664 THE JOURNAL OF GEOLOGY. 
taceous,—the reference which Professor Winchell has suggested 
for the high-level preglacial gravels of southern Minnesota. 
Assuming that the Minnesota gravel referred to is Cretaceous, 
and comparing the Devil’s Lake gravel with that of Minnesota 
alone, this would perhaps seem to be the niost natural reference 
of the Wisconsin beds. 
It is not beyond the range of possibility that some of the 
gravel beds here mentioned are Cretaceous, while others are 
Tertiary. The gravel beds of Minnesota and Crawford county, 
Wisconsin, resemble each other much more closely than either 
resembles the Devil’s Lake bluff gravel. The Devil’s Lake 
gravel, on the other hand, ts strikingly like that of the areas 
further south, and on the ground of physical likeness would be 
classed with it, rather than with the beds of Crawford county, 
Wisconsin, and of southeastern Minnesota. It is distinctly 
recognized, however, that physical likeness is not a safe criterion 
from which to draw important conclusions. Both in Crawford 
county and at Devil’s Lake, the gravel is unaccompanied by the 
clay and sand beds which are present in Minnesota, and on the 
basis of which the Cretaceous correlation was suggested. At 
various points in western and southern IJlinois, on the other hand, 
clay exists immediately beneath the gravel; the same is true in 
some parts of Missouri. So far as known, no determination of 
the age of the clay of these regions has been made, nor is it 
known that the clay and the overlying gravel belong to the same 
period. So far as association with clay is concerned, the gravel 
of southeastern Minnesota would seem to belong with the gravel 
of western Illinois (Pike and Adams counties). Manifestly little 
weight can be attached to this association, on the basis of present 
knowledge. 
There is some reason for believing that there may be both 
late Tertiary, and older (possibly Cretaceous) gravel formations 
south of the Ohio River. The gravel between the Cumberland 
and the Tennessee Rivers, exposed in the deep railway cut, may 
represent a formation much older than the gravel of Adams 
county, Illinois, or of Crowley’s ridge, Arkansas. It is not to be 
