Outer Glacial Drift. — Upham. 155 
strictly no extension in ^Montana east of the mountains and 
their immediate vicinity ; yet probably correlative and contem- 
ponaneoiis till exists beneath the later Kansan, Illinoian, 
lowan and Wisconsin drift in the Mississippi basin, where 
these divisions of the glacial series have been so fully studied 
and classified in their stratigraphic and chronologic relations 
by Chamberlin. Calvin, Leverett, and others. Eastern gla- 
ciation and drift, which overspread that region, failed to ex- 
tend into confluence with the Cordilleran glaciation south of 
the forty-ninth parallel ; but north of that line the icefields 
flowing from the east and west were confluent, with inter- 
mingling of their drift deposits on a tract of considerable width 
extending from south to north not far east of the mountains. 
In western Montana, and onward to the Pacific coast, the 
earliest and lowest glacial drift must be correlative with the 
Albertan of Dr. Dawson, which he regarded as earlier than 
the Kansan drift sheet, the lowest and oldest recognized pre- 
viously in the United States. 
Including the chief moraine belts, of Wisconsin age, trace- 
able north of the international boundary, it is probable that 
the northwestern part of the Great Plains, on the upper Mis- 
souri river and northward in the Dominion of Canada, has as 
good a representation of the entire sequence of our glacial de- 
posits, from the oncoming to the final wane of the prolonged 
and varied Ice age, as can be found ?.nywhere, not excepting 
Kansas, Iowa, and Illinois, in which states the series has been 
most elaborately studied and named. 
A general survey of our marginal drift formations, from 
the area of the glacial lake Agassiz west to the Rocky moun- 
tains and the Pacific, noting the relationships of the diverse 
courses of the glacial currents and the relative ages of their 
drift, convinces me that the view of Tyrrell,* indicating suc- 
cessive culminations of ice accumulation, first in the Cordill- 
eran region, next on the country between the Rocky mountains 
and Hudson bay, with extension south into Kansas, and last 
on the Laurentide region, stretching an ice-sheet from Labra- 
dor to the Red river valley, is erroneous. Instead of successive 
Cordilleran, Keewatin, and Laurentide or Labradorian ice- 
• /ourna/ of GeoZoA'v. vol. iv, pp. 811-815, Oct.-Nov , 1896; vol. v, pp. 78- 
sl, Ian. -Feb., 1897; vol. vi pp. 1+7-160. Feb. March, 1898. 
