Upham.] 16 [Nov. j6, 
of this marginal ice tlie jtrevioiisly superglacial sliatuiii of drift 
was chiefly amassed in driimlins. The known drumlin areas of 
New Brunswick and New England, New York, Wisconsin, and 
Manitoba, would therefore be expected to belong to the same 
stages of the closing part of the Ice age. This would imply what 
seems from other reasons not improbable, that the outermost 
moraines in the States east of Ohio and on the east side of the 
driftless area in Wisconsin correspond to some of the inner and 
late moraines in the greater ])art of the region of the Laurentian 
lakes and the uj>per Mississippi, as perhaps the exceptionally mas- 
sive Leaf Hills, Itasca, and Mesabi moraines, which are the ninth, 
tenth, and eleventh of the series in Minnesota. 
Again, in confirmation of tlie view that much drift was con- 
tained in the lower part of the ice-sheet, my studies of the very 
massive kame deposits forming the greater part of the outermost 
terminal moraine on Long Island eastward from Roslyn, of the 
large kame called the Devil's Heart, rising in a somewhat conical 
hill 175 feet above the adjoining country south of Devil's Lake in 
North Dakota, and of the esker named Bird's Hill, seven miles 
northeast of Winnipeg, seem to me to demonstrate, beyond all 
doubt, that their material, and ])robably likewise that of kanies 
and eskers generally, was supplied by supei-glacial streams from 
plentiful englacial di'ift, and could not have been brought from 
drift beneath the ice by subglacial drainage. 
In conclusion, I deem it a duty to state that this reference of 
the drumlins, terminal moraines, kames, and eskers, to rapid 
accumulation from previously englacial drift during the depar- 
ture of the ice, seems to me better accordant with the view that 
the Ice age comprised only one great epoch of glaciation, attended 
by oscillations of the ice-border, than with the alternative view, 
hitherto held by me during the past thirteen j'ears, as by most 
American glacialists, which supposes the ice-sheets to have been 
at least once and perhaps several times almost entirely melted 
away, afterward being restored by recurrent glacial epochs.^ The 
drumlins, like the terminal moraines, are effects of secular vicissi- 
tudes of climate on the border of the departing ice-sheet, which 
I think to have owed its existence to great altitude of the land at 
the beginning of the Glacial period, to have been attended when 
' See the receut article by Prof. G. F. Wright, Unity of the Glacial epocli, Aiuer. 
journ. sci., ser. 3, vol. 44, Nov. 1892, p. 351-373. 
