Geological Papers. 215 



In favor of the second theory, the size of some of the boulders, 

 viz., five to six feet in diameter, would be some evidence of glacier 

 action, rather than water, but the objection so strong against the 

 former view would be quite as cogent here. Besides, the difficulty 

 of conceiving the Kansan stage to linger long enough to erode a 

 deep valley and partially fill it again is quite serious. 



The third view harmonizes well with the possible arrangement 

 of the patches in channel-like lines, with their being on lower 

 levels than others near by which are bare of drift, and it also 

 evades easily the difficulty of the patches being so far from other 

 clear traces of glacial occupation. But it does not avoid the great 

 erosion demanded during the Kansan ice stage. Some will find 

 serious difficulty in accounting for transportation of such large 

 and so many boulders by water. This may be obviated, however, 

 by remembering the efficiency of river ice. Anchor ice may have 

 been more efficient at that time, also, and ice blocks from the glacier 

 front may have assisted. 



The fourth explanation is favored by all that favors the last, and 

 it reduces indefinitely the demand for so long a time for erosion 

 during the Kansan stage, for there would be in this case only need 

 for the cutting down of narrow divides under favorable conditions. 

 Another diifference would be that it conceives the distribution of 

 boulders by lake ice more than by river ice, though both are in- 

 volved more or less in each explanation. The conception is that 

 instead of the single valley of the Wakarusa, there were two or 

 three tributaries running northeast into the Kansas between To- 

 peka and Eudora. The Oread limestone formed the divide between 

 the eastern one and the one next west, which, as has been already 

 stated, probably followed the line of Rock, Coon and Oakley creeks 

 northward. There may have been another divide corresponding to 

 the Topeka limestone crossing the valley somewhere near Rich- 

 land. According to this view the coming of the ice blocked the 

 lower courses of these tributaries, changing the one west of the 

 Oread divide into a lake, in which for some time ice blocks from 

 the glacier front floated, distributing boulders. Eventually the 

 waters rose and broke over the divide eastward. The fall was at 

 first considerable, for the stream then occupying the present lower 

 course of the Wakarusa was not obstructed, but ran on the pre- 

 glacial level of the Kansas, which was only about 100 feet above 

 the present stream. The Three Sisters and other boulder-capped 

 hills east were probably portions of the old channel of that time. 



3. The stratified chert gravels were indefinitely older and of 



