394 R. M. Deeley — Erosion hy Rivers and Glaciers. 



angular particles of the sizes of peas, beans, or walnuts, all frozen 

 firmly together, presents in mass the properties of a viscous 

 substance like pitch, owing to a molecular action which takes place 

 between the glacier grains,^ and the glacier moves slowly down its 

 rocky bed in a manner much resembling that of a river. Again, 

 a glacier will often pass over loose deposits of bedded sand 

 and gravel without disturbing them in any way, whilst in other 

 instances it will move rock- masses, measuring hundreds of yards in 

 length for miles, and actually tear up large boulders from its rocky 

 floor. The apparently contradictory effects produced by otlier 

 agents should warn us not to assume too readily that because ice can 

 polish it cannot excavate, or because it can erode it cannot deposit, 

 or that because in one case it passes over loose sand without 

 disturbing it it cannot break up and remove hard rocks. A con- 

 sideration of the properties of ice will, however, serve to make such 

 contradictions less startling. 



Unlike the water of a river, which is stationary when actually 

 in contact with fixed objects, we have abundant proof that the ice 

 of a glacier when in contact with its floor is not always fi'ozen to it ; 

 for the glacier carries stones and sand along, and polishes and 

 scratches many of the rocks it passes over. Some writers seem 

 to take it for granted that glaciers are nowhere actually frozen 

 to their beds, and, basing their conclusions upon this assumption, 

 arrive at results which are inconsistent with the facts. But we 

 have no evidence which proves that glaciers really are always 

 separated fi'om the rocks upon which they rest by a liquid film. 

 On the conti'ary, our knowledge of the properties of ice would 

 rather seem to support the opposite view, i.e., that they are 

 frequently firmly frozen to their beds over large areas. 



When a piece of lake ice is placed between the opposing faces 

 of a press, and considerable pressure is applied, a portion of it 

 becomes liquid, and the mass assumes a stratified appearance, 

 owing to the formation of liquid layers. To produce these liquid 

 films heat is required, and as this has to come from the ice the 

 whole mass falls slightly in temperature. Indeed, experiment has 

 proved that the freezing-point of water is lowered by pressure. 

 Now beneath a thick glacier, owing to the pressure resulting from 

 its weight, the ice must be at a lower temperature by a degree 

 or so than at its upper surface, and this no matter if the pressure 

 is applied in a vertical direction only. We are here dealing with 

 the same substance in two physical conditions, but the only prac- 

 tical difference that exists between them, for our purpose, is that 

 the ice is enormously more viscous than the water. In the form 

 of water its viscosity is exceedingly small, and for all practical 

 purposes it may be considered as unable to transmit stress if 

 applied in one direction only. As ice, its viscosity being enormously 

 great, it can transmit stresses for yards or even miles in 

 a direct line. We may, consequently, have ice under great pressure 



1 Geol. Mag., Dec. IV, Vol. II, 1895, p. 152. 



