30 THE ILLINOIS GLACIAL LOBE. 



whether the water be shallow or deep, and this mar have carried stones in 

 sufficient number to have supplied the clay with the few pebbles that it 

 contains. The difficulties arising- from the great range in altitude which 

 the deposit presents may not be fatal to the hypothesis of submergence. 

 The hypothesis is, therefore, still entertained, especially since none more 

 satisfactory has suggested itself, but it can not be confidently put forward 

 as a solution. 



In considering the time relations of this clay, there is decisive evidence 

 that it was deposited at a much later date than the Kansan sheet, in the 

 fact that it also overlies the Illinoian, a younger sheet of drift. The 

 evidence is equally conclusive from its relation to terraces cut in the 

 Kansan. The gummy clay, as well as its coating of loess, is found on 

 terraces bordering' valleys cut in the Kansan sheet, and also on the uplands 

 occupied by that sheet. All the main valleys examined in southeastern 

 Iowa had been cut to a depth of 50 feet or more into the Kansan sheet, 

 and often to a width of 1 or 2 miles, prior to the deposition of this clay. 

 If, therefore, it had been examined only in districts outside the limits of 

 the Illinoian, it might have been demonstrated to be a much younger deposit 

 than the Kansan. 



Passing to the Illinoian sheet, it is found that the changes effected in 

 its surface prior to the deposition of the clay under discussion are less than 

 in the Kansan, }-et some change was apparently effected in its surface. In 

 fact, the surface of the till has often the appearance of marked atmospheric 

 reddening- prior to the deposition of the gummy clay, and there is usually 

 an abrupt change from gummy clay to the till. The till is also not infre- 

 quently leached of its calcareous material for several feet below the base 

 of the gummy clay. In places the gummy clay is mingled with the under- 

 lying till, lint it does not follow that the two deposits are contemporaneous. 

 The Illinoian till sheet was not so conspicuously channeled by streams 

 prior to the loess deposition as the neighboring portion of the Kansan, but 

 the slight channeling which took place seems to have antedated the deposi- 

 tion of the gummy clay as well as that of the overlying- loess. This clay 

 is apparently more conspicuously developed in small channels cut in the 

 Illinoian than on the bottoms or terraces of the broad channels. Not infre- 

 quently these small channels are so greatly filled by the clay that the 

 surface is nearly restored to its original planeness. The writer has found 



