266 The American Geologist. October, imi? 
Returning now again to Dubuque, we will look at the etifion 
valle_y. Including the buried portion, it has a depth of about 
800 I'eet and a width at the citj of about one mile, averaging- 
one and one-half to two miles fr()n) here southward as long as 
it remains on the Galena terrane. Comparing with the canon 
valleys of the Pecatonica and Rock rivers where they are 
trenched into Galena limestone it is seen to be somewhat larg- 
er than that of the former stream but decidedly less in width 
and slightly less in depth than tliat of the Rock river. Even 
adopting a larger factor of safety we are compelled to conclude 
that the canon valley of the Mississippi, which undoubtedly 
may be included in the same system as those of its tributaries, 
was excavated mainly by a stream decidedly smaller than that 
which now occupies it, Therefore, the inference seems reason- 
able that the origin of the Mississippi river as a large stream 
dowing i)ast Dubuque did not precede the close of the Ozarki- 
an epoch. It is also evident that a divide must have existed 
across the present course, not man}' scores of miles above Du- 
buque. I have not the data for locating even approximately 
the situation of this divide, but I would suggest that local 
observers search for it somewhere between La Crosse and Prai- 
rie du Chien, particularly where the "military ridge" is trav- 
ersed by the present river. 
Northward from this supposed divide, the present canon val- 
ley of the Mississippi must have been occupied by a stream 
flowing toward central Minnesota instead of away from it. 
Theoretically this is a strong probability for the following 
reason : Sometime in the later portion of the Cretaceous per- 
iod the Minnesota depression was occupied by the sea. The 
drainage of the land on its eastern side must have been from 
the most distant interior parts directly to and into this sea. 
It is impossible that a southwardly flowing stream in the course 
of the present upper Mississippi could then have existed. Sub- 
sequently, when the sea retreated and the early Tertiary up- 
lift occurred, the water level in the vicinity of the site of St. 
Paul, was not seriously interfered with. But an "uplift" was 
developed toward, the southeast, reaching a maximum near 
Dubuque and thence northeastward into central Wisconsin. 
Had a southeastwardly flowing stream existed on the course 
of the Mississippi previous to the institution of the uplift, it 
