iS94-] 209 [Woodwotth. 
of adjustment to gravity. It is difficult to see how the deposits 
of a superglacial stream bed could be so held in the ice ns to come 
to rest upon the ground as a distinct ridge. 
Ice-caiions do not explain the uniformity of cross-section of 
eskers. I have ah'eady referred to deposits which ajipear to have 
originated in frontal superglacial streams which have cut down to 
the base of the ice and filled their channels. Some eskers occur 
in positions where it is possible to see that an ice-caiion would be 
at a given stage in the melting of the ice-sheet an impossibility. 
An esker occurring at the l)ase of a high and steep valley-wall, 
as in the case of the Compounce Pond esker near Bristol, Conn., 
is a case in point. The site of the present pond marks an ice- 
remnant lying between the esker and the eastern base of South 
Mt. The geological conditions at this locality are represented 
Fig. 4. Generalized section across the Triassic Valley of Connecticut 
showing position of Corapounce Pond and esker. A, the Pond ; b, tiie 
eslver; c, stratified drift mostly extraglacial ; d, till-covered uplands of 
crystalline Pre-triassicand Triassic rocks; e, ice-cafion, in this paper held 
to be untenable; f, ice-mass between esker channel and the escarpment on 
the west: f, eastward extension of ice-sheet; g, g', successive stages of 
lowering ice-sheet. 
by the adjoined cross-section from west to east. Unless the ice- 
sheet were exceedingly thin when the esker was deposited, it is 
difficult to see how a cahon could be kept open with a wedge ot 
ice resting against the sloi)e of the mountain. The conditions 
are such that the inverted wedge of ice (f) would have slidden 
down the slope and eastwardly against the larger mass of ice (f), 
with the result that the canon would close and remain so, until 
the ice was too thin to obey the local slope. It is equally difficult 
to see why, if the esker were the product of a sujierglacial stream, 
the ridge should not have been formed next the valley-wall, 
without the intervening muss of ice. The drainage would have 
been lateral, with ice on one side, rock on the other. These 
facts seem explainable only on the sup-position of a subglacial 
PROCEEDINGS i;. s. N. H. Vol. xxvr U .\l'i;ir. ]8!U 
