T8 The American Geologist. Auguf-t, 1592 
centers of atmospheric disturbance were doubtless in some in- 
stances wafted far eastward, and even northward, shedding on 
the plains more moisture than they receive now. The present 
size of the interglacial gorge being somewhat larger than the post- 
glacial gorge, even after making allowances for some enlarge- 
ment by the disintegrating action of the last glacial epoch, and 
there being some reason to think the western country was wet in 
many places where now it is dry,* it is reasonable to infer that 
the river was permanently sustained at a higher stage by a greater 
precipitation. If this be allowed we shall have to admit that it 
would act more powerfully at the brink of a receding waterfall, 
and that the recession of such water-fall, for any given length of 
time, all other things being equal, would have been more rapid in 
interglacial time than it has been in post-glacial time. ^J'his dif- 
ference may have amounted to twentj^-five per centum of the post- 
glacial recession. 
In regard to the point of debouchure of the interglacial Missis- 
s,ippi into a pre-existing gorge, giving rise to a water-fall at the 
commencement of interglacial time, there are such uncertainties 
that it will not be possible to determine whether it was at St. Paul, 
opposite Dayton's bluff, or at a point in the Minnesota valley a 
few miles above Fort Snelling. This all depends on how old that 
portion of the Minnesota (now the Mississippi) valley is, between 
St. Paul and the supposed point of junction above Fort Snelling. 
There is some reason to suppose that the Minnesota river did not 
always unite with the Mississippi where it does now, nor at any 
point in the vicinity. The Minnesota itself, at Mankato, has a 
significant right-angled elbow, and at other points further north it 
shows signs, especially in glacial and post-glacial time, of having 
taken a course to the Mississippi across Rice, Dakota and Good- 
hue counties. At earlier dates its waters may have found passage 
to the sea toward the southwest and the Missouri river, or through 
some of those deep rock-cut gorges which characterize the Undine 
region in Blue Earth county, thus having had some agency in the 
pre-glacial sculpturing of those canons in which lie the curious 
'•'chains of lakes" which Mr. Upham has described in Faribault 
county. 
*Tbis difference is indicated by the existence of interglacial peat beds 
in southern Minnesota, and the discovery of many trees like cedars 
whose habitat is usually in swamps, in the upper till as well as in the 
peat beds. 
