some Tliermodynamical Relations. 35 



from the density of the gas the percentage number of mole- 

 cules decomposed. But there is no means of knowing whether 

 the dissociating substance, as such, does or does not obey 

 these laws ; hence it is evidently desirable to ascertain the 

 behaviour of substances which do not dissociate as regards 

 expansion and compressibility : so that some idea may be 

 gained as to the influence exerted by molecular attraction 

 between like molecules on the volume they occupy. An in- 

 vestigation of this nature is best carried out by ascertaining 

 the behaviour of gases when they are becoming increasingly 

 subject to molecular attraction, i. e. when they are approach- 

 ing the condition of saturated vapours. The pressure of 

 saturation is, under normal circumstances, identical with 

 the vapour-pressure ; hence the importance of a knowledge 

 of this quantity. We have before remarked, that we have 

 communicated to the Royal Society the results of experiments 

 on alcohol and ether with this view. 



Again, if the body exerting vapour-pressure be one capable 

 of partial dissociation on passage from the solid or liquid to 

 the gaseous state, its apparent vapour-pressure will really 

 consist of its real vapour-pressure, and also of the pressure 

 due to its dissociation into its constituents. It is evidently of 

 importance to determine the latter independently of the former; 

 and it is only by the employment of some law applicable to 

 all bodies, and deduced from the behaviour of stable sub- 

 stances, that an idea can be formed of the pressure due to the 

 second cause. 



In the Philosophical Magazine for April 1886, p. 299, 

 Professor Unwin also comments on our papers, and, unlike 

 Professors Ayrton and Perry, has appreciated our motive in 

 bringing forward the subject. We have to thank him for point- 

 ing out the relation of the ratio of the total heat of vaporization 

 to the heat expended in external work, deduced from the second 

 half of the thermodynamic equation. We agree with him that 

 the approximate constancy there observed lies at the basis of 

 all the relations found by us. 



We have taken the liberty of applying his formula 



t n+1 dp 

 p dt 



to alcohol, ether, and mercury, using the vapour-pressures 

 determined by us over a wide range of temperatures in the 

 first two cases ; and for mercury, the results given in the 

 Journ. Chem. Soc. 1886, p. 37. 



D2 



