THE MISSISSIPPI VALLEY. 433 



broad valley which opens southward into the Mississippi, and passes east- 

 ward across a narrow point of uplands separating this valley from the 

 Mississippi. The old valley, though covered heavily with loess, stands only 

 40 to 50 feet above the present stream bed, while the rock ridge rises 

 promptly on either side of the deflected stream to a height of about 100 

 feet above the abandoned valley. 



The deflections, it will be observed, are each situated near the margin 

 of the Kansan and also of the Iowan drift of the western ice field, as well 

 as the margin of. the Illinoian drift of the eastern ice field. It becomes, in 

 consequence, no easy matter to decide upon the influence which each 'ice 

 field may have exerted in causing these peculiar features. In the Mississippi 

 deflection at Fulton the stream has been shifted to the west, as if due to 

 obstructions on the east, while in the Elk River deflection the stream has 

 been shifted to the east, as if the obstruction were on the west. However, 

 it is not certain that the deflections are directly due to ice advances. Upon 

 examining the ground with a view to interpreting the features, it appears 

 probable that encroachment by the Mississippi through a widening of the 

 valley may in each case have so broken down a portion of the crest of the 

 narrow dividing ridge that displacement could have been brought about by 

 only a slight filling of the valleys, such a filling perhaps as the loess pro- 

 duced. It will be observed that the present course of each stream is more 

 direct than the old course. This may aid in accounting for their persistence 

 in the new lines. This interpretation, however, is not wholly satisfactory. 



DEFLECTION PAST THE LEOLAIEE OR UPPER RAPIDS. 



Immediately below Clinton the glacial deposits attain sufficient thick- 

 ness to completely disarrange the old drainage and to render it very difficult 

 to determine the course of the preglacial stream. A special study of the 

 preglacial course was made for this Survey by Prof. J. A. Udden, the main 

 results of which he has kindly furnished for publication in this report. 



The present course across the rapids has apparently been selected by 

 the Mississippi in preference to several lines which had been opened, and is 

 the most direct of these lines. The rapids proper extend from Leclaire, 

 Iowa, down to Rock Island, Illinois, a distance of nearly 15 miles. There 

 is a fall at low water of 20.4 feet, or an average slope of about 16 J inches 

 to the mile. The rapids, however, do not present a uniform slope, but 



