191 



1874.] 1-i.X [Marsh. 



that expansion was effected. The coldness of the higher atmospheric 

 regions was accounted for by reference to the expansion of the air. It 

 was thought that what we have called the "capacity for heat" was 

 greater in the case of rarefied than of unrarefied gas. But the refrigera- 

 tion which accompanies expansion is, in reality, due to the consumption 

 of heat in the performance of work by the expanding gas. Where no 

 work is performed there is no absolute refrigeration." 



A sufficient answer to both these would seem to be found in the fact 

 that the " vacuum " spoken of is stated by Joule to have been obtained 

 by means of an air pump ; whence it appears that both vessels \f eve filled 

 with air ; that in the exhausted receiver having already, during the pro- 

 cess of exhaustion, absorbed and rendered latent, all the heat necessary 

 for its expansion ; and since we have already seen that the amount of 

 latent heat in a given volume of air is almost entirely independent of 

 density, we have no reason to look for any loss or gain of latent heat by 

 the operation. The mixing of the two is quite as much a process of con- 

 densation as of rarefaction ; in one receiver the air in expanding absoi'bs 

 heat, whilst in the other the air being compressed gives out the heat which 

 it had absorbed during the process of exhaustion — the two effects counter- 

 balancing each other. The air, as a whole, just tilled the two receivers 

 at the beginning of the experiment, and it filled the same at the end ; 

 so that the effects of expansion and of condensation were completely 

 eliminated : even more so than those of mechanical power, which Joule 

 had especially in view when contriving this experiment. 



Whilst the above seems to show that this experiment ^Droved nothing 

 as to the existence or non-existence of latent heat of expansion, any one 

 who will read Joule's paper will probably be convinced that he never in- 

 tended to claim that it did. He makes no allusion to any such question ; 

 and the limit which he gives of the sensitiveness of his thermometer 

 shows that an amount of heat seventy-five times as great as that which 

 could be expected to be rendered latent by doubling the volume of only 

 two quarts of air, of any density, would be required to produce any 

 appreciable change in the temperature of the 16^ lbs. of water through 

 which, he tells us, it was distributed. He was dealing with mechanical 

 power, and took care to have it in quantities large enough to be traced ; 

 his aim seeming to be, to solve the general question of their convertibility 

 into heat, rather than to determine whether the results might, or might 

 not, have been modified in some degree, by latent heat or other disturbing 

 cause . 



These considerations seem to justify us in concluding that the question 

 of the existence of "latent heat of expansion" has not been experiment- 

 ally decided in the negative, and that we may therefore proceed to 

 inquire into its applicability to the explanation of meteoric phenomena,* 



* When this paper was read, I was not aware of the language of Joule himself on this 



subject. Part II. of the article " On the Thermal Effects of Fluids in Motion," by J. P. 



.Toule and Prof W. Thomson, (Transactions of the Roj'al Society, 1854, vol. 144 p. 337) 



speaking of the " Relations between the Heat evolved and the Work spent In Compressing 



A. P. S. — VOL. XIV. F 



