18 B. M. Deeley — Glacier Motion. 



immediately form new and less strained linkages. The higher the 

 temperature the greater will he the number of links broken in a given 

 time and the more rapid will be the distortion; also the greater 

 the stress the greater will be the distortion in the breaking of 

 each link. It does not follow that the breaking of the molecular 

 links will only take place at the freezing-point. It is generally 

 believed that the higher the temperature of the body the more 

 energetic is the motion of the average molecule or atom. Even at the 

 melting-point some of the atoms move slowly and others rapidly, and 

 as the temperature falls there are still molecules moving with sufficient 

 rapidity to cause a breakage of the linkages. This conception of the 

 cause of viscosity also explains the partial recovery of form a stressed 

 viscous substance undergoes when the load is removed. The highly 

 strained but unbroken linkages, on the removal of the external stress, 

 produce an immediate elastic recovery and also a slower recovery 

 as the strains are relieved by the constant breaking of molecular 

 attachments. 



It is clear that glacier ice is not a viscous substance in the sense that 

 water, pitch, or alcohol are ; for a crystal of ice can only be caused to 

 shear viscously in a direction at right angles to the optic axis. In 

 this case it would appear that the atomic or molecular linkages are 

 broken by the vibrations in one plane only. A mass of ice built up of 

 a number of crystalline granules, as is glacier ice, would not flow- 

 viscously if it possessed this viscous peculiarity only, for as the 

 optic axes are in all directions they would lock each other and prevent 

 motion. But it is certain that the molecules at the interfaces of the 

 crystalline granules are constantly forming new linkages, for some 

 granules are increasing in size whilst others are decreasing, there 

 being a constant transfer of molecules from granule to granule. It is 

 to this interchange of molecules between granule and granule that the 

 peculiar viscosity of glacier ice is primarily due. "When the stress 

 producing flow is small the granules are of various irregular shapes 

 and sizes and the rate of distortion is small. When the stresses 

 are great, however, the tendency is for the interfaces of the granules 

 to assume the form of shear planes, the granules then slowly sliding 

 over each other without actual fractures being produced. The granules, 

 however, are sheared also without fracture, and the coarseness of the 

 granular structure thereby reduced. No doubt the effective viscosity 

 of glacier ice depends largely upon the size and shape of the ice 

 granules. 1 Some calculations I made seemed to prove this. At any 

 rate the difference in the viscosity of the ice of several glaciers seemed 

 to be much greater than could be accounted for by errors in the data. 



I will now refer to Sir Henry Howorth's most conclusive argument. 

 He says : " Lastly, and this is conclusive, if the theory were right there 

 would not be a continually increasing differential flow from the base 

 of a glacier to its summit and from its sides towards its centre." 

 As a matter of fact there is not an increasing differential flow from the 

 base to the summit and from the side to the centre. The rate of 

 distortion is greater at the sides and bottom than it is at the upper 



1 Proc. Roy. Soc, 1908, p. 250. 



