1892.] 15 
[Upham. 
wear of its boulders and smaller rock fragments as they were car- 
ried forward with sliearing and sliding motion to the drumlin 
accumulations, and that in becoming lodged on the surface of the 
drumlins or on other and low deposits of subglacial till they 
would be further striated and planed. The previously englacial 
drift in being so transported and deposited would acquire all the 
marks of ice-wear which the till of the drumlins exhibits, and the 
pressure of from 500 to ] , 000 feet, more or less, of solid ice flowing- 
downward across it would seem adequate to produce its very hard 
and compact condition. We are thus able, as I believe, to 
account for all the differences between these deposits and the 
mostly unworn drift which fell loosely on the surface from an 
englacial or superglacial position when the ice disappeared. 
5. COMPARISON WITH TERMINAL MORAINES, KAMES, AND ESKERS. 
My study of the glacial Lake Agassiz, under the direction of 
Prof. T. C. Chamberlin, for the United States Geological Sur- 
vey and partly for that of Canada, shows that several large ter- 
minal moraines, marking pauses or re-advances interrupting the 
general glacial recession, were accumulated contemporaneously 
with the existence of that lake, while yet the whole duration of 
Lake Agassiz was apparentl}^ oiily about a thousand years. ^ '^J'he 
rapidity of formation of the moraines was thus similar to that 
of the drumlins, and both seem to have been made possible only 
by the large amount of the englacial drift. The f.ist retreat of 
the ice indicates that probably its melting border then had usu- 
ally a more steeply sloping surface than in its time of greatest 
extent to the south, and that consequently the rale of motion of 
the outer part of the ice-sheet was commonly increased during its 
flnal melting. Any pause of the retreat for even a few years 
would therefore form a moraine, though the outer belt of the ice 
may have been generally too steep to expose much superglacial 
drift. But during some stages of the recession we may conclude 
that considerable tracts of the ice-border were so thiimed by abla- 
tion that much englacial drift became superglacial, with the result 
that when again a colder climate brought a temporary thickening 
' Geol. and nat. hist. surv. Can. Ann. rept., new. ser., vol. 1, for 1888-89, p. 50, 51 
E. Amer. geol., vol. 7, April, 1891, p. 224-226. 
