H84 The American Geologist, Decembor, is'jt 
foi'iner papers.* In tlie localities thus studied and described, 
the nucleal beds were evidently of contemporaneous origin 
with the till. While the drumlin hill was being amassed, for 
a considerable time the drift forming it was stratified gravel 
and sand, brought and depf»sited by streams of water from 
the melting of the overlying and wasting ice-sheet. Other 
parts, basal and superficial, were till, being unstratitled gla- 
cial drift deposited directly from the ice, without modifica- 
tion by water action. 
Drumlins are attributed, in these papers, to convergent 
currents of the border of the ice-sheet when a layer of drift, 
having become siiperglacial, as on the Malaspina glacier, was 
enveloped by a later onflow of ice above it, being then amassed 
englacially or subglacially in these liills very near to the 
boundary of the ice, that is, within a few miles or probabl}'' in 
some cases within even less than one mile. Subse(iuently this 
opinion h':is received some confirmation or supp(U"t from Green- 
land, where, beneath the terminal cliffs of its inland ice and 
peripheral glaciers, masses of drift were seen and photo- 
graphed by Chamberlin, with the ice lamination curving drum- 
loidally over them.j 
Although the drumlins, like the Wisconsin and later mo- 
raines, belong to the general time of Late Glacial or Champlain 
retreat and final departure of the ice-sheet, this view of their 
origin ascribes their formation to substages when the reces- 
sion of the ice boundary was much slackened or even was in- 
terrupted by a temporary readvance. The edge of the ice-fields 
was then thickened by exceptional!}' abundant snowfall during 
a series of years, and by outflow of the higher ice from some 
distance back of the border. According to mj^ studies and 
mapping of drumlins and associated drift de))osits in New 
Hampshire and northern and eastern Massachusetts, it seems 
to me probable that in this region mainl}' the glacial recession 
* Boston Society of Natural History, vol. xxiv, pp. 228-242, April, 
1889; vol. XXVI, pp. 2-17, Nov., 1892, with discussion bv Profs. VV. M. 
Davis and G. H. Barton (pp. 17-25). 
Am. Geologist, vol. x, pp. 239-262, Dec. 1892, including bibliographic 
notes: vol. xiv, pp. 69-83, with sections of drumlins in Scituate, Mass.. 
containing nucleal stratified deposits, and a map of drumlins in Madison, 
Wis., which consist mainly of a nucleus of modified drift. 
tBulletin, (xeoi. Soc. Amer., vol. vi, pp. 199-220, with eight plates, 
Feb., 1895. 
