416 Tne American Geologist. juno, is'jT 
Where tracts of abundant drumlins, as in southern !New 
Hampshire, Massachusetts, and New York, occur, extending 
as they almost invariably do, approximately in parallelism 
with the great moraine belts, we may explain their origin, as 
I think, by an overflow of the upper, steeply sloping, snow- 
fed icefields upon the previously drift-covered ice of the outer 
zone, hpaping its drift in these remarkable hills.* 
Rhythm of Morainal Accumulation Independent of 
Secular Variations of Climate. 
The progress of ablation during the departure of the ice- 
sheet appears to have been mainly continuous through many 
years and centuries, excepting as it was temporarily inter- 
rupted during each winter. It was undoubtedl}^ accelerated 
or decreased in cycles of several or many years by secular va- 
riations of climate, such as now give series of years leaving 
alternately a deficiency and an excess of rainfall and snow- 
fall. But these secular climatic changes need not be appeal- 
ed to for explanation of the rhythm of morainal accumula- 
tion. AVith an entirely regular melting away of the ice- 
sheets, the recurrence of prominent moraines, with smooth 
intermorainic tracts, seems to me well explainable by the 
physical conditions dependent on the action of the glacial 
currents in amassing successive belts of the drift inclosed in 
the lower 800 to 1,500 feet, more or less, of ice that had attain- 
ed a thickness of a half mile to one mile. 
Probable Duration of the Champlain Epoch or 
Wane of the Ice Age. 
This explanation accords with my conclusion concerning 
the brevity of the existence of lake Agassiz, and with the gen- 
erally accepted time ratios and estimates of the Glacial and 
Postglacial periods, as respectively, between 40,000 and 100,- 
000 years, and between 5,000 and 10,000 years. It seems to 
me more acceptable than the reference of the successive mo- 
raines to climatic cycles of astronomic origin, by which Tay- 
lorf conjectures that a duration of 150,000 to 300,000 years 
was required for the formation of the moraines ill the north- 
ern United States, not to speak of others in Canada. 
*Am. Geologist, vol. x. pp. 339-362, Dec., 1892. Bulletin, Geol. See. 
Am., vol. VII, pp. 17-30, Nov., 1895. 
IAm. Geologist, vol. xix, p. 290, April, 1897. 
