Distribution of the Loess. — Wright. 213 
bia without its being due to some special cause. There is no 
reason why, in the ordinary advance of an ice sheet, such bould- 
ers should be accumulated as they were in the Osage valley. 
Furthermore, it is clear, from the position of these bould- 
ers, that they do not belong to the Kansan epoch of the glacial 
period. The sedimentary terrace upon which some of them are 
situated very clearly belongs to a late period in the order of gla- 
cial events. 
3. A third possibility is that the Kansan ice-sheet had ad- 
vanced beyond Topeka so as to extend over the watershed be- 
tween the Kansas and the Descygnes river, or perhaps that 
there had been a channel of overflow from a temporary Glacial 
lake in that region ; so that ice floes bearing these boulders were 
carried into the head-waters of the Osage, and floated down 200 
miles or more to their present position. But of such advance of 
the ice across the Kansas river there is no evidence; while ev- 
erything bears in the strongest manner against the supposition, 
though there are some rather indefinite reports of Glacial over- 
wash south of the Kansas river basin. 
It was reported to me, for example, that south of Kansas 
City, at Pleasant Hill, on one of the small tributaries coming 
into the Osage, a few granitic pebbles had been found. It had 
been reported, also, to professor Haworth that a rounded gran- 
itic pebble several inches in diameter had been found in Coffey 
county, 50 or 60 miles south of Topeka. This would be in the 
drainage of the Arkansas river. A personal examination, how- 
ever, of the watershed between the Kansas and the Osage river 
in the most likely place for an overflow, 20 miles south of Law- 
rence, as well as of the deposits in the Osage river from Otta- 
wa to Osawatomie, and on its northern branch at Paoli, dem- 
onstrated to my own mind about as well as negative evidence 
can, that there had been no extensive overflow from the Kan- 
san extension of the ice-sheet into the headwaters of the Osage 
river. 
4. We are therefore limited to the fourth hypothesis, name- 
ly, that the boulders at Tuscumbia were transported upon ice 
floes 64 miles up the Osage river by back water produced when 
the Missouri was subject, in the closing stages of the Glacial 
period, to annual floods of great extent from causes which did 
not affect its southern tributaries. The existence of such favor- 
