240 The American Geologist. April, 1897 
surracoin deep channels until they encountered a crevasse or a 
nioulin, when they descended to a position at or near to the 
base of the glacier, and thence they continued on their way to 
the border. This latter portion of the course was in a tunnel, 
which, near its western end, came down so low as to encounter 
the surface of rock ridges, and in at least one case reached 
deposits that probably lie buried under the present valley 
bottom. As the season progressed and melting increased, the 
tunnels were not large enough to freely carry off the drainage, 
which caused a ponding of the water in the crevasses of the 
ice at some distance from its border. The pressure of this 
ponded water for(;ed the subglacial streams through the tun- 
nels with great force, and their powerful currents eroded the 
rock ridges and other deposits which they encountered, 
sweeping them forward and suddenly dropping them at or 
near the mouth of the tunnel, where the pressure was removed. 
This is the groundwork of the theory, but many of the details 
remain to be filled in. 
The deposition of the stratified gravel and sand beds be- 
longed to the general recession of the ice-front, and rarely ever 
to its advance. Their interior shows that, except in very few 
cases, they were but little disturbed by the forw^ard movement 
of the ice. The steepness of the knolls and ridges, demonstral)!}^ 
largely original, indicates that they were not generally over- 
ridden by it. They are an upper member of the drift-sheet, 
— one of the last deposits which the ice made before abandon- 
ing the district. The erratics which constituted the englacial 
drift are frequently found upon the esker ridges, or incorpo- 
rated with the upper portion of their gravel, corroborating the 
idea that they were formed in channels arched over with ice. 
But, except in a few cases, these gravel ridges are not overlain 
by the ground moraine. One of the exceptional cases occurs 
four miles south of Freeport, where a prominent gravel knoll has 
been mantled with six feet of typical till, forming an elongat- 
ed prominence with the distinctive drumloid topography. In 
a cutting on the Illinois Central railroad, two and a half 
miles southeast of Freeport, a wedge of the boulder clay or 
ground morraine passes into the side of a stratified gravel 
ridge, plainly indicating ice-movement during the formation 
of the ridge. Just enough of such evidence occurs to prove 
