Drift in Soiitli Dakota. — Todd. 163 
We have not given the subject sufficient study to speak with 
confidence and yet it seems permissible, at least, to offer a few 
suggestions which are little more than speculations. 
If we examine a map of the region, we shall find that the 
valley of the James river in South Dakota is separated from 
the Minnesota river valley by a high table land rising over con- 
siderable of its surface, to a hight of 2,000 feet above the sea. 
Immediately north of the South Dakota line, the James river 
makes its nearest approach to the Red River of the North, the 
distance being about seventy miles. At that point the divide 
between the streams is very low and close to the former stream. 
A cut of twenty to twenty-five feet would probably turn the 
James river into the Wild Rice and Red rivers. From that 
point northward, although the divide is about 150 feet above 
the James, there is no well defined ridge, as further south. 
This brings us to a serious objection to the view that the 
course of the Missouri was down the James river valley, es- 
pecially while the mass of ice was moving up the Red river 
valley. Moreover, some have thought that the ancient drain- 
age was once northeast through the valley of the Wild Rice. 
If such were the case, the difficulties of keeping the James river 
valley open for passage of the water while the Red river vallev 
was occupied by ice, would be still more difficult to explain, if 
not inconceivable. There would be nothing to prevent the ice 
pushing directly into and across the James river vallev. It 
seems, therefore, very probable that the dividing ridge, which 
is now so well defined in South Dakota, previous to the Wis- 
consin epoch, extended further north, possibly as far north as 
Devil's Lake; though it was narrower and probably lower there 
than farther south; and that this ridge played an important 
part in hindering the advance of the ice until it had accumu- 
lated sufficiently to break through into the James river valley, 
as it did during the Wisconsin epoch. This would be the more 
easily explained if the ice sheet from the north, i. e.. from the 
Keewatin center was not so vigorous in the early stages, i. e.. 
in the Kansan and pre-Kansan stages. 
Preparatory to such a break the peripheral waters on the 
west side of the ice-lobe occupying the Red River valley were 
probably first prevented from escaping southward through the 
Minnesota valley by the growth of the ice-sheet and were 
