Tertiary and Quaternary Baseleveling. — Upham. 243 
bly fills the hollows clue to the valleys of a preglacial river 
and its tributaries, not wholly concealed hy the overlying 
drift. Within a few miles southwestward, a ridge several 
miles long and about 300 feet above the lake, singularly prom- 
inent in contrast with the surrounding moderately undulating 
or somewhat hilly surface, is reported by Prof. G-. E. Culver, 
in his exploration of that district for the Minnesota Geologi- 
cal Survey, to consist probably of Cretaceous shales beneath a 
deposit of till which has partly a smooth surface but on the 
northern slope presents a profusion of morainic knolls and 
hillocks. In a third and larger area, north and northwest of 
Red lake, where an extensive island arose above the highest 
stage of lake Agassiz, the grand topographic features seem 
likewise due to the prominence of the Cretaceous beds, there 
deeply drift-covered.* 
In North Dakota and Manitoba, west of the Cretaceous es- 
carpment, hills and buttes of the Cretaceous shales, thinly 
covered by drift, occur rarely along the Slieyenne river; in 
the vicinity of Devil's lake, one of especial prominence being 
the Big butte, about ten miles west-northwest of this lake; 
and near the Pembina river, where Star mound and Pilot 
mound in Manitoba are examples. Proceeding farther west- 
ward, and especially northwestward in Assiniboia, such com- 
paratively small hills, besides also large billy tracts similar to 
the Turtle mountain, are found more frequent. 
Direction of the Tertiary and early Quaternary drainage. 
The early Quaternary epeirogenie uplift causing the erosion 
and baseleveling of the Red river valley and Manitoba lake 
district must evidentl} r have occupied a long period as meas- 
ured by thousands of years. Its duration may well have been 
coextensive with the Lafayette period, embracing the deposi- 
tion and erosion of the Lafayette formation, which in my 
previous paper already cited I estimate t<» have comprised to- 
gether some 60,000 to 120,000 years. \ During this period the 
drainage from the entire area of the present Nelson river ba-in 
probably passed, nearly as now, toward the north and north- 
east. With the greater continental altitude id' that time a 
river system much longer than that of the Nelson and its feed- 
*Am. Geologist, vol. xi, pp. 123-425, June, 1893. 
fBulletin, Geol. Society of America, vol. v. pp. '.»;. 90. 
