SURFACE GEOLOGY. 205 
filling of the old channels of the Mississippi and the Ohio, 
as a depression of five hundred feet would bring the ocean 
nearly to Pittsburgh on the Ohio, to St. Paul on the Missis- 
sippi. 
But I think we have evidence that the continent did not 
sink uniformly in all its parts, but most at the North. Not 
to cite any other proof of this,—northern coast fiords, etc. 
—the altitude of the loess-like deposits of the upper Missis- 
sippi and Missouri (the lacustrine non-glacial sediments of 
this period of submergence), the upward reach of the Drift 
clays of the lake basin, the filling of the valleys of the 
streams flowing into the Ohio and Lake Erie, the old lake 
beaches marking the former water-level in the lake basin — 
all indicate that the continental subsidence was greatest to- 
wards the north. To this subsidence we must, as I think, 
attribute the accumulation of water in the lake basin and 
Mississippi valley to form the great inland sea of fresh-water, 
of which traces everywhere abound. It seems to me scarcely 
necessary to suppose any other barriers by which this sea 
was enclosed than the highlands that encircle it— such as are 
roughly outlined by the light tint on Professor Guyot’s map 
of North America—and the sea-water which filled the 
mouths of the two* straits by which it communicated with 
the ocean. 
Yellow Sands and Surface Boulders. I have mentioned 
that on the Erie clays are beds of gravel, sand, and clay, 
and over these again great numbers of transported boulders, 
often of large size and of northern and remote origin. 
These surface deposits have been frequently referred to as 
the direct and normal product of glacial action, the materials 
torn up and scraped off by the great ice pinen in their 
* Tf pated bat — That there ` was ome in the course of the Mississippi we oo 
nd that so lo the oth 
The eastern | sie of the lake waters may not have been by the St. saat but 
as likely through the gap between the Adirondacks and the Alleghanies. The shallow 
nnels betwee vei Thousand Islands and v" uri ent seem to indicate that 
for the lakes. 
E 
