348 The American Geologist. •^""^' ^^^^ 
cousin stage of waning glaciation, and also the courses of 
movement and the terminal moraine of a far eastwardly 
projecting outflow from the great Keewatin ice-lobe of the 
northwest. The former brought red drift, colored by the 
peroxide of iron from the red sandstones and shales of the 
Lake Superior basin, and destitute of limestone fragments ; 
while the latter, an ice invasion from the Cretaceous, Silu- 
rian, and Cambrian region of the Red river valley and 
Manitoba, brought gray drift, with abundant limestone 
boulders, pebbles, and finely comminuted rock flour. In 
the region of the St. Croix Dalles, the red Lake Superior 
drift forms the surface on the Wisconsin side, and reaches 
far westward under the gray drift, which overspreads the 
surface in Minnesota and reaches sparingly east across the 
St. Croix river, as carefully traced by Mr. Chamberlin, to a 
distance of about one to three miles into Wisconsin for 
several miles north and south of the Dalles. 
In several details I see reasons to question some of Mr. 
Ghamberlin's conclusions: (i) where he refers the gray 
drift to a later time than the red drift, supposing that the 
Lake Superior ice had melted off from this vicinity before 
the Keewatin ice advanced here ; {2) in his regarding the 
preglacial course of the St. Croix river as perhaps a few 
miles east of the Dalles, passing into the Apple river valley, 
and thence to the lower St. Croix valley; and (3) in his 
reference of the erosion of the Dalles gorge to the work 
of the river since the final retreat of the ice-sheet. All 
these questions are related to studies and conclusions pre^ 
sented five years ago in my two papers before cited in the 
Bulletin of the Geological Society of America, which appear 
not to have been considered by Mr. Chamberlin in his 
recent work. Therefore portions of those papers are used 
again in the following discussion of these several points, all 
bearing on the age of this river gorge. 
I. The Lake Superior and Keewatin ice lobes seem 
to me to have been contemporaneous, being confluent along 
an uneven and wavering line, where their glacial currents 
met, or where one overrode the other, as at these Dalles, 
and along a course trending with much irregularity and 
minor lobation in a general northwesterly and northerly 
