EQUATION AND THE NATURE OF COHESION. 55 



The value calculated from the molecular weight and with 48 

 valences is 37.82 X 10 12 in agreement with the value computed from 

 the surface tension. 



43. Cymene? C\ H l 4 . 



[a), a = 3 so 2 '* J/iV 1/3 T r /{T r — T) d u . 



M = 134. 1. T r calculated is 051. 6°. d = .862 — 



. 0008044 (/ — 11.9°)wy 2 » 3 (Morgan and Thomssen) is 



785.20 which is equal to 697.8 ergs = sv 2ls . T c — T 



is 319.15. d= .8235. /is 59.45°. ^„calculated from the 



above formula is 1.0912. With these values a = 44.44 X 10 w 



a calculated from the molecular weight, 134.1, and the valences, 



54, is 44.03 X 10 12 , which is in agreement with the value calculated 



from the surface tension. There is no indication in these last two 



calculations but what the valences are strictly additive in their effect. 



44. Dipïienylmethane. (\zH i2 . 



The past several agreements have been so good, between the 

 values calculated from the surface tension and those from the mole- 

 cular weights and valence numbers, that one might think that one 

 would always have so good an agreement. I have appended, there- 

 fore, a computation of diphenylmethane, in order to include all the 

 computations I have made and to show that one cannot always 

 expect so good an agreement Here we have G4 valences and a 

 molecular weight of 108. This would lead us to calculate by the 



formula a= 1.1763 X 10 11 (16S X G4 ) 2/:? the value a = 57.33 X 10 12 . 

 From the surface tension, however, we calculate a much higher 

 value. At 59° Morgan and Thomssen give icv 21 ' 3 =■ 1 171.26. From 

 this Ave calculate sv 213 = 1041 ergs. t r is computed by Morgan 

 and Thomssen to be 497° which Avould make T r = 770. d = 

 1.0126 — 0.0007914(/— 11)=. 9746. d„ = 1.2365. Hence by 

 the formula given in cymene we compute : a = 63.16 X 'O 12 , which 

 is about 9 °/ above the other value. However as both T c and d 

 have been extrapolated from a great distance, no great reliance can 

 be placed upon them. I may say that Dutoit and Friederich found 

 the constant G of Eötvös to be 2.25 in this substance between 108 

 and 210°. This would give the value 59.80 X 10 12 for a which is 

 only 3 — 4 % higher than the computation from M and the valences. 

 It is of interest to compare the values of a in these more com- 

 plex substances, computed in the usual way from the critical temp- 



