ÉQUATION AND THE NATURE OF COHESION. 19 



But C' is equal to a/BT c F c , and (F C /(F* C — b c )) BT e V c \a from 



p J_ ajjr-2 

 van der Waals' equation is equal to c ', •'. c . Hence 



Z — B K c a/F ( 



A similar result was arrived at by Dieterici 1 ), namely, that 

 ajV 2 =6.4 P c and in the following way. He started with two 

 results, namely, that of Young that in many normal substances S 

 is equal to 3.7. Then he proved experimentally for C0 2 , pen- 

 tane and isopentane that (dPjdT)T c = 2 BlJ' c . The only assumption 

 which will harmonize these two results is that a\V 2 = 6.4 P c . 



It might be pointed out here that Lj{L — E) c according to van 

 Laak ought not to be the same in all substances, as Ave have found 

 it to be. .He finds for example that f = BT C / (P c ( V c — b e ) = 

 1 -\- a r j P C V c 2 = Sy. Now y he thinks varies from 0.6 in H 2 and 

 0.5 in helium to 0.9 in the more complex substances. He does not 

 believe that a is independent of the temperature. The calculations 

 just made lend no support to this view of van Laar. They indicate 

 that 8 y should everywhere be equal to 7.5. 



I wish now to show that even the triatomic gases come approxi- 

 mately to this same ratio of 7.5, that is the ratio of (K c -J- P c )/P c . 

 We will start with Crompton's equation of the total latent heat 

 modified in the manner just stated, namely: 



L = U\l{V c -b,))RTLn c {dlD) (5) 



Substituting at T c the value (P c -f a/F c 2 ) for BT C /{F— -b c ) we 

 have, at the critical point: 



= p 9 r 9 + air. (6) 



Lu (d/D) 

 At other temperatures we have 



L = (Pc + ^ Fc2)Fe TLn e (d/B) (7) 



1) Dieterici: Wied. Annalen d. Physik, 12, 144 (1903). 



D 2* 



