Woodworth.l 210 [Feb. 7, 
origin of the Coiupoiiiice Pond esker. The conditions are essen- 
tiall}- repeated in the eskers on the Burrillville, \i. I., sheet, at 
the western base of \yhortleberry Hill in Sniitlitield. 
Ice-tunnels in which the glacial drainage is piped through the 
ice have been ol)served by Russell^ on the Lucia Glacier in Alaska. 
Deposits made in such a channel would naturally be limited by 
the arch so long as this remained intact. The dimensions of the 
tunnel described by Russell accord closely with the cross-section 
of some eskers. There arises, however, the difficulty of getting 
a deposit made in this situation upon the ground without destroy- 
ing that uniformity of outhne which distinguishes often consider- 
able segments of some eskers. While the ice-tunnel shows what 
limits the cross-sections of eskers, it seems doubtful if it will 
explain many of the topographic features of eskers in this district. 
The ice-arch tends to limit the cross-section of eskers to 
narrow ridges. The slopes of eskers are similar to those of 
moraine terraces, and like them indicate the deposition of the 
mass whence they have slidden upon the terrane they now 
encumber. Moreover this theory of the origin of eskers has the 
advantage of simplicity in that it supposes eskei's to have origi- 
nated where they are now found. The burden of proof i*ests, it 
seems to me, on him who would show that they were formed 
above the surface they now rest upon. 
If under a given deiJth of ice more than a given width and 
height of arch cannot be maintained, any enlargement of the 
channel will be corrected by the creeping in of the ice or the 
deflection of the sides and roof.'^ In this way it seems possible to 
see why eskers should be limited to forms of tolerably uniform 
cross-section. 
Terminals of Eskers. 
Eskers tend to be associated with other de|)Osits of water-laid 
drift, the lateral distribution of which in valleys is frequently 
synmietrical, as in moraine terraces and hillside kames, with 
' Nat. geog. mag., v. 8, p. 105-108, pis. ^12-15, 1891. 
2 Wliere the subglacial waters were under hydrostafu pressure, it is possible some 
enlargement may have been permitted. See Professor Slialer's )iaper"The eonditions 
of erosion beneath deep glaciers," etc. Bull. nius. eomp. zool.. \ . IG, no. 11, p. ^OS-'iOD, 
1893. 
