The Develoi^nient of 3Iississippi Valley. — Hershey. 267 
might have maintained its course b}'^ more rapid down-cutting 
of its channel than the crest of the "uplift" was elevated. But 
as such a stream did not exist in the Cretaceous period, it is 
not probable that it should have originated later, flowing from 
a flat lowlying plain, through deep valleys across an "uplift." 
On the contrary, the most natural direction of flowage in such 
a stream as that which originated the cailon valley of the 
Mississippi along the southeastern border of Minnesota, is 
northward. Slightly corroborative evidence of it is furnished 
by the fact that many of the streams on its western side, north 
of the "military ridge" enter it from the southwest, instead 
of the northwest as they should in a M'ell balanced drainage 
system. There may, however, liave been a different cause for 
this. 
It has been shown tliat there is sufHcient evidence of tlie 
comparative recency of the Mississippi river as a large stream 
flowing on the western border of the territory of northwestern 
Illinois, to warrant us in considering its origin as not earlier 
than a late portion of the Ozarkian epoch. This brings it so 
close to the Kansan epoch that it does not require a great 
stretch of the imagination to consider it as having resulted 
from the disturbance of other drainage systems by the accu- 
mulating northern ice. For instance, it is quite possible that 
the Kansan ice-sheet had advanced across the outlet of the 
supposed northwardly flowing ancestor of the upper Missis- 
sippi river, obstructing its flowage, and after the production 
of a great extra-glacial lake, turning the drainage of the en- 
tire region over the lowest point on the divide which inter- 
vened between it and the head waters of the southwardly 
flowing ancestor of the central Mississippi river, long before 
it glaciated the country south of the " driftless area." 
The present bluffs of the Mississippi canon valley have 
been eroded by the "larger Mississippi," and have an appear- 
ance of greater youth than those of its tributary oaiion val- 
leys. This is finely illustrated on the east side of the valley 
near Dubuque. The bluff of the great caiion is straight, and 
often precipitous, always rocky, and nearly free from timber. 
The bluffs of the small tributary valleys are timbered slopes, 
always steep, but rarely bare precipices. Southward from 
Dubuque, the canon valley has been widened in places so as 
