THE MISSISSIPPI VALLEY. 471 



the post-Illinoian, and therefore much too great to refer to the work of the 

 Des Moines River. But while there seems clear evidence that the channel 

 across the rapids was begun by the Mississippi before the culmination of the 

 Illinoian invasion, it may not follow that it was occupied by that drainage 

 from the close of the Kansan invasion. The probable effect of the Kansan 

 invasion would be to throw the stream to the east of the preglacial course. 

 Furthermore, since the ice sheet at that stage extended beyond the line of 

 the Mississippi at the lower rapids, the displaced river would naturally take 

 a course through the lowest available district east of that ice margin. This 

 being done, it may have persisted in that course for some time after the 

 culmination of the Kansan stage of glaciation, and possibly until the Illi- 

 noian invasion crowded it westward to the present line. The amount of 

 excavation preceding the culmination of the Illinoian glaciation, however, 

 favors the view that its selection of the course across the lower rapids was 

 made soon after the retreat of the Kansan ice sheet. 



The value of the lower rapids as a chronometer for determining the 

 time since the Kansan invasion, as well as the relative dates of the Kansan, 

 Illinoian, Iowan, and Wisconsin stages of glaciation were considered in the 

 paper just referred to, and the main difficulties were discussed. One great 

 difficulty lies in the fact that there was not a continuous rock excavation. 

 At both the Iowan and the Wisconsin stage some filling occurred, and at 

 the Illinoian stage also it is probable that the filling predominated over the 

 cutting. A further difficulty is found in the fact that the drainage basin of 

 the Mississippi above these rapids has suffered great changes in area since 

 the Kansan stage of glaciation, thus greatly affecting the volume of the 

 stream. The work done at the rapids may be roughly stated as follows: 



(1) The excavation (independent of filling at the Illinoian, Iowan, and 

 Wisconsin stages of glaciation, which was subsequently largely removed) 

 comprises (a) the removal of a drift deposit capping the limestone which 

 had an average depth of perhaps 25 feet, a length of 15 miles, and a width 

 of 1^ miles; (b) the removal of limestone to an average depth of nearly 100 

 feet for a length of at least 12 miles and a width of about 1^ miles. The 

 excavation of limestone is estimated to be nearly one-third of a cubic mile. 



(2) A filling of undetermined amount, but apparently not less than 30 

 feet, correlated with the Illinoian stage of glaciation, which is well shown 

 by the sections at Warsaw and at Yellow Banks below the lower end of the 



