404 'The American Geologist. December, 1894 
tion to the evidences of the formation <>f the eskers which have been 
studied in Illinois, Iowa, Minnesota, and Manitoba, by superglacial 
streams. 
Where eskers (osars) have been explored by Mr. Leveretl in Illinois 
and adjacent stales, they are round, as described by Prof. < 'hamberlin 
in the Journal of Geology (vol. i, pp. 266, 267, April-May, 1893), to vary 
in length from a few miles to aboul fifty miles, and to lie often as nar- 
row gravel and sand ridges in wide river-like channels cul into the gen- 
eral sheet of till, but to have ascents and descents over present water- 
sheds. It is thus seen thai these eskers were deposited by ice-walled 
streams, either subglacial, englacial, or superglacial. While Prof. 
Chamberlin, in the paper cited, regards them as probably subglacial, a 
different explanation seems to accord with their superglacial formation. 
From the steep slopes adjoining river valleys or canons cut into the wan- 
ing ice-sheet, its englacial and finally superficial drift, exposed by 
ablation when the greater part of the thickness of the ice there had 
been melted away, would slide down, or be washed down by rains and 
rivulets, into the stream bed, to contribute partly to the esker and to be 
partly carried onward by the stream. When the ice wholly disappeared, 
a wide shallow trough in the till and the esker ridge would have the re- 
lationship described. 
In Iowa the prominent paha or loess eskers described by McGee (l\ S. 
Geol. Survey. Eleventh An. Rep., for 1889-'90, pp. 435-471) are attributed 
chiefly to superglacial drainage, deriving the material of the paha from 
drift exposed by ablation on the ice surface. 
[n Minnesota eskers are rather infrequent. The most notable are in 
Bridgewater and Lake Johanna townships. The former is an esker 
series about seven miles long, mapped by Prof. N. H. Winchell. who 
shows its origin to have been by a superglacial stream (Geology of 
Minn., vol. i, 1884, pp. 6C>5-I5G!)); and the latter series consists of several 
parallel ridges, mapped by the present writer and similarly explained 
(vol. li. 1SSS, pp. 180, 490). 
The most conclusive evidence of the origin of the material of eskers 
from superglacial drift I find in the esker called Bird's hill, at the sta- 
tion of this name on the Canadian Pacific railway in Manitoba, seven 
miles northeast of Winnipeg. This esker had such relationship to the 
glacial lake Agassiz, there 500 .feet deep as soon as the ice-sheet re- 
treated, that its material is shown to have come from a somewhat 
greater bight in the lower part of the ice-sheet, which attained probably 
a maximum thickness of 5,000 feet or more above that region, if we 
may judge from its known thickness upon New England and New York, 
covering Mt. Washington and the Green and Adirondack mountains. 
It is further learned, by a mass of till fallen into the esker gravel and 
sand, that several feet of englacial drift existed there above the altitude 
of 500 feet in the ice. (Geol. Survey of Canada, An. Rep., new series, 
sol. iv, tor l888-'80, pp. :W-40 E.) 
The observations of the eskers of Bird's hill and the Pinnacle hills at 
Rochester, N. Y. (Proc, Rochester Acad. of Science, vol. [I, pp. 181-200, 
