52 THE TRUE VALUE OP a OP VAN DER WAALS' 



consider these substances to be real exceptions to the valence rule, 

 particularly as chlor-benzene and benzene agree so perfectly. Fluor- 

 benzene. however, was also somewhat below the calculated value. 

 It may be that these substances are real exceptions, or that they 

 have fewer valences by one or two than we think in their mole- 

 cules, but the excellent agreement of chlorbenzene makes me believe 

 that probably the critical data are not quite right. 



Van Laar's value for iodobenzene is hopelessly wrong. He gives 

 for Va the value .261; which recalculated for dynes and gram mol 

 quantities makes 29.89 X 10 12 . This is between 15 and 20 °/ 

 too low. 



39. n-Octane. C 8 // 18 . 



(a). « = 2.03 J?T,.r ( ./l.0S — P r F c 2 . T c = 569.2; P,.= 18.730; 

 «?,, = . 2327; r f = 489.9; M= 114.. a = 39.72 X N> 12 



(Jj). a = 6.5 P,.F r 2 a = 37.94 X 10 12 



(e). a = f/M 2 ^ d?\ [Ï = 92.78 a = 44.46 X 10 1 - 



(d). a = G'BT C V C . Ü = 1.858 a = 43.10 X 10 12 



(«). a = sv"- 13 3 MN'V T c 2lz /{d—D) (T r — Tf\ 

 The ca])illary constant according to 

 Schiff at 110° is 4.23. From this the 

 surface tension, s is 25.99 dynes, d is 

 .6260. sv 2 ' 3 is 834.9 ergs. D is disre- 

 garded. d is .9119. From this we 



compute a = 40.50 X I0 12 



I also computed a from Ramsay and 

 Aston's surface tension data at 15.5°, 



this gave for a the value a = 40.67 X 10 12 



There is a great variation, a variation of fully 15°/ , between 

 the highest and lowest values. The latent heat formulae both give 

 a high value. Mills has suggested that there is some decomposition 

 at the high temperatures, the critical temperature being nearly 

 300° C. The mean value is 41.14, or if Ave omit the two high 

 values of the latent heat we have 39.39. The value of a computed 

 from the molecular weight and number of valences by the usual 

 formula, taking 50 valences to the molecule, is 37.54 X 10 12 . This 

 is close to the value computed from the second of the above for- 

 mulae, but nearly 8 °/ lower than the surface tension which is 

 usually a minimum value. It will be noticed in a moment that the 

 isomeric substance, diisobutyl, is much nearer the theoretical value. 

 It hardly seems probable to me that the true value is much below 



