156 The American Geologist. s^ptombpr. i904. 
sheets, there seems to me to have been contemporaneous and. 
long continued glaciation of the entire width of the continent ; 
but during the original growth of the vast sheet of snow and 
ice covering half of North zAmerica, and again during its de- 
cline and disappearance, it doubtless outflowed predominantly 
from those three central areas." 
As the glacial drift generally extends somcAvhat beyond 
the Alissouri river, to a maximum distance of about fifty miles, 
through the Dakotas and Montana (excepting a large part of 
South Dakota, to be again noticed), up to Great Falls, whence 
the drift boundary passes across this river and runs north- 
westward to the St. Mary's river and the Canadian line at 
the base of the mountains, it is evident that during the cul- 
mination of the Glacial period, in its Kans:^n stage, all the 
rivers of these states, and also of Nebraska, were turned aside 
to the west and south, being caused to '^end their waters south- 
eastward along the border of the ice-shoet. In the Fort Benton 
folio of the Geologic Atlas of the United States, an ancient 
channel, called the Shonkin sag, eroded by the waters thus 
pouring along the ice border, has been traced by Weed about 
fiftv miles, passing across the low watersheds. He describes 
it as "a continuous depression that is clearly an old river chan- 
nel, with wide valley floor and steep walls and precipitous 
clift's." It wil be a most interesting task for amateur geolo- 
ogists of this region to trace these old watercourses, inter- 
rupted here and there by beds of lakes that adjoined the ice 
front, from the vicinity of Great Falls lor i.ioo miles east 
and south, across the Musselshell, Yellowstone, Little Alis- 
souri, Cheyenne, White, Niobrara, Platte, and Kaw rivers, to 
the Osage valley in Missouri, which doubtless for some time 
carried all the drainage of the Missouri basin, when the pres- 
ent pathway of the Missouri river was thus buried under the 
edge of the continental glacier. But my last preceding paper 
(pages 80-87 of ^''^i^ volume) has somewhat fully considered 
the relation of this majestic river to the glaciation. leaving now 
a river system probably very unlike its preglacial condition, so 
tliat further comiucnt on its changes and development during 
the Ice age will not be expected here. 
* See Maps of the Glaciation of North America, by Chamberlin in Geikie's 
Great Ice Age, third ed., 1S94-, pi. xiv, at page 724; by the present writer, U S. 
Geol. Surrev. Man. XXV, The Glacinl Lake Agassiz, 1!S95. jsls. ii and xvi; bv 
Leyerett.I/.S. Geo/. Survey. Mon. XXXVIII. The Illinois Glncial Lobe, 1899, 
pi. i; and by Russell, North America, 19()4-, pi. v, at page 315. 
