igo5~\ Mills — Molecular Artraction. 97 



working- with Biot's formula we concluded to use these values 

 and test more thoroughly the equation, 



[9] r^-^D- = C ° nStant = / - 



discussed in previous papers 1 . If the value of the ~^=- from 



equation 3 be substituted in the thermodynamical equation 2, 

 the values of the heat of vaporization so calculated are the 

 same as those obtained from equation 1. Therefore using 

 these values and substituting- them in equation 9 we obtain 

 the corrected values for the constant. The results so obtained 

 at those points nearest the critical temperature are given in 

 Table 2 under the heading "New". For comparison we give 

 under heading "Old" the values that we had previously 



8P 

 obtained for y! at this same temperature when using the ^= 



o 1 



obtained from Biot's formula in calculating the heats of 

 vaporization thermodynamically. Also we give, under the 

 heading "Mean", the average value of the constant pre- 

 viously adopted. Agreement is not of course to be expected 

 for the four associated substances last shown in the table. 

 For the other substances there is little question but that the 

 constants so obtained agree with the mean values to within 

 the limit of experimental error except in the cases of di-iso- 

 butyl, normal heptane, normal octane, and methyl acetate. 

 We are unable to explain the smaller values obtained from 

 these four substances, but the proof that equation 9 does hold 

 as near the critical temperature as it is possible for measure- 

 ments to be made is now complete for ethyl oxide, di-iso- 

 propyl, isopentane, normal pontane, normal hexane, benzene, 

 hexamethylene, fluo-benzene, carbon tetrachloride, stannic 

 chloride, methyl formate, and ethyl formate. 



i Jour. Phys. Ohem., April, 1902; June, 1904; Dec., 1904. 



