- s< » The American Geologist. November, 1893 
cessively by the head-waters of the rivers above named and 
within a few miles of them; and it is important to notice 
that it continues so close to the western limit of its drainage 
area that it has no tributaries of importance on that side for 
considerably more than half its length. If the writer's 
theory be correct, this peculiarity is due to the glacial condi- 
tions above described. When the retreating front of the 
glacier passed to the northeast side of the water-shed, the ac- 
cumulating waters found a way of escape by flowing to the 
north along' the ice-front before rising high enough to cross 
the water-shed. Flanked on one side by the elevated land 
and on the other by the wasting glacier which supplied its 
waters, the river pursued its northern course until the point 
we have designated as the great bend was reached. These 
conditions existed long enough to permit the stream to 
establish its present channel. 
An interesting feature of this part of the river's course is 
the deep gorge cut through the Carboniferous rocks at Grand 
Ledge. At this point the rocks are at or near the surface 
over a considerable area, across which the river has eroded 
its channel to a depth of about fifty feet. The writer has ex- 
amined this gorge somewhat minutely, with reference to the 
light it might throw on certain glacial problems. We believe 
that it testifies to the relative briefness of the period since 
the close of the glacial epoch. The river occupies no valle}- 
proper. The country adjoining the gorge is comparatively 
level, and the traveler reaches the brink of the cliff almost 
before he is aware of it. Though the sandstone strata are 
not especially hard, there has been practically no recession of 
the side walls of the gorge. Their base is almost at the 
water's edge, with only space for a narrow footpath along its 
margin. At some points one can almost hurl a stone from 
cliff to cliff. Though there are no data present on which to 
base an estimate of the age of this gorge, the conditions 
above described indicate its geologic briefness. It is further 
interesting to note that there were probably steep rapids at 
this place in early time. The distance from the deepest part 
of the gorge to the point down the stream where the rocks 
finally disappear under the drift is less than a mile. When 
the stream first began flowing over the rock area a channel 
