3/2 M. H. Sainte-Claire Deville on Affinity and Heat, 



meter, may be expressed by a very simple fraction of the heat of 

 contraction, if we assume that the specific heat is invariable be- 

 tween the temperature and 6, which is never the case unless 

 6 is very small. On this hypothesis it becomes 



1-|=7T. 



Considering one and the same liquid, we may, in consequence 

 of the errors introduced by delicate and numerous determina- 

 tions, suppose the temperatures to be proportional to the vires 

 vivcBj or to the square of the velocities in the molecules the move- 

 ment of which produces heat. The number tt represents, then, 

 the fraction of this vis viva which has not been expended at the 

 moment of the combination or of the solution. It is, if we like, 

 the fraction of the heat of contraction which has remained latent 

 in the product of the reaction, and tt is positiye. 



In the experiments I have terminated I have only found one 

 single exception ; and that is in the case of the formation of sul- 

 phate of soda by means of two dilute solutions, one of sulphuric 

 acid and the other of caustic soda. Sulphate of soda expands at 

 the moment of its formation, and t is greater than 6 } so that it 

 becomes negative, which is a nice point to explain. But the 

 study of the phenomena of saturation in these solutions will, I 

 hope, enable me to explain this abnormal phenomenon, like the 

 contraction of water by heat, or the expansion of bismuth at the 

 moment of its solidification, &c* 



The contraction observed on mixing with water monohydrated 

 acetic acid or saline solutions is considerable, and the cooling of 

 the combination or solution very noteworthy, as had already been 

 observed by ftudberg, if I am not mistaken. In that case /be- 

 comes negative, and it is greater than unity. Thus at the mo- 

 ment of the reaction the product at first renders latent all the 



which appears worthy of attention. In order that these temperatures 

 may be exact, not merely the variations of the specific heat C between 6 

 and t must be known, which is determined by the ordinary methods, but 



P 



also its variations for pressures comprised between V033 kil. and — , S being 



the section of the cylinder AB. It will be conceived that this second class 

 of variations might be determined by observing the velocity of cooling of 

 liquids placed in thermometer-tubes closed successively at the ordinary 



p 

 pressure and the pressure — . 



Under these conditions the mechanical equivalent would be obtained by 

 replacing the compression of gases by the compression of liquids ; the me- 

 chanical agent would be simply replaced by the chemical action. 



* Since writing these lines, M. J. Regriaiild has made observations on 

 this subject which lead to a system of explanation far superior to my own. 

 (Journal de Pharmacie, 4 ser. vol. i. p. 401.) 



