I O ERA NK LE VERE TT 



the valley until forced westward by the advance of the Labrador 

 ice-field. But if a great erosion took place in this part of the 

 valley prior to the Illinoian stage of glaciation there would seem 

 good grounds for supposing that the stream assumed its present 

 course soon after the Kewatin ice-field made its final with- 

 drawal. 



Examining into this question it is found that after this drift 

 was deposited by the Kewatin ice-field, an erosion so great took 

 place that it was removed throughout the greater part of the 

 width of the valley down to a level scarcely fifty feet above the 

 present stream at the mouth of the Des Moines, and to an 

 equally low level at Hannibal. The depth of cutting appears, 

 therefore, to have been about 100 feet at the mouth of the Des 

 Moines, and perhaps twenty-five feet at Hannibal. It seems 

 safe to assume an average depth of fifty feet for the entire 

 section and a width of five or six miles, making an erosion of 

 nearly three cubic miles of drift in the fifty miles below the 

 mouth of the Des Moines River. It is scarcely necessary to 

 raise the question whether this erosion could have been accom- 

 plished by the Des Moines and other tributaries of the Missis- 

 sippi below the rapids, for it is evidently out of proportion to 

 the work which these small streams would be able to accomplish 

 since the Kansan stage of glaciation. It seems certain that the 

 Mississippi River is responsible for the principal part of the 

 erosion. This makes necessary the opening of the new channel 

 across the rapids, for the old channel west of the rapids was not 

 utilized by the river after the Kansan stage of glaciation, and no 

 other line of drainage could have been adopted by the river 

 that would pass through the portion of the valley below the 

 rapids. 



Evidence is found within the new channel, of an erosion such 

 as the interpretation just given demands. In the south part of 

 Keokuk, between the foot of Main Street and the mouth of Soap 

 Creek the rock bluff rises but fifty to sixty feet above low water, 

 and is capped by a bed of bowlders about twenty feet in depth. 

 Attention was called to this bed some thirty years ago by Mr. 



