424 G. K. Warren— Valley of the Minnesota and Mississippi. 
the great mass of hard rocks and other material brought from 
regions far to the north. There seems a probability that much 
of the present Upper Mississippi basin had previously been for 
long ages exposed only to erosions of streams and of the atmos- 
phere, so that it was probably much cut up and fissured, as we 
see in regions farther west, where no glacial action has occurred. 
t must have been an easy matter then for the glacier to have 
thoroughly filled up all the valleys and ravines, leaving the 
surface everywhere of the well-known rounded hill and basin 
forms of the drift regions. Wherever the glacial scratchings are 
preserved, their uniform directions indicate a massive move- 
ment to the southwest quite independent of all influence o 
underlying inequalities. The water which flowed from them 
would seek the first lowest line and excavate its course without 
regard to the nature of the older stratified rocks buried beneath 
the glacial deposits, and such seems to have been the case, for 
the valley takes a great variety of courses, running about north- 
east at St. Paul, due west at Rock Island, and its direction’ fill 
every azimuth in different parts from northeast around by sout 
to west. To the old stratified rocks its course seems to have 
no relation, now cutting across an anticlinal, then following the 
strike in one direction and again in the opposite one. wee 
How the valley was formed.—At St. Paul, on the Mississipp!, 
and in the Minnesota above, are the banks of an ancient water- 
course when at such higher level than now that the river-bed 
was the magnesian limestone rock, the same as that of the Mis- 
sissippi, just above the Falls of St. Anthony. The existing 
channel of the ancient valley has probably been formed by a 
cataract in the great river, similar to that at St. Anthony. This 
view is sustained by the high islands of rock in the valley of 
the Minnesota, being remains of strata once continuous across 
it. These high islands also exist below in the Mississippi, such 
as Barn Bluff, at the head of Lake Pepin, and the Trempeleau 
hills. Some of these detached bluffs may have been formed 
by bends approaching each other by erosions gradually forming 
a neck and cutting it off. One such, nearly completed, is seen 
in the Dalles of the Chippewa River. The period which must 
have elapsed in doing this work was long, but it is probable 
that the volume of water, during the melting of the glaciers 
north of it, was tly in excess over that of the present drain- 
age of the Winnipeg basin. The period may have been some- 
what shortened by the new watercourse regaining in places 
some ancient one, filled only with glacial débris. 
If we look at the valley shown on the map from Lake Trav- 
erse to Rock Island, we see that it gradually widens and con- 
tracts along its course, but, as a whole, widens as we descend. 
It widens where the rocks on the banks are soft, and narrows 
where they are harder and capable of resisting atmospheric ero- 
