DRAINAGE OF THE GLACIAL PERIOD. 333 



the 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 ; huge cones 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 Rocky Mountains. Here 

 the conditions were somewhat peculiar. The natural drain- 

 age is down the flanks of the Rocky 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. 



* "Proceedings of the American Association for the Advancement of Sci- 

 ence," vol. xxxiii, 1884, p. 391. 



