500 PROCEEDINGS OF SCIENTIFIC SOCIETIES. 
















water from the glacier, and this limit was the Missouri river, and that 
was the river formed thereby. It cut along this glacial limit because all 
the streams west of it came from the mountains toward it, and there their 
old course was terminated. We see what lakes must periodically have 
n the glaciers began to retire to the north-east, as long as the general 
slope of the continent was towards the glacial mass, successive rivers 
were marked out by it along its western face, and all have a parallel- 
ism, and are close to each other, and have short tributaries or parallel 
branches if any. They are, besides minor streams, the James river, the 
ig Sioux river, the Des Moines river, the Iowa and Cedar rivers, and 
run directly from it, and such we see is the direction of all the tributaries 
of the Mississippi on the east side. This direction corresponds with that 
of the preglacial rivers, and it is probable that some of them, es. as the 
St. Croix river, Chippeway river, Wisconsin river, etc., byes 
regained their old beds, for so their appearance cue indic 
The bend of the great valley along the Minnesota river, tal Man- 
kato and St. Paul, being at right angles to the main course of all these paral- 
lel glacial rivers, would seem to disprove the view that the formation of 
this river took place along the glacial margin. But it probably is the bed 
of one of these preglacial rivers, as it lies in the proper fold of the Silurian 
rocks and seems to have been formed in an ancient valle Ema % : 
southward, however, the present course of the Mississippi is cu 
across this fold in the rocks, and the glacial action is the only ae 
of it. 
The manner in which the glacial action produced these valleys was not 
by abrading the strata with the grinding power of rocks embedded in the 
ice, but after the manner in which a block of marble is sawed. e gla 
cier supplied an immense power in the der water, and into this water 
it was constantly dropping Apa rocks and s 
The waters issuing from a lake have Se abrading power, for they 
have comparatively little ie material to operate with. Thus it wi 
that the waters issuing at the old southern outlet of Lake Mba cou 
have drained it, nor would the banks then have been as high as bove the 
as agape at the Bijou hills, which are eight hundred feet a ete : 
of the river. The slope of the Missouri is more than dou 

