Eskers of the Kanscin Epoch. — Hershey. 248 
For it is difficult to understand liow the streams, after leaving 
fclie tunnels, could have flowed down over these sometimes al- 
most perpendicular slopes without destroying the ridges or 
carrying much material with them. This difficulty, liowever, 
disappears when we remember that all the deposits which we 
have been discussing occur in basins which, being open only 
toward the east, were at that time obstructed by ice, and quite 
naturally we should expect them to be occupied by extra- 
glacial lakes. The streams, on issuing from the tunnels, 
flowed into these lakes, immediately dropping their coarser 
material, but carrying their loess-like silt to a greater distance 
and spreading it over the lake bottom. That the existence of 
these lakes is not merely theoretical is proven by the deposits 
of one of them situated in the Yellow creek valley in western 
Stephenson county, Illinois, and the Pecatonica river valle}' in 
Green county, Wisconsin.* 
The signiflcance of these extra-glacial lakes lies not soleh' 
in the explanation of the abrupt western ends of the gravel 
ridges, but they strongly negative the supposition of a west- 
ward tilting of the district, and so their bearing on the " tun- 
nel and hydrostatic pressure" question is of the greatest im- 
portance. This may be briefly stated as follows: Deposits 
of gravel were eroded by strong currents at a level of 100 feet 
or more below the surface of the lake into which these streams 
flowed, only a few miles distant. Hydrostatic pressure oper- 
ating on a current of water closely confined, as in a tunnel, is 
the only reasonable theory that can be advanced in explana- 
tion of this phenomenon. 
The parallel esker ridges which together constitute the 
larger belts, may probably be explained by supposing that, 
where the stream course was determined by a broad valley, the 
exact situation of the tunnel varied from time to time, perhaps 
*The deposits referred to consist of a loess, averaging a few feet in 
thickness, but increasing in places to eight or ten feet. This is confined 
to such basins as would l)p formed by the closure on the east of certain 
valleys at the time wheu the ice-front stood about twelve miles back or 
east from the outer glacial border. The surface of the loess has been 
formed, during the Aftonian epoch, into a black soil varying from one 
to three feet in thickness. There are also indications of some erosion of 
the loess at this epoch. During the ensuing lovvan epoch, the black 
soil was covered by eight to ten feet of Upland loess, giving the now 
buried black soil an interloessial position. I shall refer to the lower 
loess in Stephenson county as the product of Lake Lena. 
