Hirer Vallejis of the Ozark Plafean. — Ilershcji. 353 
the south slope of the uplift, we find a precisely similar de- 
posit beginning to appear at intervals along the banks of the 
streams, growing more and more distinct from the 1,000-foot 
level down, forming low terraces, and finall}^ as I have been 
informed, connecting with the undoubted Columbia deposits 
along the borders of the Mississippi embayment region. 
This deposit is invariably a reddish brown, light-brown, or 
sometimes butf-colored, sandy clay, nearly free from pebbles 
except in its lower portion. Fossils occur scattered through 
the deposit, mostly terrestrial species, generally land snails. 
It is usually semi-massive, although occasionally Avell strati- 
fied. It occurs on the sides of the lower troughs or present 
stream vallej^s as remnants of a formation M-hich once com- 
pletely filled the valleys to the level of the present uppermost 
deposits, but which has been almost completely removed. Al- 
though falling far short of the Lafayette level, this Columbia 
level is much above the present highest flood-plain deposits. 
This formation was of fluvial origin, when the streams ran at 
a much higher level than they ever succeed in reaching at the 
present time. The streams, however had already cut their val- 
leys to nearly their present depth, so that their Columbia 
high level was produced either by excessive rainfall or bj' a 
subsidence of the land with the consequent general raising of 
the water level. The former hypothesis is disposed of b}' the 
nature of the deposits, which contain few pebbles and indicate 
deposition in sluggish currents. Hence it was a raising of 
the water level consequent on a subsidence of the entire Ozark 
plateau, which caused a flooding of the streams by great 
diminution of the drainage gradient, and resulted in the de- 
position of the Columbia clays in the valleys of the region. 
The color and texture of the deposit indicate either a small 
amount of vegetation, insuttlcient to form a black soil, or a 
rather rapid gathering of the material frf)m tlie subsoil clay 
of the ridges b}'^ increasing precipitation of rain or snow. 
Probably both these conditions were present togethci-. and 
were caused by the proximity to a vast ice sheet on the n<u-th. 
the lowan glacial stage or epoch being then at its climax. 
The synchronism of these claj'^s with tiie Columbia dei)osits 
of the lower Mississippi and Missouri region seems reason- 
ably certain, for not only istlierc apparent continuity between 
