Generalizations o/Ynn der Waals. 173 



rough approximations to the truth, but the deviations of the 

 ratios from constancy are in most cases much too large to be 

 attributed to error of experiment. 



3. The three Alcohols and Acetic Acid. — The majority of the 

 generalizations do not hold good at all ; the deviations of the 

 ratios from constancy are, however, not very great in the case 

 of the molecular volumes of liquid at corresponding pressures 

 and temperatures. 



Discussion of the Generalizations. 



1. Corresponding Temperatures and Pressures. — It must, I 

 think, be concluded that the statement that " if the absolute 

 temperatures of various substances are proportional to their 

 absolute critical temperatures their vapour-pressures will be 

 proportional to their critical pressures '' has not been proved 

 by experiment to be true except in a very limited number of 

 cases; indeed, when the alcohols and acetic acid are compared 

 with fluorbenzene, the statement is quite wide of the mark. 

 It follows, therefore, that in the comparison of the molecular 

 volumes the ratios at corresponding pressures must differ from 

 those at corresponding temperatures. 



2. Molecular Volumes of Liquid. — The deviations of the 

 ratios from constancy are smaller in this case than in any of 

 the others, but are certainly not within the limits of experi- 

 mental error. The comparisons at corresponding pressures 

 seem to be somewhat better on the whole than at corresponding 

 temperatures, but the difference is not very marked. 



3. Molecular Volumes of Saturated Vapour.— At corre- 

 sponding pressures the deviations from constancy are within 

 the limits of experimental error in the case of ether and 

 stannic chloride, and are relatively small for benzene and 

 carbon tetrachloride. 



There can be no doubt that with the saturated vapours the 

 comparisons at corresponding pressures are very much better 

 than at corresponding temperatures ; and it may therefore be 

 concluded that it is better to compare the molecular volumes 

 both of liquid and of saturated vapour at corresponding- 

 pressures than at corresponding temperatures. 

 Saturated Vapours. 



It has been pointed out (Trans. Cliem. Soc. 1891, p. 137) 

 that if the generalizations of Van der Waals were strictly true, 

 it would follow that the ratios of the actual densities of the 

 saturated vapours of different substances to their theoretical 

 densities should be equal at corresponding pressures : and as 

 this method of comparison is a very convenient one I have 

 thought it worth while to tabulate the ratios, although it has 

 already been shown that the relation cannot be strict lv true. 

 The values for benzene and its halogen derivatives have 



