88 IOWA ACADEMY OF SCIENCES. 



Whether the deposition took place by water or by wind, 

 there seems to have been a suspension of erosion on the lower 

 rapids, and the length of this suspension must certainly be 

 sufficient to affect materially their duration. An estimate of 

 the time involved seems at present impossible, there being 

 fewer data for an estimate than in the filling which occurred 

 at the Illinoian stage. 



EROSION FOLLOWING THE LOESS FILLING. 



After the deposition of the loess, the valleys throughout 

 much of the Mississippi basin experienced a marked deepen- 

 ing, which brought their bottoms to a lower level than before 

 the loess filling. In the ^portion of the Mississippi valley 

 which lies within and near the rapids the deepening seems to 

 have proceeded continuously to a level nearly as low as the 

 present stream, or fifty to seventy -five feet below the excava 

 tion which occurred in the interval following the Kansan 

 glaciation. This excavation, in the section embraced within 

 the rapids, was mainly rock, for the loess and alluvium had 

 built up the channel scarcely thirty feet above the rock floor 

 of the post- Kansan erosion. But for some distance, both 

 above and below these rapids, the excavation was largely in 

 till. The channel across the rapids was opened to a width 

 but little greater than the stream, or about one mile. Else- 

 where the channel is three to six times the width of the 

 stream. 



This erosion seems to have continued until the early part of 

 the Wisconsin glacial stage, when, as indicated below, another 

 filling occurred. The extent and depth of the erosion which 

 took place prior to the Wisconsin filling, is well shown in the 

 broad portion of the valley above the rapids. Numerous wells 

 indicate that the till had been removed nearly to present river 

 level, over the greater part of the width of the valley, before 

 that filling set in. 



The amount of erosion in the Mississippi valley seems to 

 have been nearly as great in this interval as in the post- 

 Kansan interval of erosion. It is doubtful, however, if the 

 time involved was so great as in that interval, for the gradient 

 appears to have been higher. To properly estimate the time 

 involved, it is necessary also to know the volume of water dis- 

 charged through the valley at each interval, a matter concern- 

 ing which very little is yet known. 



