﻿956 W. Sukhodski on the 



respect to K and T. In the absence of anything better, 

 I would tentatively suggest as a working hypothesis — 



il BK = 1 W r9V „ 



K dT I d'T w 



This working hypothesis cannot be considered as con- 

 venient at least for a great interval of temperature, for upon 

 integrating equation (2) we obtain f 



K = CV (3) 



After substituting (3) in (1) Ave obtain farther 



°> K = T (!?)„ w 



K=/(»)T C , (5) 



Z=(V(«)T* (6) 



It is difficult to consider these (3), (5), (6) equations as 

 true for the sole reason, that at T K (critical temperature) 1 = 0, 



but K— — jj, which is not zero. 



V K 



But the introduction of any working hypothesis may be 

 superfluous. This will be obvious, if we remember the 

 deduction of equation (1)J. The fundamental assumption 

 of this deduction is that b is independent of temperature. 

 If b is dependent on temperature, we obtain on differentiating 

 the van der Waals' equation with respect to T, keeping the 

 volume constant, 



Mr. McC. Lewis has neglected ——, (^ \v compared with 1. 



It is true that van der Waals assumes l^J = §, but he 

 supposes also (s^p) = 0* 



The van der AVaals' equation does not respond to the 

 observed facts with sufficient precision within large limits. 



* Trans, of Faraday Soc. vii. p. 113. Cf. Phil. Mag. xxii. p. 194. 



t In the following C 1; 2 , etc. mean constants. 



X Phil. Mag. xxii. p. 269. 



§ Lehrbuch der Thermodynamik, von v. der Waals, 1908, p. 78— ". . . lasst 

 sich zeigen das die Einfiihrung von a als Temperaturfunktion mit der 

 innern latenten Verdampfungswarme in Konflikt kommt." Ibid. p. 80— 

 "Eine Abhangigkeit von b von der Temperatur wird durch diese Ueber- 

 legung unwahrscheinlich gemacht."' 



