796 T. C. CHAMBERLIN 
increases. In some instances the loess mantle rises to the divide 
and connects with the similar deposit of an adjacent valley, but 
the law of progressive fineness and thinness still holds. This 
relationship is such as to create a very strong conviction that 
the deposit of the loess was in some vital way connected with the 
great streams of the region. 
2. The second significant feature is the distribution of the 
loess along the border of the former ice-sheet at the stage now 
known as the Iowan. (Strictly speaking there was more than 
one stage of loess formation, but for convenience only the main 
stage will be here discussed.) The elaborate paper of McGee 
made us familiar some years ago with this relationship in eastern 
Iowa. The studies of Calvin and his colleagues, Bain, Beyer, and 
Norton, of the Iowa Survey, of Winchell and Upham of the Min- 
nesota Survey, of Todd of the South Dakota Survey, and of Salis- 
bury, Leverett, Udden, Buell, Hershey, and the writer of the 
United States Survey, have greatly extended the evidence of this 
relationship. It has recently been much advanced by the Iowa 
geologists and by Leverett and Hershey in northwestern IIlinois. 
Next the border of the ice-sheet the loess is thick and typical, 
but graduates away with increasing distance from the ice border 
ina manner similar tothe graduation away from the river valleys. 
On the border next the ice there are developed the formations 
designated by McGee paha, elongated domes of quasi-drumloidal 
_ contours which are mantled by loess. This superficial loess grad- 
uates downwards into loess of coarser and coarser texture until 
it often passes into a nucleus of sand. Below this there is often 
an embossment of till. These pahas seem to be ice border phe- 
nomena. Whatever their special mode of formation their distri- 
bution seems to connect them in some more or less direct genetic 
relationship with the ice. 
It has been affirmed by several independent observers that 
the loess graduates into glacial clays and glacial till and this 
relationship further tends to confirm the association of the loess 
with glacial action. 
It has been shown by the microscopical examinations of 
