104 The Americaii Geologist. February, 1900 
forced to traverse the divide westward, particularly over its 
narrowest portion northwest of the present lake Traverse. 
During the Kansan and lowan epochs, we may suppose, they 
excavated so wide and deep valleys, that the Wisconsin ad- 
vance found comparatively easy access into the James valley, 
then occupied by the Missouri as already stated. 
From a general consideration of the extent of the so-called 
Kansan till as compared with the Wisconsin, we may infer that 
the natural center during the former stage was further east: 
probably northeast of lake Superior. In fact, we may con- 
ceive that some of the higher points north of lake Huron were 
the first to receive a permanent ice cap. As the region became 
more chilled, the zone of accumulation would naturally extend 
along the more elevated suiface of the ice and then the great- 
est accumulation would lie naturally near the edge of the zone 
and advance slowly outward. In this way, we may perhaps 
account for the greater vigor of the streams passing down 
lake Michigan and lake Superior during the Kansan stage or, 
as some would say, the latter during the Kansan stage and the 
former during the Illinoian stage. If we believe the ice to 
have here pushed forward southwest in the axis of lake 
Superior basin, it is not difificult to conceive that its course 
would lie diagonally across the state of Minnesota, being con- 
fined in a broad, shallow canal between the highlands about 
Itasca and the region of central Wisconsin, that it was directed 
to the Minnesota valley and across it against the high trans- 
verse ridge of the "east coteau", the high divide separating the 
Minnesota from the James, which now has an elevation of 
1,700 to 2,000 feet. From the shape of the land and the course 
of the stream, it seems not unlikely that the highest elevations 
were along the axis of this stream. As the Des Moines valley 
to the south offered an easier slope, we may conceive the ice 
sheet to have expanded more rapidly in that direction and to 
have spread out during the Kansan stage, from that valley 
westward and south into northwestern Missouri. We may 
account for its failure to press westward over into the James 
valley by the elevation of the Coteau region and by the divert- 
ing influence of the Big Sioux valley, which we may suppose 
had greater effect upon the thinner edge of the ice which there 
lav in the zone of ablation. 
