500 THE GLACIAL LAKE AGASSIZ. 



wliicli outcrop in the Ijottom ot" the valley alony a distance of several miles 

 next below Big Stone Lake. 



Changes in levels of the heaehes only a partial measure of the ice weiglit. — If 

 the thickness of the ice-sheet upon the area of Lake Agassiz was a half 

 mile to IJ miles, as seems probable, increasing from south to north and 

 northeast, the crustal uplift measuring the inflow to this area of an equal 

 weight of plastic or molten rock would range from 880 feet, or a sixth of 

 a mile, at Lake Traverse, on the south, to 2,640 feet, or a half mile, at 

 the north end of Lake Winnipeg, on the assumption that the density of the 

 inflowing rock or magma were that of the upper portion of the earth's 

 crust. The density of ice is taken as 0.9, water being 1.0; and that of the 

 rocks forming the earth's surface is assumed to average 2.7, the earth's 

 mean density, determined by three independent methods Avith closely 

 accordant results, lieiug about 5.5. But the mobile stratum next beneath 

 the solid crust is surely somewhat heavier than the crust. Comparison of 

 the earth's superfcial and mean densities indicates for this magma a proba- 

 ble density of 3.5 or more, which would reduce the computed uplift to 680 

 feet at the south, with increase of about 2 feet ])er mile northward to 2,040 

 feet, in round numbers 2,000 feet, at the mouth of Lake Winnijieg and 

 along the Nelson River, or less than these amounts if the density of the 

 uplifting magma was greater than 3.5. It is very ])robable that tlie sub- 

 sidence caused by the ice-sheet, depressing the crust below its jjreglacial 

 height, was more than would be thus strictly proportionate to the weight 

 added by the ice accumulation; but on the other hand it seems probable, 

 as shown by the northward ascent of the beaches of Lake Agassiz, that 

 only a minor fraction, perhaps nowhere within this basin exceeding one- 

 fourth, of the weight of ice removed was compensated by the differential 

 uplift of the land. 



But could we well explain the facts of glacial striation and dritt trans- 

 portation by assmning for the ice-sheet a less thickness, as one-third of a 

 mile to 1 mile from soiith to north upon the lacustrine area, which may, 

 indeed, be nearer the truth, tlu' rate of ascent of the shore-lines within the 

 area of my survey, I'esulting apparently from the departure of the ice, 

 would be closely in accordance with the hypothesis that the earth's crust is 



