404 Royal Society :— 



the effect of pressure in lowering the freezing-point of water, affords 

 a perfect explanation of various remarkable phenomena involving 

 the internal melting of ice, described by Professor Tyndall in the 

 Number of the ' Proceedings ' which has just been published. I 

 wish now to show that the stratification of vesicular ice by pressure 

 observed on a large scale in glaciers, and the lamination of clear ice 

 described by Dr. Tyndall as jiroduced in hand specimens by a 

 JBrahniah's press, arc also demonstrable as conclusions from the 

 same theory. 



Conceive a continuous mass of ice, with vesicles containing either 

 air or water distributed through it ; and let this mass be pressed 

 together by opposing forces on two opposite sides of it. The vesi- 

 cles will gradually become arranged in strata perpendicular to the 

 lines of pressure, because cf the meltiny of ice in the localities of 

 f/reatest pressure and the revelation of the water in the localities of 

 least pressure, in the neit/hdourhood of groups of these cavities. For, 

 any two vesicles nearly in the direction of the condensation will 

 afford to the ice between them a relief from pressure, and will occa- 

 sion an aggravated pressure in the ice round each of them in the 

 places farthest out from the line joining their centres ; while the 

 pressure in the ice on the far sides of the two vesicles will be some- 

 what diminished from what it would be were their cavities filled up 

 with the solid, although not nearly as much diminished as it is in 

 the ice between the two. Hence, as demonstrated by my brother's 

 theory and my own experiment, the melting temperature of the ice 

 round each vesicle will be highest on its side nearest to the other 

 vesicle, and lowest in the localities on the whole farthest from 

 the line joining the centres. Therefore, ice will melt from these 

 last-mentioned localities, and, if each vesicle have water in it, the 

 partition between the two will thicken by freezing on each side of it. 

 Any two vesicles, on the other hand, which are nearly in a line per- 

 pendicular to the direction of pressure will agree in leaving an aggra- 

 vated pressure to be borne by the solid between them, and will each 

 direct away some of the pressure from the portions of the solid next 

 itself on the two sides farthest from the plane through the centres, 

 perpendicular to the line of pressure. This will give rise to an in- 

 crease of pressure on the whole in the solid all round the two cavi- 

 ties, and nearly in the plane perpendicular to the pressure, although 

 nowhere else so much as in the part between them. Hence these 

 two vesicles will gradually extend towards one another by the melt- 

 ing of the intervening ice, and each will become flattened in towards 

 the plane through the centres perpendicular to the direction of press- 

 ure, by the freezing of water on the parts of the bounding surface 

 farthest from this plane. It may be similarly shown that two vesi- 

 cles in a line oblique to that of condensation will give rise to such 

 variations of pressure in the solid in their neighbourhood, as to make 

 them, by melting and freezing, to extend, each obliquely towards the 

 other and. from the parts of its boundary most remote from a plane 

 midway between them, perpendicular to the direction of pressure. 



The general tendency clearly is for the vesicles to become flattened 



