Histonj of Kaic Lake. 193 



and it may have begun before the ice advanced. Possibly some 

 of the down-cutting may have preceded the influence of the 

 ice, but the more plausible explanation of the facts may be 

 ascribed to the flooding of streams from the melting of the ice 

 sheet as it advanced from the north. The first to enter the 

 basin of Kaw Lake, we figure, was from the Big Blue, and 

 when the ice front had reached Pawnee county, Nebraska, or 

 later, the northern part of Nemaha county, Kansas, Rock 

 creek would have begun its share of the work when the ice had 

 reached northern Pottawatomie county, twenty-five miles 

 away. It may be that Vermilion river may have come into the 

 basin and begun similar work at a considerable earlier date, 

 but we have no evidence of it. Meanwhile tributaries of the 

 Kansas further east would have caused similar erosion of the 

 trough of Kansas river further east, in its lower course, so 

 that before the ice had reached its valley we may believe that 

 the down-cutting may have been going on quite rapidly. 

 Velocity of the ice and rapidity of melting would have accel- 

 erated erosion, and the length of time it took for the ice to 

 advance from the divide to the channel of the river would de- 

 'termine the amount of excavation. 



3. The Formation of Kaw Lake. 



^^'e do not know the rate at which the ice advanced nor what 

 fluctuations it experienced, but we have found no trace of its 

 edge pausing anywhere long enough to form a distinct mo- 

 raine, or rows of drumlins. 



We judge, from the distribution of bowlders, that when at 

 the maximum extent the edge of the ice lay from the junction 

 of Camp and Rock creeks southeast across the river and over 

 the divide south into the valley of Mill creek, and thence east- 

 ward, with its southern edge five to ten miles south of the 

 river, and recrossing to the north side near Lecompton, 



The evidence for this is the abundance of bowlders scat- 

 tered north of that line on higher points, and traces of 

 bowlder-strewn channels leading away south of that line, ac- 

 cording to topography. It has not been fully explored, only 

 from Lawrence to Dover, and from Paxico to McFarland. 



The Kansas valley was not so deep as now, by 75 or 80 feet. 

 We can imagine that the river struggled with the ice for the 

 possession of its channel, probably forming rapids and erod- 



13— Sci. Acad.— 2163 



