1890.] Observations on Pure fee. 115 



plasticity of ice, in the mass, at similar temperatures. Mr. Trotter 

 further remarks that " the supposition that, while ice at C. is sensibly 

 viscous, the viscosity diminishes rapidly with the temperature is in 

 complete accordance with the facts of the changes which take place 

 in a glacier during the winter." The comparatively great contracti- 

 bility in ice observed at considerably reduced temperatures, see my 

 paper on " Observations on Pure Ice and Snow " (' Roy. Soc. Proc.,' 

 vol. 40, 1886, p. 544), accounts for the great reduction in its plastic 

 properties. This is in full accord with the practical cessation of 

 motion in glaciers during the cold of winter. I believe that the 

 plastic properties of ice in the mass are variable, and are also, to 

 some extent, influenced by the rapidity or otherwise of its crystal- 

 lisation. Thus, a mass of water rapidly frozen at an intensely low 

 temperature would, no doubt, crystallise into a larger number of 

 smaller crystals than those which would result from a more slow 

 solidification of the mass of water at a comparatively higher tempera- 

 ture of freezing. 



4th. It will be noticed, on comparing the results of Diagrams I and 

 II, that the plasticity of the naturally frozen pond ice was manifestly 

 greater than that of the artificially prepared pure ice, and the 

 difference in results may, to a certain extent, probably be accounted 

 for by difference of composition of the respective frozen waters and 

 ice (the block ice being frozen from pure distilled water, the com- 

 position of the pond water and ice frozen therefrom being given on 

 Tables I and II) ; I think, also, the difference was, to some extent, 

 owing to the direction from which the pond ice was frozen, viz., from 

 the surface only. Further, this comparative difference in the 

 behaviour of the pond ice was doubtless owing to a portion of the 

 saline constituents of the water interspersing, during congelation, 

 between the faces of the individual crystals of ice, thereby tending 

 to reduce the cohesion of the mass as a whole and increasing its 

 plasticity. Mr. J. Y. Buchanan, F.R.S., has shown that in Arctic ice, 

 contrary to expectation, the whole of the salts do not separate from 

 sea- water ice during congelation, but that they remain interspersed 

 amongst the interstices of the crystalline mass ("Ice and Brine," 

 ' Edinburgh Roy. Soc. Proc.,' vol. 14, 1888, p. 129). My experiments 

 on Table II also show that the composition of pond or natural river 

 ice is affected in a similar manner, and afford confirmation of the 

 views held by Buchanan on the composition of sea- water ice in Arctic 

 regions. The fact of the pond ice having been slowly crystallised 

 would further tend to modify its physical properties, compared with 

 ice rapidly crystallised by an intensely low temperature, simul- 

 taneously acting only from the bottom and sides of the mould or 

 tank. Measurements of the expansibility of ice may also be affected 

 accordingly as such measurements are taken either longitudinally or 



