238 Tlic American Geologist. April, isas 
Some of these drumlins are niiich elongated, to an extent of 
two or three miles; and their longer axes trend toward an ad- 
jacent looped marginal moraine, betokening their deposition 
and moulding by the ice-sheet at the same stage, during its 
general retreat, when the moraine was formed. In Sweden, 
so far as Baron Dc Geer has observed during very extensive 
explorations, drumlins are almost entirely absent. In my 
journeys through Holland, Germany, Denmark, Norway to 
Trondhjem, and thence east into Sweden and south in that 
country to Stockholm and Goteborg, no drumlins were found : 
and I fully agree with De Geer that probably they are no- 
where well develo])cd on the Scandinavian peninsula. 
Why and how did the ice-sheets form so abundant drum- 
lins in some limited parts of New England, as notably about 
Boston and Worcester, also in New York and southeastern 
Wisconsin, in the lar-Connaught district, and in the Clyde 
valley at Glasgow, while other areas apparently not less fa- 
vorably situated are destitute of such hills? and in what way 
could the broad, deep sheets of slowly moving land-ice heap 
up these prominent masses of the unmodified glacial drift? 
It is entirely easy to account for the retreatal moraine hills 
of our continental ice-sheets, of which we have the counter- 
parts at the ends of now existing glaciers. Scarcely more diffi- 
culty is encountered in ascertaining the mode of formation of 
kames and eskers, which are knolls and ridges of modified 
drift gravel and sand deposited by streams walled in part by 
ice and therefore left by its melting in high ])innacles and 
ridges. Drumlins, on the other hand, differ from any ob- 
served product of the present puny glaciers of the Alps and 
other mountain districts; but Chamberlin's observations and 
photographs of the borders of the Greenland ice-sheet give 
some suggestions of their origin.* 
The view which appears to me to afford the fullest explana- 
tion of the origin of drumlins. in a brief statement, refers then- 
accumulation to convergent currents of the irregularly in- 
dented and channelled border of the ice-sheet during its re- 
treat, when a layer of drift, having become superglacial, as 
on the Malaspina glacier, was enveloped by a later onflow of 
♦Bulletin, Geol. Soc. Amer., Vol. VI, pp. 199-220, with eight plates, 
Feb. 1895. 
