93 Prof. Tyndall on Ice and Glaciers. 



star-shaped masses are formed exactly similar to the liquid spaces 

 produced by the sunbeam in my experiments. The process of 

 liquefaction is simply an inversion of the process of solidification ; 

 and to me it seems perfectly natural that the phjenomena should 

 be of this complementary character. 



Prof. Thomson also objects to my conclusion, that it might 

 be inferred from the experiments on internal liquefaction that ice 

 was a uniaxal crystal*. I look at the matter purely from a phy- 

 sical point of view ; and thus regarded, nothing, I think, can be 

 more certain than the conclusion referred to. A line may be 

 drawn through a mass of ice, round which the particles are 

 arranged with perfect symmetry ; and there is but one direction 

 through the substance round which this symmetry prevails. To 

 me it appears certain that this line will be the optic axis of the 

 crystal, and that the crystal will possess but one such axis. The 

 liquefaction experiments prove the facts ; and all that we know of 

 optical phrenomena confirms the conclusion. 



In the last paragraph but one, Prof. Thomson writes as fol- 

 lows : — "I believe the theory I have given above contains the ex- 

 planation of one remarkable fact observed by Dr. Tyndall in 

 connexion with the beautiful set of phsenomena which he dis- 

 covered to be produced by radiant heat, .... the fact namely 

 that the planes in which the vesicles extend are generally jmrallel 

 to the sides when the mass of ice is a flat slab ; for the solid will 

 yield to the ' negative ' internal pressure due to the contractility 

 of the melting ice, most easily in the direction perpendicular to 



the sides Hence the vesicles of melted ice, or of vapour 



caused by the contractionof the melted ice,must, as I have shown, 

 tend to place themselves parallel to the sides of the slab." 



Now the fact is, that the melting of the ice is totally inde- 

 pendent of the sides of the slab. If the sides do not coincide 

 with the surfaces of freezing, the planes of the flowers will not 

 be parallel to the sides. The effect is not due to any mechanical 

 weakness dependent on the form of the mass operated on; but the 

 interior melting is solely dependent on the interior crystallization. 

 No matter how irregular in form the fragment of ice may be, the 

 sending of a calorific beam through it reveals at once its planes 

 of freezing ; for the flowers always form parallel to these planes, 

 no matter what the direction of the beam through the mass may 

 be. This I have distinctly stated. I may add, that in the 

 single case recorded in my paper to which Mr. James Thomson's 

 theory is applicable, I have applied that theory myself. 



Prof. J. D. Porbes has recently described before the Koyal 

 Society of Edinburgh some experiments that he has made on 



* Phil. Trans, part 1, 1868. Phil. Mag. Nov. 1858, p. 33G. 



