( 37 ) 



given by Keksom, had already been drawn up by me (May 25, 1905, 



p. 652) in the identical form : 



RTv* = 2 [a (1 -.<•) (ar — \/ay + a (p-6) 1 ], 

 where « = ^/n, — |/a, and (3 = ft 2 — 6 l . 



2. The answer to the question whether the plait extends from 

 C„ to C\ with or without double point in the spinodal curve, i. e. 

 with or without minimum plaitpoint temperature, in other words the 

 answer to the question whether the plait passes from C\ to C\ un- 

 divided, or whether two plaits extend on the if'-surface, one starting 

 from C a , the other from C„ which meet at the minimum tempera- 



T t T, 



ture — depends on the value of 6 = — (on which also — - depends) 



1 



for given value of at = — . The condition for this I derived in 



/'i 

 Aug. 17, 1905, p. 150, and Jan. 25, 1906, p. 581. In the summer of 

 1906 I calculated the place of the minimum itself (Cf. Oct. 25, 1906, 

 234, line 18 — 16 from the bottom), but seeing that the paper, which at 

 that time had already been completed and sent to the editor of the 

 Arch. Teyler, has not yet been published (it may be even some time 

 before it is), I think it desirable to publish already now the calcu- 

 lation in question. 



Like the calculations of Keesom, Verschaffelt and others, it starts 

 from the supposition that a and b do not depend on v or T, and 

 that these quantities may be represented by 



a x = [(1-a) y'a, + .v \/a,f ; b x = (1—x) b, + a b x . 



So in conformity with Berthelot and others we assume that 

 a,, = Va x a s . Some time ago Prof, van der Waals raised his voice 

 against this supposition '), and it seems to me that there is really 

 much to be said in favour of a„ being in general not = V'a^a^, 

 But as a first approximation the equation put may be accep- 

 ted, the more so as also the variability of b with v and T is 

 neglected. That in consequence of the assumption « 15 = Va x a % the 

 left region, mentioned by van der Waals, would be compressed to 

 an exceedingly small region, can hardly be adduced as an argument 

 against this supposition ; rather the fact that the attractions are 

 specific quantities, and that therefore e 12 need not be = ^i*,- 



For the calculation of the minimum we start from the equation 

 of the spinodal curve, derived by us (loc. cit.) ; 



!) These Proc, March 2ö, 1907, p. 630—631. 



