54 THE TRUE VALUE OF a OF VAN DER WAALS' 



valences is much better than with the isomeric octane. The results 

 frojn the latent heat are not so divergent here. It may be noted 

 that the diisobutyl upon which these determinations were made, 

 does not appear to have been a pure substance, or else it slightly 

 decomposed on standing as Young and Mills have remarked. The 

 critical data, however, were probably not seriously affected by this 

 circumstance. Van Laar gives for diisobutyl no value, but one 

 computes from his formula the value a ,/2 = .288; which translated 

 into the other units would be 36.39 X l^ 12 which it will be 

 noticed is only 3 or 4 °/ below that required by the computation 

 from the molecular weight and valences. It is, however, more 

 divergent than mine from the values actually found. 



4 1 . Diisoa my I. C\ H 2 2 . 



(a), a = 3 MN ilA so''" T, 1/A /(d- D) (T r — Tf A . T r = 075. T c — 

 T= 402. ö?at 0°a = .7413. s= 23.072 dynes, sv 1 :i = 

 707.3 ergs. 1) disregarded. M= 142.Ï8. Surface ten- 

 sion by Morgan and Owen a = 52.79 X 10 12 



I could not find the critical pressure and density of this sub- 

 stance. This value of a is not very different from that calculated 

 from the molecular weight of 142.2 and the valences, there being 

 62 valences in the molecule. The value of a required by the latter 

 computation is 50.20 X Ï0 12 , which is only some 4 % below the 

 value computed from the surface tension. This is the most complex 

 substance I have calculated. This gives a total range of valences 

 from 2 in hydrogen to 62 in diisoamyl, and a range of molecular 

 weio-hts from 2 to 260.8 in stannic chloride. I think this is a 

 sulticient diversity of substances and a sufficient range to establish 

 the general applicability of the law relating cohesion to the valences 

 and molecular weight. I have, however, calculated from the surface 

 tension three other substances which are also complex. 



42. Mesitylene. C Q H~ V2 . 



{a), a = 3 w- :i MJY i/B TJ(T C — T) </„. From Morgan and Thomssen, 

 wd^ at 23.5° is 812.61. This gives for sv 2l?J the value 

 722.2. ergs. T r is calculated to be 640.7°. M is 120.1. 

 d x calculated by the formula: d= .8746 — .00081^ 

 is .8556. d by the same formula is calculated to 

 be 1.0958. T r — T is 344.2. This gives for a the 

 value a . = 37.41 X I0 ia 



