70 The American Geologist. August, 1892 
advance of the same glacial sheet, requiring a hundred, or two or 
three hundred, ^xars ; and as such if it should be accepted finally 
it will remove from the field of future controversy one of the ex- 
treme views of the shortness of that interval, and will open up a 
more clear, because more restricted, field for further investiga- 
tion. 
In the sixth annual report (1877) of the Minnesota survey (p. 
70) allusion was made to an old gorge in Ramse}^ county, in the 
following words: 
There is some appearance of the former extension of the valley of 
Rice creek much further southward, and it is no unreasonable sugges- 
tion that the great Mississippi itself may have once occupied this valley, 
entering the great gorge again where it becomes remarkably widened, 
at St. Paul; but the evidence is entirely topographical. Such as it is it 
is perhaps overbalanced by a confusion of hills and high drift ridges 
nortb of St. Paul, which render it improbable that the Trenton is any- 
where entirely cut through from the Rice creek valley to St. Paul, as 
would have been the case if the Mississippi ever passed through there. 
Also, on p. 85 of the same report is the following statement: 
There is also a significant change in the direction [of the Mississippi], 
and one the more significant as it seems not to have been due to any 
rock formation existing at St. Paul, but directly contrary to the rock 
sculpturing that exists there favorable to the continuance of the river in 
any pre-occupied valley running in the same direction. Allusion has 
been made to a possible ancient gorge through the Trenton north of St. 
Paul in describing the surface features of the county, but on the geologi- 
cal map of the county no such gorge is represented, because it never has 
actually been discovered, and its hypothetical location would perhaps be 
of no service. * * * * 
These anomalous and significant facts can all be reasonably explained 
on the supposition that the Mississippi river was diverted from its 
ancient valley-gorge, north of St. Paul, by the ice and drift of the first 
glacial epoch, and that it was driven- into that which has been described 
in the report on Hennepin county, toward the west further, and joined 
the Minnesota valley at some point above Fort Snelling, but between 
that point and Shakopee, without passing over or through the Trenton 
limestone at all. Their united waters then formed the river which 
excavated the gorge between Fort Snelling and St. Paul (unless the 
Minnesota alone had already done it) between the first and second glacial 
epochs. 
Since that report was written several results have been attained 
which diminish the obstacles to the hypothesis which then existed. 
The extent and distribution of the moraines of the state have been 
established by more field work. The ancient discharge of the 
