Generalizations of Van der Waals. 175 



already been published (loc. cit.) ; but there are one or two 

 obvious small irregularities, and I have therefore mapped the 

 ratios against temperature and constructed curves for each 

 substance (Plate I.). The smoothed values read from the 

 curves are given in this paper, and the molecular volumes of 

 saturated vapour have also been recalculated from the density 

 ratios. The same method of procedure has been adopted with 

 the other substances. 



The close similarity in the behaviour of the halogen deri- 

 vatives of benzene and the approximate agreement in the case 

 of the members of Group II. is very clearly shown in Table 

 XVII. 



It is generally admitted that gaseous acetic acid at low 

 temperatures contains molecules more complex than those 

 corresponding to the ordinary formula C 2 H 4 2 , and from the 

 very high density at the critical point it would appear that 

 many of these complex molecules have escaped dissociation 

 even at this high temperature. 



It will be seen that with the alcohols the differences from 

 the other substances become generally more marked as the 

 critical point is approached. At the highest pressure the 

 vapour-density of methyl alcohol is very distinctly higher 

 than that of any member of the first or second group at the 

 corresponding pressure ; and this fact appears to favour the 

 conclusion of M. Guye and others, that some of the molecules 

 of methyl alcohol at the critical point and in the liquid state 

 at all temperatures are more complex than in the ordinary 

 gaseous state. On the other hand, by a comparison of the 

 densities of the saturated vapours of acetic acid with those of 

 the alcohols and ether, Dr. Ramsay and I were led to the con- 

 clusion that the molecules of ordinary substances, including 

 the alcohols, are not more complex in the liquid than in the 

 gaseous state. This conclusion, so far as the members of 

 Groups I. and II. are concerned, is strengthened by the results 

 given in the preceding Table ; but it is certainly weakened 

 to some extent in the case of the alcohols, or at any rate of 

 methyl alcohol. 



There is no doubt that these bodies do show marked differ- 

 ences in many of their properties from the majority of com- 

 pounds; and the most plausible explanation of these differences 

 seems to be the existence of complex molecules in the liquid 

 state, although there is ample proof that they are not present 

 in the saturated vapours at low temperatures. 



As the alcohols differ so widely from other substances, it is 

 of interest to find whether the relations of Van der Waals hold 

 good when they are compared among themselves. I have 

 therefore calculated the ratios of the absolute temperatures, 

 pressures, and volumes of ethyl and propyl alcohol lo methyl 

 alcohol and of propyl alcohol to ethyl alcohol. The results 

 arc given in the following Tables: — 



