734 ABSTRACTS OF PAPERS 



are elsewhere completely covered by scorings made by the later Labradorean 

 glacier. To the west of the WiDiiipeg moraine the striations on the rocks 

 have been formed exclusively by the Keewatin Glacier from the north and 



northwest. 



The lake itself, at all events for the greater portion of its existence, was 

 bounded to the east by the front of the Labradorean glacier, and to the north 

 by the front of the Keewatin Glacier. 



The relationship of these two glaciers is beautifully shown by a section at 

 the mouth of the Saskatchewan River, where 12 feet of stratified Lake Agassiz 

 clays lie between the till of the Keewatin Glacier and that of the Labra- 

 dorean glacier. 



Lake Winnipeg is the diminished representative of Lake Agassiz, and even 

 now it has but a precarious existence, for it is bounded to the north by a clifiC 

 of lacustrine clays. The outlet from the lake accidentally began where this 

 clay was shallowest, near its northeastern extremity, and over a bottom of 

 hard granitoid rocks, but if at some time the outlet should change to a posi- 

 tion farther west, where the clays are thicker and the underlying rock is at a 

 greater depth, the lake might very easily be more or less completely drained 

 and we would have the Red River flowing down through the bottom of the 

 basin that is now covered by the waters of Lake Winnipeg. 



Mr. Waeren Upham contributed the following remarks, which were pre- 

 sented to the Soc-iety at the meeting by Prof. Lawrence Martin: The con- 

 clusions of Mr. Leverett differ so much from my general view of the history 

 of the glacial Lake Agassiz, presented in Monograph XXV of the U. S. Geo- 

 logical Survey, that I am moved to endeavor here to present friendly com- 

 ments showing how I regard some of his views and arguments. He discrim- 

 inates, as the abstract states, three sheets of the drift in northern Minnesota 

 of different ages; but, instead, I consider them as practically contempo- 

 raneous in deposition, representing a tendency toward lobation where con- 

 tiguous parts of the broad ice-sheet, confluent together, had respectively 

 broad glacial currents from the northwest, north, and northeast, each of these 

 somewhat lobate contemporaneous ice-fields bringing drift from the areas 

 whence its currents came. Therefore the drift from the northwest is calca- 

 reous, while the drift from the north and northeast, the latter colored reddish 

 by the red sandstone and shales of the Lake Superior basin, has little or no 

 limestone. W^hen the ice boundary was melted back at the end of the Glacial 

 period, its marginal moraines, formed at any times of halt or readvance in- 

 terrupting the geueral departure, were made up by whatever drift was in the 

 adjoining past of the ice-sheet, so that our series of marginal moraines, in 

 crossing Wisconsin. :Minnesota. and the Dakotas. passes continuously across 

 the several diverse drift sheets. Much of the glacial drift was borne along in 

 the basal quarter or fifth part of the thickness of the ice-sheet, up to heights 

 of 1,000 feet or more above the land surface, as shown by the esker of Bird's 

 Hill, near Winnipeg, described in my Geological Society of America pai>er 

 two years ago. Decaying granitic and other feldspathic rocks, deeply weath- 

 ered before the Ice Age. yielded much drift and abundant boulders and 

 smaller rock fragments, shaped by the glacial erosion and held in shape while 

 frozen, forming a large proportion of the older drift and of the lower and 

 most abundant englacial boulders, small fragments and fine drift, to be worn 



