214 Kansas Academy of Science. 



from the ice front. The boulders accumulated as the strength of 

 the current declined from tte subsiding of the waters from some 

 cause, either by diminution of the rate of melting or the diversion 

 of the flood down the Kansas valley from the melting back of the 

 ice front. That temporary floods still occupied the older channel 

 from time to time is inferred from the deep silt capping overlying 

 the gravel and boulders. 



2. The problems furnished by the bouldery patches and the 

 higher bouldery terraces are much more complex and difficult. 

 We think best to treat them separately, though there are reasons 

 for thinking them possibly contemporaneous. The bouldery patches 

 are a mere veneering of boulders over elevated points not quite up 

 to the general upland level. There is not enough of clay or gravel 

 assomated with them to show whether they are the deposit of land 

 ice or of shallow streams or irregular lakes. It seems that they 

 may be arranged into channel-like strips, but the patches are so de- 

 tached that the appearance may be illusive. Four possible expla- 

 nations present themselves: First, that they are the work of a 

 pre-Kansan ice sheet. Second, the work of a more advanced stage 

 of the Kansan sheet, postulating that that sheet lingered long 

 enough afterward to erode channels 100 feet deep and to fill them 

 as contemplated under the head already considered. Third, they 

 may be the work of shallow streams wandering over the plain in 

 front of the Kansan sheet before Wakarusa valley had been begun. 

 A fourth view may be that that they are the deposits of a tempo- 

 rary glacial lake formed in the valley of a small tributary of the 

 Kansas, which may have flowed northward along the line of Coon 

 and Oakley creeks, and which was dammed by the advent of the 

 ice sheet. 



In favor of the first supposition, it may be urged that traces of 

 such an advance of the ice have been found as near as southwest- 

 ern Iowa, and that it would easily afford adequate time for the 

 erosion of such a valley as we have found in existence during the 

 Kansan stage in the terrace already discussed. Against such a 

 conclusion, however, is the apparent impossibility of conceiving an 

 ice sheet covering these areas south and southeast of Clinton, with- 

 out leaving traces of its presence over wide areas where no trace is 

 found. For example, no traces are found on the higher points of 

 the divide north of the Wakarusa from north of Stull eastward, 

 and no traces south of the Wakarusa on this level west of Rock 

 creek. 



