GRANULATION. 211 



melted under pressure promptly freezes again when free from })ressure. 

 We owe to the same investigators the law that freezing is facilitated by 

 the presence of frozen surfaces in close juxtaposition. We owe to Tyndall 

 the doctrine that isolated i)articley or ])oint8 of ice melt more freely than 

 others from lack of support on either hand. Here, therefore, is a group 

 of agencies which favor melting under certain conditions and freezing 

 under other conditions. Now, all of these conditions affect the individual 

 granule as it occurs in the mass of a glacier. It has its points of contact 

 and pressure, its points of free surface, and its capillary interspaces. It 

 is suliject to pressures, torsions and tensions, according to the stresses 

 imposed upon it by neighboring granules. It is always under the influ- 

 ence of gravity acting directly upon it, and also indirectly through sur- 

 rounding granules. The combined effect is a resultant pressure urging 

 motion down the slope, but with every yielding of the granules, by melt- 

 ing or otherwise, there is a new adjustment of pressures, torsions and 

 tensions," and hence new susceptibilities to melting and freezing. Now 

 these being the conditions of the granules, it seems only necessary that 

 there pass over them an agency capable of acting upon the different sus- 

 ceptibilities of their different parts to produce loss here and gain there, 

 and hence to determine the growth of some parts of each granule and 

 the decadence of other parts. In other words, a granule may continually 

 change its form by partial melting and freezing, by loss in one part and 

 gain in another, and through this may either move itself or permit motion 

 in its neighboring granules, or both. 



Now, every warm day sends down into the glacier a wave of heat- 

 energy. This enters the upper surface as sensible temperature, but for 

 the most part it is soon changed to potential heat-energy in the form of 

 melted ice. We should not fail to see that the sheet of melted ice that 

 creeps down between the granules of the glacier as the result of a day's 

 sun-action is as trulv a wave of heat-energ}" as if it remained in the form 

 of sensible temperature. With what freedom the day's heat is conveyed 

 below by the melted product is not accurately known, but there is good 

 evidence that it is large. Ui)on the Igloodahomyne glacier we observed 

 at midday that the dust-wells were covered with thin films of ice, from 

 which the water below bad shrunk away to an average distance of jxt- 

 liai)s two inches. The suggestion was that this was the amount of aljsorj)- 

 tion of water wliich had taken place since the freezing of the film during 

 the preceding night, or, in other words, the absorption of perhaps twelve 

 hours. Circumstances did not permit the careful watching of an indi- 

 vidual well, and this inference was not verified, but it is certain tliat 

 wells of the largest sizes become entirely etnptied of their water within a 

 few days after cold weather cuts off their supply. The moisture whicli. 



