1891 -J 
241 
[Upham 
The idea that drainage from the waning ice-fields would be 
subglacial is most naturally suggestedby the river born at the foot 
of the Alpine glacier, and by the subglacial rivers of the much 
larger Alaskan glaciers and of the Greenland ice-sheet wherever 
its outlets of ice end, whether in the sea or above it. But the 
only large glacier known which has a drift-covered surface, simi- 
lar to that of the ice-sheets of the Glacial period during the 
latest stage of their melting, is the Malaspina glacier, where Rus- 
sell reports large rivers flowing partly in open channels and often 
with a considerable thickness of ice beneath them. In the sum- 
mers of the Champlain epoch the progress of ice-melting was very 
rapid, as is known by the rate of deposition of the modified drift 
enclosing lakes ; and for the reasons before stated I believe that 
then the drainage from the ice-sheet was mostly in channels on its 
surface, there depositing the osars, or eskers, and kames, while 
much of the finer gravel and sand was spread just outside the 
mouths of these glacial streams, where they descended to the 
land. 
If the osars were subglacial, we should expect them to be often 
covered wholly or partly with the englacial drift, as boulders and 
loose deposits of till, which would be permitted to fall upon them 
when their ice-roof was melted away. Such a roof would be 
more or less overspread with the drift that had been contained in 
the higher portions of the ice-sheet and was exposed on its sur- 
face by ablation. Sections indeed are occasionally found, where 
subglacial beds of modified drift have become covered by subgla- 
cial and englacial till but these usually differ widely in their 
character from the torrential osar and kanie deposits, which very 
rarely contain or bear upon their surface any considerable abun- 
dance of boulders or other drift materials that have not evidently 
been transported, worn, and assorted by water. In nearly all 
the localities where I have observed boulders or masses of till 
imbedded within osars or lying on their surface, the most proba- 
ble explanation of their derivation has been by falling from the 
enclosing ice-walls of channels open to the sky, or by being 
brought while frozen in ice-floes . 2 At only one place, in Dover, 
1 Geology of N. H., vol. iii, pp. 108, 131-137, 289-291. Geol. and Nat. Hist. Survey 
of Minnesota, Eighth annual report, for 1879, pp. 113,114: Final report, vols. i and ii. 
Proceedings of this Society, vol. xxiv, 1889, pp. 231-5, 237-9. 
2 Geology of N. H., vol. iii, pp. 43, 46, 85, 88, 90, 92, 127, 145, 148, 158, 160, 162, 
Geology of Minn., Final report, vol. ii, p. 550. Geol. and Nat. Hist. Survey of Canada. 
Annual report, vol. iv, pp. 40-42 E. 
PROCEEDINGS B. S. N. H. 
VOL. XXV 
16 
June 1891 
