142 



Dr. Tyndall's statement, that " the hazy surfaces produced by the 

 compression of the mass were observed to be in a state of iatense 

 commotion, which followed closely upon the edge of the surface as it 

 advanced through the solid. It is finally shown that these surfaces 

 are due to the liquefaction of the ice in planes perpendicular to the 

 pressure." 



There can be no doubt but that the " oscillations " in the melting- 

 point of ice, and the distinction between strong and weak pieces in 

 this respect, described by Dr. Tyndall in the second section of his 

 paper, are consequences of the varying pressures which different por- 

 tions of a mass of ice must experience when portions within it 

 become liquefied. 



The elevation of the melting temperature which my brother's 

 theory shows must be produced by diminishing the pressure of ice 

 below the atmospheric pressure, arid to which I alluded as a subject 

 for experimental illustration, in the article describing my experi- 

 mental demonstration of the lowering effect of pressure (Proceedings, 

 Roy. Soc. Edinb. Feb. 1850), demonstrates that a vesicle of water 

 cannot form in the interior of a solid of ice except at a temperature 

 higher than Cent. This is a conclusion which Dr. Tyndall ex- 

 presses as a result of mechanical considerations : thus, " Regarding 

 heat as a mode of motion," " liberty of liquidity is attained by the 

 molecules at the surface of a mass of ice before the molecules at the 

 centre of the mass can attain this liberty." 



The physical theory shows that a removal of the atmospheric 

 pressure would raise the melting-point of ice by 7 f ^ths of a degree 

 Centigrade. Hence it is certain that the interior of a solid of ice, 

 heated by the condensation of solar rays by a lens, will rise to at 

 least that excess of temperature above the superficial parts. It appears 

 very nearly certain that cohesion will prevent the evolution of a 

 bubble of vapour of water in a vesicle of water forming by this process 

 in the interior of a mass of ice, until a high " negative pressure" has 

 been reached, that is to say, until cohesion has been called largely 

 into operation, especially if the water and ice contain little or no air 

 by absorption (just as water freed from air may be raised consider- 

 ably above its boiling-point under any non- evanescent hydrostatic 

 pressure). Hence it appears nearly certain that the interior of a 

 block of ice originally clear, and made to possess vesicles of water by 



