496 Rev. J. F. Blake — On the Motion of Glaciers. 



The crystal will not form to suit the cavity, the cavity must be made 

 ^0 contain the crystal." Here the idea appears to be that the main 

 mass of the '' molecule " is in the interstice, which would, of course, 

 involve that ice should be only about half the density of water, 

 unless we resort to'the hypothesis of the interstices being of different 

 sizes, or occurring every now and then, and the results happening 

 only when the melted " molecule " happened to fall into a cavity 

 large enough to nearly hold it all. But it is of no use to follow 

 this further ; the whole matter is founded in a confusion about what 

 a molecule is, and such a molecule as Mr. Croll appears to conceive 

 has never been yet heard of in physical science. 



But going back to where we left off in strict language. Suppose a 

 number of molecules were to be loosed from their crystalline arrange- 

 ment by heat, and so occupy, on the whole, less space, on re-crystal- 

 lizing what is there to prevent them going back to their own place 

 again ? Unless some new force is brought to play on the instant, 

 whatever old forces they originally overcame in the act of crystal- 

 lizing the first time, they can certainly overcome again. But what 

 new force is brought into play ? None that Mr. Croll mentions. 

 Oravity always acted, and the molecular forces of crystallization 

 are just as competent to push back the molecules against it as they 

 Avere before when first the ice was formed. 



It is to be observed that there is nothing in Mr, Croll's theory to 

 distinguish a glacier from an ordinary piece of ice, and if one will 

 flatten out as he supposes, the other ought to do so too. But who 

 ever saw a block of ice bulge out under the influence of heat ? If 

 any one has ever seen such a thing, or has made any experiments 

 upon it, it would be far more to the point than theory ; or if these 

 molecular changes could go on, even in a large mass of ice, without 

 any vis-a-tergo, surely some tendency to a definite shape ought to 

 have been observed in icebergs, which should, " as the mass widens 

 out, grow thinner and thinner." If, therefore, we could substitute 

 for Mr. Croll's theory, as explained by himself, some other, physically 

 correct, embodying the same idea of molecular changes under the 

 influence of heat, it would, even then, prove too much. 



Hitherto I have assumed, with Mr. Croll, that heat acting on a 

 mass of ice could produce these meltings and re-solidifying of small 

 portions of it as distinct from those other surrounding portions, or 

 rather that such changes might somehow be brought about, and have 

 shown that thej'' can have no such effect as supposed. But I now 

 ask, What is the real result of heat acting on a mass of ice ? Can 

 there be such a thing as a small melted portion (I will no longer 

 use the word " molecule " in a wrong sense) passing on its heat to 

 " the next," by which means that becomes melted, and so passes on 

 the heat again ? It is certainly an ingenious idea, and one cannot 

 help admiring its boldness ; but then it has nothing whatever to do 

 with the real state of the case when a solid body such as ice is 

 subject to heat. We hear nothing of its latent heat, and of its con- 

 ductivity — two of the essential elements of the question. Before a 

 small mass, or, as we may call it, particle, of ice can melt, it must 



