( 172 ) 



qnjiiitily of the coexistiiii»' Üiiid pluiso, the volimic remaiiung coiistiiiil. 



It may be said ot' tliis (iiiaiitit) that it is negative as a rule, and 

 so, tiiat during the process mentioned no loss of energy, but increase 

 of energy takes place, and that this negati\e value is generally so 

 great, that it far exceeds the first term of the second member of 

 equation (4), wliich is generally negative, but which can also be 

 positive, so that ir^f is negative. 



Yet this is not nl/ntj/s the case. In the first place an exothermal 

 chemical reaction may take |)lace during the solution of the solid 

 phase in ('onse(pience of which the energy at constant volume does 

 not inciu'asi', but decrease, so that (a../),, becomes positive, and when 

 this term exceeds the negative value of the first term, //",,/• is positive. 



As the reactiou is here an exotheruial one, the cheuiical ecpiilibrium 

 will be shifted in the endothermal direction with rise of temperature, 

 in consequence of which (?>;/•),; will become less strongly positive, and 

 can finally become negative again. A consequence of this it will be that 

 Wsf passes through zei'O at rise of temperature, aftei- which it becomes 

 negative again. 



In the second place, as van dkk Waals 8r. ') showed, it may be 



fdr\ 



derived trom the couj'se ol the isobars I that the line 7'_s/= 



gets outside the connodal line in the neighboui-hood of the plaitpoint 

 (Z/ = 6' -}- S). As now this quantity is negative outside the locus 

 v,ƒ=:0, whereas it has a positive value inside it till the line 



— =z is reached, where it increases up to x and becomes ayain 



negative inside this latter locus, we can predict that in the neigh- 

 bourhood of the mentioned plaitpoint the quantity ?;.,ƒ can assume 

 rather high positive values in the stable region. 



In this case the term 





i\,; will begin to preponderate, 



and this too may bring about that the quantity iv^f becomes zero, 

 and then negati\e, hut this reversal of sign is confined to the critical 

 neiyhbourhood, irltereas the revt^rsal discussed in the first place is only 

 possible, when the phenomenon of solution is accornpanied by an 

 ed'othermal chemical reaction. 



Let ns once more consider after this introduction how to explani 

 the phenomena observed for the system H,0 — Na.^S04. 



1) These Proc. Uct. 1903, p. 280, Nov. 1903, p. 357. 



