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



pulling a thick, stiff liquid down the lee side of an embossment, but of a 

 rigid body thrusting itself over its crest. 



It is not easy to escape the influence of these observations if we push 

 inquiry back to the cause of movement. Competency to thrust and 

 measurable ability to individualize itself in layers seem to be requisites. 

 A general force might perhaps so individualize itself, but the phenomena 

 naturally lead us to seek an agency acting within the layers. The limits 

 of this address will not permit me to enter far upon the mooted question 

 of the cause of glacial motion, but the hypothesis that has come to be 

 dominant as the result of the summer's observations may be briefly indi- 

 cated. 



Granulation. — Back of the stratification of ice lies the phenomenon of 

 granulation. A glacier starts with snow-crystals or snow-pellets ; thence 

 there is a growth into shot-like granules, and thence into larger and 

 larger accretions. Drygalski places the limit at the size of walnuts. So 

 far as macroscopic study goes, this progressive growth of granules con- 

 stitutes the most essential change through which the ice passes. This 

 invites the inquiry whether the essence of glacial movement does not lie 

 in the changes which the granules undergo. If for a moment we enter- 

 tain the quite erroneous supposition that all the granules of a given 

 horizon grew from the smallest to the largest dimensions, it would appear 

 that the expansion of the layer and the movement at its free end would 

 be very great. Every doubling of the diameters of the granules would 

 push the foot of the layer forward a distance equal to the whole length 

 of the glacier. The western slope of the Greenlandic ice-field in the 

 northern tract is probably 500 miles. Three or four doublings of the 

 constituent granules would push its foot, if unmelted, over into Alaska ; 

 but it is obvious that each original grain does not develop into a larger 

 one — some are sacrificed for the growth of others. The hypothetical 

 case is introduced to emphasize this and to illustrate the possibilities of 

 motion involved in changes of the constituent granules. It would be 

 exceedingly helpful if we knew the laws which determine the destruction 

 of some granules and the growth of others. The process of progressive 

 granulation obviously involves the melting of some particles and the freez- 

 ing of the water in new relations. The vital question is : At what points 

 do melting and freezing respectively take place and what are the results? 

 We owe to James Thompson the law that pressure lowers the melting 

 point of ice. Whatever incompetency this may have as a sole agency, it 

 may be an extremely efficient factor in determining the precise points 

 at which melting will take place when both heat-energy and differential 

 pressure are present. I suppose the converse of Thompson's law also 

 holds true. We owe to Faraday and Tjaidall the observation that ice 



