The Upland Loess of Missouri. — Hcrshey. 371 
stroyed both, it and the loess. The Missouri and Mississippi 
valleys must have been excavated to equally as great depth 
to form an outlet for the waters of the tributary valleys, and 
hence no such rock barrier as claimed could have existed in 
the lowan epoch. 
The "creeping" hypothesis might be pertinent in a mountain 
region where slopes are long and steep, but it can have a very 
limited application in such a country as the Mississippi basin. 
Certainly, in northern Missouri it cannot be charged with the 
task of transporting the loess down the slopes from the up- 
lands so generally that a larger part of the surface of the drift 
is obscured by it. However, there is positive evidence against 
it. When superficial material is sliding on a slope, there is 
a constant working of the under portion outward and down- 
ward, a kind of kneading process, which thoroughly mixes the 
Txiass. In the case under discussion, some of the drift under 
the loess should have become intermingled with the fine silt, 
which has not generally occurred. The reddish brown 
weathered layer is invariably in place under the loess in the 
valleys, and on their slopes shows no tendency to penetrate 
the latter. 
In the valley of the BonneFemme river which flows into 
the Missouri from the north near Boonville, the loess of the 
uplands descends the slopes and form terraces in the valley. 
These terraces of thick loess are remnants of a flat loess-plain 
which formed the valley-floor at the end of the loess-deposit- 
ing period. This great body of loess has not crept down the 
slight slopes from the neighboring low uplands, but is, with- 
out doubt, in place in its original condition. These terraces 
are finely developed at Fayette in. I believe, Howard county. 
They occur in less characteristic form in many other valleys 
of northern Missouri. They are the equivalents of the loess 
terraces in the Pecatonica and other valleys in northwestern 
and west-central Illinois. Indeed, there is a close parallelism 
between the distribution and development of the loess in these 
latter districts and in northern Missouri. In Illinois, the 
occurrence of the loess in situ in the valleys has never been 
questioned, and had the Quaternary geology of northern Mis- 
souri been more thoroughly studied, it never would have in 
that region also. 
