506 Determination of the Critical Volume. 



volumes of the twelve substances, adding, for the sake of 

 convenience, the critical temperatures and pressures. The 

 critical densities are those calculated by M. Mathias ; but in 

 the case of the alcohols I give also the values which I have 

 obtained by the same method, taking into account the cur- 

 vature of the " diameter.-" The critical volumes of a gram 

 and molecular volumes are calculated from the densities. 

 The last two columns contain the ratios of the new critical 

 molecular volumes to that of fluorbehzene, together with the 

 ratios given on p. 184 of my previous paper and obtained by 

 the graphical method there described. It will be seen that 

 these ratios show satisfactory agreement. 



As regards the alcohols it will be noticed that, if the cur- 

 vature of the diameter is taken into account, the critical 

 densities are lower and the volumes higher ; also that the 

 ratios of the critical molecular volumes to that of fluor- 

 benzene agree better with those obtained by the graphical 

 method. 



It may be pointed out that if the generalizations of Van 

 der Waals were strictly true, the angular coefficients (a) of 

 the diameters of different substances should be directly pro- 

 portional to their critical densities and inversely proportional 

 to their absolute critical temperatures, or that for any sub- 

 stance 



, D «T 

 a = const, x Tpj or yr- = const. 



I have calculated the values of this " constant '' for the 

 substances mentioned in this paper (with the exception of 

 methyl and propyl alcohol) and for the five esters studied by 

 Mr. G. L. Thomas and myself. 



Substance. — . Substance. — . Substance. — . 



Fluorbenzene '932 Carbon Tetrachloride '899 Methyl Formate... -997 



Chlorobenzene 953 Stannic Chloride .. -988 Methyl Acetate ... 1- 04 7 



Broraobenzene -957 Ether -972 Ethyl Formate ... 1061 



Iodobenzene '957 Acetic Acid -993 Ethyl Acetate 1-055 



Benzene '921 Ethyl Alcohol * 1030 Methyl Propionate 1*021 



It can hardly be said that there is more than a rough 



aT 



approximation to constancy in the values of w. 



I have recalculated the ratios of the actual to the theoretical 

 critical densities, and the results are given in the following 

 table :— 



* M. Mathias' value of a. 



