( 690 ) 



oiu' plan is easily compared with the latter for the small volume 

 part. To find how our vapour plait will be connected with this 

 portion, it is only necessary to use pi. Ill fig. 2. Tlie chief difference 

 is seen in the different forms of the spinodal line. Ours is given 

 by the dot-dash line of fig. 6 and Maxwell's in fig. 7. Ours consists 

 of two portions which remain apart to the lowest temperature 



Fig. 6. Fig. 7. 



(T=0), while Maxwell's shows only a secondary loop. Quite 

 improbable are Maxwell's isotherms. They show in the vapour plait 

 a point of inflection in the isotherms. Thereby they differ entirely 

 from the equation of state of van der Waals, which is certainly 

 qualitatively correct for vapour and liquid. 



II. The Gibbs' surface for CO^. 

 {The general model). 



This model was constructed by the help of the empirical equation 

 of state given in Comms. Nos. 71 and 74 with the assistance of the 

 thermodynamical formulae obtained in No. 66. 



^oT — 



Tk vk Tk V T 



nvT — r^]c 



whence 



c^{T — Ti) — ^u Vk 



pkVk b, + 2 1-3 + 4 b, 



Tk vk Ik V ' T 



b, + 2(>3 + 4b, pkvk c,+2c3 4-4c, 



^+4 + 4 

 Pk v^• t t' 



+ 



Pk Vk" 



2 v^ 





+ 



(1) 



