182 J.] Causes of Calorific Capacity, Latent Heat, fyc. 461 



The numerical value of latent heat is not influenced by the 

 proportion of the quantities of steam and vapour only, but like- 

 wise, and even in a more eminent degree, by the temperature. 

 Other things being alike, the higher either or both of the temper- 

 atures are the greater will be the quantity oflatentheat. If the va- 

 pour be -^th of the water and its true temperature 1200, as I have 

 taken it before, that of the water being 1160 or 197-89° Fahr.the 

 latent heat will be 1218-78°. In a preceding example, when the 

 higher temperature and proportion of water to vapour were the 

 same, but the true temperature only 1000, the latent heat 

 appeared not to exceed 966-72°, that is, full 250° lower. Unhap- 

 pily I know of no experiments with which I can compare this 

 conclusion ; for the temperatures of the water in all the experi- 

 ments I know of are so near together, and the influence of the 

 inequality of weight in the water so decidedly overbalances that of 

 the trifling difference of temperature, that the effect of the latter 

 cannot be detected. There are, however, some philosophers, I 

 have no doubt, whose zeal for the advancement of science will 

 put this to the test of experimental proof, and by this means have 

 the honour of for ever putting at rest the merits of a doctrine 

 which, it has often appeared to me, reflects more credit on Black 

 for boldness of conception than soundness of judgment. 



It is proper to observe, before any experiments are made to 

 confirm these inferences, that the values of latent heat deduced 

 from the preceding theorem presuppose the admission of our 

 general law of temperature for simple mixtures. If this be 

 denied, it will require a different expression for the amount of 

 latent heat. For instance, taking from our third table t' to cor- 

 respond with F,, and t with F, the general expression for the 



latent heat in Fahr. will be ( " f '* " *' ~ ?" + »£ x -*5_. If 



u> + 1 100000 



we put in this expression I' = 1200 and t = 1000 as before, the 

 value of the latent heat when id = 9 comes out 1009-92°, and 

 when w = 99, it is 947-71°. So that here we have not only the 

 inference verified which I have drawn from my general theory, 

 and confirmed by the experiments of Ure and Rumford, but even 

 verified in a more striking degree ; for the difference is here 62^-° 

 instead of 58-±-°. 



However, though the proportion of weight has as great an 

 influence in this theorem as in the last or a greater, the variation 

 of temperature has by no means so^reat. For instance, if we 

 retain w = 99 and t = 1000 (32° Fahr.) the latent heat will be 

 978° by putting t' = 1250, which corresponds with 302° Fahr. 

 The increase of latent heat, therefore, otlier things being alike, 

 by putting the vapour at 302° instead of 243°, or 59° higher, is 1 1°. 

 Wow by Mr. Southern's experiments, an augmentation of 66° in 

 the temperature of the steam, that is, from 229° to 295°, gave an 

 increase of 8° in the latent heat. I need make no observation on 

 this new accordance of our theory with observation. Philoso- 



