igoj] Mills — Molecular Attraction. ( ) ( > 



and to a less degree for ethyl formate, the agreement with 

 the mean values of /u.' is excellent and we may therefore regard 

 the truth of equation 9 as having been established at the 

 critical temperature itself for fifteen of the nineteen sub- 

 stances under consideration. (Ethyl formate decomposes 

 slightly at the higher temperatures which is sufficient to 

 account for the divergence, three per cent., there observed.) 

 We have already shown that normal heptane and normal 

 octane give constant values for i*.' in equation 9 over a range 

 of more than 200° C in temperature. We have unpublished 

 results showing the same to be true for methyl acetate. Di- 

 isobutyl did give a variation of several per cent, at low tem- 

 peratures, a divergence that we think was sufficiently 

 explained 1 . It would seem probable therefore that the diver- 

 gences shown by these four substances at the critical temper- 

 ature must be due to some change or decomposition taking 

 place in the substance at that temperature and that equation 

 9 is applicable for all normal substances quite up to the criti- 

 cal temperature. 



RATIO OF THE THEORETICAL TO THE ACTUAL CRITICAL 

 DENSITY. 



In a preceding paper 2 we showed that the molecular attrac- 

 tion at unit distance, /x, was equal to Cfi'f / ~m~. Therefore we 

 have from equation 12, 



[13] p = c fJTV \ 9 * 4T - -0 4 955 P V j. 



In the same paper on the assumption that the critical tem- 

 perature was the point where the kinetic energy of the mole- 

 cules was just balanced by the molecular attraction we 

 derived equation 24 of that paper, viz. : 



1 Jour Phys. Ohem., 8, 595 (1904). 



2 Jour. Phys. Ohem., 8, 630 (1904). 



