702 
GEOGRAPHY: W. M. DAVIS 
from the main glacier could have compelled its narrow branch to ad- 
vance 100 miles against such discouragements: yet the ice not only did 
advance 100 miles up Clark-fork valley, but advanced with such insistent 
pressure that it tore off the resistant rock of the valley-side spurs. 
The sublacustrine glacial erosion thus attested takes its place as the 
last term of a series of unanticipated processes. Seventy years ago, the 
fiords of Norway, the sea lochs of Scotland, and other similar embay- 
ments were interpreted, by those who then accepted Dana's principle 
of shoreline development, as submerged river valleys. Forty-five years 
ago, the opinion gained ground that fiord troughs were largely the work 
of glacial erosion, but the erosion was supposed to have taken place 
above sea level; the occupation of the troughs by arms of the sea was 
explained as the result of later submergence. This view was gradually 
modified by recognizing that a great glacier might erode a trough, if it 
eroded at all, somewhat below sea level; but the extreme depth of such 
submarine erosion was placed at about six-sevenths of the thickness of 
the glacier: at greater depths, the ice would be buoyed up so that it 
could not erode the trough bottom. Then about twenty years ago 
Gilbert suggested, as a result of observations that he made in Alaska 
when a member of the Harriman expedition, that heavy glaciers must 
press so heavily on their trough beds that water could not enter beneath 
them; hence such glaciers could erode as well below as above sea level; 
but it was not supposed that they could be immersed for a score of miles 
or more. Now the Clark fork branch-glacier seems to have done its 
visible erosive work on the valley-side spurs — and presimiably a con- 
siderable amount of invisible work on the valley bottom also — although 
it must have been wholly immersed in Lake Missoula for two or three 
score if not for four score miles. It seems impossible for a glacier to 
perform erosional work under such conditions, yet the erosional work is 
undeniably visible. Perhaps the conditions of its performance were 
other than those here indicated, but if so, I have not been able to discover 
them. 
iDavis, W. M., these Proceedings, 1, 1915, (626-628). 
2 Davis, W. M., Geogr. Rev., 2, 1916, (267-288). 
3 Pardee, T. G., Chicago, J. Geol. Univ. Chic, 18, 1910, (376-386). 
