238 Correspondence — R. M. Deeley. 



other, so that the Becke effect is mainly due to the relation between 

 their indices of refraction. The author discusses in detail the 

 effect of the refractive indices of the different directions of vibration 

 on the result. 



COE.E.E SIP OISriD E3ITO E - 



SCANDINAVIAN ICE - SHEETS AND BRITISH GLACIAL DRIFTS. 



Sir, — I thank Professor Bonney for the very kind way in which he 

 refers to my letter on the Scandinavian ice-sheet. The question 

 should, I agree with him, be approached without shirking difficulties. 

 It would not be possible within the limits of this letter to discuss all 

 the points he raises. I will, therefore, merely deal with the question 

 as to whether, if sufficient ice passed off the Scandinavian Peninsula, 

 it could flow across the deep JS'orwegian Channel and shallow jS'orth 

 Sea and invade England. Professor Bonney says the main point for 

 consideration "for our present purpose is the inadequate 'ramming' 

 power of the ice from the Scandinavian upland, because by far the 

 greater part of the journey to England would have been over land, not 

 by water ". If glacier ice be really viscous, the flow would not be the 

 result of ' ramming ' from behind. The movement would result from 

 the absence of support in front. Again, he says, "Thus it must 

 either have been forced up and over an undulating floor for over 

 350 miles before it reached the English coast, or its movement have 

 been practically restricted to the upper layers." If such a form of 

 flow took place, it would not be viscous flow. There would be a more 

 or less steady gradient of the upper surface of the glacier rising from 

 England to the Scandinavian uplands. The stresses due to the 

 existence of this upper slope would act on every portion of the ice- 

 sheet, and the bottom of the ice would drag the floor in the direction 

 of motion even though the motion of the ice at the bottom were 

 uphill. Ice will flow like water until its upper surface has become 

 practicalljr horizontal ; but as the former has a viscosity of only 

 O'Ol ^ c.g.s. 793 units, whilst the viscosity of glacier ice" is about 

 125 X 10'^ c.g.s. units, water — if we do not consider the effects of inertia 

 — would flow cceteris paribus as far in one second as glacier ice would 

 in about 200,000,000 years. Now, if the water in the Norwegian 

 Channel were raised about 100 feet, as Prof. Bonney suggests, and then 

 suddenly released, it would invade England and carry material along 

 the sea bed with it. Ice is not 'rammed' forward any more than 

 water is rammed along a river bed. The spreading out in great sheets 

 over the Antarctic Seas of the glaciers coming down from tlie Antarctic 

 uplands is proof, I consider, that glacier ice in bulk behaves as a viscous 

 substance, i.e. that the rate of shear is proportional to the stress even 

 when the stress is exceedingly small. A great many very capable 

 field geologists hold that there is ample evidence that an ice-sheet did 

 invade England from the east and north-east, and I always understood 



1 Phil. Mag., 1909, p. 518. 



- Roy. Soc. Proc, vol. Ixxxi, p. 251. 



