THE EXTENT OF THE CORDILLERAN ICE-SHEET 
CHARLES A. STEWART 
Through the courtesy of members of the United States Forest 
Service I was enabled, in the summer of 1912, to make some 
observations upon glaciation in the Kaniksu National Forest in 
northern Idaho. Because of the inaccessibility of a part of the 
region and the lack of suitable maps, the work was not complete; 
but in view of our slight knowledge of the extent and character of 
the continental glaciation in the northwestern United States, and 
because the nature of the country makes it extremely unlikely 
that it will be more carefully studied in the near future, it seems 
advisable to publish the facts noted. Moreover, there was dis- 
covered evidence of more complete reworking of glacial deposits 
by stream action than has yet been described. 
It is known that the front of the Cordilleran ice-sheet was 
marked by a series of marginal lobes occupying the north-south 
intermontane valleys of the northwestern United States. The 
line of crosses in Fig. 1, based on data published by Professor 
Salisbury,’ shows the probable extent of some of these lobes. It 
is believed that the Pend d’Oreille lobe was continuous with that 
in the Colville valley, but its connection with the Kootenai lobe 
was well north of Bonners Ferry, and probably north of the Inter- 
national Boundary. In the paper cited nothing is said of the 
eastern boundary of the Pend d’Oreille lobe, but my observations 
show that the ice must have covered the divide between the Pend 
d’Oreille and Priest Lake valleys. 
This divide is a range of forest-covered hills having a maximum 
elevation of 6,500 feet, and an average height a thousand feet less. 
The former presence of ice on these hills is unmistakably shown 
by the rounded topography, by roches moutonnées, and by glacial 
striae—in one instance on bedrock, and in others on float. The 
striae on bedrock strike N. 10° W. Looking east across Priest 
t Jour. Geol., IX, 721-24. 
427 
