Digitized by the Internet Archive 
 in 2017 with funding from 
 
 University of Illinois Urbana-Champaign Alternates 
 
 https://archive.org/details/discussionofclimOOupha 
 
t < 
 
 • £- a 8V < 7y //a / 
 
 
 / 
 
 The President, Prof. F. W. Putnam, in the chair. 
 
 The President announced that the evening would be devoted to 
 a discussion of the Climatic Condition of the Glacial Period to be 
 considered in relation to the existence of man at that time. 
 
 He called upon Prof. G. Frederick Wright, who had obtained 
 important information in relation to the “ Nampa Image,” to open 
 the discussion. 
 
 Professor Wright read letters relating to the Nampa Image, 
 and spoke of the probable climatic conditions at the time of the 
 deposition of the deposits. from which the image was obtained. 
 
 The following abstracts have been prepared by the speakers. 
 
 THE GROWTH, CULMINATION, AND DEPARTURE OF 
 THE QUATERNARY ICE-SHEETS. 
 
 BY WARREN UPHAM. 
 
 As the character of a man is discerned by his life and work, so 
 we may learn what were the climatic conditions of the glacial pe- 
 riod by the study of their results in the formation of ice-sheets 
 whose farthest extent and stages of recession are known by the 
 limits of glacial striation and deposits of till or boulder clay and 
 by the course of terminal moraines. In this inquiry we may prof- 
 itably consider the climatic changes producing glaciation in their 
 manifestation successively by the growth, culmination, and de- 
 parture of the ice-sheets. The most interesting and difficult cli- 
 matic problem presented in all the geologic record is that of its 
 latest period, immediately preceding the present, to discover the 
 causes of the accumulation of its vast sheets of land-ice. So wide 
 and careful investigations have been bestowed on the glacial drift 
 that we are enabled to outline definitely the boundaries of the ice- 
 slieets both at their time of maximum extent and during pauses or 
 times of re-advance interrupting the general retreat. Further- 
 more, evidence is found both in America and Europe showing, as 
 many geologists believe, that there were at least two principal 
 epochs of glaciation and between them a very long interglacial 
 
 l 
 
 ( 450 ) 
 
Upham.] 
 
 451 
 
 [Jan. 1, 
 
 epoch when the ice was restricted as now to alpine glaciers and 
 polar regions. 
 
 The fossil floras of Greenland and Spitzbergen indicate that 
 those far northern latitudes enjoyed a temperate climate in the 
 Miocene period ; and from the absence of glacial drift through the 
 great series of Tertiary and Mesozoic formations we infer that 
 climates as mild as those of the present day had prevailed during 
 long eras before the ice age. With suddenness, geologically speak- 
 ing, there came at the close of the Pliocene period a great refriger- 
 ation of the climate of northern regions, overwhelming the Siberian 
 herds of mammoths, and covering the surface of the northern half 
 of North America and of northwestern Europe with snow and ice 
 which increased to thousands of feet in depth. The conditions 
 that seem requisite for the formation of these ice-sheets are long 
 continued rather than an excessive cold and an abundant supply 
 of moisture by storms, giving plentiful precipitation of snow dur- 
 ing more of the year than now, so as to include in the time of snow 
 accumulation not only the present winter but also the autumn and 
 spring months. The summers too were probably cooler than now, 
 for their heat was not sufficient to melt away the accumulated 
 snow, which gradually increased in thickness from year to year, 
 its lower part being changed to ice. When large portions of con- 
 tinents became thus ice-covered, the storms sweeping over them 
 would be so rapidly cooled that the greater part of their snowfall 
 would take place upon the borders of the ice-sheet, within prob- 
 ably fifty to two hundred miles from its margin ; but the snow- 
 fall during the advance of the ice was probably in excess of 
 the amount of evaporation and melting over the whole ice-cov- 
 ered area. In New England and New York the average as- 
 cent of the surface of the ice was twenty-five to thirty feet per 
 mile for the first one hundred to two hundred miles from its 
 boundary. Toward its centre the slope diminished, as on the in- 
 terior ice of Greenland ; but the ice-sheet enveloping the north- 
 eastern part of North America probably attained, as estimated 
 by Professor Dana, a maximum thickness of about two miles on 
 the Laurentian highlands between the river Saint Lawrence and 
 Hudson bay. 
 
 When the Quaternary ice-sheets occupied their greatest area, at 
 the culmination of the effects of the extraordinary climatic condi- 
 tions of this period, the southern border of the ice crossed the northern 
 

 1890 .] 
 
 452 
 
 [Upham* 
 
 United States from Nantucket, Martha’s Vineyard, Long Island, 
 and northern New Jersey, through Pennsylvania into southwestern 
 New York, thence west-southwesterly to southern Illinois and 
 Saint Louis, thence westward nearly to the junction of the Re- 
 publican with the Kansas river, thence northward through eastern 
 Nebraska and north-northwest through South Dakota, bending 
 from this course about thirty miles west of Bismarck, thence pass- 
 ing westerly through northern Montana, Idaho and Washington, 
 reaching the Pacific ocean not far south of Vancouver island. It ex- 
 tended beyond the Ohio river only for short distances in the vicinity 
 of Cincinnati ; but the Missouri river lies mainly within the gla- 
 ciated area. On the Mississippi river three hundred to four hun- 
 dred and fifty miles north of the boundary of the ice-sheet where 
 it reached farthest south, a large driftless area including southwest- 
 ern Wisconsin and parts of adjoining states escaped glaciation. In 
 the Rocky mountains, the Cascade range, and the Sierra Nevada, 
 ice-fields of great extent were accumulated along distances of 
 seven hundred to eight hundred miles south from the border of the 
 continental ice-sheet, to latitude 37° S. ; but no evidences of such 
 local glaciation south of the ice-sheet are found in the Appalachian 
 mountains. 
 
 Upon British America the directions of the glacial striae and 
 o transportation of the drift show that there were two general ice- 
 ^ sheets, one reaching from Newfoundland and Labrador to the 
 "i Rocky mountains and the Arctic ocean, having its greatest thick- 
 et ness over the Laurentian highlands and James bay, with outflow 
 thence to the east, south, west, and north ; and the other west of 
 ^ the Rocky mountains, covering British Columbia, attaining a max- 
 imum thickness of about one mile, and outflowing south into the 
 United States, west into the Pacific ocean, and northward to the 
 upper part of the Yukon basin. The glaciers of the Rocky moun- 
 tains were doubtless confluent with these ice-sheets, so that at the 
 time of maximum extent of the ice it was continuous from the At- 
 lantic to the Pacific, covering approximately 4,000,000 square 
 miles of this continent. 
 
 Half as large an area was ice-covered in Europe, the principal 
 center of outflow being the plateau and mountains of Scandinavia, 
 whence the ice moved west and north into the Atlantic, southward 
 over northern Germany, and eastward over a large part of Russia. 
 Smaller ice-sheets were formed upon Scotland and Ireland, and 
 these became confluent with each other and with the Scandinavian 
 
Upham.] 
 
 453 
 
 [Jan. 1, 
 
 ice which crossed the present bed of the North Sea to the borders 
 of Great Britain. Glaciers also were far more extensive than now 
 in the Alps, Pyrenees, Caucasus, and Himalayas ; but no large 
 portion of Asia is known to have been overspread by ice. 
 
 The most anomalous feature of the accumulation of these ice- 
 sheets was their absence from Siberia and northern Alaska, while 
 so heavily massed in the same and more southern latitudes of 
 British America, the northern United States, the British Isles, and 
 northwestern Europe. Within the past year Dali’s observations 
 of the absence of drift over large portions of Alaska have been 
 confirmed and extended by the explorations of Russell and Mc- 
 Connell, whose communications read last week before the Geo- 
 logical Society of America define the limits of our glaciated area 
 by a line that crosses the upper Yukon in British America near 
 lat. 62° N., long. 135° W., and thence extends northward to the 
 Arctic ocean somewhat west of the mouth of the Mackenzie. Rus- 
 sell also states that no glaciers now exist on the mountains of 
 Alaska north of the Yukon, though they are grandly developed 
 on the south along the high ranges of the Pacific coast. Prob- 
 ably the study of the present climate of that country will contri- 
 bute much toward explaining why the ice-sheet failed to extend 
 over the lower Yukon region, and will herewith help greatly to- 
 ward the solution of the broader and very complex problem of the 
 whole earth’s climatic conditions during the ice age. 
 
 But the course of the southern glacial boundary in the United 
 States shows important points of resemblance between the climate 
 of the glacial period and that of the present time. Where rainfall 
 now is deficient in the interior, on the arid plains east of the 
 Rocky mountains, the precipitation of snow in the Ice age was 
 similarly deficient, causing the great northward deflection of the 
 ice-margin from the latitude of St. Louis to that of Bismark ; and 
 Dana has shown that the driftless area of Wisconsin now receives 
 less rainfall than the contiguous regions that were ice-covered. 
 The courses of storms and precipitation of moisture having been 
 thus nearly the same as now, the great accumulations of ice on 
 the continents bordering the North Atlantic suggests that a large 
 element in the causation of the cold was probably a subsidence of 
 the Isthmus of Panama, allowing the equatorial oceanic current 
 to pass west into the Pacific, instead of being turned northward 
 in the warm Gulf Stream. 
 
 To this cause of the refrigeration of climate upon the glaciated 
 
1890 .] 
 
 454 
 
 [Leverett. 
 
 portions of the globe, we may add the influence of their elevation 
 as continental plateaus much above their present height. The 
 fiords of these regions indicate that they had generally stood 
 higher than now during probably the whole Pliocene period ; and 
 the submerged fiord of the Hudson shows that when this uplift 
 culminated, being as I believe the most important part of the 
 causes of the ice-accumulation in the glacial period, its extent on 
 our coast at New York was about 2,800 feet. But after the 
 earth’s crust became heavily loaded by the ice-sheets it sank until 
 the coast from Boston northward was depressed below its present 
 level, submerging the border of northeastern Massachusetts, New 
 Hampshire, Maine, and New Brunswick, and carrying the sea 
 inland along the valleys of the St. Lawrence, the Ottawa, and lake 
 Champlain. This depression of the submarine Hudson fiord was 
 so rapid that a bar 1,600 feet in height was formed across its 
 mouth while the deep fiord behind the bar remained unfilled by the 
 alluvial sediment of the river. 
 
 The greater part of the drift deposits exposed to observation, 
 including the moraines of recession, their accompanying kames 
 and overwashed frontal plains, the osars and their associated plains, 
 the valley drift and its terraces, the drumlins, the contour of the 
 general sheet of till, and even the glacial striae, tell us far more 
 of the retreat of the ice-sheet than of its advance and action dur- 
 ing its maximum extension. A very great change of climatic con- 
 ditions restored again prevailingly mild and warm temperatures, 
 melting away the vast ice-sheets ; but the returning warmth and 
 the glacial cold several times contended for the mastery, recessions 
 of the ice being succeeded by stages of re-advance, spreading till 
 in many places over interglacial soil and fossiliferous beds. 
 
 From my studies of the glacial lake Agassiz for the geological 
 surveys of Minnesota and the United States, I am much impressed 
 with the suddenness of the departure of the ice-sheet. This lake, 
 which grew to be six hundred miles long, probably as large as the 
 combined areas of the Laurentian lakes, and seven hundred feet 
 deep, confined in the northwardly sloping valley of the Red river 
 of the North and lake Winnipeg by the barrier of the waning 
 ice-sheet, is recorded by very definite beaches and occasional low 
 escarpments of shore-erosion ; but the total amount of wave action 
 on its shores is far less than that shown by the extensive deposits 
 of dune sand and the high cliffs of till bordering lake Michigan. 
 Now the whole of postglacial time, during which these effects have 
 
Leverett.] 
 
 455 
 
 [Jan. I, 
 
 been produced on lake Michigan, cannot be much more, according 
 to the investigations of N. H. Winchell, Andrews, Gilbert, and 
 Wright, than 7,000 to 10,000 years. Only a minor fraction of 
 this time could have been occupied in the retreat of the ice from 
 lake Traverse, at the mouth of lake Agassiz on the west side of 
 Minnesota, to Hudson baj', finally allowing this glacial lake to be 
 drained by the Red and Nelson rivers flowing northward as now 
 to the ocean. During this short time, however, several prominent 
 moraines were formed on the country adjoining lake Agassiz. It 
 thus appears scarcely less difficult for the meteorologist to explain 
 the climatic causes of the rapid departure of the ice than of its be- 
 coming accumulated so thick upon so large areas of the earth’s 
 surface. 
 
 CHANGES OF CLIMATE INDICATED BY INTERGLA- 
 CIAL BEDS AND ATTENDANT OXIDATION AND 
 LEACHING. 
 
 BY FRANK LEVERETT. 
 
 A study of the climate of the glacial period involves not only a 
 consideration of the conditions necessary to produce glaciation, 
 but also of the changes of climate that occurred. Owing to the 
 complexity of the phenomena that evidence these changes, their 
 extent has not been demonstrated to the satisfaction of all students 
 of the subject. As is well known, opinion is divided on the ques- 
 tion of the unity of the period, many now considering the evidence 
 to be conclusive that there were tw T o distinct invasions of the ice, 
 with an interglacial period of great length ; others still maintain- 
 ing the occurrence of one continuous glacial period. 
 
 The main evidence brought in support of the former theory is 
 (1) a sheet of carbonaceous soil, called by Dr. Newberry the forest- 
 bed, and fossiliferous interglacial sands and clays, all of which 
 are found extensively between sheets of till, both in America and 
 in Europe ; (2) the great erosion of the older drift (£. e., of the de- 
 posit belonging to the so-called first glacial epoch) previous to the 
 deposition of the newer, shown b}^ the unconformity of the newer 
 drift sheet, covering as it does great valleys and eroded ridges in 
 the older drift ; also a more eroded and more aged surface in the 
 uncovered portion of older drift lying south of the margin of the 
 
 newer. 
 
1890 ,] 
 
 456 
 
 [Leverett. 
 
 In defence of the latter theory it is maintained that the forest- 
 bed was rapidly formed and covered during the numerous oscilla- 
 tory retreats and advances, and the erosion of the so-called older 
 drift was effected by glacial floods attending the retreats ; the more 
 aged appearance of the drift near its southern margin being ac- 
 counted for by its supposed larger derivation from residual ma- 
 terial gathered by the glacier as it first pushed across the country. 
 
 Whether the forest-bed was formed rapidly or not ; whether pre- 
 glacial leaching or postglacial age was the more important factor 
 in the production of the aged appearance of the southern margin 
 of the drift ; and how far the great erosion is attributable to glacial 
 torrents and how far to smaller streams of longer action ; — are prob- 
 lems doubtless possible of solution. Already much has been done 
 toward solving them, but much remains to be done before it will 
 be definitely known to what extent each theory is applicable. 
 
 I wish to present here a line of evidence in support of the theory 
 of two distinct epochs, drawn from studies made under the direc- 
 tion of Pres. T. C. Chamberlin, in the eastern part of the Missis- 
 sippi basin, — one which I think has not yet received due attention, 
 namely, the depth to which the carbonates were leached from the 
 drift surface during an epoch of deglaciation in the midst of the 
 glacial period. 
 
 There are in Illinois, Indiana and Ohio, a series of about ten 
 more or less distinct moraines, all having at surface a fresher drift 
 than that lying outside their southern limit. Several of these have 
 at some depth a till corresponding in character to, and seemingly 
 a continuation of, the outlying drift. The southernmost moraine 
 has been traced from southwestern Illinois eastward to the Scioto 
 river, its further distribution being as yet undetermined . 1 
 
 The till is soft and fresh in appearance to a depth of at least 
 twenty feet in this southernmost of the series, and to a depth of 
 one hundred feet or more beneath later moraines in central Illinois 
 and Indiana. Below it is a hard, partially cemented till, fre- 
 quently capped by a dark colored soil a few inches in thickness, 
 and oxidized to a depth of several feet. The oxidized portion is 
 of a brownish yellow color, differing from the oxidized portion of 
 the surface tills of the same region which are yellowish gray. This 
 
 J The following cities are situated near it : in Illinois, Litchfield, Hillsboro’, Pana, 
 Shelbyville, Mattoon and Paris; in Indiana, Terre Haute, Rockville, Greencastle, Ed- 
 inburgh, Columbus, Greensburg, Connersville and Brookville ; in Ohio, Lockland, Leb- 
 anon, Hillsboro’, Bainbridge and Chillicothe. 
 
Leverett.] 
 
 457 
 
 [Jan. 1, 
 
 hard till below the newer drift is frequently exposed in ravines in the 
 moraines toward the south, and can be traced continuously back 
 from the district south of the outermost moraine until the fresher 
 drift becomes too thick for ravines to reach it. Along these ra- 
 
 vines the following section is frequently exposed : 
 
 1. Surface soil and leached subsoil, 2-6 feet. 
 
 2. Yellowish-gray till, 6-10 46 
 
 3. Blue till, soft and fresh in appearance, 0-50 “ 
 
 4. Soil and leached sub-soil, 3 “ i 
 
 5. Brown till, hard, dry, and partially cemented, 5-10 “ 
 
 6. Blue till, hard, dry and partially cemented, trav- 
 
 ersed by brown streaks, 5-50“ 
 
 Where ravines fail to reach the lower drift sheet it has fre- 
 
 quently been struck in wells, so that its occurrence is known as far 
 north as Mendota, Illinois, the vicinity of the Kankakee river and 
 in central Indiana, a distance of not less than two hundred and 
 fifty miles from the southern margin of the drift in Illinois, and 
 nearly one hundred and fifty miles from the southern moraine of 
 the newer drift in that state. 1 
 
 It is not the great extent of retreat and subsequent advance of 
 the ice nor the mere presence of a buried soil that I wish to em- 
 phasize, but the amount of oxidation and leaching which took place 
 in and below the buried soil during this retreat. Near Mendota, 
 Illinois, I examined specimens of the buried soil, subsoil and un- 
 derlying till from several borings, the depth to the buried soil rang- 
 ing from ninety-four to one hundred and twenty-eight feet. The 
 soil and subsoil resemble the underlying till in containing com- 
 mingled sand, clay and pebbles, but they lack the limestone frag- 
 ments and calcareous rock-flour which characterize the till. The 
 leaching is so perfect that no response with acid was obtained with- 
 in three feet of the surface of the buried soil, but below this depth 
 the till became within a foot or two very calcareous. Similar re- 
 sults followed the test applied along the ravines in southern Indi- 
 ana and southwestern Ohio, both in exposures beneath the newer 
 drift, and in the district south of its margin, the depth of leaching 
 being little, if any, greater than at Mendota. 
 
 The surface of the newer drift of this region is leached to various 
 
 Juried soils are reported from various places in northeastern Illinois and southern 
 Wisconsin, but I have had no opportunity to examine them, and have been unable to 
 ascertain whether or not the leached subsoil is found in connection with them. 
 
1890.] 
 
 458 
 
 [Leverett 
 
 depths, the variation depending largely upon drainage facilities. 
 Where drainage is good and denudation moderate it is two to six 
 feet, but beneath marshes there is little leaching below the black 
 muck, such places having frequently marl beds at the surface of 
 the till. In the level tract near Chicago (not marshy) leaching has 
 in many places scarcely reached a foot in depth. Few analyses 
 of the unleached tills of this region have been made from which the 
 percentage of carbonates may be ascertained. Two analyses, by 
 R. B. Riggs, 1 of a glacial and glacio-lacustrine till, near Milwaukee, 
 Wisconsin, show about forty per cent, of carbonates. Three analyses 
 of glacial sands from hillocks of angular gravel in Ohio, made by R. 
 B. Smith, 2 show a wide difference in the amount of carbonates, one 
 giving 70.364 percent. ; the second, 75.604 percent, soluble in hydro- 
 chloric acid, the greater part being in the form of carbonates of lime 
 and magnesia ; the third giving 38.964 percent, of carbonates. My 
 observations lead me to think that forty per cent, represents ap- 
 proximately the amount of calcareous material in the till from 
 which the buried soil is derived. There is, therefore, if we may 
 reason from the few analyses made, and if we take no account of 
 surface denudation which goes on more or less rapidly wherever 
 drainage is good, a probability that fully two-fifths of the upper 
 five feet of the surface of the older drift sheet have been removed by 
 leaching leaving the three feet of leached drift now associated with 
 the buried soil. To accomplish this would certainly require the 
 lapse of a long interval of time. 
 
 The land surface south of the margin of the newer drift in Illi- 
 nois, Indiana, and Ohio, is covered by two to five feet or more of 
 silt, showing that it was extensively submerged after the soil had 
 formed on the old drift surface. This silt is itself leached but has 
 probably prevented the leaching from extending to greater depth 
 beneath the old soil, it being very compact. 
 
 The question may be raised whether the absence of carbonate of 
 lime in the soil and subsoil and its presence in the underlying till 
 of the older drift may not be due to original difference in structure. 
 It is my own opinion based upon numerous comparisons of the soils 
 with the till that the material forming the soil has the same origi- 
 nal structure as the oxidized till, both being commingled drift con- 
 
 1 Sixth Annual Report, U. S. Geological Survey, p. 250. 
 
 2 Am. Journ. of Sci., May, 1884, pp. 383-384. 
 
 2 
 
Shaler.] 
 
 459 
 
 [Jan. 15, 
 
 taining pebbles, sand and clay. The change from noncalcareous 
 subsoil to the calcareous till is usually gradual, seldom, if ever, so 
 abrupt as should be expected on the theory of the superglacial ori- 
 gin of the one and the subglacial origin of the other, — not so ab- 
 rupt as is usually the change from the oxidized to the unoxidized 
 till. If we accept, with some, the theory that the oxidized till is 
 all superglacial, the extent and character of the leaching have still 
 the same significance in denoting a lapse of time as on the theory 
 of its subglacial origin. 
 
 The forest bed occurs at several other horizons than the one in- 
 dicated above. It is found to a limited extent between the sheets 
 of newer drift, but here it has a calcareous till immediately below 
 it, and this till is not perceptibly older than that above the soil, 
 there being little or no oxidation. Similarly, in the older drift there 
 are frequently exposed thin carbonaceous beds between the oxidized 
 and unoxidized tills of which it is composed, and also interstrati- 
 fied with the unoxidized tills, but accompanied by little or no leached 
 subsoil. Such soils may have been formed and buried during periods 
 of oscillatory retreat of short duration and of limited extent. 
 
 In brief, the evidence from this district, so far as its buried soils 
 and leached tills have a bearing, seems to teach that there were two 
 main epochs of glaciation, each characterized by oscillatory retreats, 
 between which was an epoch of deglaciation of considerable length. 
 
 NOTE ON GLACIAL CLIMATE. 
 
 BY N. S. SHALER. 
 
 [Published by permission of the Director of the U. S. Geological Survey.] 
 Every fragment of evidence, which can serve to show us the 
 character of the climatal conditions during the last glacial period, 
 is so important that I venture to present certain facts which so far 
 as I am aware have hitherto escaped attention. The evidence I 
 mean to discuss is found in this country and Europe in the regions 
 immediately south of the glaciated areas of the two continents. It 
 is a well-known fact that in the present condition of the climates 
 of the earth, the decrease in temperature as we rise above the sea 
 is about 3° F. for each 1000 feet of altitude. Local circumstances 
 may considerably affect this variation, but the range is not great. 
 
460 
 
 [Shaler. 
 
 If glaciers were by the refrigeration of the climate restored to the 
 surface which they occupied during the last ice period, we should 
 expect to find the line of perpetual snow rising as we went south- 
 ward about 3030 feet for each degree of latitude. 
 
 If on an inspection of the areas glaciated during the recent ice 
 epoch we should find that this principle in the distribution of the 
 glacier did not hold, we should apparently be justified in the sup- 
 position that the glacial climate was not due to greater cold than 
 that which exists at present. Any departure from the normal rate 
 of ascent of the perpetual snow line in the region south of the gla- 
 cier would be likely to throw some light on the climatal conditions 
 prevailing during the time when the continental ice sheets were 
 developed. 
 
 Beginning our inquiry with the Appalachian section of eastern 
 America, we find there a region in many ways well suited for the 
 determination we seek to make. The principal front of the ice 
 stretched across the continent on a line which is now well deter- 
 mined. It is unmistakably evident that it crossed the valley of the 
 Ohio at Cincinnati and extended a little distance south of that 
 stream into Kentucky. I have recently reexamined the evidence 
 which goes to show the presence of the ice at the above-named point 
 and have no doubt as to the goodness of the determination. At 
 this point the surface of the country lies at a height at no point ex- 
 ceeding 900 feet above the level of the sea. From this position the 
 level of the country gradually rises in a southerly direction until in 
 the synclinal mountains near Cumberland Gap it attains a height 
 probably exceeding 3500 feet. From this elevation the profile de- 
 scends in the broad valley of the upper Tennessee to about a thou- 
 sand feet above the sea level ; thence it again rises until in the 
 mountains of North Carolina, we enter a field where many peaks 
 rise to more than 6000 feet in height. From the front of the ice 
 sheet near Cincinnati to the central part of the North Carolina 
 mountain district is about 200 miles. It is to be observed that the 
 whole of this district is within the same great valley and in a re- 
 gion where the isotherms at the present time follow each other with 
 normal curves. We may therefore fairly conclude that under the 
 usual conditions of climate such as prevailed in North America, 
 the ice line should be found in the mountains of North Carolina 
 at the height of 2000 feet above the base of the glacier at Cincin- 
 nati, or say at 3000 feet above the level of the sea. From that 
 
Shaler.] 
 
 461 
 
 [Jan. 15, 
 
 level to the top of the North Carolina mountains or say for the 
 height of 3500 feet we should have indications of glacial conditions. 
 A tolerably careful investigation of this country has shown me no 
 evidence of ice action whatsoever and all the other students of the 
 subject who have visited this area have failed to find any facts 
 which might afford even a supposition of glacial work in that field. 
 I am therefore compelled to assume that the slope of the snow line 
 rose so rapidly from the ice front at Cincinnati southward that it 
 passed above the summits of these mountains. 
 
 If the elevation of western North Carolina was in the form of 
 an isolated peak, we might have less confidence in this indication. 
 But the district of land which should have lain much above the 
 snow level is some thousand square miles in area, a field sufficiently 
 great to have developed very extensive glacial areas in case the 
 peaks lay above the line of perpetual snow. The same considera- 
 tions, though in a less accented way, are met when we examine the 
 highlands of the Blue Ridge in Virginia or the Alleghany moun- 
 tain district on the uplands of Virginia and West Virginia. A 
 large part of the Blue Ridge in Virginia is high enough to have 
 been the seat of glaciers, provided the snow line were anywhere near 
 the level of the glacial sheet where it crossed the existing Atlantic 
 coast. The traces of glacial work in the Blue Ridge are extremely 
 scanty. At the western extremity of Rock Fish Gap, immediately 
 south of the Chesapeake and Ohio Railway, near its junction with 
 the Shenandoah Valley Railwa}^, there are accumulations which 
 apparently are to be classed as glacial. This point is about 1 600 
 feet above the level of the sea. If this accumulation be really of 
 glacial origin, it apparently establishes the height of the ice front 
 in the Shenandoah, but as yet I must regard the indication as 
 somewhat questionable. In the Alleghany Mountains west of Cov- 
 ington, Va., there are deposits which I am disposed to consider of 
 a glacial nature. At this point the deposits lie about 2000 feet 
 above the sea level. These are the southernmost points at which 
 I have found any satisfactory indications of glacial work, in the 
 region south of the Potomac, and until further investigated, both 
 of these deposits must be regarded as of doubtful character. 
 
 In Europe in the region south of the Alps, we find the facts sim- 
 ilar in their character to those existing in North America. Dur- 
 ing the last glacial period the ice sheet extended down on to the 
 Italian plains, unquestionably attaining levels less than 1000 feet 
 
462 
 
 [Shaler. 
 
 of altitude above the level of the sea and probably occupying po- 
 sitions not more than 500 feet above that level. From my obser- 
 vations on the field I am disposed to think that the general mantle 
 of the ice covered the southern face of the Alps down to within a 
 few hundred feet above the sea. From one hundred and fifty to 
 two hundred miles south of the Alps in the mountains of Tuscany, 
 we have an extensive surface rising 4000 or 5000 feet above the 
 sea. A careful search over much of this field showed me no evi- 
 dence of occupation by ice. At the present rate of rise in the 
 perpetual snow line in Switzerland we should expect an ascent of 
 that plane about 1500 feet in passing from the foot of the Alps to 
 the Apennine mountains north of Florence. We have thus a case 
 similar to that we find in the North Carolina mountains in which 
 there are elevations just south of the continental glaciers of a suffi- 
 cient height to have been covered by ice under normal circum- 
 stances, but where the evidence of such coating is conspicuously 
 wanting. 
 
 I have endeavored to apply the same considerations to the gla- 
 cial phenomena of the Rocky Mountains, but the facts are as yet 
 so imperfectly in hand that I have not been able to determine the 
 relative attitude of the sheet in a satisfactory manner. This, how- 
 ever, may be said, the distinct glacial accumulations in Colorado 
 probably do not extend below the level of 6000 feet. As this re- 
 gion is about on the parallel of the mountains of western North 
 Carolina, may perhaps indicate that the snow line lay through- 
 out the southern parts of the United States above the summits of 
 the Carolina mountains. It seems to me, however, that in the 
 existing state of our knowledge of the distribution of the glacial 
 sheet in the Cordilieran section, we cannot attach much importance 
 to this evidence. 
 
 We have now to consider the possible explanation of the facts 
 above adduced. Assuming that the relative height of the surface 
 occupied by the glacier, when it crossed the Ohio River and that 
 of the region within two hundred miles south of it, even during the 
 ice epoch, were what they are at the present day, it at first sight 
 seems necessary to suppose that there was a rapid change in the 
 temperature in passing from the ice front towards the Gulf of 
 Mexico. Before we adopt this consideration, however, we must 
 bear in mind the fact that the ice sheet of the last glacial period 
 probably advanced for a considerable distance south of the perpet- 
 
Shaler.] 
 
 463 
 
 [Jan. 15, 
 
 nal snow line, in substantially the same way in which an Alpine 
 glacier descends in many cases to the depth of a thousand feet or 
 more below the fields of enduring snow by which it is fed. Ac- 
 cepting the elevation of the continents as they now exist and al- 
 lowing 3° of temperature for each 1000 feet of altitude, it seems 
 likely that the snow line just touched the summit of the Carolina 
 mountains and came to the surface of the sea near the southern 
 end of Hudson’s Bay. In other words, the protrusion of the ice 
 to the south of this glacial snow line carried it at a distance of 
 near 1000 miles south of the gathering ground. This supposition, 
 however, is of little value, for the reason that the level of the con- 
 tinent was clearly much disturbed during the glacial period, the 
 surface declining to the northward within the glacial envelope and 
 probably rising to the southward of the ice front. 
 
 It seems to me most likely that during the occupation of the 
 northern part of the continent by glaciers, the southern portion of 
 the continent was considerably elevated. All the streams which 
 discharge into the ocean south of the former ice front between New 
 York and the Rio Grande show in their lower parts onty moderate 
 accumulations of alluvium which has been deposited since the close 
 of the glacial period. They generally enter bays which appear to 
 be the lower parts of gorges which were formed during the period 
 when the area was more elevated than it is at the present time. 
 These facts make it probable that if the mountains of North Caro- 
 lina varied in elevation from the present height, they 5 were more 
 elevated than at this day. All the facts are against the supposi- 
 tion that we can explain the absence of glaciers in their highlands 
 by supposing that the summits were lower during the ice period 
 than they now are. 
 
 It seems to me we are compelled to suppose that the climate in 
 the mountains of North Carolina and probably in the great portion 
 of the Apennine section south of the Alps had during the glacial 
 period a temperature not much if any lower than they have at the 
 present time. As far as it goes the evidence is thus opposed to the 
 supposition that the glacial period was brought about by a general 
 refrigeration in climate of the continents occupied by the sheet. 
 
 Within the basin of the Ohio, especially in the valleys of the 
 upper Tennessee system of waters, we find certain phenomena 
 which lead us to the conclusion that the rainfall in a recent pe- 
 riod, probably contemporaneous with the glacial epoch, was more 
 
1890 .] 
 
 464 
 
 [Upham. 
 
 considerable than at the present day. In many valleys which I 
 have observed in that section the debris built into the imperfect 
 alluvial plains is of a much coarser nature than that now brought 
 down by the rivers. The channels bear the aspect of having re- 
 cently been the seat of more voluminous streams than now occupy 
 them . This evidence gained from many points in the southern Appa- 
 lachians leads me, independently of the hypothesis I am now suggest- 
 ing, to the conclusion that during the last glacial epoch the rainfall 
 of this country was much greater than it is at present. At Big 
 Bone Lick in Kentucky which lies within a few miles of the south- 
 ern boundary of the ice sheet excavations made by me in 1868 show 
 embedded in the deposits formed by the springs, an abundant set 
 of herbivorous mammals, including the mastodon and elephant, 
 an extinct species of buffalo and a musk ox kindred to our arctic 
 species but of much larger size, a species of carabou, indistinguish- 
 able from our living American forms. The conditions of this de- 
 posit led me to suppose that these animals were probably not more 
 ancient than the glacial period, and that they most likely occupied 
 the surface during the time of abundant rainfall when the marshes 
 were more extensive than at present, a period which if not ex- 
 actly coincident with the extreme advance of the ice must fall 
 within the glacial epoch. 
 
 The abundance of these large herbivora, the relatively great size 
 of the species, point also to the coincident occurrence of a rather 
 abundant vegetation. If the period indicated by the massive grav- 
 els of the torrential streams and the herbivora of Big Bone Lick 
 be identical, and if this period coincides with the glacial period, as 
 it appears to do, then we may fairly assume that the climatal con- 
 ditions immediately to the south of the glacial sheet were not those 
 of extreme cold. This evidence has nothing like the sure founda- 
 tion, as that obtained by the lack of glaciers in the mountains of 
 North Carolina, but as far as it goes it confirms the results of those 
 observations. 
 
 It is not my purpose, however, in the present writing to con- 
 sider the perplexing question as to the cause of glacial climate. 
 I desire only to call attention to the extent to which our glacial 
 streams appear to have advanced, by what we may term forced 
 marches, far to,the soutli of the line of perpetual snow. Although 
 the value of the evidence above noted cannot be determined until the 
 matter has been more carefully brought together and abundantly 
 
Crosby.] 
 
 465 
 
 [Jan. 15, 
 
 discussed, the facts seem to me to militate against any h} 7 pothesis 
 which seeks to account for the glacial period on the supposition 
 that the climate in the glaciated regions was cooler than at present. 
 
 In the subsequent discussion, Mr. Upham spoke of the very ex- 
 ceptional character of the climate of the glacial period. He be- 
 lieves, with Nordenskiold and Wallace, that there was no wide- 
 spread glaciation at any time during very long preceding geologic 
 eras. The ice-brought blocks in Miocene deposits south of the 
 Alps, and other local glacial formations of Tertiary and Mesozoic 
 age are so infrequent that they seem to be best explained by ref- 
 erence to alpine glaciers at times of great uplift of neighboring 
 mountains. But more widely distributed evidences of glaciation 
 occur in the Carboniferous and Permian series, boulder-bearing 
 deposits of so remote age, closely like the Quaternary till, and 
 also striation of the underlying rock, being found in Natal near 
 latitude 30° S., and in India only 20° N. of the equator. Appar- 
 ently contemporaneous glacial deposits are also known in south- 
 eastern Australia, in Great Britain, and elsewhere. Croll has 
 shown that the coal measures were laid down under cool temperate 
 climates, attended by frequent oscillations of the land, like those 
 of the Quaternary period ; and it is very probable that man} 7 Car- 
 boniferous and Permian conglomerates were formed by ice-sheets. 
 
 Looking for causes of glaciation which could have acted effi- 
 ciently at the close of Palaeozoic time and again after the Tertiary 
 era, upon the threshold of the present, Mr. Upham thinks that 
 great earth movements at these times elevated portions of the 
 globe to such heights that they had cool climates with most of 
 their precipitated moisture as snow instead of rain. Deformation 
 of the earth was doubtless the means of its relief from the strains 
 due to its continuous contraction during the ages of quietude be- 
 tween epochs of mountain-building ; and the extent of the defor- 
 mation would be greatest, elevating extensive areas as high pla- 
 teaus, just before the plication of sections of the earth’s crust 
 and their upthrust as mountain ranges. These culminations of 
 the effects of terrestrial contraction Mr. Upham finds, associated 
 with the Carboniferous and Permian glaciation, in the folded Ap- 
 palachian ranges of eastern North America and the Sinian moun- 
 tains system of eastern Asia, both formed in the Permian period 
 and constituting parts of the grandest disturbances of the earth’s 
 
466 
 
 [Crosby. 
 
 crust between the end of the Archaean era and the beginning of 
 the Quaternary. Again, the latest and probably most extraordi- 
 nary glacial period that our earth has passed through is known to 
 have been attended by uplifts of the Cordilleran ranges in both 
 North and South America, and especially by mighty throes of 
 mountain-building in the Himalayan chain and considerable dis- 
 turbances along its western continuation in the Caucasus, Alps, 
 Pyrenees, and Atlas mountains. 
 
 Thus the chief cause of the climatic changes producing ice- 
 sheets and local glaciers seems to be found in the elevation of 
 broad areas for the former, and of mountain districts for the lat- 
 ter, to heights much above their present levels, bringing cooler 
 temperatures throughout the year. On the other hand, the return 
 of warmth and departure of the ice were coincident, at least in the 
 case of the Quaternary ice-sheets, with subsidence of the glaciated 
 portions of the earth’s crust, which indeed appears to have been 
 due to the weight of the ice and to have become in turn the princi- 
 pal influence leading to amelioration of climate and the final gla- 
 cial melting. 
 
 Prof. W. O. Crosby referred briefly to some examinations of the 
 till or bowlder clay in the vicinity of Boston which he has made 
 during the past year with a view to determining the normal pro- 
 portions of mechanical detritus (bowlders, pebbles, sand and rock 
 flour) and the residual product of chemical decay (clay) in its 
 composition. The rock flour, which must be ascribed mainly to the 
 scoring and grinding action of the ice-sheet, proves to be the most 
 prominent constituent of the till, amounting usually to from forty 
 to fifty per cent of the whole mass ; while only from ten to fifteen 
 per cent of the till is found to be true clay, the remainder consist- 
 ing of sand and coarser materials of mechanical origin. This small 
 proportion of clay is no more than it is necessary to regard as be- 
 longing to the preglacial sedentary soil of this region. Therefore, 
 we are forced to the conclusion that, although the ice age is sup- 
 posed to have lasted many thousands of years, the chemical action 
 of air and water upon the rocks during that time must have been, 
 probably, on account of the intense cold, almost nothing. 
 
 It was also pointed out that while the exposed surfaces or nat- 
 ural ledges of even the more acidic rocks, like granite, show appre- 
 ciable or considerable decomposition in postglacial times, this 
 
Crosby.] 
 
 467 
 
 [Jan. 15, 
 
 action is usually less marked or entirely inappreciable where the 
 rocks have been covered by the till, and that even the fragments 
 and minute grains of feldspar disseminated through the till, being 
 hermetically sealed in the well compacted clay, have almost en- 
 tirely escaped subsequent chemical change. The conclusion is thus 
 reached that the glacial climate was not only very cold, but that 
 free access of warm meteoric waters is essential to th apid chem- 
 ical decay of the silicate rocks. 
 

 
 
 
 
 
y 
 
 }