KANSAS GLACIATION 503 



to 20 feet deep lying similarly on interglacial beds which include a thin 

 layer of peat with decaying fragments of wood next beneath the till. 

 More distant northwestward, at Barnesville, in Clay County, on the 

 southern part of the area of the Glacial Lake Agassiz, a well penetrated 

 through 12 feet of till into quicksand "containing several branches and 

 trunks of trees, thought to be tamarack, up to eight inches in diameter, 

 which lie across the well, and which, together with the inflow of water, 

 prevented further digging." This well is in the till area of the village 

 of Barnesville and about 80 feet below the highest beach of Lake Agassiz, 

 which passes from south to north about three miles east of this village. 



In 1888, at the time of my survey of that region, I considered the 

 occurrence of this interglacial bed within the area of the glacial lake 

 as good evidence that the ice-sheet had been melted back at least far 

 enough to give outflow into Hudson Bay from the Red Eiver valley, 

 thus draining off the forerunner of Lake Agassiz. Subsequent studies, 

 however, lead me to hold now the different view that probably an inter- 

 glacial lake here may have cut its southern outlet, at the site of Brown's 

 Valley, to a lower level than the well at Barnesville ; or that the attitude 

 of the land was unlike what it is now, having then such an ascent from 

 south to north that the Barnesville locality during the interglacial time 

 was above the general surface of the region at Brown's Valley, into 

 which the River Warren, outflowing from Lake Agassiz, cut its deep, 

 continuous valley. So I now think that we have in the section of this 

 well only evidence of an ice-retreat (that is, a departure of the outer- 

 part of the ice-sheet) northward a little more than halfway between the 

 south and north boundaries of Minnesota. 



On the extensive area of the three chains of lakes and at the more north- 

 ern places where the thickness of drift above the old interglacial land 

 surface is known, the Kansan, Iowan, and Wisconsin glaciation, to the 

 end of the Ice Age, effected little or no erosion and made only scanty 

 deposits of till beneath the ice-sheet. Each of the three recorded sec- 

 tions, in New Ulm, Rushseba, and Barnesville, appears to have received 

 its upper till, lying on the interglacial beds, from the drift gathered up 

 into the ice-sheet, borne along in its basal portion, and laid down when 

 the ice in which it had been held was melted away. This englacial drift 

 appears, therefore, to have been equivalent generally to a deposit 10 to 20 

 feet thick. 



The Yarmouth intekglacial Stage 



During its maximum Kansan extension the east margin of the Kee- 

 watin icefield probably coincided nearly with the present course of the 



