216 T. C. CHAMBERLIN — GLACIAL STUDIES IN GREENLAND. 



of supposing that the material was borne from lower to higher altitudes 

 by the streams that formed the kames and eskers, only to be let down 

 again, we are compelled to locate their origin at the bottom of the ice- 

 sheet. This appears to confine hypotheses to the question whether 

 accumulation took place in tunnels under the ice or in channels cut back 

 from its edge. 



Drumlins. 



No drumlins were seen in process of formation, nor were any seen in 

 the abandoned territory, unless we force interpretation in a few doubtful 

 cases. The observations, however, seem to have some important bearings 

 upon the elucidation of drumlins. The limitation of the debris to the 

 basal layers of the ice limits the horizon of drumlin-making, as in the 

 case of eskers and kames. The observations which showed the weakness 

 of the marginal ice in comparison with the resistance of its own debris 

 furnish ground for comprehending the accumulation of masses of drift 

 beneath the edge of the ice. 



In describing the behavior of ice in passing over embossments of rock 

 it may be recalled that the laminse were found to start on the frontal side 

 of the embossment and to curve gradually upward and backward and at 

 length downward in the lee, the trend of this curve being quite similar 

 to the profile of a drumlin. I suspect that this is the true drumloidal 

 curve, and that it represents the balance or the accommodation between 

 the force of onthrust on the part of the overriding ice on the one side 

 and the friction and resistance of the ice and debris against the emboss- 

 ment on the other. I suspect that the progressive tendency in such a 

 case is toward the accumulation of debris below this drumloidal line, 

 which was apparantly a line of shearing, and that the result of such an 

 accumulation would be a drumlin. Why this particular curve should 

 be assumed is a problem the precise mechanics of which I do not profess 

 to understand, but seeing the curve developed repeatedly I infer that it 

 must be in conformity with the dynamics of ice-motion under these con- 

 ditions, and that nothing remains requisite to the formation of a drumlin 

 but the lodgment of drift below the drumloidal curve. (Figures 11 and 

 12, plate 8.) 



Rate of Movement of the Ice. 



Lieutenant Peary has commenced a series of observations upon the 

 movements of glaciers of the Inglefield gulf region, both by instruments 

 and by photographs taken at intervals. He found the daily movement 

 of the Bowdoin glacier, the most active in the immediate vicinity of his 

 headquarters, during the month of July to be four-tenths of a foot at the 



