696 
GEOGRAPHY: W. M. DAVIS 
results from loading. The thrust modulus, E, computed from triplets 
of data for definite steps of pressure (1, 2, 1), (2, 3, 2), etc., kg. (i.e., 
15, 25, etc., kgm. per square centimeter), are given in figure 3. They 
increase in marked degree with the load. Turning the rods down to 
smaller diameters successively and testing them in turn, no essential 
difference in the results was apparent. With rods of high rigidity like 
glass, brass, steel, only about one-half of the probable modulus can be 
reached with rods of the above dimensions. The remainder is lost in 
the small dislocations within the apparatus. These rods^ must not be 
more than 1 or 2 mm. thick and enclosed in corresponding sheaths, to 
be available in an apparatus-like figure 1 . Tentative'' as the results are, 
however, they are interesting, inasmuch as the dependence of the elas- 
tics of a rod on its molecular instabilities will most probably be clearer 
in case of bodies of light structure like the organic bodies. The whole 
phenomenon is very much like the condensation of a vapor, requiring 
higher pressures to condense the instabilities and lower pressures for 
their release or evaporation, as it were. Deformation proceeds at a 
rapidly retarded rate through infinite time.^ 
1 These Proceedings, 3, 1917, (412). 
2 Shown in the side elevation, figure la, with the offsets removed. The fibres d and di 
are tightly stretched. 
3 Thus in case of steel rods like the above, per kg of load, A7V/AP = 44 x 10""^ cm., 
which is too small for any micrometer. 
*I have thus far been unable to arrive at a trustworthy distinction, except in magni- 
tude, between the deformations within the apparatus and those of the rods themselves. 
5 From a report to the Carnegie Institution of Washington, D. C. 
SUBLACUSTRINE GLACIAL EROSION IN MONTANA 
By W. M. Davis 
DEPARTMENT OF GEOLOGY AND GEOGRAPHY, HARVARD UNIVERSITY 
Communicated November 7. 1917 
The mountains of northwestern Montana and northern Idaho are 
characterized by tv^o classes of deglaciated forms. The forms of one 
class are the work of relatively small, local glaciers, and are limited to 
the loftier ranges in which cirques, excavated in the higher slopes, lead 
down through well-scoured troughs to terminal moraines on the moun- 
tain flanks or on the open ground of interment basins. The forms of 
the other class are the work of great Canadian glaciers and are limited 
to the sides and floor of the larger valleys. Two such glaciers crossed the 
international boundary, as shown in figure 1, truncated the side spurs 
