line of demarkation between the Tertiary formation and the Quaternary, or recent. In making chronological calcula- tions from the vast amount of erosion spoken of in the, pre- ceding paragraph, Le Conte warns us, however, to note the prodigious rapidity with which erosion now proceeds in con- nection with hydraulic mining. " In the North Bloomfield mine, the pebble-loaded torrent resulting from the incessant play of the hydraulic jet against the cliff, though working but eight months per year, has cut in four years a channel three feet wide and fifty feet deep in solid slate." f * Le Conte's "Elements of Geology," 1888, pp. 555, 585. f ''American Journal of Science," March, 1880, toI. cxix, p. 179. 562 TEE ICE AGE IN NORTH AMERICA. Unfortunately, the evidence that human remains have been taken from beneath these lava-capped mountain-ridges is neither of recent date nor that of professed geologists. We are compelled to depend upon the testimony of plain miners, exhibiting what they found and recounting what they saw several years ago. This fact, which needs expla- nation, is said to arise from the wholesale changes in the methods of mining introduced by hydraulic processes. By present methods terraces are washed down into a promiscu- ous heap by jets of water forced against them with great ve- locity, so that there is little hope of finding the archaeologi- cal specimens they may have contained. The golden days for the archaeologist in California have passed by. Still, so many independent witnesses from different localities have testified to the facts which we now relate, and circumstantial evidence so fully corroborates the statements of the witnesses, that Professor Whitney and his associates think they are be- yond question.* As early as 1863 Dr. Snell, of Sonora, began a systematic collection of animal and human remains from the mines in his vicinity. In his collection were several objects marked as " From under Table Mountain," among which was a human jaw. Dr. Snell's collection was destroyed by fire, and he died in 1869 ; but Professor Whitney and Mr. Voy had repeatedly examined it and conversed with him. A stone utensil, apparently used for grinding, was the only one which Dr. Snell claims to have taken with his own hands from the dirt as it came out of the tunnel under the moimtain. In 1857 Hon. Paul Hubbs, of Vallejo, Cal. (subsequently a State Superintendent of Pubhc Instruction), picked a por- tion of a human skull out of the dirt as it was brought from the Valentine shaft, under Table Mountain, near Shaw's Flat. This skull was given to Dr. C. F. Winslow, who soon after (October 7, 1857) divided it, and sent one piece to the Philar * A few paragraphs are here substantially reproduced from my " Studies in Science and Religion," p. 285 et seq. MAN AND THE GLACIAL PERIOD. 563 delphia Academy of Sciences and the other to the Boston Society of Natural History, where it still remains, and in whose " Proceedings " (vol. vi, p. 278) Mr. Winslow's original communication may be found. Ten years after, Mr. Hubbs more fully detailed the cir- cumstances of the discovery, and Professor "Whitney and Gorham Blake, Esq., made special examination of the locality and careful inquiries of the owners of the mine, and satisfied themselves that the bone really came from under the basaltic covering of Table Mountain. Mr. Walton, one of the owners, did not remember anything about the bone, but did remem- ber that a " mortar " had been found in the tunnel, near the same situation. In 1870 Mr. Oliver W. Stevens gave to Mr. Voy a large stone bowl, which was incrusted with sulphuret of iron, and which he makes affidavit that he picked with his own hands, in 1853, from a load of dirt which came from a tunnel under Table Mountain, two hundred feet in, at Shaw's Flat. Mr. Llewellyn Pierce also makes affidavit that a certain stone mortar which he gave to Mr. Voy was taken, in 1862, from under Table Mountain, eighteen hundred feet from the mouth of the tunnel. All this is preliminary to the famous Calaveras skull, the facts about which are as follows : In February, 1866, Mr. Mattison, one of the owners of a claim on Bald Mountain, near Altaville, says he took from a tunnel under the basaltic capping of the mountain an ob- ject which, on account of incrusted earthy and stony ma- terial, he thought at first to be the petrified root of a tree, but which he discovered to be a portion of a skull. He took it to Mr. Scribner, agent of the express company, who, after seeing the importance of the discovery, passed it over to Dr. William Jones of Murphy's, a physician of extensive practice and scientific tastes. Both these gentlemen were well known to Professor Whitney, and their veracity is vouched for by him. The skull was forwarded by Dr. Jones to the office of the State Survey on the following June (1866). Mr. Mat- 564 TEE WE AGE IN^ NORTH AMERICA. tison has been repeatedly interviewed, and his testimony is uniformly coherent and explicit, to the effect that he took the skull with his own hands from gravel underneath a cap- ping of forty feet of black lava and in connection with drift- wood. The appearance of the skull in every way corrobo- rates his statement. The original incrustation shows that it was not taken from a cave. The late Dr. "Wy man, of Harvard College, and Professor Whitney together carefully removed the incrustations from the skull. Fragments of bones and gravel and shells were so wedged into the cavities of the skull as to satisfy them that there could be no mistake as to the character of the situation in which it was found. Chemi- cal analysis showed that organic matter was nearly absent, and the carbonate of lime had largely displaced the phosphate ; i. e., it was in a fossilized condition. We omit mention of a large number of human remains found at great depths in the ancient higher-level gravel where not covered with lava, though some of them are doubtless of the same age with those from under Table Mountain. According to Professor Whitney, the evidence " all ; points in one direction, and there has never been any. at- i tempt made to pass off on any member of the Survey any- ; thing out of keeping or, so to speak, out of harmony with { what has been already found, or might be expected to be i found. It has always been the same kind of implements ' which have been exhibited to us — namely, the coarsest and the least finished which one would suppose could be made, and still be implements at all." This result, he cogently re- marks, would hardly be possible where so many parties are concerned in furnishing the evidence, if the objects were not genuine, and shows to his mind that the evidence has not been got up to deceive. As might be expected, strenuous efforts have been made to discredit these facts. With reference to the Calaveras skull, we read in Dr. Southall's " Kecent Origin of Man " (p. 558) that " Dr. Andrews informs us [Dr. Southall] that the Eev. R. W. Patterson, D. D., of Chicago, tells him that MAN- AND THE GLACIAL PERIOD. 565 he was informed by the Kev. W. W. Brier, a reliaUe minister of Alvarado, Cal., that his [Brier's] brother, a miner, was one of two men who took the so-called Calaveras skull from a cave in the side of the valley, and placed it in the sliaft, where it was found, and that the whole object was a prac- tical joke to deceive Professor Whitney, the geologist." Whether this is probable can be judged from the foregoing statement of facts as since detailed by Professor Whitney. At any rate, it would have been the proper thing for this renegade brother of the P.ev. Mr. Brier to have submitted himself to closer cross-examination from competent parties than he seems to have done. Sir William Dawson and others have questioned whether these human remains might not have been introduced at a period subsequent to the deposition of the gravel and the overflow of the lava. They have suggested that the Indians, in searching for gold, may have run horizontal shafts into the gravel underneath ; or, since the lava is not compact but tufaceous in its character, it does not seem impossible that, in some places, pits may have been sunk from the surface. The most formidable opposition to Professor Whitney's conclusions comes, curiously enough, from evolutionists, so that, upon this question, they are now found "among the prophets." The thorough-going evolutionist believes that early man was ape-like in his features, and that he invariably passed through a stage in which he used rough stone imple- ments before learning to polish them. But the Calaveras skull, which, if genuine, far antedates anything human which has been discovered in Europe, is not of a particularly- inferior order, and the implements purporting to come from under Table Mountain are not of the palasolithic type, but, though exceedingly coarse and rude, correspond to those of the smooth stone period in Europe. Professor Putnam, however, suggests,* and Professor A. Winchell is ready to * " Report of the United States Geological Surveys west of the One Hun- dredth Meridian," toI. vii, pp. 10-15. 566 THE ICE AGE IN NORTH AMEBIOA. admit * that man wandered into California long before he entered Europe, and attained there the higher state of devel- opment reached by palaeolithic man in other parts of the world at a much later date. The objection to Professor Whitney's inferences arising from the possibility that the aboriginal inhabitants of that region themselves carried on mining operations for the sake of obtaining this gold is presented, in a convincing manner, by Dr. James Southall. According to him,t Bancroft, in his " Native Eaces of the Pacific States," refers to it as a well-known fact that mining operations were carried on in Mexico to a great extent, opening galleries into the solid rock, in some cases two hundred feet or more in depth ; and Schoolcraft, in his " Archaeology," % describes one of these ancient shafts, which was discovered in 1849. This was two hundred and ten feet deep, and its mouth was situated on a high mountain. "The bones of a human skeleton were found at the bottom. There were also found an altar for worship and other evidences of ancient labor." It is to be observed that, in the quotation from School- craft made by Bancroft, it is stated that " no evidence has been discovered to denote the era of this ancient work. There has been nothing to determine whether it is to be regarded as the remains of the explorations of the first Spanish advent- urers, or of a still earlier period. The occurrence of the re- mains of an altar looks like the period of Indian worship." Professor Putnam, however, writes me : " I think there is a strong objection to the ancient mining theory, inasmuch as we do not know of gold among the Californian Indians. The Mexicans had it, but did they mine it in California ? The stone mortars we find in the California gravels are not of Mexican type, but of Californian type, the same form used in recent times." * " Pre-Adaioites," p. 428. f "Pliocene Man in America," p. 1. X Vol. i, p. 105. For a f uU summary of facta see Bancroft's " Native Races of the Pacific States," vol. iv. MAN AND THE GLACIAL PERIOD. 567 It must be confessed, also, that the neolithic character of the implements found is trying to one's faith in their ex- treme antiquity, since it betokens a progress in civilization which did not elsewhere originate until a much later date. On the whole, it can not be said that the facts implying extreme antiquity for man in California have been suffi- ciently well attested to receive unqualified credence. Re- specting them, Professor W. Boyd Dawkins writes as fol- lows : " The auriferous gravels offer no evidence upon the question, because none of the human remains hare been proved to be contemporaneous with them. The human re- mains belong to the class of relics left behind in California, Arizona, and New Mexico by the ancestors of the present native tribes, and imply a rude civilization of the same kind. They have, in my opinion, either been imbedded in the gravel by the action of streams or of slips from the mount- ain-sides in modern times, or are the result of interments or of the mining operations which Dr. Southall describes, car- ried on by the native tribes in modem times and not in the Pliocene age." To the same effect is the testimony of Professor Joseph Le Conte in the last edition of his " Elements of Geology," whose opinion, from his long residence in California, is entitled to more than ordinary weight : " Doubts as to this extreme antiquity of man (in California) are of three kinds, viz. : 1. Doubts as to the Pliocene age of the gravels — they may be early Quaternary. 2. Doubts as to the authenticity of the finds, no scientist having seen them in situ. 3. Doubts as to the undisturbed condition of the gravel ; for auriferous gravels are specially liable to disturbance. Tlie character of the implements found gives peculiar emphasis to this doubt, for they are not palseolithic but neolithic." * CoNCLirsioN. — In the discovery of man's relation to the Glacial age, interest in the subject of this volume has culmi- * Le Conte's " Geology," p. 698. 568 THE ICE AGE IN NORTH AMEBIC A. nated. In that discovery archaeology and geology are united in an indissoluble union, and the study of every class of gla- cial phenomena becomes invested with all the higher interest of historical research. Signal changes were introduced into the world's history by the conditions which accompanied the Glacial epoch. In America as well as in Europe this advent of northern cold greatly disturbed the conditions of animal life, and, we may well suppose, directly led to the extinc- tion of many animal species. In North America the camel, the hippopotamus, the rhinoceros, the tapir, the mammoth, the horse, the mastodon, were abundant at the opening of the Quaternary age. Their complete extermination is one of the most startling facts in geology. But, as Darwin has so well shown, the effects of a glacial advance are by no means limited to the region directly reached by the ice. In pushing southward the plants and animals of the northern part of the continent, the struggle for life in the more crowded quarters of the decreasing congenial portions of the country became more and more intense, and thus doubtless was brought about much of the extinction of species which the geologists have to record as having taken place in the early part of the Quaternary period. The evidences of man's existence in North America before the close of the Glacial period would indicate that he too shared in the sharp struggle which ensued with the new and rapidly changing conditions of that time. Did he also, like so many of his companions among the larger animals, share in this extinc- tion ? The sharpness of the transition from the paleeolithic to the neolithic type of implements, as we pass out from the Trenton gravel into the shallow soil above it, would seem to indicate an absolute distinction between the two succeeding races. But even so, whether the first became extinct from natural causes, and the other simply came in later as colonist, or whether the latter as conqueror exterminated the first, may always remain a doubtful question. It is possible that the Eskimo is the lineal descendant of preglacial man in America, and the conditions of life with which the Eskimo MAN AND THE GLACIAL PERIOD. 569 is so passionately in love would seem to resemble closely those which evidently surrounded palaeolithic man in New Jersey, Ohio, Indiana, and Minnesota. But, on the other hand, such human remains as we have from the Trenton gravel are regarded by Professor Putnam as belonging to a race distinct in type from the Eskimo. A closing remark is in place with reference to the date of man's appearance in America. In the first place it should be observed that, to say man was here before the close of the Grlacial period only fixes a minimum point as to his an- tiquity. How long he may have been here previous to that time must be determined by other considerations. Secondly, with our present knowledge of glacial phenomena, the date of the close of the Glacial period is regarded as much more modern than it was a few years ago. Sir Charles Lyell's estimate of thirty-five thousand years as the age of the Niag- ara gorge, which is one of the best measures of post-glacial time which has yet been studied, is greatly reduced by what we now know of the rate at which erosion is proceeding at the falls. Ten thousand years is now regarded as a liberal allowance for the age of that gorge. But, finally, the term " close of tlie Glacial period " is itself a very indefinite ex- pression. The Glacial period was a great while in closing. The erosion of the Niagara gorge began at a time long sub- sequent to the deposit of the gravel at Trenton and at Madi- sonville. Between those two events time enough must have elapsed for the ice-front to have receded a hundred miles or more, or all the distance from New York to Albany ; since only at that stage of retreat would the valley of the Mohawk have been freed from ice so as to allow the Niagara River to begin its work. The deposits at Trenton, Madisonville, and Medora, took place while the ice-sheet still lingered in the southern water-shed of New York and Ohio. "When, there- fore, the age of the mound-builders of Ohio is reckoned by centuries, that of the glacial man who chipped these palaeo- lithic implements must be reckoned by thousands of years. As is evident from the description of Mr. TJpham, the 570 THE ICE AGE IN NORTE AMERICA. gravel at Little Falls, Minn., is considerably more recent than that in the more southern localities, since the gravel in Min- nesota could have been deposited only when the ice-front had retreated some hundreds of miles from its farthest extension, while the first-named deposits occur near the very margin of the glaciated area. Most of those who have taken pains to read the preceding pages through from the beginning have doubtless been sur- prised at the wide range of questions involved in the subject under discussion. The movement of ice itself brings up for consideration one of the most singular and obscure of physical problems. A wide field of investigation is still open to the physicist in determining how it is that brittleness and mo- bility can so unite in one substance as to produce the phe- nomena of motion observed in living glaciers. The majesty of the ice-movement, as brought to light in the study of the glaciated area in North America, is equaled only in the movement of the forces of astronomy, or in that of those which have elevated the mountain-ranges on the surface of the earth. Almost every human interest in the northern part of the United States and in British America is likewise seen to be profoundly affected by the ice-movement which we have been permitted to study. During the great Ice age the old lines of drainage were obliterated, and new lines established, crooked places were made straight, and rough places plain. The change in the river-courses produced by the obstruction of glacial deposits has given rise to the innu- merable waterfalls where have grown up the flourishing manufacturing and commercial centers of New England and the interior. The Great Lakes are in the main the result of similar glacial obstruction. The vast internal commerce of the lake region avails itself of slack- water navigation result- ing from the ice-movements of the Glacial age. The innu- merable lakes of smaller size which adorn the surface of the northern part of the continent are also the result of glacial action. The anomalous distribution of insects and plants can likewise, in many cases, be traced to the same cause. The MAN AND THE GLACIAL PERIOD. 571 arctic butterflies and the Alpine flowers upon the summit of Mount Washington, as well as the gigantic forests of Califor- nia and some of their more distant relatives on the Atlantic coast, were fugitives from the arctic regions in glacial times, who have since become naturalized citizens of the lower lati- tudes. And, finally, man himself is connected with the clos- ing centuries of the Glacial period in the United States. American scholars who are ambitious to carry on archaeo- logical investigations need no longer go to the valley of the Euphrates or the Nile, or to the languages of central Asia, to find the oldest relics of man in the world, or the surest means of determining the greatness of his antiquity. A boundless, comparatively unworked, most promising and most interesting field lies before the American investigator in the glacial problems of his own country. Nowhere else in the world did the ice of the Glacial period deploy out upon so wide a margin of dry land, and leave so inviting and easy a field of study. Every river rising within the glacial boundary and emerging from the glaciated region presents a problem worthy of the life-long attention of any investigator. Every glacial waterfall and every glacial lake holds out the possibil- ity of yielding up an important clew to chronological ques- tions of absorbing interest. The ingenuity of Professor Asa Gray and others in tracing out the efPects of the great Ice age upon the distribution of plants and animals, has only introduced us to subjects which need yet to be worked out in endless detail. The object of the present treatise will be largely accomplished if it serves to stimulate and guide the host of local investigators which the subject is sure to in- terest. APPENDIX. A. PEOBABLE CAUSES OF GLACIATION. By WAEKEN UPHAM. An examination of the evidence of changes in the relatiye heights of land and sea in various parts of the world during Quaternary time has led me to an explanation of the causes of the Glacial period, which, in this application of its fundamental principle, seems to be new, while in its secondai-y elements it combines many of the features of the explanations proposed by Lyell and Dana and by Oroll. Briefly stated, the condition and relation of the earth's crust and interior appear to be such that they produce, in connection with contraction of the earth's mass, depressions and uplifts of extensive areas, some of which have been raised to heights where their precipitation of moist- ure throughout the year was almost wholly snow, gradually forming thick ice-sheets ; but under the heavy load of ice sub- sidence ensued, with correlative uplift of other portions of the earth's crust ; so that glacial conditions may have prevailed alternately in the northern and southern hemispheres, or in North America and Europe, and may have been repeated after warm interglacial epochs. Quaternary oscillations of land and sea in glaciated regions have been discussed by Croll and Geikie on the assumption that the earth was so rigid that its form would not be changed by the load of the ice-sheet nor by its removal, which seemed more probable because of the well-known physical and mathe- matical researches of Hopkins, Thomson, Pratt, and Professor G. H. Darwin, who conclude that the earth is probably solid, 574 TEE ICE AGE IN NORTH AMERICA. with not less rigidity than that of glass or steel. In def- erence to their researches, this conclusion is accepted and taught in recent text-books of geology by Le Conte and A. Geikie ; but in similarly recent text-books Dana and Prestwich teach that the earth probably consists of a comparatively thin crust underlaid by a molten interior, which may change within a moderate depth to a great nucleal solid mass. Among other geologists and physicists who have discussed the condition of the earth's interior, King* and Shaler f believe it to be solid ; while Whitney,! Button,* Powell, || Wadsworth,^ Crosby, t) Gilbert, J Claypole, % Airy, % Fisher,** and Jamieson,f f believe that it is molten, or, at least, is surrounded by a molten layer, and that the earth's crust floats in a condition of hydrostatic equilibrium upon the heavier liquid or viscous mobile interior or layer enveloping the interior, subject, however, to strains and resulting deformation because of the earth's contraction. The thickness of the crust, according to this hypothesis, is variously estimated to be from twenty to fifty miles, or possibly a hundred miles or more. It must be confessed that we have only a very inadequate knowledge of the conditions which would result from the enor- * " United States Geological Exploration of the Fortieth Parallel," vol. i, "Systematic Geology," 1878, pp. IIY, 696-'725. f " Proceedings of the Boston Society of Natural History," 1866, vol. xi, pp. 8-15 ; 1868, vol. xii, pp. 128-136 ; 18M, vol. xvii, pp. 288-292. " Memoirs of the Boston Society of Natural History,'' 1874, vol. ii, pp. 320-340. " Scribner's Magazine," vol. iii, pp. 201-226, February, 1888. X "Earthquakes, Volcanoes, and Mountain Building," 1871, pp. 77-87. * "Penn Monthly," vol. vii, pp. 364-378, and 417-431, May and June, 1876. " United States Geological Survey, Fourth Annual Report," pp. 183-198 ; " Sixth Annual Report," pp. 195-198. II "Science," vol. iii, pp. 480-482, April 18, 1884. ^ " American Naturalist," vol. xviii, June, July, and August, 1884. "Proceedings of the Boston Society of Natural History," 1883, vol. xxii, pp. 443-485. "Geological Magazine," II, vol. x, 1883, pp. 241-252. J "American Journal of Science," III, vol. xxxi, pp. 284-299, April, 1886. $ "American Naturalist," vol. xix, pp. 257-268, March, 1885. "American Geologist," vol. i, pp. 382-386, and vol. ii, pp. 28-38, June and July, 1888. J " Nature," vol. xviii, pp. 41-44, May 9, 1878. ** " Physics of the Earth's Crust," 1881, pp. 223, 270, etc. tt " Geological Magazine," III, vol. iv, 1887, pp. 344-848. APPENDIX. 575 mous pressure and high temperature of the earth's interior, and wide diversity in speculations on this subject will probably long continue. Professor Shaler, while holding that the earth is mainly solid throughout, perhaps haying in its most mobile layer beneath the crust "a rigidity such as belongs to the metals of average resistance to compression," yet is one of the earliest and most decided advocates of the opinion that the weight of an ice-sheet may depress the area on which it lies, and that the departure of the ice would be attended by re- elevation. In comparison, however, with the physical condi- tions and laws familiar to us upon the earth's surface, the subsidence and elevation of extensive areas, as of nearly all glaciated regions, seem to demonstrate a mobility of the earth's interior as if it were fused rock. The same conclusion is in- dicated by volcanoes, which are probably the openings of molt- en passages that communicate downward through the crust to the heavier melted interior, thence deriving their supply of heat, while their outpoured lavas consist largely or wholly of fused portions of the crust, the phenomena of eruption being caused by the access of water to the upper part of the molten rock near the volcanic vent. But the great plications of the strata in the formation of mountain-chains evidently involve only the upper part of the earth's crust, crumpled into smaller area in adapting itself to the diminishing volume of the lower portion of the same crust, which, with the nucleus, is under- going contraction on account of the gradual loss of its heat, and perhaps also on account of progressing solidification and compression. There is in this process no dependence on the molten condition. of the interior, except as that seems to be necessary for distortion of the earth, both of the crust and nucleus or mobile layer enveloping the nucleus, whereby con- siderable shrinkage of volume can take place before the ac- cumulated strain becomes sufficient for the formation of a mountain-chain. At the present time depressions and eleva- tions, probably caused by accumulating strains, are slowly changing the relations of land and sea upon many parts of the earth's surface. In the same way the downward and upward movements which would be caused by the burden of the ice- sheet and its removal are doubtless in many places complicated 576 THE ICE AOE IN NORTH AMERICA. by concomitant or subsequent moyements thus due to defor- mations under strains, by which the elevation attributable to the departure of the ice-sheet may be augmented or partly or wholly counteracted, giving much irregularity to the glacial and post-glacial oscillations of the land. Jamieson appears to have been the first, in 18G5, to suggest this view, which I receive from him, that the submergence of glaciated lands when they were loaded with ice has been caused directly by this load pressing down the earth's crust upon its fused interior, and that the subsequent re-elevation was a hydrostatic uplifting of the crust by underflow of the inner mass when the ice was melted away.* Two years later Whit- tlesey published a similar opinion. f In 1868 Shaler referred the subsidence of ice-covered areas to a supposed rise of iso- geothermal lines in the subjacent crust, operating, in conjunc- tion with the ice-sheet, to produce downward flexure ; X but in 1874 and later he regards the depression as due directly to the weight of the ice, and the re-elevation as due to its removal.* The same view is advanced also by Chamberlin to account for the basins of the Laurentian lakes, where he believes a con- siderable part of the glacial depression to have been permanent. || *" Quarterly Journal of the Geological Society," vol. xxi, p. 178. Later discussions of this subject by Mr. Jamieson are in the " Geological Magazine," II, vol. ix, pp. 400-40'? and 457-466, September and October, 1882; and III, vol. iv, pp. 344-348, August, 1887. In the article last cited, he applies this ex- planation to the changes of the beaches of Lake Agassiz, which up to that time I had attributed mainly to ice attraction. The same principle, however, was brought forward by Herschel in 1836, and had been advocated by Professor James Hall, of New York, in 1859, in attributing to the weight of sediments the long continued subsidence of the areas on which they have been deposited in great thickness. t " Proceedings of the American Association for the Advancement of Science," vol. xvi, pp. 92-97. X " Proceedings of the Boston Society of Natural History," vol. xii, pp. 128-136. * " Proceedings of the Boston Society of Natural History," vol. xvii, pp. 288-292; "Memoirs," ibid., vol. ii, pp. 335-340; "American Journal of Science," III, vol. xxxiii, pp. 220, 221, March, 1887; " Scribner's Magazine," vol. i, p. 259, March, 1887. II " Geology of Wisconsin," vol. i, 1883, p. 290; " Proceedings of the Ameri- can Association for the Advancement of Science," vol. xxxii, 1883, p. 212. The APPENDIX. 5Y7 Accompanying the subsidence of ice-loaded areas, there were doubtless uplifts of contiguous regions, perhaps sometimes in- cluding outer portions of the country glaciated. For example, the Quatenary elevation of which Le Conte finds evidence in the Sierra Nevada and northward may have been contempora- neous and correlative with depression of the northern parts of the continent beneath its ice-sheet. Furthermore, instead of being wholly offset by deformation of the crust, the glacial depression may have produced also extensive extravasation of lava, as is suggested by Jamieson * and Alexander Winchell,t for the vast Quaternary lava-flows of California, Oregon, Wash- ington, and a large adjacent region. As Jamieson well re- marks, this result would tend to cause a permanence of part of the depression of the ice-covered area. However it may have been caused, probably such permanent Quaternary subsidence is true for the coasts of many glaciated countries, as shown by fiords, and for the basins of the Laurentian lakes, which, ex- cepting Erie, are depressed several hundred feet below the level of the ocean. One of the most interesting fiords of North America is that of the Saguenay, tributary to the St. Lawrence. Along a distance of ahout fifty miles the Saguenay is from 300 to 840 feet deep below the sea-level ; its adjoining cliffs rise abruptly in some places 1,500 feet above the water ; and the width of its wonderfully sublime and picturesque gorge varies from about a mile to one mile and a half. J This fiord, like the many which indent our Eastern coast from Maine to Labrador and Greenland, and our Western coast from Puget Sound to the Arctic Ocean, was eroded by a stream that flowed along the bottom of the gorge when it was above the sea ; and this erosion was probably going forward in the epoch immediately preced- problems of ice attraction and of deformation of the earth's crust have been further discussed by President Chamberlin before the Philosophical Society of Washington, March 13, 1886 ; and, jointly with Professor Salisbury, in the " Sixth Annual Report, United States Geological Survey," pp. 291-304. * " Geological Magazine," II, vol. ix, 1882, p. 405. ■f " American Geologist," vol. i, pp. 139-143, March, 1888. X "J. W. Dawson, "Notes on the Post-Pliocene Geology of Canada," 1872, p. 41. 37 6Y8 THE IGE AOE IN NORTH AMERICA. ing the Ice age, for earlier subsidence during any period of much length, geologically speaking, would have caused the submerged valley to be filled with sediments. The preglacial elevation of the Saguenay region therefore appears to have been at least 1,000 feet greater than now; and it seems to be similarly proved by fiords that nearly the entire extent of the conti- nental glaciated area was considerably higher before than after glaciation. There is also evidence that part of the Atlantic coast of the United States close south of the limits of glaciation was at least for a short time preceding the Glacial period uplifted much above its present height. The submarine Hudson Eiver fiord * indicates that the vicinity of New York and Philadelphia then stood 2,800 feet above the sea, and that it afterward slowly sank 1,600 feet, while a bar of that height was formed by coast- wise wash across the mouth of the fiord. In this remarkable preglacial elevation, and in its being more or less shared by the whole northern half of the continent, the formation of the North American ice-sheet seems to be explained. If this was the cause of glaciation, probably the formerly greater height of about 1,000 feet on the Saguenay was not exceptional. In- deed, the elevation there and over large portions of the vast territory of Canada, bounded on the east, north, and west by fiord-indented coasts, may have been much more than is measured by the depth of the Saguenay River. Going a step further back, we may regard this northward elevation as a distortion of the earth's form in the storage of energy of lateral pressure which culminated, with the introduc- tion of the new factor of northward depression by the ice weight, in the Quaternary uplifts of the Western plains, the Eocky Mountains, the Sierra Nevada, and the Great Basin. These important changes in the elevations of great areas during the comparatively short Quaternary period seem to be consistent only with the hypothesis that our globe has a comparatively thin crust and a molten interior. In the Glacial period significant changes of the sea-level * A. Liadenkohl, " American Journal of Science," III, vol. xxix, pp. 476-480, with plate, June, 1886. APPENDIX. 6Y9 were caused : first, by abstraction of water from the ocean and its deposition on the land as snow, which under pressure made the vast ice-sheets ; and, second, by ice attraction of the ocean, lowering it still further, except in the vicinity of glaciated lands. An area of about 4,000,000 square miles in Iforth America, and another of about 3,000,000 square miles in Europe, were covered by ice-sheets, which in their maximum extent had probably an average thickness of a half or two thirds of a mile, or perhaps even of one mile. Assuming that these ice-sheets were contemporaneous, but disregarding ice-fields of smaller extent, which probably existed in parts of Asia and of the southern hemisphere, as also the glaciers of mountain dis- tricts, the lowering of the ocean surface, which covers approxi- mately 145,000,000 square miles, would slightly exceed 100 feet, if the mean depth of the ice accumulation was half a mile. More probably the sea over the whole globe was thus depressed fully 150 feet, which would correspond to an average of about 3,600 feet of ice on the glaciated areas of ISTorth America and Europe. For the second factor in causing such changes, Mr. R. S. Woodward's computations* indicate that gravitation toward the ice would further depress the ocean probably twenty-five to seventy-five feet within the tropics and in the southern hemisphere, while it would raise the level enough near the borders of the ice-sheets to counterbalance approxi- mately the depression due to the diminution of the ocean's volume, and would lift portions of the North Atlantic and of the Arctic Sea perhaps two or three hundred feet higher than now. Stream erosion while the sea was lowered to supply the ice of the Glacial period may explain the indentations of the southeastern coast of the United States, as Pamlico and Albe- marle Sounds, besides similar inlets in many other parts of the world ; but the excavation of Chesapeake and Delaware Bays seems more probably referable, at least in part, to the time of preglacial elevation, with the channeling of the now sub- merged Hudson fiord. When the ice-sheet of the last Glacial epoch finally re- * " United States Geological Survey, Sixth Annual Report," pp. 291-300 ; and " Bulletin No. 48," " On the Form and Position of the Sea-Level." 580 THE lOE AGE IN NORTH AMEBIOA. treated, the land which it had covered stood mostly lover than now, as is shown by the occurrence of fossiliferous marine deposits overlying the glacial drift up to considerable eleva- tions. Near Boston, and northeast to Cape Ann, the coast seems to have been submerged to a slight depth, probably not exceeding ten to twenty -five feet. Proceeding toward the north and northwest, the elevation of the marine beds lying on the glacial drift increases to about two hundred and twenty- five feet in Maine, about five hundred and twenty feet in the St. Lawrence Valley at Montreal, and four hundred and forty feet at a distance of one hundred and thirty miles west-south- west of Montreal ; but eastward, along the St. Lawrence, it decreases to three hundred and seventy-five feet opposite the Saguenay, and does not exceed two hundred feet in the basin of the Bay of Chaleurs ; while these marine deposits are want- ing in Nova Scotia and Cape Breton Island.* This subsidence accords well with the explanation that it was due to the press- ure of the ice -weight, which was greatest on the highlands between the St. Lawrence and Hudson Bay. Along the East Main coast of Hudson Bay and on Hudson Strait raised beaches are conspicuous, according to Dr. Eobert Bell, up to heights of at least three hundred feet, f In the region draining from the southwest to James Bay, Dr. Bell reports marine shells in stratified beds overlying the glacial drift along the Moose, Mattagami, and Missinaibi Eivers up to about three hundred feet above the sea ; X along the Albany * A. S. Packard, Jr., " Memoirs of the Boston Society of Natural History," vol. i, pp. 231-262. J. W. Dawson,' " Notes on the Post-Pliocene Geology of Canada " ; and " American Journal of Science," III, vol. xxv, 1 888, pp. 2n0-202. C. H. Hitchcock, " Proceedings of American Association for the Advancement of Science," Portland, 1873, vol. xxii, pp. 169-175; "Geology of New Hamp- shire," vol. iii, pp. 279-282 ; and " Geological Magazine," II, vol. vi, 18Y9, pp. 248-2S0. R. Chalmers, "Transactions of the Royal Society of Canada," sec. iv, 1886, pp. 139-145. W. Upham, " Proceedings of Boston Society of Natural History," vol. xxiv, pp. 127-141, December, 1888; "American Journal of Sci- ence," May, 1889. t " Geological and Natural History Survey of Canada, Report of Progress for 1877-'78," p. 32 C; for 1882-'83-'84, p. 31 DD. X " Geological and Natural History Snrvev of Canada, Report of Progress for 1875-'76," p. 340; for 1877-'78, p. 7 C. APPENDIX. 581 and Kenogami Elvers up to a height of ahout four hundred and fifty feet ; * and on the Attawapishkat to about five hundred feet above the sea.f It is also evident that the shores of Hud- son Bay are still undergoing elevation4 unlike the eastern coast of the United States and Canada, where the post-glacial uplift- ing has ceased, and there is now in progress a very slow sub- sidence of the land from New Jersey to the Gulf of St. Law- rence. Scantier but yet conclusive proofs of the uplift of British Columbia after glaciation are found in the valley of the Fraser Eiver, and on the Pacific coast, in Vancouver Island and the Queen Charlotte Islands. Lamplugh has observed recent ma- rine shells in a railway cutting on the west bank of the Harri- son River, near its junction with the Fraser, at an elevation not less than one hundred feet above the sea.* At New West- minster, on the Fraser, near its mouth, raised beaches inclos- ing fragments of marine shells are reported by Bauerman about thirty feet above the river. || Fossiliferous marine deposits found in the vicinity of Victoria and Nanaimo, in the south- east part of Vancouver Island, at small elevations above the sea, are believed by Dr. Gr. M. Dawson to have been formed at or near the wasting edge of the ice-sheet ; -^ and near the mid- dle of the northeast side of this island two distinct deposits of till occur, with intervening beds of loess-like silts, from which * " Geological and Natural History Survey of Canada, Report of Progress for 18'71-"72," p. 112; for ISTS-'^B, p. 340; "Annual Report," vol. ii, for 1886, pp. 34 and 38 G. f " Geological and Natural History Survey of Canada, Annual Report," vol. ii, p. 2^ G. X " Geological and Natural History Survey of Canada, Report of Progress for 18'?7-"78," pp. 32 C and 25 CC; for 1878-"79, p. 21 C; for 1882-'83-'84, pp. 26 and 30 DD ; " Annual Report," vol. i, for 1885, p. 11 DD. * " Quarterly Journal of the Geological Society," vol. xlii, 1886, pp. 284, 285. II " Geological and Natural History Survey of Canada, Report of Progress for 1882-'83-'84," p. 33 B. •^ " Geological and Natural History Survey of Canada, Annual Report," vol. ii, for 1886, p. 99B; "Quarterly Journal of Geological Society," vol. xxxiv, 1878, pp. 97, 98, and vol. xxxvii, 1881, p. 279. Compare also Mr. 6. W. Lamp- lugh's observations of glacial shell-beds at Esquimault, near Victoria, " Quar- terly Journal of Geological Society," vol. xlii, 1886, pp. 276-284- 37 582 TEE IGE AGE IN NORTE AMERICA. this author infers two periods of glaciation, separated by an interglacial epoch, in which the land was submerged from one to two hundred feet.* Again, in the northeast part of the Queen Charlotte Islands Dr. Dawson finds evidence of sub- mergence to the amount of two or three hundred feet, while the glacial conditions still endured, f In Europe the glaciated area stood at a greater height be- fore the Ice age, as is shown by fiords ; it was similarly de- pressed while loaded with the ice-sheet ; and since then it has been partially re-elevated. Its maximum post-glacial uplift known in the British Isles, so far as it has not been counter- acted by subsequent depression, is five hundred and ten feet, near Airdrie, in Lanarkshire, Scotland;! and in Scandinavia it is about six hundred feet.* As in all the North American districts noted, these upward movements seem attributable to the rise of the earth's crust, upborne by inflow of a molten magma beneath. But the derivation of the floras of the Faroe Islands, Ice- land, and even Greenland from the flora of Europe, demon- strates, according to Professor James Geikie, that the portion of the earth's crust extending from Britain and Scandinavia to Greenland was uplifted in early post-glacial times about * " Geological and Natural History Survey of Canada, Annual Report," vol. ii, p. 105 B. f " Geological and Natural History Survey of Canada, Report of Progress for 1878-"79," p. 95 B. Further important notes of recent changes in level of the coast of British Columbia, and of Washington Territory and Alaska, are given by Dr. Dawson in the " Canadian Naturalist," new series, vol. vlii, pp. 241 -248, April, 187Y. He concludes that this area had a preglacial elevation at least about nine hundred feet above the present sea-level, during part or the whole of the Pliocene period, this being indicated by the fiords ; that it was much depressed in the Glacial period ; and that in post-glacial time it has been re-elevated to a height probably two or three hundred feet greater than now, followed by subsidence to the present level, the latest part of this oscillation being a somewhat rapid depression of perhaps ten or fifteen feet during the latter part of last century — a movement which may still be slowly going on. X " Quarterly Journal of the Geological Society," vol. vi, 1850, pp. 386-888, and vol. xxi, 1865, pp. 219-221 ; " American Geologist," vol. ii, pp. 87l-3'79, December, 1888. «" Geological Magazine," I, vol. ix, 1872, p. 309; and II, vol. ii, 1876, p. 390. APPENDIX. 5i83 three thousand. feet higher than now;* and the same author shows that m interglacial time tropical animals passed from Barbary into Spain upon land where now the Strait of G-i- braltar has a depth of one thousand feet, f These changes in the relations of land and sea can not be ascribed to glacia- tion, but seem to be distortions of the earth's form, such as may be attributed to the action of strain upon the crust by which the earth can become reduced in volume through the subsidence and elevation of extensive areas during intervals between epochs of mountain-building. In the same class of changes are also to be included, wholly or in part, the post- glacial elevation of Grinnell Land and the northwestern coast of Greenland, one thousand to sixteen hundred feet ; X post- Pliocene upward movements of two thousand feet or more in Jamaica and Cuba ; * the recent uplift of the coast of Peru at least twenty-nine hundred feet, || which in diminished amount seems to extend along the whole range of the Andes ; its prob- able connection with the upheaval of the Cordilleras of North America, where Le Conte believes that the elevatory move- ments reached their greatest intensity in early Quaternary time, causing a rise of several thousands of feet in the Sierra Nevada ; ^ and the apparently correlative subsidence of a great * "Prehiatoric Europe," pp. 513-522, and 568, with Plate E. t " Prehistoric Europe," pp. 325, 337-339 ; " Quarterly Journal of the Geo- logical Society," vol. xxxir, 18'78, p. 505. X " Quarterly Journal of Geological Society," vol. xxiiv, 1878, p. 66; "Geo- logical Magazine," III, vol. i, 1884, p. 522. * J. G. Sawkins, " Reports on the Geology of Jamaica," 1869, pp. 22, 23, 307, 311, 324-329 ; W. 0. Crosby, "On the Mountains of Eastern Cuba," " Ap- palachia," vol. iii, pp. 129-142. Compare William M. Gabb's memoir, "On the Topography and Geology of Santo Domingo," " Transactions of the American Philosophical Society," vol. xv, pp. 103-111. II A. Agassiz, " Proceedings of the American Academy of Arts and Sciences," vol. xi, 18Y6, p. 287: and "Bulletin of the Museum of Comparative Zoology, at Harvard College," vol. iii, pp. 287-290. Above this height, at which corals are found attached to rocks, recent elevation of much greater amount seems to be indicated by terraces, by saline deposits, and by the presence of eight species of AUorchestes — a genus of marine Crustacea, in Lake Titicaca, 12,500 feet above the sea. 4 "American Journal of Science," III, vol. xxxii, pp. 167-181, September, 1886. 584 THE IGE AGE IN NORTH AMERICA. area dotted with coral islands in the Pacific. The Quaternary uplifts of the Andes and Rocky Mountains and of the West Indies make it nearly certain that the Isthmus of Panama has been similarly elevated during the recent epoch. On the line of the Panama Railway the highest land rises only two hundred and ninety-nine feet above the sea, and the highest on the Nicaragua Canal is about one hundred and thirty-three feet, while the isthmus nowhere attains the height of one thousand feet.* It may be true, therefore, that the submergence of this isthmus was one of the causes of the Glacial period, the con- tinuation of the equatorial oceanic current westward into the Pacific having greatly diminished or wholly diverted the Gulf Stream, which carries warmth from the tropics to the northern Atlantic and northwestern Europe. In view of the extensive recent oscillations of land and sea both in glaciated and unglaciated regions, it seems a rea- sonable conclusion that, while some of these movements have resulted directly from the accumulation and dissolution of ice- sheets, more generally, when the whole area of the earth is considered, they have been independent of glaciation. May not such movements of the earth's crust, then, have elevated large portions of continents, as the northern half of North America and the northwestern part of Europe, to heights like those of the present snow-line on mountain-ranges, until these plateaus became deeply channeled by fiords and afterward cov- ered by ice-sheets ? For the recentness of the latest glaciation, believed to have ended in the northern United States not more than ten thousand to six thousand years ago,f forbids our re- * Charles Ricketts, " The Cause of the Glacial Period, with reference to the British Isles," "Geological Magazine," II, vol. ii, 1875, pp. 673-580. A. R. Wallace, " The Geographical Distribution of Animals," vol. i, p. 40. f N. H. Wiuchell, "Geology of Minnesota," "Fifth Annual Report," for 1876, and "Final Report," vol. ii, pp. 313-341; "Quarterly Journal of the Geological Society," vol. xxxiv, 1878, pp. 886-901. E. Andrews, "Transactions of the Chicago Academy of Sciences," vol. ii. James 0. Southall, " The Epoch of the Mammoth and the Apparition of Man upon the Earth," 1878, chaps, xxii and xxiii. G. F. Wright, " American Journal of Science," III, vol. xxi, pp. 120- 123, February, 1881 ; "The Ice Ago in North America," chap. xx. G. K. Gil- bert, "Proceedings of the American Association for the Advancement of Sci- ence," vol. XXXV, 1886. APPENDIX. 685 f erring the glacial climate to conditions brought about by a period of increased eccentricity of the earth's orbit from two hundred and forty thousand to eighty thousand years ago, which has been so ably maintained by Croll and Geikie ; and some other adequate cause or causes must be sought for the successive Quaternary glacial epochs of these great continental areas and other districts of smaller extent both in the northern and southern hemispheres ; also for the rare occurrence of gla- cial conditions in various portions of the earth during past geologic ages, especially in the Carboniferous and Permian periods. The principal cause of all these epochs of glaciation seems to the writer to be probably found by the clew supplied in the relations already stated of the earth's crust and interior, whereby they become somewhat distorted from the spheroidal form while the process of contraction goes forward, the lateral pressure bearing down some portions of the earth's surface, and uplifting other extensive areas. Protuberant plateaus, swept over by moisture-laden winds, would be the gathering- grounds of vast ice-sheets, which would probably wax and wane with the changes of the earth's attitude toward the sun, by which the earth's place in any season, as summer, alternates from aphelion to perihelion, and back to aphelion in cycles of twenty-one thousand years. A similar explanation of the Gla- cial period was long ago proposed by Lyell and Dana, but without referring the elevatory movements to the earth's de- formation by contraction and accumulating lateral pressure while approaching an epoch of mountain -building, which fundamental principle was first suggested to me in an article from the pen of Professor W. 0. Crosby, on the origin and relations of continents and ocean basins.* During the periods immediately preceding gi-eat plications and shortening of segments of the earth's crust involved in the formation of lofty mountain-ranges, the broad crustal move- ments causing glaciation would be most wide-spread and attain their maximum vertical extent. The accumulation of ice- sheets may have brought about the depression of their areas, * "Proceedings of the Boston Society of Natural History," 1883, vol. xxii, pp. 4S5-460. 586 TEE ICE AGE IN NORTH AMERICA. with corresponding elevation of other plateaus, which in turn would become ice-covered, so that the epochs of glaciation of the northern and southern hemispheres may have alternated with each other ; * and this may have been several times re- peated, because of crustal oscillations due to ice-weight and its removal, the effects of elevation and depression of the land being re-enforced by climatic influences arising from the revolu- tion each twenty-one thousand years in the place of the seasons. When the building up of a great range of mountains ensued, which may have been initiated and accelerated by the repeated depressions under ice-weight and consequent transfers of the earth's deformation from one region to another, the accumu- lated strain in the earth's crust, with development of immense lateral pressure, would be diminished below the limit of its competency to cause glaciation. Such Quaternary mountain-building is known to have oc- curred on a most massive scale in Asia, where the Himalayas, stretching fifteen hundred miles from east to west, and tower- ing twenty thousand to twenty-nine thousand feet above the sea, are known to have been formed, at least in great part, and perhaps almost wholly, during this latest geologic period,f contemporaneously with the glaciation of North America, Europe, and portions of the southern hemisphere. Within the same time the great table-land of Thibet, J and much of central and northwestern Asia, have been uplifted ; the tract extend- ing from the Black and Caspian Seas northeast to the Arctic Ocean has risen to form a land-surface ; and the deep basin of Lake Baikal has been probably formed in connection with these crustal movements. Accompanying the formation of the Himalayas, there has been doubtless much disturbance by faults, local uplifts, and here and there plication of strata * Compare the opinions of Button, cited in A. Geikie's " Text-Book of Geolo- gy," second edition, p. 912, that the former greater extension of glaciers in New Zealand was caused by an increase in the elevation of the land, and that it be- longed to a much earlier time than the Ice age in the northern hemisphere, probably to the Pliocene period. t " Manual of the Geology of India," by H. B. Medlioott and W. T. Blanford, Calcutta, 18'79, Part I, pp. Ivi, 372; Part II, pp. 569-571, 667-669, 672-681. X Ibid., Part II, pp. 585, 686, 669-672. APPENDIX. 587 along the whole complex east to northwest and west mountain system of Oceania, Asia, Europe, and Northern Africa, from New Guinea, the Sunda Islands, Anam, and Siam, to the Cau- casus, Carpathians, Baliians, Apennines, Alps, Pyrenees, and Atlas Mountains, stretching quite across the eastern hemi- sphere ; but the greater part of the relief from the previously existing deformation of the earth was doubtless along the cen- tral part of the belt, in the colossal Himalayan range. In like manner the North American Cordilleras and the Andes, reaching in one continuous mountain system from the Arctic Circle to Cape Horn, have experienced within the same period great disturbances, as already noted, similar to those of the mountains of Southern Europe and the adjacent part of Africa. With this American orographic belt is also probably to be associated the mountain system, consisting largely of volcanoes now active, which forms the Aleutian Islands, Kamtchatka, the Kurile Islands, Japan, Formosa, the Philippines, Borneo, and Celebes, lying nearly in the same great circle with the Andes and Kocky Mountains, and with them continuous in an arc of about two hundred and forty degrees. Along two lines transverse to each other, one having an extent of half and the other of two thirds of the earth's circumference, the great lateral pressures of the earth's crust, which probably caused the elevation and glaciation of extensive areas during the Quaternary period, have been relieved by plication, faults, and uplifts, in the processes of the formation of mountain- ranges.* Combined with oscillations of the earth's crust, which are here regarded as the primary cause of the growth and decline of ice -sheets, many other concomitant conditions, notably changes in aerial and oceanic currents, and the earth's cycles of twenty-one thousand years through precession and nutation, enter into the complex causation of recurrent glacial and inter- glacial epochs, and serve to intensify or to mitigate the severity of the glaciation due to elevation. The influences of these conditions would be nearly the same that are claimed for them * See Prestwich's " Geology," vol. i, chap, xvii, treating of the relative ages of the principal mountain-ranges of the world. 588 THE lOE AGE !¥ NORTH AMERICA. in the luminous glacial theory of CroU, but their origin and effectiveness toward causing a glacial climate are here referred to extensive crust oscillations instead of eccentricity of the earth's orbit. The prolonged warm interglacial epoch, or sev- eral such epochs, of which evidence is obtained in the Qua- ternary deposits of Europe and North America, preceded and followed by severe arctic climate and ice-sheets, meet an ade- quate and consistent explanation in the view here taken ; and, indeed, the same reasoning that is presented by Croll in the details of the secondary elements of his theory seems equally applicable if these depend primarily on crustal elevation. The principal interglacial epoch in the United States, un- der this view, may well have been several times longer than either the previous or subsequent epochs of glaciation, or than the whole of post-glacial time, as claimed by McGee ; * but it does not follow that an exact parallelism will be found in the glacial history of Europe. Former extension of vast glaciers in the Eocky Mountains and Andes, the Pyrenees and Alps, the Atlas range, the Caucasus, the Himalayas, and elsewhere, far exceeding the glaciers of the present time, may be due to the uplift of these mountains much above their present height, followed by subsidence f with retreat of the ice ; but these os- cillations and resulting alternations of climate were not proba- bly synchronous everywhere. The highest mountain-ranges in four grand divisions of the world — namely, Asia, Europe, and North and South America — were doubtless largely uplifted and plicated, with formation of deep adjoining lakes, during the early part of the Quaternary period. Twice upheavals of the whole district of the Alps seem to have covered the region with great accumulations of ice, which each time may have depressed the area, first to be succeeded by the formation of interglacial deposits with lignite, and during each depression to send down floods, spreading loess along the Ehine, the Rhone, and the Danube.! After the later epoch of subsidence and glacial re- * " American Journal of Science," III, vol. xxxv, pp. 463-466, June, 1888. f A. Geikie's " Text-Book of Geology," second edition, p. 934. X J. Geikie's " Great Ice Age," second edition, chapters xxxiii and xxxiv, and his " Prehistoric Europe," chapters ix and xi. APPENDIX. 589 cession, there seems to have been a renewal of elevation, as shown by the height and slopes of the loess. Asia had no extensive ice-sheet like those of Europe and North America, probably because a sufficient elevation was not attained there until the Himalayas and Thibet were uplifted in the Glacial period. Their southern latitude and the position of Thibet and Mongolia in an arid and partly rainless belt, which stretches thence west to the Sahara, forbade their glacia- tion ; but from these recently uplifted Asiatic table-lands and mountains the most extensive Quaternary deposits of the world have been brought down by rivers and spread in the vast low plains of Siberia, eastern China, and northern India, sloping gently toward the sea, into which the finer part of this allu- vium is carried. All the puzzling features of the Chinese loess formation,* reaching to great elevations with such thickness and slopes of its surface that it could not be so accumulated as alluvium of flooded streams under the present conditions, seem to be readily explained by referring its deposition to annual floods from immense snow-melting, during the European Gla- cial epochs, upon the gradually rising central part of the Asi- atic Continent, which consists largely of easily erosible strata, and had in preglacial time become extensively disintegrated by weathering under a dry climate. With this reference of glaciation primarily to oscillations of land, a new element of Quaternary history is introduced, which seems to help much in accounting for peculiarities in the areal distribution of identical or closely allied species of animals and plants that have doubtless sprung from a common source but have now become widely separated. Not only are we able to follow Gray in his tracing the origin of the big trees of California, of the species in the flora of the eastern United States — which are the same with species of Japan, China, and the Himalayan region, or are represented there by closely re- lated forms, though unrepresented in Europe — and of mount- * Baron Richthofen, "Geological Magazine," 11, vol. ix, 1882, pp. 293-305.. J. D. Whitney, "American Naturalist," vol. xi, pp. '705-'i'13, December, 18'7'7. E. Pumpelly, " American Journal of Science," III, toI. xvii, pp. 133-144, Feb- ruary, 1879. E. W. Hilgard, "American Journal of Science," III, toI. xviii, pp. 106-112, August, is'yg. 590 THE lOE AGE IN NORTE AMERIOA. ain plants identical with those of the Arctic zone ; * but also we may now more satisfactorily bridge over the tropics and equa- tor, by uplifts and subsidences of mountain-ranges, so that species incapable of enduring a torrid climate could sometimes become dispersed even to sach distances as from north tem- perate latitudes to Tierra del Fuego and the Cape of Good Hope, f It seems probable that the rate of the earth's contraction has been somewhat uniform throughout the vast ages known to us by the researches of geology ; but the corrugation of the earth's surface in mountain-building has been much more rapid in some epochs than in others, and between the times of for- mation of great mountain-ranges there have been long intervals of quietude. % The slowly progressing contraction of the globe has been uninterrupted, and in some way the cooled outer part of the crust which has not shared in this diminution of volume has been able to accommodate itself to the shrinking inner mass. As stated on a previous page, this has probably resulted in distortion of the earth's form, both of the whole thickness of the crust and of the probably molten interior, within mod- erate limits during the periods of quiet, until so much lateral pressure has been accumulated as to compress, fold, and uplift the strata of a mountain-range. In attributing the severe climate of glacial epochs to great uplifts of the areas glaciated through such deformation preparatory to the process of mount- ain-building, it is distinctly implied that the Quaternary period has been at first exceptionally marked by such broad crustal movements, and has since gained comparative rest from the lateral stress to which they were due by equally exceptional * " Sequoia and its History," " Proceedings of the American Association for the Advancement of Science," Dubuque, vol. xxi, 1872, and " American Jour- nal of Science,"' III, vol iv ; " Forest Geography and Archaeology," " American Journal of Science," III, vol. xvi, ISTS ; " Characteristics of the North Ameri- can Flora," " Report of the British Association for the Advancement of Science," Montreal, 1884, and "American Journal of Science," III, vol. xxviii. f Darwin's " Origin of Species," chapter xi. Wallace's " Geographical Dis- tribution of Animals," chapter iii, and his " Island Life," chapter vii. X Dana's " Manual of Geology," third edition, p. 795 ; Prestwich's " Geology," vol. i, chap. xvii. APPENDIX. 591 plication, uplifts, and faults in the birth and growth of mount- ains. Further, it is implied also that stress in the earth's crust had been gradually increasing through long previous time, while the processes of mountain-building failed to keep pace with contraction, but were still sufficient to keep the earth's deformation less than is required to produce glaciation ; for no evidences of intense and widely extended glacial conditions are found in the great series of Tertiary and Mesozoic formations, representing the earth's history through probably ten or fifteen millions of years. And indeed these conclusions, drawn from the Quaternary Glacial period and the absence of glaciation through vast eras preceding, accord well with the known age and stages of growth of mountain-ranges that have been formed during these eras. No epoch since the close of Palaeozoic time has been more characterized by mountain-building than the comparatively short Quaternary, whose duration may probably be included within one hundred thousand or two hundred thousand years. The continuation of the earth's faunas and floras, with only very slight changes of species and exceedingly rare instances of extinction through the Quaternary period, not- withstanding its remarkable vicissitudes of climate and changes in the relative heights of land and sea, which seem especially adapted to produce both modifications and extinctions of or- ganic forms, bears indisputable testimony of the brevity of this period, when compared with those of Tertiary and Mesozoic time. As we extend our investigations backward in the geo- logic record, the species now existing are found in decreasing numbers until we come to the beginning of the oldest. In their places very difEerent species, genera, orders, and groups tenanted the earth before them ; and the gradual and doubtless very slow evolution of the present from the past must have re- quired duration almost incommensurable by years and centu- ries. But the total of mountain-building that has taken place during the Mesozoic and Tertiary eras is disproportionately small in comparison with that of the Quaternary period, even when ample allowance is made for long and very great denu- dation. N"ot until we go back to the Permian and Carboniferous periods are numerous and widely distributed proofs of very 692 THE ICE AGE IN NORTH AMERICA. ancient glaciation encountered. The atmosphere had been purified by the formation of Palaeozoic limestones of great thickness, and by the storing up of the principal coal-deposits of the world ; and these changes in the air had quite surely produced greater diversities of climate than before existed, especially in respect to the range of temperature in the seasons and in the several zones. Alternating beds of coal, shales, and sandstones, which form the coal-measures, indicate oscillations of level and climatic conditions much like those of the Qua- ternary period ; * and bowlder-bearing deposits, sometimes closely resembling till and including striated stones, while the underlying rock also occasionally bears glacial grooves and striae, are found in the Carboniferous or more frequently the Permian series in Britain, France, and Germany,! Natal,J India,* and southeastern Australia.! In Natal the striated glacier floor is in latitude 30° south, and in India only 20° north of the equator. During all the earth's history previous to the Ice age, which constitutes its latest completed chapter, no other such distinct evidences of general or interrupted and alternating glaciation have been found ; and just then, in close relationship with extensive and repeated oscillations of the land, and with widely distant glacial deposits and striation, we find a most remarkable epoch of mountain-building, sur- passing any other time between the close of the Archaean era and the Quaternary. The Appalachian Mountain system of the United States, with its grand plications and upheaval of the whole Palaeozoic group of rocks, belongs to this epoch, and the same line of disturbance extends by faulting and up- lifts northeastward to Gasp6 and Newfoundland. In Europe * CroU's " Climate and Time," chap. xxvi. f " Climate and Time," chap, xviii ; Wallace's " Island Life," chap. ix. X " Quarterly Journal of the Geological Society," vol. xxvi, 18Y0, pp. 514- 517; vol. xxvii, 1871, pp. 57-60. * " Manual of the Geology of India," Part I, pp. xxxv-xxxviii, 102, 109-112, 229. II " Quarterly Journal of the Geological Society," vol. xliii, 1887, pp. 190- 196. " Die oarbone Eiszeit," by Dr. W. Waagen, " Jahrbuch d. k. k. geol. Reich- enstalt," Vienna, 1888, vol. xxxvii. Part II, pp. 143-192; reviewed in the "American Geologist," vol. ii, pp. 836-340, November, 1888. APPENDIX. 593 the Permian period ended with disturbance and mountain- building along a somewhat irregular west-to-east course through southern Ireland, Wales, England, northern France, Belgium, Germany, and southern Kussia ; * and it is to be remarked that this European orographic line lies approximately in the same great circle with the Appalachian ranges, both being in- cluded by an arc of ninety degrees. Transverse to this circle the Sinian Mountain system of eastern Asia was formed in the same epoch, stretching from southwest to northeast along the border of the Old World as the Appalachians similarly bound our own continent, f Each of these mountain systems was perhaps much longer than the extent now remaining, and each has been reduced by erosion to only moderate heights ; but it is not improbable that their altitude originally was like that of the loftiest ranges of the world, some of which have been formed and the others much uplifted during the last geologic period. The shortness of the time that has elapsed since the latest glaciation of North America, according to the observations and computations of N. H. Winchell, Andrews, Wright, and Gilbert, shows that this cold epoch was not coincident with the period of eccentricity of the earth's orbit, which is regarded by Croll, Geikie, and Wallace as the primary cause of the Ice age. Eccentricity, therefore, had no share in producing this most recent glaciation, which was more intense and severe, and probably more sudden and brief, than the earlier very cold epoch of the Ice age, as indicated by comparison of the mo- rainic and other drift deposits. Furthermore, it seems proba- ble that the Quaternary Glacial period, including its two or more epochs of glaciation, each subdivided by episodes of tem- porary retreat and readvance of the ice, besides the principal interglacial epoch of warm or temperate climate, and perhaps complete departure of the ISTorth American ice-sheet, was wholly subsequent to the maximum eccentricity which the * Prestwicb's " Geology," vol. i, p. 300. f " Geological Researches in China, Mongolia, and Japan," by Raphael Pum- pelly, " Smithsonian Contributions," vol. xv, 1867, chap. vii. The conclusions reached by Pumpelly concerning this mountain system are fully confirmed by the subsequent grand work of Baron Richthofen on the geology of China. 594 THE IGE AGE IN NORTH AUERIGA. earth's orbit attained two hundred thousand years ago. Through all the past ages of geology, also, the earth has from time to time passed through similar stages of increased eccen- tricity, sometimes having a still higher maximum,* which we should expect, in accordance with Croll's theory, to find re- corded by deposits of glacial drift intercalated in the Tertiary, Mesozoic, and Palaeozoic strata of circumpolar and temperate regions. But the recentness of the Quatern^y glaciation, and the general absence of earlier drift formations,! excepting within the Carboniferous and Permian periods, seem to demon- strate that eccentricity has not been the primary cause of gla- ciation, either with the concurrence of climatic conditions and changes of the course of winds and of oceanic currents such as might attend its slight modifications of the seasons, while the present arrangement and relative heights of land and sea were unchanged, or, as Wallace suggests,:|: with much greater elevation of the areas glaciated, which he thinks to have been necessary, seconding the effects of eccentricity, for the accumu- lation of ice-sheets. Indeed, we may well doubt that eccen- tricity has exerted any determining influence in producing unusual severity of cold either during the Quaternary or any former period. Elevation of broad areas, as half of North America and half of Europe, either synchronously or in alternation, to such heights that their precipitation of moisture throughout the year was nearly all snow, forming gradually ice-sheets of great thickness, seems consistent with the conditions of the earth's crust and interior, which are indicated by the changes in the levels of glaciated countries. A molten magma beneath the solid crust appears, in connection with contraction of the earth * CroU'a " Climate and Time," chap, xix, with plate iv, representing the variations in the eccentricity of the earth's orbit for three millions of years before A. D. 1800, and one million of years after it. f Nordenskiold reports that, in sections observed by him in Spitsbergen and Greenland, including all formations from the Silurian to the Tertiary, and occu- pying in the aggregate, as he estimates; not less extent than a thousand English miles, he has never observed erratic blocks nor any evidence of glacial action. — "Geological Magazine," II, vol. ii, ISTS, pp. 626-532, and vol. iii, 1876, p. 266. X " Island Life," chaps, viii, ix, and xxiv. APPENDIX. 695 and the formation of mountain-ranges, to afford an adequate explanation of glaciation. It is probable that the great up- lifts which are thus supposed to have caused ice-accumulation were very slow in their progress, and that their effect upon extensive continental areas was so distributed that the maxi- mum changes in slope on their borders would nowhere exceed twenty or thirty feet per mile, while perhaps some portions of the uplifted region would receive no change of slope. And the subsidence beneath the weight of accumulated ice was probably equally slow and similarly distributed, no limited district being greatly changed. Excepting the rare instances where disturbances of mountain-building or extraordinary ris- ing or sinking of mountain-ranges were associated with these movements, the contour of the country, with its valleys, hills, and mountains, has remained in general the same from pre- glacial time through the Ice age to the present with only changes of slope, small in any limited tract, which in long distances allowed great upheavals and depressions. The ele- vation of the central part of glaciated areas, with downward slopes on all sides, would favor the outward flow of the ice-sheets and their erosion and transportation of the drift. But mount- ains and hills jutted upward in ridges and peaks within the moving ice-sheet, as they now stand forth in bold relief above the lowlands ; and the ice with its inclosed drift was pushed around and over them, some portions being deflected on either side, and usually a larger part being carried upward across their tops. Katahdin, the White Mountains, the Green Mount- ains, and the Adirondacks, stood directly in the pathway of the ice outflowing southeastward from the Laurentian high- lands. Its thickness in northern New England and northern New York seems to be measured approximately by the eleva- tion of the highest of these summits above the adjoining low- lands, about one mile ; but northward the ice-sheet evidently was somewhat deeper upon the valley of the St. Lawrence, and Professor Dana's estimate seems still reliable, that its maxi- mum depth, lying on the water-shed between this valley and Hudson Bay, was probably about two miles. B. ON THE GLAOIATION OF EASTEEN CANADA. Since the preceding chapters went to press, there has come to hand a very valuable paper by Mr. Eobert Chalmers, of the Canadian Geological Survey, on the glaciation of Eastern Canada,* which should not be passed without notice. His conclusions will best be given in bis own words, which are : That the glaciation of Eastern Canada has been effected by local glaciers on the higher grounds and drift-ice or icebergs on the lower coastal areas. In their movements, the glaciers, generally speaking, followed the slopes of the land or the drainage-channels. They seem to have had extensive gathering-grounds upon the more elevated parts of the country, where snow-flelds and niiie ice existed. Whenever motion began, these became converted into glacier-ice. Upon those areas where the snow never underwent change into ice, no striation of the rocks is found. Some of the glaciers appear to have been quite large, and those from adjacent drainage -areas may have coalesced on the lower grounds and become confluent; at all events, the slopes and coastal tracts are, generally speaking, more glaciated than the interior and higher grounds. Each area of central dispersion has, however, had its own glacier or glaciers. In Nova Scotia there was a shedding of the ice from the Cobequid Mountains northward and southward ; and probably the elevation known as the South Mountain likewise sent glaciers down its slopes on either side. In New Brunswick the low water-shed running across it from northwest to southeast sent ofi gla- ciers in opposite directions, or northeastwardly on the northern slope and southeasterly on the southern, these courses being deviated from in a greater or less degree, however, according as the ice was influenced by local topographic features. The Schickshock or Notre Dame Range in Quebec, and its continuation southwestwardly along the interna- tional boundary, likewise shed the ice in both directions at about right angles to the main axis of the chain— that is, nearly southeast- ' Canadian Record of Science," April, 1S89, APPENDIX. 697 ward and northwestward — while the Archaean area north of the St. Lawrence and Great Lakes sent sheets of ice down its slopes in all directions around its circumference. On the east side of Hudson Bay the ice moved directly westward into its basin, according to Dr. R. Bell and Mr. A. P. Low. Considerable areas of rock-surface in the interior and more elevated portions of Eastern Canada, where gathering-grounds for glaciers may be supposed to have existed, are without striae or other evidence of glaciation, the decomposed rock lying undisturbed, except from subae- rial action, and bowlder-clay being absent. Occasional smaller patches of similar character are met with near the coast. These, during the Ice age, were probably covered by snow only, or by ice which had little or no motion. The evidence of the action of icebergs or floating ice observed by Sir William Dawson * and the writer t are chiefly in the St. Lawrence Valley and on the Bale des Chaleurs coast. In the former, the mark- ings produced by these occur, so far as I have observed them, only on rock-surfaces below the three hundred and fifty to three hundred and seventy-five-foot contour-line above sea-level, while on the coast of the bay referred to they were not seen higher than two hundred feet above its surface. Icebergs or drift-ice played an important part in striating the ledges on these lower levels and in transporting bowlders. On the Isthmus of Chiegnecto the striation of some rock-surfaces is attributable to them. I These facts do not, as some might suppose, bear seriously against the theory of general glaciation advocated- throughout the present Tolume ; for, in that theory, abundant allowance can be made for the diversity in direction of the striae over the region under consideration. In the breaking up of the Glacial period, the movement would be more and more modified by the local topography, and the separate centers of radiation would more and more dominate the direction; From the * " Acadian Geology," second and third editions, " Notes on the Post-Pliocene Geology of Canada," 1872, " Canadian Naturalist," etc. f " Annual Report of the Geological Survey of Canada," 1886, vol. ii, Part M. "Transactions of the Royal Society of Canada," 1886, in a paper on "The Gla- ciation and Pleistocene Subsidence of Northern New Brunswick and Southeast- em Quebec." X " Annual Report of the Geological Survey of Canada," 1885, vol. i, Part G. G. 598 THE IGE AGE IN NORTH AMERICA. nabure of the case also, the striae now remaining upon the rocks would represent the very last movement of the ice over them. It is not strange, therefore, that, from pronounced axes of elevation, such as cross Nova Scotia and New Bruns- wick, there should have been at this time an outward move- ment in different directions, and that the signs of that move- ment should now be more observable than those of any other. As to the movements spoken of as descending northward into the valley of the St. Lawrence from the axis of elevation run- ning along the international boundary between Maine and Quebec, I should be inclined to regard it as a motion in the opposite direction (i. e., toward the southeast), when the ice from the Laurentian highlands pushed completely over the Notre Dame range, as we know it did over the Green Mount- ains, transporting bowlders from the northwest side com- pletely over the range and depositing them in the valley of the Connecticut.* The signs of this upward movement of ice can be observed on the stoss side of almost any rocky elevation within the glaciated area. I have noticed them on many of the mountains of New England, where there was no possibility for an accumulation of ice upon the summit sufficient to pro- duce local glaciers. By recurring to pages 233 and 343, the reader will see illustrations, on a small scale, of deep furrows in hard rock which have been formed by an upward move- ment of ice. On page 343 the camera was pointed downward about 20° toward the water. Nearly all the facts concerning the variety of movements in the ice referred to in Mr. Chalmers's paper will be found to have been incorporated in the map on page 392. In my own experience I have come to pay less and less attention to the direction of glacial strise as indicative of the general course of the -ice-movement ; for, as already said, these strisB represent, in the main, the last part of the movement during the decay- ing stages of the ice-sheet, when the local topography largely dominated the direction. I well remember an illustration of this upon a mountain near Glacier Bay, east of Muir Inlet. The first tributary glacier to the right of the inlet, shown upon * See page 217. APPENDIX. 599 the map on page 40, is moving to the north, joining the main body of ice some distance above the head of the inlet. Of course, the glacial striae produced by that branch are from south to north ; but, at a recent time, the Muir Glacier ex- tended much farther south than now, filling the whole inlet, and rising over the mountains on either side of it. Of this I found abundant evidence, both in transported material upon their summits and in glacial striae as high as thirty-seven hun- dred feet above the inlet, which must have been produced by ice moving from the north to the south. Here, therefore, within the distance of a single mile, are glacial striae parallel to each other, but running in exactly opposite directions, and, if this whole region should be depleted of its ice, the striae would remain to puzzle and perhaps confound the observer. The limitation of supposed iceberg-action in the valley of the St. Lawrence to the three-hundred-and-seventy-five-foot contour- line above the sea-level, spoken of by Mr. Chalmers, confirms our theory, discussed on page 314, that the depression in the vicinity of Montreal need not be supposed to have been more than about five hundred feet. The unglaciated areas referred to by Mr. Chalmers, on page 597, would seem to be analogous to the larger unglaciated area in southwestern Wisconsin spo- ken of on pages 130 and 184, and appearing upon our general map of the glaciated area. II^DEX. Aar Glacier, 167, 225, 226. Abbeville, France, 513, 514. Abbott, Dr. C. C., finds palaeolithic implements at Trenton, N. J., 457, 509, 511, 515 et seg., 530, 533, 556; on early man in Ohio, 529. Abbott, Richard, 518. Ackley, Pa., 137, 138, 286. Adams county, Ohio, 142. Adelphi, Ohio, 141, 212. Adirondacks, 211, 317, 318, 358, 437. Africa, glaciation in South, 436. Agassiz Glacier, 30. Agassiz, A., on recent elevation of Peru, 583. Agassiz, Louis, discovers glacial mo- tion, 2; on depth of glaciers, 167; on terminal moraines in Ammonoo- suc Valley, N. H., 193; on parallel roads of Glen Eoy, 324. Aire Glacier, 397-399. Airy, Sir George B., 574. Akron, Ohio, 313. Alaska, islands of, 15, 23, 39, 51, 161 ; glaciers of, 25 et seg., 36 et seg., 160 406, 438, 440 ; condition of northern, 33; buried forests of, 58; rainfall of, 64, 162. Alaskan Peninsula, glaciers on, 32. Albany, N. Y., 289. Aletsch Glacier, 92, 201, 324. Alexander Archipelago, 160, 164. Alleghany Mountains, 119, 212, 269, 322, 331, 442. -Alleghany River, 330, 494; valley of erosion, 198, 270; preglacial drain- age of, 276 ; glacial drainage of, 280, 336 ; terraces on, 288. Alpine glaciers, terminal moraines of, 16, 19; velocity of, 72; size of, 73, 76; depth of, 92, 167, 173, 396; re- cession and advance of, 93, 409, 490 ; subglacial streams of, 224, 22S ; for- mer extension of, 588. Alps, bowlders of, 206, 207, 219 ; lakes in, 324, 225, 414; erosion of, 201, 224, 225, 248 ; plants on, 379, 386 ; insects of, 388. Ameralik Fiord, 91. Amherst, Ohio, 110. Andes, existing glaciers in, 96 ; Quater- nary uplifts of, 583, 687, 588 ; ancient glaciers in, 688. Andover, Mass., bowlders in, 208 ; kames in, 297 ; kettle-holes in, 472. Andover, Me., 309. Andrews, Dr. E., on date of Glacial pe- riod, 471, 649, 584, 593. Androscoggin River, 193, 302, 309. Animals, extinction of, during the Gla- cial period, 386, 667. Animal remains in glacial deposits, near Morgantown, W. Va., 338 ; in valley of the Somme, 514; at Tren- ton, N. J., 527 ; valley of the Little Miami, 523 , in California, 560. Antarctic Continent, glaciers of, 92, 100 ; icebergs of, 100 et seg. ■ depth of ice in, 173. Apennines, 379. Appalachian Mountains, 199, 229, 269, 322, 437 ; correlated with Carbon- iferous or Permian glaciation, 437, 592. Archibald, Pa., 291. Arkansas River, 149, 197. Armstrong, Dr. A., on transported bowlders in the Hudson Bay region, 215. Aroostook River, 303. Asbury, Pa., 134. Asia, 440; glaciers of, 92, 95, 96; 602 TEE ICE AQE IN NORTH AMERICA. loess in, 369, et seg. ; existing plants of northeastern, 372, 376 ; tempera- ture of northern, 428 ; glaciation of, 434, 435 ; absence of general glaci- ation in, 589. Assiniboin River, 293, 295. Aughey, Prof., discovers paliBolithic im- plements, 568. Auk Glacier, 27. Aunt Ann's Kun, Ohio, 490. Australia, Carboniferous or Permian glacial deposits in, 692. Babbitt, Miss Franc, palaeolithic dis- coveries of, 509, 611, 538, 539, et seq. Baffin Bay, 84, 443. Bainbridge, Ohio, 212, 286, 332, 467. Bakewell, Mr., on rate of recession of Niagara Falls, 456. Baldwin, Mr. Prentiss, 36. Baltimore and Ohio Railroad out, 661. Bancroft on the antiquity of man, 666. Barnstable county, Mass., 123. Bassett's Creek, 469, 460. Bauerman, H., 581. Beaches, raised, 366, 367, 539, 648, 676, 680, 681. Beach Haven, Pa., 119, 134, 286. Beardslee, Commander, 89. Beardslee Islands, 39. Bears, 386. Beaver county. Pa., 212. Beech City, Ohio, 286. Beech Flats, Ohio, 333, 334, 344. Belcher, Sir Edward, visits Kotzebuc Sound, 33. Bell, Dr. R., 580, 597 ; on transported bowlders in the Hudson Bay region, 216. Bellevue, Pa., 335, 336, 344. Belvidere, N. J., 128, 286, 526. Benton, Mr. E. R., on the Richmond train of bowlders, 209, 210. Benton, Pa., 134. Bessels, Dr., on transported bowlders in the Hudson Bay region, 215. Bethlehem, Pa., 524, 525. Big Beaver River, 270, 280, 286-288. Bigger township, Ind., 491. Big Sandy River, 270, 342, 344. Big Stone Lake, 282, 314, 366. Birtle, Manitoba, 295. Bismarck, Dakota, 146, 214, 442. Bison, 33, 886, 567, 560. Black River, Ohio, recession of falls of, 466. Blake, Mr. William P., on glaciers of the Stickeen River, 25, 27. Blandford, W. T., 686. Blennerhassett Island, 287. Block Island, 123. Blue Mountains, N. J., 119, 130. Blue Ridge. See Kittatinny Mount- ain. Bolivar, Ohio, 286. Bonney, T. G., on cirques, 244. Boone county, Ky., terminal moraine in, 142; bowlders in, 218, 327, 328, 346, 846. Borden, Professor, reports vegetal de- posits, 492. Boston, Mass., and vicinitv, 109, 123, 209, 251 et seq., 401, 412" 488, 580. Bottomley, Mr., experiment by, 5. Boussingault, M., experiment by, 4. Bouv6, T. T., 301. Bowlder-clay. See Till. Bowlders, 118, 119; size of, 146, 152, 157, 206 et seq., 434, 436 ; elevation of, in ice, 169, 216 et seq. ; transpor- tation of, 206 et seq., 411, 412, 434. Bradford, Mass., 207. Bralnerd, Minn., 544, 545. Branner, Professor J. C, on depth of ice in Pennsylvania, 169. British America, 407 ; terminal mo- raine in, 119; Missouri coteau in, 121, 148, 187 et seq., 214 ; depth of ice in, 171-173 ; transposition of bowlders in, 148, 174, 214 et seq. ; erosion in, 229 ; glacial drainage in, 293 et seq. ; preglacial plants of, 375. British Columbia, 148, 248 ; glaciers of, 16, 125, 151, 163 ; islands of, 23 ; rivers of, 25; wells in, 157; oscilla- tions of, 581. British Isles, maximum post-glacial up- lift, 163, 582. Brooklyn, N. Y., 126, 320. Brown county, Ind., 142, 218, 491. Brown county, Ohio, glacial boundary in, 142, 442 ; bowlders in, 212. Brown, Robert, on Vancouver Island, 157, 158, 160. Brown's Valley, 282. Brunswick, Me., 303. Brush Creek, 332 et seq., 344. Buffalo, N. Y., 351, 354, 456. Burbank, Mr. L. S., on preglacial ero- sion, 231. Burchill, Mr., reports wood in a well, 491. Buried channels, in trough of the INDEX. 603 Ohio, 270, 2*71, 532 ; in Knox county, Ohio, 2*71 ; connecting Beaver Creek with Grand River, 272 ; in Clark county, Ohio, 272 ; connecting Lake Ontario with the Hudson, 273 ; in Cuyahoga county, Ohio, 274; be- tween Lakes Huron and Ontario, 275 ; in Michigan and Illinois, 276 ; in Pennsylvania, 276, 277, 292 ; in New York, 275, 277 ; in Canada, 277 ; re- lation of, to mining interests, 292 ; from Whirlpool to St. Davids, 453 ; in Minnesota, 469 «< seq. Buried forests, at Muir Glacier, 58 et seq., 203, 204, 482 ; theory of, 476, 479 , in Ohio, 482, 483. Bute Inlet, 158. Butler county, Ohio, striae in, 481 ; or- ganic remains in glacial deposits in, 482, 488 et seq., 491, 493. Butterflies, Alpine species of, upon the White Mountains, 388. Cairo, III., 198. Calaveras skull, 563 et seq. California, 361 ; existing glaciers in, 13 et seq. • ancient glaciers in, 148, 149, 151, 152; terminal moraines in, 194; prehistoric man in, 369, 558; flora of, 376; river-beds of, 558 et seq. Camel, 386, 560. Campbell county, Ky., 142, 327. Canada, 211, 279, 285; depth of ice in, 173 ; of till in, 228 ; buried channels in, 274, 277; lake ridges of, 354; salt deposits of, 408; submergence in, 413 ; glaciation of eastern, 596 et seq. Canadian Pacific Railway, 25. Canton, Ohio, 140, 286. Cape Agassiz, 89. Cape Breton Island, absence of Quater- nary marine beds on, 580. Cape Cod, 123, 125, 126, 179, 207, 208, 309. Cape Dudley Digges, glacier near, 78. Cape Forbes, 89. Cape St. Roque, 422. Carboniferous glacial epochs, 437, 585, 691, 594. Caribou, 557. Carll, Mr., on preglacial drainage, 272, 276, 277. Carmichaels, Pa., 493. Carr, Mr. Lucien, finds palaioliths, 517. Carroll county, N. H., 311. CarroUton, N. Y., 276. Carver visits Falls of St. Anthony, 463. Cascade Range, 15, 19, 23, 121, 149, 150, 158, 163, 165, 369. Cat, 386. Catskill Mountains, 133, 167, 170. Cattaraugus county, N. T., 137 276, 334. Caucasus Mountains, 95, 379. Cause of the Glacial period, chapter on, 405-447; not strictly a geological question, 405, 445 ; cold not a suffi- cient, 405 ; theories of the, enumer- ated, 407 ; not a shifting of the poles, 407 ; not solely excessive moisture, 408 ; changes in the distri- bution of land and water, a possible, 408 et seq. ; continental elevations, a probable, 410 et seq. ; variations in the temperature of space and the heat of the sun, a possible, 416 ; CroU's theory of, 416 et seq., 445 ; precession of the equinoxes, a, 416 et seq. ; changes iu eccentricity of the earth's orbit, a, 418 ; changes in the Gulf Stream, 419 «< seq. ; relation of trade-winds to, 420 ; of Cape St. Roque to, 422 et seq. ; relative coolness of the southern hemisphere, a, 424, 425 ; Woeikof's explanation, 425, 426 ; ignorance of the laws govern- ing the distribution of the heat upon the earth, 427, 446; inequality of this distribution, 428, 445 ; diather- mancy of the atmosphere, 429, 446 ; defective evidence of former glacial periods, 431 et seq., 446 ; number of possible causes, 441 ; local centers of dispersion, 441 et seq., 447 ; indi- rect influence of Greenland ice, 443 ; present ignorance of, 444, 447 ; ap- pendix on, 573-596. Cedar wood in glacial deposits, 482, 488-490, 493. Centers of radiation, 151, 179 et seq., 194, 443, 596, 598. Central Pacific Railroad, 497. Chaleurs, Bay of, 597 ; marine beds overlying glacial drift in, 680. Chalmers, Robert, 580; on depression in lower St. Lawrence, 413; on gla- ciation of eastern Canada, 696 et seq. Chamberiin, President T. C., 576; on the position of bowlders in clay, 116 ; on rock-scoring, 119; discovers ter- minal moraine, 120; on loess, 144, 364, 365 ; attenuated border, 146 ; 604 THE 10 E AGE IK NORTH AMERICA. on glacial boundary in Dakota, 147 ; west of Dakota, 148 ; on moraine of central New York, 178 ; on mo- raine west of the AUeghanies, 1 79; on the Kettle Range, 183, 184, 193, 195, 444 ; on depth of till, 228 ; on drum- lins, 256, 262 ; on CroU's theory of successive glacial epochs, 439; 475 ; on post glacial denudation in Wis- consin, 470 ; on two glacial epochs, 475, 478. Champaign county, Ohio, 227, 228, 272. Chappaquiddick Island, 123. Charleston, W. Va., 339, 494. Chatham Strait, 38, 55. Cherryfield, Me., 308. Cherry Talley, Pa., 167 el seq., 'ill. Chesapeake and Ohio Railroad cut, 339. Chespeake Bay, 135, 501. Chewtown, Pa., 286. Cheyenne River, 146, 147, 214, 295. Chili, glaciers in, 92, 96, 97. (;hilkat, Alaska, 27, 29. Chillicothe, Ohio, 212, 286, 467, 492. China, loess in, 359 et seq. ; flora of, 376, 377. Chippewa Creek, 458. Christianshaab Glacier, 91. Chugatch Alps, Alaska, glaciers on, 32. Cincinnati, Ohio, 142, 270, 271, 287, 327, 482, 530 ; glacial dam at, 326 et seq., 358, 493. Cirques, description of, 242 ; formation of, 244 et seq., 315; prevalence of, 247. Clark county, Ohio, 272. Clarksburg, W. Va., 338. Claymont, Del., palaeolithic implement discovered at, 551 et seq. Claypole, Professor E. W., 574; on depth of drift, 227; on Cincinnati ice-dam, 345-347 et seq. ; on the Deluge, 388. Cleavering, explorations by, 90. Clermont county, Ohio, glacial boundary in, 142; bowlders in, 212; vegetable deposits in, 492. Cleveland, Ohio, 274, 327. CHmate, preglaeial about the north pole, 374, 380, 387. Close, Mr. M. H., on drumlins, 260, 261. Coast Range, 15, 23, 158, 165, 379. Colchagua, 97. Cole's Spring, N. Y., 276. CoUett, Professor, reports vegetable deposits in Indiana, 493. CoUomb, M. E., on subglacial streams in the Alps, 225. Colorado, 194, 244, 391; former gla- ciers in, 148, 149. Colorado River, 198. Columbia deposits, 135, 410, 411, 501, 502, 522 et seq., 553. Columbiana county, Ohio, fringe in, 139 ; terminal moraine in, 140 ; bowlders in, 212, 286. Columbia River, 198. Columbus, Mount Vernon, and Akron Railroad cut, 141. Concord, N. H., 290. Conewango Creek, 137-139, 276, 286. Connecticut, depth of ice in, 167, 171 transported bowlders in, 208, 209 depth of till in, uncertain, 229 drumlins in, 255. Connecticut River, 253, 302, 303, 305 et seq., 544 ; glacial drainage of, 290, 291, 436, 524. Connell, Mr. J. M., reports wood in a well, 492. Contocook River, 290. Contractions of the earth, probably causing glaciation, 573, 575, 585, 590. Cook, Captain, 31 ; on icebergs of the Southern Ocean, 103, 104. Cook, Professor George H., 527, 557 ; discovers terminal moraine, 120, 136, 193; moraines of retrocession, 179 ; on transported bowlders, 211. Copper River, 32. Coral islands indicating subsidence, 584. Cornish, N. H., 311. Coteau des Prairies, 184, 186, 322, 367, 546. Coulee, 191, 294. Cowlitz River, 21. 'Crenate character of the glacial margin, 118, 119,442 etseq. Cresson, Mr. H. T., palseolithic discov- eries of, 609, 512; at Medora, Ind., 533-537; at Claymont, Del., 551, 553 et seq., 557. Crevasses. See Fisspres. CroU, Mr.. J., 673, 686,. 588, 592-594; theory of glacial motion, 6 ; on ice- bergs in Southern Ocean, 105 et seq. ; cm depth of ice in the Antarctic Continent, 173 ; theory of cause of Glacial period, 416 et seq., 448, 450, 468 ; on date of Glacial period, 474, 494 ; on post-glacial erosion- in Scot- INDEX. 605 land, 470 ; on succession of Glacial periods, 431 el seg., 475 ; on rate of denudation, 503. Crosby, Professor W. 0., 574, 583, 585. Cross Sound, 29, 38, 39, 55-57, 161. Crust, relation of the earth's, and in- terior, 574, 578, 594. Cuba, post-glacial elevatiQn of, 583. Cuyahoga River, 273. ,.-^ Cypress Hills, 171, 17£ Dakota, glacial boundary in, 119, 144 et seg., 441, 442 ; marginal drainage in, 146, 147, 292, 295; fringe in, 146 ; Coteau des Prairies in, 184, 186, 322, 367; Missouri coteau in, 187, 189, 292 ; bowlders in, 213, 214; lakes in, 322, 357; loess io, 368. Dall, W. H., explorations of, in Alaska, 29, 161 ; in Kotzebue Sound, 33 ; in northern Alaska, 34. Dana, Professor J. D., 287, 444, 673, 574, 586, 590, 595 ; on depth of ice, 171, 173; on transported bowlders, 209 ; on kames in the Connecticut Valley, 306; on the floods of the Connecticut Valley, 306, 524. Danville, Ohio, 141. Darrtown, Ohio, 488. Darwin, Charles, 568, 690; on glaciers of South America, 97 et seq. ; on dust- storms, 361 ; on earthworms, 370. Darwin, Professor G. H., 573. Date of the Glacial period, 336, 628, 549, 584, 693 ; chapter on, 448-506 ; astronomical evidence insufficient, 448 ; calculations of, affected by uni- formitarianism, 449 ; tendency to make more recent, 449 ; question of, geological, not astronomical, 450 ; calculated from the erosion below Niagara Falls, 452 el seg.- below the Falls of St. Anthony, 468 et seg. ; from falls in northern Ohio, 466; from Paint Creek cut-off, 467 ; from erosion of Raccoon Creek, 468 ; of Plum Creek, 469 ; from general amount of denudation, 470; from the shores of Lake Michigan, 471 ; from deposition in kettle-hole, An- dover, Mass., 472 ; bearing of theory of successive Glacial periods upon the subject of, 475; freshness of vegetable deposits indicates a receilt, 482 et seg. ; buried peat-beds no bar to theory of a recent, 484 et seq. ; evidence from Lakes Lahontan and Bonneville concerning, 496 et seg. ■ Shaler on, 601 ; McGee on, 501 ; Prestwich on, 502 et seq. Davidson Glacier, 27. Davis, Professor WiUiam M., on drum- !in, 264, 265 et seg., 260 et seq. ; on glacial lakes, 324, 326. Davis Strait, 77, 397. Dawkins, Professor Boyd, on the origin of bowlder-clays of Britain, 396, 397 ; visits Trenton, N. J., 609, 518, 619 ; on prehistoric man in California, 567. Dawson, G. M., 151, 581, 682 ; discovers terminal moraine, 120 ; on Missouri coteau, 148, 189 et seg. ; on glacial phenomena of Bute Inlet, 158 et seg. ; expedition to the Yukon, 162; on depth of ice, 171 ; on transportation of bowlders, 171, 174, 214, 215 ; on the glacial theory, 172 ; on junction of two ice-streams of the Northwest, 185 ; on cirques, 248 ; on glacial drainage, 293. Dawson, Sir William, 577, 580, 597; on submergence in Canada, 413 ; on prehistoric man in California, 665. Dead Sea, Palestine, gravel deposits about tlie, records of the Glacial pe- riod, 500. Deblois, Me., 308. Delaware, palaeolithic discoveries in, 551 et seg. Delaware River, 119, 122, 127, 130, 131, 136, 198, 285, 286, 318, 411, 460, 468, 621 et seg., 524, 526, 627, 651, 653, 557. Delaware Water-Gap, 132, 168, 169, 198, 217 et seg., 230, 312, 622, 525. Delta terraces, 286 et seq. ; in New Hampshire, 290 ; in Minnesota, 294 ; in Maine, 308, 309 ; on Cape Cod and Long Island, 309 ; in Wisconsin, 310 ; south of the glacial limit, 411, 412. Deluge, the, perhaps connected with the Glacial period, 388. Depth of ice during the Glacial period, 365, 579, 580, 596 ; in Yellowstone Park, 149 ; in California, 162 ; in Maine, 166, 167; in New Hampshire, 166; in Vermont, 166; in Massa- chusetts, 167; in Connecticut, 167, 171 ; in New York, 167, 231 ; in Pennsylvania, 167_, 169, 170; in New Jersey, 170, 623 ; in British America, 171, 172; over the region of the Great Lakes, 172, 178 ; in Greenland, 606 THE ICE AGE IN NORTH AMERICA. 173, 174; in the Alps, 173, 396; in the Antarctic Continent, 173 ; in Labrador, 174 ; in Norway, 395 ; on the Harz Mountains, 395 ; in Scot- land, 396 ; in England, 398 ; effect on snow-fall, 444. Desor, on depth of glaciers, 167; on erosion in the Alps, 225 ; on rate of recession of Niagara Falls, 455. Devil's Lalie, 183. Diamond Peak, 19. Diller, Mr., on glaciers of the Cascade Range, 19. Disco Bay, 67. District of Columbia, 553. Dixon's Entrance, 30. Dog, 386. Dolager's nunataks, 73. Dolfus, on erosion in the Alps, 224. Dolphins, 386. Driftless area of Wisconsin, 120, 365, 444. Drumlins, chapter on, 251-267 ; local ity of, 251, 254 et seq. ; description of, 22, 252 et seq. ; Upham on, 252, 259, 262 ; direction of axes of, 2BS, 254, 257, 259, 262 ; Davis on, 254 H seq., 260 et seq. ; in northeastern Massachusetts, 254 ; Percival on, 255 ; Johnson on, 256, 261 ; Cham- berlin on, 256, 262 ; theory of, 257 et seq ; absence of kettle-holes in, 258 ; earlier than kames, 258 ; Shaler on, 258 ; King on, 258 ; Hitchcock on, 260, 262 ; Close on, 260 ; Geikie on, 261. D'Urville, explorations of, in the Ant- arctic Continent, 102. • Dust-fogs, 361, 362. Dutton, Captain, C. E., 574. Eagle, Wis., 183, 310. Earthworms and loess, 370. Easton, Pa., 128, 522. Eccentricity of the earth's orbit, varia- tions of, 418, 594 ; not coincident with the latest epoch of glaciation, 685, 593 ; probably not influential in causing glaciation, 594. Elephants, 386, 387, 514. Elevation of bowlders, in ice, in Penn- sylvania, 169, 211, 217 et seq.\ in British America, 172, 214; in Mas- sachusetts, 207, 209, 210; in New Hampshire, 216, 217 ; in Vermont, 217 ; in Maine, 217 ; explanation of, 220 et seq. Elevation probably causing glaciation, 573, 584 et seq. ; due to removal of. the ice-sheet, 675, 676, 582. Elizabeth Islands, 123, 126, 175, 320. Elliott, H. W., on Alaskan glaciers, 30, 32. Ells, Mr. R. W., on buried channels in Canada, 277. Elyria, Ohio, 466. Emmons, Mr. S. P., ascent of Mount Tacoma, by, 22. England, 199, 450; glaciers in, 394, 396 et seq. ; changes of level in, 401 ; palseoliths in, 506, 514 ; preglacial man in, 460, 668. Equador, 96. Erosion, by water, 196 et seq., 249, 268 et seq., 332, 438, 462 et seq. ; com- pared with that of ice, 197, 315 ; chemical, 199, 470; glacial, 200 et seq. ; preglacial, 196 et seq., 231, 268 et seq. ; post-glacial, 268, 452 et seq. Eschscholtz Bay, 32, 33. Eskimos, 215, 390, 513, 529; lineal descendants of preglacial man, 388, 668, 569. Essex county, Mass., 266, 259. Europe, existing glaciers of, 92 et seq., 98 ; glacial erosion in, 201, 202, 224, 225, 280, 248, 264 et seq. ; transport- ed bowlders in, 206, 207; drumlins in, 260, 261; kames in, 297, 299; glacial lakes in, 324, 326 ; loess in, 362 ; preglacial plants of, 378 et seq. ; during the Glacial period, chap- ter on, 393-404 ; successive Glacial periods in, 433-436 ; interglacial man in, 613-615; Quaternary oscil- lations in, 582. Evans, Mr., 514. Ewing, Professor A. L., on chemical erosion in the Nittany Valley, 199, 200. Fairfield county, Ohio, glacial boundary in, 141 ; bowlders in, 212. Falconer, Dr., 514. Falls of Minnehaha, post-glacial, 273. Falls of Niagara, 857, 475 ; post-glacial, 273 ; formation of, 452 ; rate of re- cession of, 462 et seq., 602, 549, 668. Falls of St. Anthony, 475, 689 ; post- glacial, 273, 458 et seq. ; rate of re- cession of, 463 et seq., 602, 649. Falmouth, 176 et seq. Fargo, Minn., 294. Faroe Islands, 682, INDEX. 607 Fillmore county, Minn., 495. Finger Lakes, New York, 178, 312, 313, 318, 323, 357. Fiords, 577, 578, 582. Fisher, Rev. 0., 574. Fisher's Island, 126. Fishing Creek, 134. Fissures in glacial ice, 8, 19, 46, 85, 87, 88. Flathead River, 13. Floods at the close of the Glacial pe- riod, 280 et seq., 306 et seq., 347 e( seq., 464, 623-525 et seq. Floras of Faroe Islands and Iceland, 582. Florence, Ky., 327. Fondalen glacier, 94, 203. Foote, Professor H. C, on sediment of subglacial streams, 64. Forbes, Professor J D., discovers gla- cial motion, 2, 80, 86 ; on transported bowlders in the Alps, 207. Forel, M., on recession and advance of Alpine glaciers, 93. Fort Snelling, 281, 549 ; contrast be- tween bluffs above and those below, 459, 460. Fort Wrangel, 27, 35, 122, 160. Foshay, Mr. P. Max, on buried chan- nels, 272. Fossiliferous marine beds overlying glacial drift, 580, 581. Four-mile Creek, Ohio, 488, 490. France, 437 ; dust-shower in, 362 ; pa- Iseoliths in, 506, 513, 514; intergla- cial man in, 558. Frankfort, Ohio, 492. Franklin, Pa., 286, 288. Franz- Josef Land, 95. Fraser River, 25, 198. Frederickshaab Glacier, 326 ; slope of, 173. Freehold, Pa., 488. French Creek, Pa , 286, 288. Frere, Mr. John, 614. Fringe, the, of the glaciated area, 135, 139, 146, 204, 214. Gabb, Mr. William M., 583. Gastaldi, B., on cirques, 244 ; on gla- cial deposits in Italy, 436. Geikie, Professor A., 574, 686, 588 ; on the extent of glaciation in Eu- rope, 395, 396 ; in Scotland, 433. Geikie, Professor James, 673, 582, 585, 588, 593 ; on drumlins, 261 ; on gla- cial erosion, 264 et seq, ; on composi- tion of kames, 297, 299 ; on Glacial period in Great Britain, 393, 397 ; theory of cause of Glacial period, 4:16 et seq. ; on glaciation in Scotland, 434. Genesee River, 273. Germantown, Ohio, 482, 483, 487, 488, 490. Gibraltar Island, Ohio, 236. Gibraltar, Strait of, migrations across, 583. Gietroz Glacier, 325. Gilbert, Mr, G. K., 574, 584, 593 ; on the moraine of Maumee Valley, 179 et seq., 236 ; on lake ridges, 361 ; on subsidence west of the Mohawk, 413 ; on CroU's theory of successive glacial epochs, 439, 475 ; on recession of Niagara Falls, 456, 458 ; on Lake Bonneville, 499 ; on date of Glacial period, 549. Glacial looundary in North America, serrate character of, 118 ; crenate character of, 118, 119; chapter on, 120-165; discoverers of, 120; not always a terminal moraine, 121 ; south of New England, 123 et seq. ; across New Jersey, 126, 127 ; across Pennsylvania, 131 el seq. ; limits of error respecting, 134 et seq. ; in New York, 137 et seq.; in Ohio, 139 et seq. ; in Kentucky, 142 ; in Indiana, 142 ; in Illinois, 142, 281 ; beyond the Mississippi, 144 ; beyond the Missouri, 146 ; in British America, 148 ; beyond the Rocky Mountains, 148 et seq. ; in California, 151 ; in Washington Territory, 154 ; in Brit- ish Columbia, 165; in Alaska, 161 ; cause of irregularity of, 441. Glacial dams, 315, 357, 451, 549; in the Conewango, 138; across Kansas and Platte Rivers, 144, 366 ; in the head-waters of Platte and Arkansas Rivers, 149 ; in Leevining River, 194 ; in New Hampshire, 323 ; in New York, 323 .; in New Jersey, 323 ; in the Alps, 324 ; in Scotland, 324 ; in Greenland, 326 ; in the Ohio, at Cincinnati, 326 et seq., 358, 493 ; in England, 399. Glacial deposits, contrasted with those of water, 115 et seq.; factors deter- mining extent of, 121, 175, 196; cause of stratification in, 128. Glacial drainage, terraces produced by, 138, 283 et seq.; chapter on, 279- 608 THE WE AGE IN NORTH AMERICA. 296 ; floods of, 279 et seq. ; from the Ked River basin, 279, 281, 282 ; in the valleys of the Mohawk and Hud- son, 289 ; from the Oontocook Val- ley, N. H., 290 ; in Grafton county, N. H., 291 ; in eastern Pennsylvania, 292 ; in Dakota, 147, 292, 295 ; from the valley of the Saskatchewan, 293 ; in southwestern Manitoba, 294 ; in southern Minnesota, 296 ; marked by kames, 300, 308 ; in the valley of the Connecticut, 306 et seq., 624 ; in the Little Miami, 532, Glacial erosion, in Washington Terri- tory, 22 ; manner of, 200 et seq. ; variation of, 202 ; small near the margin, 203 et seq. ; in Greenland, 223, 226 ; in the Alps, 224, 248 ; in British America, 227, 248 ; in Penn- sylvania, 168, 230 ; in Scandinavia, 230 ; as affected by secular disinte- gration, 231 et seq. ; at the west end of Lake Erie, 232 et seq. ; of rock- basins in the high Sierra, 237 ; in the Tosemite Valley, 240, 241 ; of cirques, 241 et seq. ; Russell on, 242 ; Lorange on, 246 ; of fiords, 248 ; of lake-basins in Europe, 248 ; summary concerning, 249 et seq. ; J. Geikie on inequalities of, 264 et seq. Glacial lakes, 316 et seq., 367, 368, 549 ; on Pocono Mountain, 134; in the Conewango River, 139; in Kansas and Platte Rivers, 144, 366 ; in the Platte and Arkansas, 149; in Lee- vining River, 194 ; in the Sierra, 239 ; in Contocook Valley, N. H., 290 ; kinds of, 319 ; occupying ket- tle-holes, 319 e< seq. ; caused by dams of glacial dSbris, 322 et seq. ; Lake Winnepesaukee, 323 ; in New York and New Jersey, 323 ; caused by dams of ice, 324 et seq., 414 ; in the Alps, 324 ; in Scotland, 324 ; in Greenland, 326; in the Ohio, 330, 346 ; in the Red River Valley, 366 ; in England, 399. Glacial periods, supposed succession of, 418, 431 et seq., 438, 448; lack of evidence concerning, explained, 431 ; in Great Britain, 433, 434 ; in South Africa and India, 434, 435 ; in Switz- erland and Italy, 436 ; in North America, 436 et seq. Glacial striie. See Rock-soomno. Glacial theory, confirmations of, 116, 118, 119, 129, 156, 210, 283, 286 et seq. ; objections to, by Canadian geol- ogists, 172. Glaciation, probable causes of, 573 et seq. Glaciation, signs of former, chapter on, 108-119; scratches on the rocks, 109 et seq. ; unstratified deposits, 115 et seq. ; distribution of bowlders, 118, 119. Glacier Bay, Alaska, 27, 35, 39, 40, 51, 67, 122, 160, 482. Glaciers, ancient, in the Rocky Mount- ains, 148 ; in the Sierra Nevada, 149 et seq. ; in the British Isles, 393, 394, 896 et seq., 433 ; in Scandinavia, 395 ; in Germany, 402 ; in India, 434, 435 ; in South Africa, 435 ; in the Andes, 688. Glaciers, existing, in the Rocky Mount- ains, 13 ; in the Sierra Nevada, 13 et seq. ; in southern California, 13 ; on Mount Shasta, 15 ; on Mount Hood, 19; on Mount Tacoma, 2 1 ; on Eraser River, 25 ; on Stickeen River, 25; in Taku Inlet, 27; of Lynn Canal, 27 ; of Mount St. Elias, 30; of Chugatch Alps, 32 ; of the Alaskan Peninsula, 32 ; in Glacier Bay, 36 et seq. ; in Greenland, chap- ter on, 67-91 ; in the Alps, 92, 93 ; in Scandinavia, 92, 94 ; in Spitzber- gen. Nova Zembla, and Franz-Josef Land, 95 ; Iceland, 96 ; in Asia, 96 ; in South America, 96 ei seq. ; in New Zealand, 100 ; in the Antairctic Con- tinent, 100 et seq. Glaciers, movements of, 2, 72, 266; semi-fluidity of, 3, 4, 73 ; structure of, 7; slope of, .18, 43, 69, 70, 77, 80, 87, 96, 170, 178, 174, 392; ve- locity of, 3, 50, 70, 72, 74, 93, 174; thickness of, 19, 22, 43, 69, 71, 73, 77, 83, 85, 92, 101, 102, 106, 107, 173, 174. Glen Roy, parallel roads of, -324. Godfrey's Ridge, Pa., 169, 217. Gould, Dr. D. T., on buried channels in Ohio, 274. Grafton county, N. H., 290, 291. Grafton, W. Va., 830. Grand Tower, 111., 143, 281. Granville, Ohio, 468, 493. Gravitation of the ocean toward the ice-sheet, 579. Gray, Professor Asa, 689 ; list of Alas- kan plants identified by, 66 ; on dis- tribution of plants in the northern hemisphere, 372 et seq. INDEX. 609 Great Lakes of North America, 386, 409, 413 ; depth of ice over, 172, 173 ; erosion in beds of, 230 et aeq. ; formation of, 232, 316 et seg., 676, 577 ; Newberry on, 232, 316 ei seq. ; preglacial outlets of, 274 et seq. ; gla- cial drainage of, 279, 280; depths of, 316, 410 ; influence of ice-bar- riers on, 850 et seq., 357. Great Miami River, 142, 270, 271, 283, 286, 287, 532. Great Salt Lake, 496, 498. Green Bay, Wis., 183, 184, 444. Greenland, the glaciers of, chapter on, 67-91 ; area of, 67, 85 ; interior condition of, 67, 76, 395 ; Norden- skiold on, 69 ; Kink on, 69 et seq. ; thickness of ice In, 69, 71, 74, 77, 173, 174, 503; rate of motion of, 70, 72, 74 ; explorers of, 72 ; Hel- land on, 72 ; subglacial streams of, 74, 225 ; icebergs from, 74, 77, 87, 206 ; Whymper on, 76 et seq. ; Kane on, 78 et seq., 85 et seq., 205 ; Hayes on, 81 et seq. ; north of 79°, 84 ; on the eastern coast, 89 ; Nansen on, 90 ; erosion in, 223, 226, 226 ; Marr on, 326 ; preglacial plants of, 374, 375, 380, 406 ; preglacial climate of, 380, 407 ; insects of, 389 ; cause of glaciers in, 407 ; during the Glacial period, 443 ; post-glacial elevation of, 683. Green Mountains, 166. Grinnell Expedition, 78. Grinnell Land, post-glacial elevation of, 583. Grooves. See Eock-scoring. Grote, Mr. A. R., on the White Mount- ain butterfly, 389 et seq. Ground moraine. See Till. Gulf of Mexico, 368, 422. Gulf of St. Lawrence, 413. Gulf Stream, possible changes in, 409, 584 ; effect of, on the Atlantic, 419 ; cause of, 420 et seq. Guyandotte River, 389, 341, 342, 494. Guyot Glacier, Alaska, 30. Hall, Professor James, 576 ; on eleva- tion of bowlders, 218 ; on drift hills of New York, 256 ; on rate of reces- sion of Niagara Falls, 455. Hamilton county, Ohio, 212, 492. Hamilton, Ohio, 114, 117, 490, 532. Hanover, N. H., 306. 39 Hartwig, G., on glaciers of Magdalena Bay, 95. Haughton, Professor S., on transported bowlders in the Hudson Bay region, 215. Haverhill, Mass., 298. Hayes, I., explorations of, in Green- land, 78, 81 et seq. Haynes, Professor H. W., on palseoliths, 509 et seq., discovers palseoliths, 518, 519. Helland, Mr. A., 393 ; on rate of move- ment of Greenland glaciers, 72, 74, 174 ; on slope of Jakobshavn Gla- cier, 173; on erosion in Greenland, 225 ; in Scandinavia, 230 ; on depth of drift in Germany and Russia, 230 ; on formation of cirques, 244. Hennepin discovers Falls of St. An- thony, 463. Herschel, Sir John, 576. Hicks, Professor, on Raccoon Creek, 468, 469. Higginsport, Ohio, 142. Highland county, Ohio, 142, 481, 492. Hilgard, Professor E. W., 589 ; on loess, 362, 363, 367; on depression in Mississippi Valley, 410, 411. Hillsborough county, N. H., 290. Himalayan Mountains, 96, 325, 377 ; Quaternary uplift of, 586, 589. Hind, Professor, on glacial drainage, 296. Hippopotamus, 560. Hitchcock, President E., on size of bowl- ders, 206 et seq. ; on the Richmond train of bowlders, 209 ; on kames in Andover, Mass., 297. Hitchcock, Professor Charles, 580 ; on elevation of bowlders, 166,216, 217; on depth of ice, 174 ; on lenticular hills, 251, 263, 260, 262; on buried kame in Hanover, N. H., 306 ; on CroU's theory, 474. Hocking River, 270, 283, 286. Holkham Bay, 27. Holmes county, Ohio, terminal mo- raine in, 141 ; bowlders in, 212 ; glacial terraces in, 286. Hopkins, Professor W., 573. Horse, 386, 388, 560. Horseshoe Fall, Niagara, recession of, 466. Hudson Bay, 171, 216, 314, 316, 366, 410, 496, 597. Hudson Bay and Strait, raised beaches of, 580, 610 THE ICE AGE IN NORTH AMEBIOA. Hudson River, 122, 198, 2'76, 289,312, 317, 318, 410; submarine fiord of, 27B, ^78, 579. Humboldt Glacier, 83 et seq. Humboldt Kange, 149. Hunt, Professor Sterry, on preglaoial erosion, 231. Hunter, Captain, on depth of Muir Gla- cier, 41. Huntingdon Mountain, 134. Hurricane Creek, W. Va., 341, 342. Button, Captain, 586. Icebergs, on coast of Alaska, 31 ; from Muir Glacier, 36, 46, 47, 63 ; from Greenland, 77, 87 ; in Straits of Ma- gellan, 99 ; of the Antarctic Conti- nent, 100, 102 et seq.; formation of, 205. Iceberg theory, 112, 119, 129, 168, 172 et seq., 210, 413. Ice, characteristics of, 1 et seq., 202 ; is but compressed snow, 6. Iceland, 96, 380, 391, 582. Ice-pillars, 10, 45, 66. Ice-sheet, advance of, 195, 279 ; retreat , of, 125, 138, 193, 195, 280, 281, 282, 303, 309, 312, 347 el seq., 364, 464, 513, 538, 545; causing depression of the earth's crust, 368, 369, 573, 576, 586 ; probably due to eleva- tion, 573, 686, 587. Icy Bay, 30. Illinois, driftless area of, 120; depth of ice in, 173 ; glacial boundary in, 142-144, 441, 442, 444; bowlders in, 213 ; depth of drift in, 227 ; loess in, 367, 480 ; stri^ in, 481 ; buried wood in glacial deposits in, 491. Illinois River, 280, 314. Implement-bearing gravels in Europe, 506, 514 et seq. ; at Trenton, N. J., 520 et seq., 551, 568 ; at Madison- ville and Loveland, Ohio, 531 et seq., 551, 568; at Medora, Ind., 533 et seq., 551, 568 ; at Little Falls, Minn., 540 et seq., 551, 568; at Claymont, Del, 551 et seq. India, 437, 586, 589, 692 ; glaciers of, 96, 325, 414; Permian glacial de- posits in, 435, 692. Indiana, glacial boundary in, 142, 179, 441, 442, 444; bowlders in 213, 536; depth of drift in, 227, 491, 492; wells in, 270, 491,492; loess in, 367, 480 ; striae in, 481 ; vegetal deposits in, 491, 492, 493; palseo- Uthic implements in, 533 et seq. Indianapolis, Ind., 142. Indian Ridge in Andover, Mass., 297. Insects, migrations of, during the Gla- cial period, 388 et seq. ; Alpine spe- cies of butterflies upon the White Mountains, 388. Interglacial deposits, how preserved, 265 et seq. ; theory of, 476 ; in Ohio, 482, 488; in Minnesota, 495, 496. See Vegetable Remains in Glacial Deposits. Interglacial epoch, evidence favoring, 475 ei seq., 496 et seq. ; duration of, 588 ; probable causes of, 573, 588. Interglacial man, in France, 506, 613 ; in England, 506, 614; at Trenton, N. J., 515 et seq. ; at Madisonville and Loveland, Ohio, 530 ei seq. ; at Medora, Ind., 533 et seq. ; at Little Falls, Minn., 637 et seq. ; at Clay- mont, Del., 551 et seq. ; in Nebraska, 667 ; in the bed of Lake Lahontan, 558 ; in Tuolumne county, Cal.. 569 et seq. ; Le Conte on, 568 et seq., 566 ; Dawkins on, 666. Interglacial migrations from Africa to Europe, 583. Interior moraines, 126, 176, 179 et seq. Interior, probable condition of the earth's, 674, 578, 594. Iowa, 322, 441, 444, 545 ; driftless area of, 120, 366; depth of ice in, 173; bowlders in, 213 ; lakes in, 321 ; loess in, 365-368 ; vegetal deposits in, 496. Ireland, 11 297, 320, 321, 893, 394, 396. Irish Sea Glacier, 394, 397-399. Isblink, 71. Italy, glaciation in, 436. Jackson county. 111., 143, 144, 213. Jackson county, Ind., 142, 491, 533. Jackson, Dr. Sheldon, on depth of Muir Glacier, 41. Jackson,' Mr. C. T., cited, 291. Jacobus Creek, 312. Jakobshavn Glacier, 70, 72; rate of discharge, 74, 77 ; slope of, 173, 174 ; Bubglacial streams of, 225 ; glacial dams of, 326. Jamaica, post-glacial elevation of, 583. James, Professor Joseph F., on buried channels in the Ohio, 271. James River, Dakota, 186, 187, 296. INDEX. 611 James River, Va., 412. Jamieson, T. P., 574, 576, 577. Japan, flora of, 373, 376, 381, 382, 384 ; dust-fog in, 361. Jefferson county, Ind., 492. Jennings county, Ind., 491. Jensen, A. D., 72, 73, 173. Johnson, Dr. L., on drumlins in New York, 256, 261. Johnsonville, Pa., 312. Jtivdliarsuk Glacier, 71. Juneau, Alaska, 27, 66. Kamchatka, 380. Karnes, in Rindge, N. H., 290; Win- chendon, Mass., 290; chapter on, 297-314; Edward Hitchcock's de- scription of, 297 ; J. Geikie's de- scription of, 297 ; structure of, 299 ; compared with terminal moraines, 299; theory of, 11, 16, 62, 300; forming at the Muir Glacier, Alaska, 62, 302 ; system of, in New Eng- land, 302 ; buried, 305 ; in the Con- necticut Valley, 306 ; Dana on, 306, 307; at Stroudsburg, Pa., 308, 312; deltas of, 308, 312 ; in Rangeley Lakes, 309; location of, prognosti- cated, 309, 310; of backward drain- age, 310 et seq.; in Schoodic Lake, 310; near Ossipee Lake, 311; near Portland, Pa., 312 ; near the Finger Lakes, N. T., 312; in Ohio, 313; absence of, in the Northwest, 313 ; compared to a skeleton, 314. Kanawha River, 270, 339, 341, 343, 494. Kane, Dr., explorations of, in Green- land, 78, 79, 84 et seq. ; on formation of icebergs, 205. Kansas, glacial boundary in, 144 ; loess in, 366, 367. Kansas City, Mo., 144, 364. Kansas River, 144, 366. Karajak Glacier, 70. Kelly's Island, 232. Kenai Peninsula, glaciers on, 32. Kennebec River, 193, 302. Kenton county, Ky., 327. Kentucky, 142, 173, 213, 327, 328, 835, 345, 346, 437. Kettle-holes, defined, 11 ; formation of, in front of Muir Glacier, 63, 62 ; in Plymouth county, Mass., 125, 320; explanation of, 129 ; resemble sink- holes, 129 ; formation of, 130, 176. 319, 320, 472 ; on Pocono Mount- ain, 134; near Ackley, Pa., 138; in Ohio, 140, 141, 487 ; abundance of, in terminal moraine, 176, 320, 321 Professor Koons on, 176 el seq. ; di rection of longer axis of, 176, 177 in New England, 320, 321 ; age of, estimated, 472 et seq. ; near Free- hold, Pa., 488. Kettle Range, Wis., 120, 130, 179, 183, 184, 193, 195, 321, 367, 443. King, Mr. Clarence, 674; on glaciers of Mount Shasta, W et seq. ; discov- ers terminal moraine, 120, 193; re- port on fortieth parallel, 149 et seq. ; on Yosemite Valley, 241 ; on drum- lins, 258 ; on dust-fogs, 361. Kinnahan, on drumlins, 260. Kittatinny Mountain, 119, 132, 198; elevation of bowlders on, 168, 217 «< seq. ; erosion on, 230. Knox county, Ohio, glacial boundary in, 141 ; buried channel in, 271 ; ter- races in, 286. Koldewey, Captain, on east coast of Greenland, 90. Koons, Professor B. F., on kettle-holes, 176 et seq. Kryokonite defined, 9. Kuro-Siwa, 408. Labrador, depth of ice in, 174 ; insects in, 389-391; submergence in, 412; a center of snow-fall, 443, 597. Lake Agassiz, 358, 466, 538, 539 ; delta terrace in, 294 ; beaches of, 366, 357, 539, 548, 576 ; vegetable deposits in bed of, 493, 496. Lake Bonneville, 498-500. Lake Calhoun, 461. Lake Champlain, 318, 323. Lake Chautauqua, 277, 323, 357. Lake Erie, 173, 314, 361, 368, 409; islands of, 232 el seq. ; preglacial outlet of, 273 et seq., 452 ; glacial drainage of, 280, 356 ; formation of, 316 et seq., 323, 367. Lake Harriet, 461. Lake Huron, 173, 327; preglacial out- let of, 274 et seq. ; glacial drainage ■of, 280 ; origin of, 316 et seq. Lake Humber; 396, 399, 400. Lake Itasca, 539. Lake Lahontan, 498, 600, 568. Lake Lindeman, 29. Lake Michigan, 183, 276, 314, 321,441, 444, 549 ; preglacial outlet of, 276, 276 ; glacial drainage of, 280 ; origin 612 THE ICE AGE IN NOBTH AMERICA. of, 316 et seq.\ post-glacial erosion in, 471. Lake Minnetonka, 183, 184, 321, 357. Lake Mono, California, 243, 49B. Lake Ontario, 134, 351, 354, 358, 452; preglaoial drainage of, 274 et seq. ; origin of, 316 et seq. Lake ridges, around Erie, 361, 353 ; around Ontario, 351. Lake Superior, 171, 173, 184, 185, 285, 327, 444, 548; preglaoial outlet of, 275, 276 ; glacial drainage of, 280 ; origin of, 316 et seq. ■ vicinity of, a center of snow-fall, 443. Lake Traverse, 314, 356, 357, 282. Lake Washington, 153. Lake Wiunepesaukee, 193, 323, 357. Lake Winnipeg, 538, 639. Lakes formed in mountain-building, 586, 588. Lamplugh, G. W., 581. Lanark county, Canada, 413. Lang, Mr. C, 93. La P6rouse, 38, 39. Lattas, Ohio, 492. Laureutian lakes, basins of the, 576, 677. Laurentian mountains, 171, 214, 216, 229, 317, 443. Lava-beds in western United States, 369, 370. Lawrenceburg, Ind., 286, 287. Lawrence county, Pa., 286. Lawrence, Mass., 303, 310. Le Conte, Professor Joseph, 574, 677, 583 ; on succession of glacial epochs, 440 ; on deep placer deposits in Cali- fornia, 558 et seq. ; on antiquity of man in California, 567. Lehigh River, 286, 624. Lehigh Water-Gap, 169. Lenox, Mass., train of bowlders in, 209, 210. Lesley, Professor J. P., on depth of the ice, 167-169; on elevation of bowl- ders, 167, 218 et seq.\ on depth of drift in Pennsylvania, 229 ; on ero- sion in northern Pennsylvania, 230 ; on Cincinnati ice-dam, 831. Lesquereux, Leo, on rate of accumula- tion of peat, 484 et seq. ; cited, 660. Lewis, Professor H. Carvill, 412, 609, 518, 519, 624; begins the survey of Pennsylvania, 121, 134, 187 ; discov- ers moraine on Pocono Mountain, 133; death of, 135; on tnmsporta- tion of bowlders, 169, 217 et seq ; on depth of ice in eastern Pennsylvania, 169 ; on erosion in northern Penn- sylvania, 230; reports drumlins in Pennsylvania, 257 ; the Philadelphia red gravel and brick-clay of, 410, 525, 553 ; on marginal kames in east- ern Pennsylvania, 312 ; on glaciated area of Great Britain, 896, 397 et seq. ; on kettle-hole at Freehold, Pa., 488. Lewiston, Me., 303. Licking county, Ohio, 141, 468, 493. Licking River, Ky., 327, 346-347. Licking River, Ohio, 270. Lindenkohl, A., 578. Little Falls, Minn., 509, 537 et seq. Little Falls, N. Y., 289. Little Miami River, 287, 458, 530, 632. Little Sandy River, 19. Lituya, 38, 39. Llama, 386, 387. Loess, 144, 226; chapter on, 359-371; Pumpelly on, 369, 360 ; composition of, 369, 364; in China, 360, 689; Baron Richthofen on, 360; in Eu- rope, 362, 588; in America, 362; Hilgard on, 362, 363 ; altitude of de- posits of, 362, 365, 368; Chamber- lin on, 364 ; theory of, 366 et seq., 480 ; extent of subsidence indicated by, 366, 368, 411; Upham on, 367; earthworms and the, 370 ; in Ohio, 632, Long Island, terminal moraine on, 126, 130, 176, 179; bowlders on, 209; over-wash gravel on, 309; kettle- holes on, 320. Long Level, W. Va., 341, 494. Long, Mr , on Paint Creek, 468. Lorain county, Ohio, 361, 469. Lorange, on cirques, 244, 246. Louisville, Ky., 142, 270. Loveland, Ohio, palaeolithic implements found at, 458, 530, 532. Low, Mr. A. P., 597. Lowell, Mass., 808-305. Lubbock, Sir John, 514. Luzerne county, Pa., 136. Lycoming county. Pa., kettle-holes in, 187; terminal moraine in, 187; bowlders in, 212. Lycoming Creek, 137, 286. Lyell, Sir Charles, 673, 585 ; observa- tions of, in Nova Scotia, 112, 113; on Richmond train of bowlders, 209, 210 ; theory of the cause of the Gla- cial period, 408; principle of uni- INDEX. 613 formity, 449 ; on rate of recession of Niagara Falls, 4D5, 568 ; on an- tiquity of man, 506, 514. Lynn Canal, Alaska, 27, 39, 55, 57, 160. Maohias River, 302, 303. Mackenzie Rirer, 216. Madisonville, Ohio, paloeolithic imple- ments found at, 458, 530, 532. Magdalena Bay, 95. Maine, 123, 391, 549 ; depth of ice in, 166, 167; bowlders in, 217; depth of drift in, 229 ; absence of drum- lins of large size in, 256 ; kames in, 309-311, 313; lakes in, 321, 357; submergence in, 413 ; marine beds overlying glacial drift, 580. Mammoth, 388, 560. Mammoth Cave, 199. Man, antiquity of, in America, 388, 450, 527, 528; and the Glacial pe- riod, chapter on, 506-570. Manchuria, 377, 382. Mandan, 280. Manitoba, 314, 357, 295. Manomet Hill, 126. Marginal drainage, 147, 293 et seg., 311, 312. Marietta, Ohio, 283, 287. Marine shells, elevation of, in glacial deposits in Wales, 401, 412 ; near Boston, 401, 412. Marr, Mr. J. E., on glacial dams, 325; on erosion in Greenland, 226. Marshall, Ohio, 492. Martha's Vineyard, 123, 130, 175, 207, 501. Mary Minturn River, 84. Massachusetts, terminal moraines of, 123, 125; depth of ice in, 167; depth of drift in, 229 ; transported bowlders in, 207-209 ; drumlins, in, 253-255, 258; kames in, 290, 301, 313 ; kettle-holes in, 320, 321, 475 ; lakes in, 321. Mastodons, 386, 387, 527, 533, 557, 660. Mattawamkeag, Me., 303. Mattmark Sea, 3 2 5 . Maumee River, terminal moraine in valley of, 179 ec seg., 236. McConnell, Mr. R. G., on the Cypress Hills, 171, 172; on depth of drift in British America, 229. McGee, Mr. W J, 538 ; on subsidence on Atlantic coast, 136, 411, 412; on refrigeration under the ice-sheet, 415; Columbia period of, 501, 526, 553, discovers implements, 558. MoKeesport, Pa., 339. Medial moraines, east arm of Kettle Range, a, 183; Coteau des Prairies, a, 186. Medial moraines, formation of, 9 ; of Muir Glacier, 39, 45, 206 ; of White River Glacier, 22. Mediterranean Sea, 383. Medlicott and Blandford on signs of glaciation in India, 434, 435, 686. Medora, Ind., paleolithic implements found at, 633 et seg. Megalonyx, 386. Megatherium, 386. Melville Bay, 84. Mer de Glace of Switzerland, 3, 4, 10 ; of Greenland, 79, 81, 82. Merjelen Sea, 324, 414. Merrimack River, 290, 291, 303, 304, 310, 644. Mesozoic era, absence of glacial epochs during the, 591, 594. Metuchen, N. J., 128. Metz, Dr. C. L., iinds palaeolithic imple- ments at Loveland and Madisonville, Ohio, 468, 509, 412, 530, 532, 533. Michigan, buried channel in, 276 ; salt deposits of, 408. Middle Bass Island, 232. Migrations of plants and animals, 372 et seg., 649, 682, 589. Milan, Ind.,. 492. Milk River, 294. Mill Creek, 270, 271, 530, 532. Minneapolis, 183, 184, 273, 281, 321, 468, 460, 539, 644. Minnesota, drif tless area of, 120 ; depth of ice in, 173 ; terminal moraines in, 183-186, 441, 501, 539, 645 et seq.; varieties of drift in, 186 ; depth of drift in, 228, 461 ; drumlins absent in, 257; waterfalls in, 273, 458 ci seg.; delta terraces in, 294 ; occasional kames in, 313; lakes of, 321, 367; loess in, 368 ; preglaeial man in, 388, 537 et seg. ; post-glacial erosion in, 458 et seg. ; buried channel in, 459 et seq. ; wells in, 461, 495, 496 ; vegetable deposits in, 495 ; palaeo- lithic discoveries in, 609, 537 et seg. Minnesota River, 186, 281, 282, 295, 314, 321, 322, 356, 460-462, 464, 465. Mississippi River, 122, 143, 175, 314, 317, 318, 321, 349, 369, 443, 523, 639, 614 TEE ICE AGE IN NORTE AMERICA. 540, 549; preglaoial erosion of, 19S, 268, 281, 410 ; loess of the valley of, 226, 362 et seq., 367 ; glacial drain- age of, 279, 280; terraces on, 281, 367, 464, 538 ; post-glacial erosion in, 281, 458 et seq., 549; subsidence in the valley of, 411, 414 ; rate of transportation by the, 469, 470. Missouri ooteau in British America, 121, 148, 189 et seq. ; source of, 214. Missouri ooteau in Dakota, 187, 189, 292, 293, 295. Missouri, glacial boundary in, 144, 367 ; bowlders in, 213. Missouri River, 144-147, 187, 192, 213, 214, 295, 366-368. Mohawk River, 167, 198, 273, 275, 280, 312, 313, 317, 318, 351; ice- dam in, 354, 358, 462, 455. Mohegan Rock, Conn., 208. Mono county, Cal., 13, 194. Monongahela River, 270, 330, 332, 337, 338, 344, 494. Monroe county, Pa., terminal moraine in, 132, 134; depth of ice in, 167, 170; buried channels in, 277. Montague, Canada, 413. Montana, former glaciers in, 13, 148 ; glacial boundary in, 119, 147 ; bowl- ders in, 213. Mont Blanc, 92, 97. Montgomery county, Ohio, organic re- mains in glacial deposits in, 482, 483 ; depth of till in, 482. Montreal, marine beds overlying gla- cial drift, 580. Montreal Mountain, 413. Moose, 657. Moraines, formation of, described, 9. Moraines of retrocession, 53, 179, 194, 195. Moreau River, 146, 147. Morgan county, Ind., 491. Morgantown, Ind., 491. Morgantown, W. Va., 330, 337, 338, 493. Morris county, N. J., 211. Morrison county, Minn., 537, 539. Moulins, defined, 9 ; on Mount Tacoma, 22 ; on Muir Glacier, 44. Mountain-building, probably correlated with epochs of glaciation, 575, 578, 583, 685 et seq. Mount Baker, 23, 152, 164. Mount Crillon, 38, 39, 64. Mount Dana, 15. Mount Desert, Me., 123, 166, 167. Mount Fairweather, 31, 38, 39, 64. Mount Hood, 19, 164, Mount Jefferson, 19. Mount Katahdin, Me., 167, 217, 308, 391. Mount Kelertingouit, Greenland, 77, 78. Mount Lyell, 15, 239, 243. Mount Rainier. See Mount Tacoma. Mount Shasta, 15 et seq., 164. Mount St. Elias, 30-32, 38. Mount Tacoma, 21, 122, 152, 164, 165. Mount Washington, 166, 216, 389-391. Mount Wrangel, 32. Mower county, Minn., 495. Mud River, W. Va., 341. Muir Glacier, cross-section of, 4, 35; kames forming in, 11, 62, 302 ; dis- covery of, 35 ; chapter on, 36-66 ; accessible to tourists, 36 ; difBoul- ties of measurement, 37, 48; sur- rounding mountains, 37-39 ; vegeta- tion in the vicinity of, 37, 53 ; view of, 38 ; former extension of, 39, 55, 161 ; relation to Glacier Bay, 41 ; width of, 41 ; terminal moraine of, 41, 53 ; transported bowlders on, 41, 206 ; thickness of, 43, 204 ; slope of, 43 ; extent of, 43 ; moulins of, 44 ; nunataks of, 43, 71, 161; medial moraines of, 39, 45, 52, 206 ; ice-pil- lars of, 45 ; surface of, 46 ; front of, 46 ; icebergs from, 46, 47, 63 ; sub- glacial streams of, 47, 51, 53, 57, 62 et seq , 302 ; buried forest near the, 49, 57 et seq., 203, 482 ; velocity of, 51, 72; daily discharge of, 51, 74 ; view of surface of, 52 ; kettle- holes in front of, 53, 54, 62, 130, 176; retreat of, 53; former extent of, 55 ; Vancouver on, 56 ; view of stumps of cedar-trees, 58, 59 ; move- ment over gravel deposits, 62, 203 ; sediment of, 64 ; temperature in front of, 66 ; striae of, 699. Muir Inlet, 37, 41, 161, 164; list of plants about, 65. Muir, Professor, explorations in Alaska, 35, 64. Muskingum River, 270, 280, 283, 287. Musk-ox, 657. Mylodon, 386. Naaman's Creek, 557. Nansen, Dr. F., on east coast of Green- land, 90. Nantiooke, Pa., 292. Nantucket, 123, 125, 175. INDEX. 615 Natal, Permian glacial deposits in, 592, Naushon, 176 et seq. Nebraska, glacial boundary in, 144 ; bowlders in, 213 ; loess in, 363, 366, 367, 480, 558. Nelson River, 279. Nevada, former glaciers in, 148; salt lakes in, 496. Neve field, in Greenland, 80; on the Sierra Nevada, 18, 151, 243. New Alexandria, Ohio, 140. Newark, Ohio, 141, 286. Newberry, J. S., 437 ; on glaciers of Cascade Range, 19 ; on depth of gla- cial deposits, 227 ; on glacial erosion, 231 ; on formation of the Great Lakes, 232, 316 et seq. ; on buried channels in Ohio, 270, 271 ; on pre- glacial drainage through Ontario, 275. New Brunswick, 596. New Burlington, Ohio, 492. Newburyport, Mass., 303. Newcomb, Professor Simon, 164, 426. New England, 125, 318, 354, 437, 441 ; glacial drainage of, 122 ; kettle-holes in, 125, 320, 321 ; depth of ice in, 166, 167, 173 ; terminal moraines of, 120, 175, 321 ; transported bowlders iu, 207, 208 ; depth of drift in, 229 ; glacial drainage in, 290, 291, 302, 303 ; kames in, 290, 303 el seq. ; lakes in, 321, 323, 357; peat-bogs in, 320. New Hampshire, 125, 549; depth of ice in, 166 ; transported bowlders in, 166, 207, 216; terminal moraines in, 193 ; depth of till in, 229 ; drumlins in, 253, 255, 258 ; delta terraces in, 290; kames in, 299, 300; lakes in, 321, 323. New Harmony, Ind., 142, 286. New Jersey, terminal moraine in, 119, 120, 126, 128, 130, 178; fringe in, 136; depth of ice in, 170; trans- ported bowlders in, 211 ; lakes in, 323 ; interglacial man in, 388 ; pale- olithic discoveries in, 509, 510, 515 et seq. New Mexico, 244. New Richmond, Ohio, 327. New York, 317, 354, 549; glacial boundary in, 119, 441 ; depth of ice in, 167, 231; terminal moraine in, 137, 178; bowlders in, 212; drift- hills in, 256 ; buried channels in, 275, 277 ; kames in, 312," 313 ; lakes in, 323, 357; lake ridges in, 351; salt deposits in, 407. New York Central Railroad, terraces along the, 289. New York city, 317, 354, '410, 678. New Zealand, glaciers in, 92, 100, 438, 586 ; cirques in, 244. Niagara gorge, length of, 269, 313, 336, 452 ; a glacial chronometer, 269, 452 et seq. Niles, Professor, on erosion by the Aletsch Glacier, 201, 202. Niobrara River, 144. Nisqually River, 21. Nittany Valley, Pa., 199, 200. No Man's Land, 123. Nordenskiold, A. E., 694, on influence of meteoric dust, 9 ; expedition to Greenland, 67 et seq. ; on subsidence of Sahara, 383. Northampton county. Pa., terminal mo- raine in, 132. North Bass Island, 232. North Bend, Ohio, 142, 327. North Carson Lake, Nevada, 496. Northern Pacific Railroad crosses the glacial boundary, 146, 147, 186. North Sea Glacier, 394, 397. Norway, glaciers of, 92, 94, 98, 202, 203, 393 ; erosion in, 248 ; depth of ice in, 395. Nova Scotia, 112, 113, 696; absence of Quaternary marine beds in, 580. Nova Zembla, 95, 438. Nunataks, 43, 71, 161, 224, 326. Oahe, Dakota, 146. Oberlin, Ohio, 469. Ocean lowered to supply ice, 579 ; raised by ice attractiba, 579. Ohio, fringe in, 139 ; terminal moraine in, 140 et seq., 179, 332 ; kettle-holes in, 140, 141 ; wells in, 140; 141, 227, 228, 492; bowlders in, 212, 213; depth of till in, 227, 228,1482, 483, 488-490 ; preglacial erosion in, 269 et seq. ; buried channels in, 270 et seq., 274, 532 ; glacial drainage in, 280 ; absence of terrace^ in ungla- ciated part, 283; kame^ in, 313; limit of glacier in, 317, 318; glacial erosion in, 318 ; glacial boundary in, 327, 441, 442 ; lake ridges in, 361 et seq., 356 ; interglacial man in, 388, 528 et seq. ; post-glacial erosion in, 466 ; striae in, 481 ; (Vegetable de- posits in, 482, 483, 488 et seq., 492,. 616 THE IGE AGE IN NORTH AMEEICA. 493; loess in, 532; palaeolithic im- plementa in, 630 et seq. Ohio River, 142, 180, 198, 313, 318, 523, 530; glacial drainage of, 280, 313 ; ocoupioa a preglacial gorge, 269 et seq., 344; glacial dam in, 326 el seq., 358 ; terraces on, 330 et seq. Oil Creek, Pa., 286. Olympian Mountains, 152, 164. Ontario, Canada, 413,,549. Oregon, 194, 369, 316 ; glaciers in, 19, 149, 163, 164, 496. Orton, Professor Edward, on depth of till, 227 ; on buried channels in Ohio, 272 ; on Paint Creek, 466 ; on vege- tal deposits in Ohio, 482, 492. Oscillations of land and sea, RIS et seq. Ossipee Lake, 311. Over-wash gravel, 176, 308 et seq. See Delta Tskraces and Kames. Oxford, Ohio, 490. Packard, Prof. A. S., 580; reports trans- ported bowlders on Mount Katahdin, Me., 217. Paint Creek, 286, 332 et seq., 344; 466 et seq. Palaeoliths, characteristics of, 506 el seq. ; Haynes on, 509 et seq. ; occur- rence of, in glacial gravels in France and England, 513-515 ; at Trenton, N. J., 515 et seq.; at MadlSonville and Loveland, Ohio, 530 et seq. ; at Medora, Ind., 533 et seq. ; at Little Falls, Minn., 537 et seq. ; at Clay- mont, Del., 661, «< seq. Palestine, Ohio, 140. Panama, Isthmus of, Quaternary oscil- lations of, 684. Paris Crossing, Ind., 492. Parker Creek, 16. Parker, Pa., 494. Passadumkeag, Me., 303. Patagonia, 92, 438. Patterson Glacier, 27. Patton, Rev. J. L., 36. Payer, Mr., on east coast of Greenland, 90. Peabody Bay, 84. Peat-beds, formation of, 320, 484 et seq. ; age of, 473, 474, 487, 488 ; in southern Ohio, 482, 483; in Penn- sylvania, 488 ; in Minnesota, 495. Peekopee Lake, 294. Pelee Island, 232. Pelican Lake, 294. Pembina River, 293, 294. Penck, Dr. A., on cirques, 245, 248. Pendleton county, Ky., 142. Pennsylvania, 386, 437 ; terminal mo- raines in, 119, 132 et seq., 178, 336, 442; kettle -holes in, 134, 488; fringe in, 135, 136, 139 ; transported bowlders in, 136, 211, 212; terraces in, 138, 33'J-332, 335, 339, 344; depth of ice in, 167 et seq. ; direction of striae in, 218, 219; depth of till in, 229 ; glacial erosion in, 230 ; wells in, 230, 494; few drumlins in, 257 ; buried channels in, 276, 277, 292; kames in, 308, 312; marginal drainage in, 312; vegetable deposits in, 494. Penobscot River, 302, 303. Percival on drumlins, 255. Permian glacial epochs, 434, 435, 686, 591, 594. Perry county, Ohio, 141, 286. Perth Amboy, N. J., 126. Perthes, Boucher de, palaeolithic dis- coveries of, in France, 474, 613, 614. Peru, post-glacial elevation of, 683. Petermann Peak, 90. Philadelphia, 653 ; and vicinity, pre- glacial elevation of, 578. Philadelphia red gravel and brick-clay, 411, 522 et seq., 563, 656. Pickaway county, Ohio, 141. Pike county, Ohio, 332, et seq., 334. Pike county, Pa., 177. Pilgrim Fathers, 123. Pipestone county, Minn., 313. Piscataqua River, 302, 312. Pittsburg, Pa., 330, 335, 344, 358. Plainfield, N. J., 126, 128. Plains, the Western, uplifted in the Quaternary period, 678. Plants, migrations of, during the Gla- cial period, chapter on, 372-391 ; preglacial distribution of, 372, 380 ; Prof. Asa Gray on, 372 et seq. ; re- semblance of, in Japan and eastern North America, 373, 382 ; manner of migration of, 374 ; of the Pacific coast of North America, 876, 386 ; of Japan and north China, 377, 384 ; of Europe, 377 ; of the Tertiary pe- riod, 378, 380 ; effect of Glacial pe- riod upon, 379, 381, 383 ; the birth- place of, 382; on oceanic islands, 386. Platte River, 144, 149, 366, 868. Plum Creek, Ohio, 469, 470, 475. Plum Island, 126. INDEX. 617 Plymouth county, Mass., 123, 125, 208, 320. Plymouth Rock, Mass., 209. Pocono Mountain, terminal moraine on, 119, 132, 134, 442 ; depth of ice over, 170; bowlders on, 211. Pohlman, Dr., on the Niagara gorge, 456. Point Pleasant, Ohio, 327. Point Wilson, Washington Territory, 154.. Polar ice-cap, 151, 162. Pomp's Pond, Aiidover, Mass., 472 et seq. Port Foulke, 81, 84. Port Jervis, 280. Portland, Me., 309. Portland, Ore.. 21, 35, 36, 121. Portland, Pa., '312. Port Townsend, 152, 154. Post-glacial erosion in gorge of Niag- ara, 269, 452 et seq. ; of Mississippi, 281, 458 et seq. ; of Paint Creek, 467, 468 ; in Raccoon Creek, Ohio, 468, 469 ; in Plum Creek, Ohio, 469,470 ; in Lake Michigan, 471. Post-glacial osoUlations of level, 580- 682. Pot-holes in Grafton county, N. H., 291 ; in Lackawanna county. Pa., 292. Potomac River, 411, 412, 553. Potter county. Pa., terminal moraine in, 137 ; depth of ice in, 167. Powell, Major J. W., 574. Pratt, Archdeacon J. H., 573. Precession of equinoxes, 417, 418, 585, 587. Preglacial climate, 407. Preglacial drainage, chapter on, 268- 278; lower levels of, in the valley of the Ohio, 270 et seq. ; erosive ef- fect of, 270 ; buried channels of, 271 et seq. ; in the vicinity of Cincin- nati, 271 ; in central Ohio, 272 ; in the vicinity of Minneapolis, 273 ; be- tween Lakes Erie and Ontario, 273 ; in southern tributaries to Lake Erie, 274 ; between Lake Huron and Lake Ontario, 274 ; in the Mohawk Valley, 275 ; in Illinois, 276 ; in western Pennsylvania and western New York, 276, 277 ; in eastern Pennsylvania, 277 ; in the province of Quebec, 277. Preglacial elevation, 410, 414, 572, 578. Preglacial erosion, extent of, 196 et seq., 231, 249, 268 et seq., 313, 456, 458. Preglacial terraces, 332. Preglacial weathering of strata, 583. Prehistoric wooden structures of Naa- man's Creek, 557. Prestwich, Professor Joseph, 514, 574, 587, 590, 593 ; on length of Glacial period, 602 et seq. Prince WiUiam Sound, 32. Proviuoetown, Mass., 123. Puget Sound, 122, 148, 152, 164, 165. Pumpelly, Professor Raphael, 589, 593 ; on preglacial erosion, 231 ; on loess, 359 et seq. Putnam^county, W. Va., 339, 494. Putnam, Professor F. W., 530 ; on his discoveries of palaeolithic imple- ments, 517 ; on antiquity of man in California, 666, 666. Puyallup River, 21. Pyramid Lake, Nev., 496. Pyrenees, 379, 383, 386, 397 ; ancient glaciers in, 588. Qu'Appelle River, 293, 296. Quebec, 596. Queechee, Vt., 217. Queen Charlotte Islands, 582. QueenstOn, 452 et seq. Raccoon Creek, Ohio, 284, 468, 476, 493. Ramsay, Sir A. C, theory of rock- basins, 239 ; on early glaciation, 433 et seq. Randolph, N. Y., 137. Rangeley Lakes, 302, 309. Rappahannock River, 412. Ravenna, Ohio, 313. Read, Mr. M. C., on buried channels in Ohio, 271. Recession, rate of, of Falls of Niagara, 462 et seq. ; Falls of St. Anthony, 458 et seq. ; falls in Black River, Ohio, 466 ; of tributary to Paint Creek, 467, 468. Red Bank, Ohio, 532. Red River of the North, 171, 192, 228, 282, 293, 294, 314, 321, 366, 496, 538. Reindeer, 529, 557. Rensselaer Harbor, 81. Revillo Gigedo Island, 160. Rhine River, 362, 364, 395. Rhinoceros, 560. 618 THE ICE AGE IN NORTH AMERIGA. Rhode Island, terminal moraine in, 126, 176 ; transported bowlders in, 209 ; depth of drift in, 229. Rhone Glacier, 22o. Richardson, Mr. James, 159, 160. Richardson, Sir J., on transported bowlders in the Mackenzie Valley, 216. Richmond, Mass., train of bowlders in, 209-211. Richthofen, Baron, 589, 593 ; on loess, 360, 361. Ricketts, Dr. Charles, 584. Riggs, Rev. Thomas L., lil. Rink, H., on results of Greenland ex- ploration, 69 et seq. Ripley county, Ind., 492. Ripley, Ohio, 142, 327, 442. River that turns, 295, Rivifere des Lacs, 294. Roanoke Rirer, 412. Roches moutonn^es, Y6, 155, 157, 159, 197, 201, 265, 433. Rock-basins in the Sierra, 237 et seq., 315. Rock-scoring in Glacier Bay, Alaska, 40, 599 ; described, 109 et seq. ; near Victoria, Vancouver Island, 165 et seq. ; on islands in Lake Erie, 232 et seq. ; on the Sierra, 237 et seq. ; in South Africa, 435. Rocky Mountains, 13, 172, 214, 243, 367-369, 379, 391, 406, 496 ; an- cient glaciers in, 147 - 149, 588 ; Quaternary uplifts of, 578, 583, 887, 588. Rocky River, 274. Boss county, Ohio, 332; terminal mo- raine in, 141, 142 ; transported bowl- ders in, 141, 212 ; vegetable depos- its in, 492. Ross, Sir J. C, explorations of, in Ant- arctic Continent, 100 et seq. Russell, Mr. I. C, on glaciers of the Sierra Nevada, 151 ; on moraines of Mono county, Cal., 194 ; on glacial erosion, 237; on Yosemite Valley, 241 ; on cirques, 241 et seq. ; on Lakes Lahontan and Bonneville, 497, 499 ; on the gravel deposits about the Dead Sea, 500. Sabine, E., explores east coast of Green- land, 90. Saco River, 193, 302, 311. Saguenay, fiord of the, 577. Sahara, Desert of, 383, 409. Salisbury, Professor R. D., 577 ; on loess, 144 ; disoaveries in northern Germany, 400, 402. Sandusky, Ohio, 232, 350. Sandy River, 19. Saskatchewan River, 171, 192, 214, 229, 293 et seq., 442. Sawkins, J. G., 583. Scandinavia, glaciers of, 92, 94, 393, 394 ; erosion in, 230 ; maximum post-glacial uplift, 582. Schenectady, N. Y., 289. Schoodic Lake, 310. Schoolcraft on the antiquity of man, 566. Schuylkill Water-Gap, 169. Schwatka, F., explorations in Alaska by, 29, 30, 161. Scioto River, 141, 270, 280, 283, 286, 287, 335, 467. Scoresby, explores east coast of Green- land, 90. Scotland, 437, 450 ; glacial erosion in, 248; kames in, 297; glaciation in, 393, 394, 433, 434 ; post-glacial ero- sion in, 470 ; maximum post-glacial uplift, 582. Scratched stones, by moving ice. 111 ei seq. ; by land-slides, 113. Scudder, Mr. Samuel, on the existence of Alpine insects on the White Mountains, 388, 389. Sea cow, 386. Sea-level, changes of the, 679. Seattle, Washington Territory, 21, 121, 152-154. Second Glacial epoch, 179, 195, 228, 464, 475, 479, 491, 551. Seaman, G. B., visits Kotzebue Sound, 33. Serrate character of glacial margin, 118, 441 e< seq. Seville, Ohio, 313. Seymour, Ind., 491. 1/Shaler, Professor N. S., 574-576 ; on transported bowlders in Rhode Isl- and, 209 ; on depth of the drift in New England, 229 ; on drumlins, 258, 260 ; on glacial action in the Con- necticut Valley, 436; in Appala- chian region, 437 ; on length of the Glacial period, 501 ; on the Trenton gravel, 520. Shawshin River, 298. Sheep, 386. Sheyenne River, 293, 294. Shingle Creek, 459, 460. INDEX. 619 Ship-canal, possible, between Lakes Su- perior and Michigan, 276. Ship Rock, Mass., 207. Siberia, 426, 428, 429. Sierra Nevada Mountains, 361, 379, 496, 499 ; glaciers in, 13 et seq., 149, 151, 152, 243; rock-basins in, 237 et seq. ; cirques in, 242 et seq. ; Quater- nary uplifts of the, 577, 578, 583. Sinian Mountain system, 593. Sink-holes, frequent in limestone re- gions, 129. Sitka, Alaska, 64, 161 Skeena Biver, 25. Slickenside, 115. Smith Sound, 78, 84, 215. Smock, Professor J. C, on depth of ice in New Jersey and Pennsylvania, 170. Somme River, 474, 513, 514. Souris River, 294. Southall, J. C, 584; on antiquity of man In California, 564, 565. South America, 387, 428 ; glaciers of, 92, 96 et seq., 406 ; relation of con- tour of, to the Gulf Stream, 422, 423. South Bass Island, 232, 236. South Brewster, Mass., 125. Spencer, Professor J. W., on the erosive power of glaciers, 202, 203 ; on pre- glacial drainage, 272, 274 ; on de- pression in the upper St. Lawrence, 413, 414. Spitzbergen, 92, 95, 374, 380, 407. Split Rock, Ky., 345, 346. Springdeld, Ohio, 272. Squier, Mr., 345. Stainmoor Glacier, 397-399. Stark county, Ohio, 141, 286. Staten Island, 126. St. Davids, Ontario, 453 et seq. Steenstrup, K. J. V., on rate of move- ment of Greenland glaciers, 72. Stickeen River, 25, 27, 122, 198. St. Joseph River, Ohio, 180, 181. St. Lawrence River, 275, 280, 314, 317- 319, 354, 413, 697. St. Louis, Mo., 120, 143, 281, 481. St. Marie's Biver, 180, 181. Stockbridge, Mass., train of bowlders in, 209, 210. Stone, Professor George H., on moraines in Maine, 193 ; on depth of drift in Maine, 229 ; on drumlins in Maine, 256 ; on kames in Schoodio Lake, Me., 310, 311. St. Paris, Ohio, 227, 228, 272. St. Paul, Minn., 281, 459, 460, 539, 544. Strait of Georgia, 158, 160, 164. Strait of Juan de Fuca, 154, 164. Straits of Magellan, 99. Striae, direction of, on the west coast of North America, 155, 158 et seq., 599 ; in Maine, 1 67 ; on Hudson Strait, 216; in western Pennsylvania, 218, 219; in Lake Erie, 235, 236; in Scandinavia, 394; in India, 434; in Ohio and Indiana, 481 ; in Illinois, 481, 482; in Canada, 596 et seq. Stroudsburg, Pa., 132, 312, 625. Subglacial streams, source of, 74 ; of the Alps, 18, 225 ; on Mount Shasta, 18 ; Mount Hood, 19 ; Mount Ta- coma, 21 ; milky-white color of, 18, 21, 121, 122, 224; of Muir Glacier, 47, 51, 68, 63, 64, 302 ; of Greenland, 74, 225, 226 ; transporting power of, 63, 64, 74, 121, 122, 224 et seq. Submergence during the Glacial epoch in North America, 172, 365, 411 ; in Wales, 401, 412. Subsidence on the Atlantic south of the 'glacial limit, 135, 411, 412, 414; in Rocky Mountain region, 172 ; theory of, to account for lake ridges, 363, 354 ; necessary to account for loess, 365; theory of, 365, 368; due to weight of ice-sheet, 368, 369, 673, 576, 586, 595 ; of Sahara and cen- tral Asia, 383 ; in the Mississippi Valley, 411 ; on the Atlantic, north of glacial limit, 412, 413, 624; in Vermont, 413, 624 ; in Dominion of Canada, 413, 624 ; general, of the glaciated area, 481 ; in the Delaware Valley, 524-627. Sullivan county. Pa., 167. Superglacial streams, 22, 44, 67, 82. Susquehanna River, 119, 122, 134-136, 170, 198, 211, 285, 312, 411, 412. Sverdrup, Lieutenant, on east coast of Greenland, 90. Sweden, 92, 297. Switzerland, 407, 409 ; erosion in, 201, 202, 244 ; glaciers of, 393, 396. Syracuse, N. Y., 413. Table Mountain, 562. Tacoma, 122. Taku Inlet, 27. Tapir, 386, 387, 560. Teazes Valley, 339 et seq., 494. Temperature of the southern hemi- 620 THE ICE AGE IR NORTH AMERTOA. sphere lower than that of the north- ern, 425 ; of the interior of South America, 428 ; of northeast Siberia, 428 ; not altogether dependent on the heat received from the sun, 427 ; variations of, in different latitudes not understood, 429; causes affect- ing, 430. Tennessee, conglomerates in, 437. Terminal moraines, 221 ; conditions favoring the formation of, 121, 175, 195 ; characteristics of, 128 ; on Nantucket and Martha's Vineyard, 123, 175 ; on the Elizabeth Islands, 123, 175 ; in Plymouth county, Mass., 125; in Long Island, 126, 176; in New Jersey, 128, 178 ; in eastern Pennsylvania, 130 et seq., 178; on Pocono Mountain, 132 ; in valley of Fishing Creek, 134 ; in valley of Conewango Creek, 138 ; in Ohio, 140 et seq., 179 et seq. ; in Dakota, 146 ; in Washington Territory, 162 ; in Wisconsin, 183 ; south of Maine, 167 ; south of New England, 175 ; in central New York, 178 ; in Arn- monoosuc Valley, 193 ; in Great Britain, 396 et seq. ; in Germany, 395, 402 ; compared with those of America, 402 et seq. ; lobate contour of, 118, 179, 194, 444, 446, 545. Terraces, glacial, on the Conewango, Ohio, 138 ; on the Mississippi, 281, 367, 463, 538 ; on the Ohio and its tributaries, 283, 330 et seq. ; in west- ern Pennsylvania, 331 ; on Brush Creek, Ohio, 332 ; at Bellevue, Pa., 335, 344 ; near Morgantown, W. Va., 337 et seq., 493 ; at Clarksburg and Weston, W. Va., 338; at McKees- port. Pa., 339; in Teazes Valley, W. Va., 389 et seq., 494 ; on the Big Sandy, W. Va., 342; and Guyan- dotte, 342 ; on Lake Agassiz, 356, 357. 539, 548, 576 ; on the Minne- sota, 464 ; on the White, Ind., 491 ; on Raccoon Creek, Ohio, 493 ; on the Somme, France, 513, 614; in south- ern England, 514, 515 ; on the Dela- ware, 521 et seq.; on the Little Miami, 532, 683. Tertiary era, absence of glacial epochs during the, 591, 594. Thames River, 395, 397. Thibet, Quaternary uplift of, 586, 589. Thompson, Mr. Gilbert, on the glaciers of Mount Shasta, 15. Thomson, Sir William, 573. Thornville, Ohio, 141. Three Sisters, 19. Tidioute, Pa., 276. Tierra del Fuego, 97, 98. Till, defined, 109, 115 et seq.; depth of, 227 et seq. ; in British Columbia, 157 ; in Minnesota, 185, 228, 461, 495, 496 ; in Canada, 228 ; in Indi- ana, 227, 491, 492 ; in Illinois, 227 in Ohio, 227, 228, 482, 483, 489, 490 ; in New England, 229; in Penn- sylvania, 229, 230, 494 ; in Europe, 230 ; average in North America, 250. Tioga county. Pa., 167. Titusville, Pa., 286. Todd, Professor J. E., on glacial dams across the Kansas and Platte Rivers, 144 ; on Coteau des Prairies, 186 ; Missouri coteau, 187, 189 ; on glacial drainage, 292. Tonawanda Creek, 458. Topeka, Kan., 144. Torsukatak Glacier, 70. Towson, Mr., 105, 106. Trade-winds, motions of, in the Atlan- tic, 420 ; relations to the Gulf Stream, 422 ; cause of the predominance of the southeast, 423 et seq. Transportation by subglaoial streams, 121, 122, 224 ei seq. ; in Alaska, 63, 64, 122 ; in Greenland, 74, 226 ; in the Alps, 224, 225. Transportation of bowlders, 206 et seq. ; in Alaska, 41 ; direction of, 168, 208 et seq. ; in Europe, 206 ; in New Eng- land, 207 et seq. ; in New Jersey, 211; Pennsylvania, 211 ; Ohio, 212; Indiana and Illinois, 213 ; beyond the Missouri River, 213 et seq. ; in British America, 148, 214 et seq. ; on northeastern Mackenzie River, 215 ; upward, 216 et seq. ; south of glacial limit on the Atlantic coast, 41 1, 41 2 ; in Scotland, 434 ; in Africa, 435 ; in India, 435. Transportation of till, 227 et seq. ; in Ohio, 227 ; in the Red River region, 228 ; in New England, 229 ; in Penn- sylvania, 229 ; in Europe, 230. Transporting power of rivers, 525, 526. Trenton, N. J., 289 ; palaeolithic imple- ments at, 457, 609, 510, 615 ei seq. ; 533, 563. Trenton gravel, 520-522, 526 et seq., 553. INDEX. 621 Tuckemuclc Island, 123. Tundra in Alaska, 32, 33. Tuolumne county, Cal., 13, 152, 559 et seq. Tuscarawas Eiver, Ohio, 2*71, 286. Two glacial epochs, 475 et seq., 496 et seq., 551. Tyndall, Professor John, measurements of glaciers, 2, 3, 31, 73. Tyndall Glacier, 79. Tyrol Mountains, 92, 325. Uintah Mountains, 150, 151. Unalaska, Island of, 32. Unlformitarianism, 448, 449. Upham, Mr. Wavren, 573, 580 ; discov- ers terminal moraine, 120; traced moraine through Iowa and eastern Dakota, 183, 184 ; on depth of drift, 228, 229; on drumlins, 252, 253, 267, 259, 262; on Brown's Valley, 282 ; on glacial drainage in the Northwest, 290 et seq. ; in British America, 296 ; on kames in the Con- necticut Valley, 306 ; on the grayel plains north of Portland, Me., 309 ; on kame-like ridges in Minnesota, 313; on lakes of Minnesota, 321, 322 ; on Lake Agassiz, 356, 357 ; on loess, 367, 368 ; on Lewis's work in England, 396, 400; on shells near Boston, Mass., 401, 412; on age of lake-basins, 471 ; on vegetable de- posits in northern Minnesota, 496 ; on palaeolithic implements, 638 et seq. ; on probable causes of glacia- tion, 673 et seq. Ural Mountains, 96, 395. Utah, former glaciers in, 148, 149. Vancouver on glaciers of Alaska, 22 el seq., 30 ; on Glacier Bay, 55 et seq. Vancouver Island, 156, 157 et seq., 164, 581. Vegetable remains in glacial deposits, near Muir Glacier, 57 et seq., 203, 482; at Point Wilson, 154, 166 ; at Morgantown, W. Va., 337, 338, 344, 493 ; in Teazes Valley, 341, 494 ; the- ory of, 476, 479; in Montgomery county, Ohio, 482, 483 ; in Butler county, Ohio, 482, 488, et seq. ; in Morgan, Jackson, Jennings, and Jef- ferson counties, Ind., 491, 492 ; in Hamilton, Highland, Ross, and Lick- ing counties, Ohio, 492, 493 ; in south- western Indiana, 493 ; in Pennsyl- vania, 494 ; in Minnesota, 495. Vegetation in the vicinity of Muir Gla- cier, 37, 65. Veins in glacial ice, 7. Vermont, 166, 217, 413, 624, 549, 598. Vessel Rock, N. H., 206. Victoria, glacial grooves near, 155, 156. Viscosity of ice, 80, 88. Volcanic outflows probably correlated with glaciation, 577. Volcanoes indicating molten interior of the earth, 575. Waagen, Dr. W., 592. Wabash River, 142, 270, 280, 283, 286, 313, 314. Wadsworth, Professor M. E., 574. Wahsatch Mountains, 149-151, 244. Wakefield, N. H., 311. Wales, 394, 398, 401, 412, 434. Wallace, A. K., 366, 584, 590, 592- 594. Wallenpaupack Creek, Pa., 277. Walrus, 386, 388, 529, 557. Warner, L. C, 94. Warping rafts, 311. Warren county. Pa., terminal moraine in, 137, 138; preglacial drainage in, 276. Warren, Pa., 137, 138, 276; kettle- hole in, 488. Washington county. Pa., 494. Washington, D. C, 411, 412. Washington Land, 78, 84, 86, 89. Washington Territory, 121, 150, 152, 153 ; glaciers in, 19 «i seq., 149, 163, 194. Watch Hill, 126. Water, peculiarities of its freezing- point, 1, 4, 6. Waterfalls, scarcity of, in unglaciated regions, 322. Watertown, N. Y., 431. Wells, in Ohio, 140, 141, 227, 228, 492; in British Columbia, 157; in Pennsylvania, 230, 494 ; in Indiana, 270, 491, 492 ; near Morgantown, W. Va., 337 ; in Red River Valley, 366 ; in Minnesota, 461, 495, 496. Wells, Mr. David A., 208. Wensleydale Glacier, 397, 398. Weston, W. Va., 338. West Virginia, 494; terraces in, 337 et seq., 493, 494; vegetable remains in glacial deposits in, 337 et seq., 622 THE lOE AGE IN NORTH AMEMICA. 493, 494 ; wells in, 33*? ; transported bowlders in, 339, 342. Whirlpool in Niagara River, 452 et seq. White Mountains, Alaska, 39, 5Y. White Mountains, N. H., 208, 379, 388 et seq., 695. White, Professor I. C, on the Cincin- nati ice-dam, 169, 330, 331, 344, 493 ; reports wells in Wyoming county. Pa., 230 ; reports buried channels in Pike and Monroe counties. Pa., 277 ; on terraces in West Virginia, 337 etseg., 342-344. White River, Dak., 146. White River, Ind., 491, 533. White River, Oregon, 19. White River, Washington Territory, 21—23 Whitney, Professor J. D., 674, 589 ; ascent of Mount Sliasta by, 18 ; cited 96, 148, 150; on glaciers of New Zealand, 100; on formation of Yo- semite Valley, 152, 241 ; on cause of Ice age, 408 ; palseolithic discoveries of, 517 ; on antiquity of man in California, 668, 562 et seq. ; on deep placers of California, 559, 560. Whittlesey, Colonel Charles, 576 ; on formation of kettle-holes, 130 ; re- ports wells in Ohio, 492. Whymper on Greenland glaciers, 74 et seq. ; cited, 96. Williamson county. 111., 142, 143. Willoughby Island, 40, 57, 161. Wilson, Mr. Thomas, 513. Winchell, Professor A., 577 ; on the effect of pressure of a glacier, 369, 370 ; on antiquity of man in Cali- fornia, 665. Winchell, Professor N. H., 487, 684, 593 ; on rate of recession of Falls of St. Anthony, 458 et seq., 549 ; palae- olithic discoveries of, 537, 640, 544 Wind River Mountains, 18. Wind, transporting power of, 360, 3 -. Winslow, Dr. C. P., 562. Wisconsin, 441 ; driftless area of, 120, 194, 366, 366, 443 ; depth of ice in! 173 ; kettle-range in, 120, 130, 182^ et seq., 193, 195, 321 ; drift hills in! 256; lakes in, 321, 357; loess in 365, 366 ; post-glacial erosion in. 470. Woeikoff, on Croll's theory of the cause of the Glacial period, 424 ; on glaciers of the southern hemisphere, 425, 426 ; on distribution of heat over the earth, 428, 429 et seq. Wood's HoU, 123, 125, 177, 178. Woodward, Mr. R. S , 265, 679 ; on rate of recession of Niagara Falls, 456, 458. Wright, Professor G. F., 549, 584, 593. Wright, Professor W. E. C, 309. Wyman, Dr., 563. Wyoming county. Pa., 230. Wyoming, former glaciers in, 13, 148. Yellowstone River, 147, 149, 198. Yosemite Valley, 13 ; formation of, 162, 241. Young, Rev. S. H., explorations of, in Alaska, 35, 64. Yukon River, 27, 29, 162, 215. THE END. D. APPLETON & CO;S PUBLICATIONS. Great Ice Age, and its Relation to the Antiquity of IVIan. By James Geikie. With Maps and Illustrations. 12mo. Cloth, |2.50. Weather: A Popular Exposition of the Nature of Weather-Changes from Day to Day. By the Hon. Ralph Abercromby, Fellow of the Royal Meteorological Society, London. No. 58 of the International Soientifle Series. With numerous Diagrams. 13mo. Cloth, |1.75. Earthquakes and other Earth Movements. By John Milne, Professor in the Imperial College of Engineering, Tokio, Japan. Inter- national Scientiiic Series. With 38 Illustrations. 13mo. Cloth, $1.75. An attempt is made in this volume to give a systematic account of various Earth Movements. These comprise Earthquakes, or the sudden violent movements of the ground; Earth Tremors, or minute movements which escape our attention by the smallness of their amplitude ; Earth Pulsations, or movements which are overlooked on account of tlie length of their period ; and Earth Oscillations, or movements of long period and large amplitude. Volcanoes : What they Are and what they Teach. By J. W. Judd, Professor of Geology in the Royal School of Mines (London). With 96 Illustrations. 13mo. Cloth, $3.00. Climate and Time in their Geological Relations: A Theory of Secular Changes of the Earth's Climate. By James Croll, of H. M. Geological Survey of Scotland. With Maps and Illustrations. 13mo. Cloth, $2.50. Discussions on Climate and Cosmology. By James Croll, LL. D., P. R. S. With Chart. 13mo. Cloth, $2.00. CoNTEKTS : Misapprehensions regarding the Physical Theory of Secular Changes of Climate. — The Ice of Greenland and the Antarctic Continent not due to Elevation of the Land. — Mr. Alfred K. Wallace's Modification of the Physical Theory of Secular Changes of Climate. — The Physical Cause of Mild Polar Climates. — Inter- glacial Periods and Distribution of Flora and Fauna in Arctic Eegions. — Tempera- ture of Space and its Bearing on Terrestrial Physics. — Probable Origin and Age of the Sun's Heat^ etc. Stellar Evolution and its Relation to Geological Time. By James Croll, LL. D., P. R. S. 13mo. Cloth, Outlines of a Mechanical Theory of Storms. Containing the True Law of Lunar Influence, with Practical Instructions to the Navigator to enable him approximately to calculate the Coming Changes of the Wind and Weather for any Given Day, and for any Part of the Ocean. By S. Bassnett. ISmo. Cloth, $2.00. D. APPLETON & CO., Publishers, New York.