and on the Conservation of. Force. 168 



of the gradual liquefaction of ice*, and assumes that ice must 

 be essentially colder than ice-cold water, i. e. the water in con- 

 tact with it. 



I find no difficulty in thinking it would be easy to arrange 

 a mixture of water and snow in such a manner that it might 

 be kept for hours and days without any transition of heat either 

 to or from it ; but I find great difficulty in thinking that the par- 

 ticles of snow, small as they may be made, would remain for the 

 whole of the time at a lower temperature by 0°"3 F. than the 

 particles of water intermingled with them. Still, admitting for 

 the present the possibility that Prof. Forbes's view may be cor- 

 rect, and also the truthfulness of Mr. Thomson's principle, and 

 its possible action in regelation, I wish to say a few words on 

 the other principle already referred to, which was originally 

 assumed by myself, which, in relation with the mechanical theory 

 of heat, has been adopted by Dr. Tyndall, and which, after all, 

 may be the sole cause of the effect. 



The principle I have in view being more distinctly expressed 

 is this : — In all uniform bodies possessing cohesion, i. e. being 

 in either the solid or the liquid state, particles which are sur- 

 rounded by other particles having the like state with themselves 

 tend to preserve that state, even though subject to variations 

 of temperature, either of elevation or depression, which, if the 

 particles were not so surrounded, would cause them instantly 

 to change their condition. As water is the substance in which 

 regelation occurs, I will illustrate the principle by the phae- 

 nomena which it presents. Water may be cooled many degi'ees 

 below 33° Fahr.f and yet retain its liquid state for, as far as 

 we know, any length of time without solidification ; yet, intro- 

 duce a piece of the same chemical substance, ice, at a higher 

 temperature, and the cold water freezes and becomes warm. It 

 is certainly not the change of temperature which causes the 

 freezing, for the ice introduced is warmer than the water. I 

 assume that it is the difference in the condition of cohesion 

 existing on the different sides of the changing particles which 

 sets them free and causes the change. The cold water par- 

 ticles would willingly, as to temperature, have solidified without 

 the ice, but were held fluid by the cohesion with them of other 

 like fluid particles on all sides. 



In the other direction, Donny's experiments have taught us 

 that the cohesion amongst the particles of water is so great 



* Comptes Rendus, 1850, xxx. 626. 



t Water may be cooled to 22° F. It is probable that if it were perfectly 

 freed from air it would remain fluid at a much lower temperature j for the 

 air is excluded at the freezing-poiut, aud the occurrence of this exclusion 

 would break cohesion. 



