( 439 ) 



1 , 2 



small. So / — 1 is equal to - e.g. for;/ = 2. For» = 3,/ — 1 = — 



40 15 



But this is no longer the case for high value of n. We need 



not fear in any case, however, that / will become so great that 



o, + fl,-2a„ would become <0. That c be > 0, the following 



equation must hold : 



2a lf <a, -f a, 



or 



or 



Ol' 



or m our ease 



or 



2Z^/a 1 a,<a 1 +^ s 



1 



2/ < n -f - 



// 



(/-1)< V 



Now 



(/-1) = 



2n 

 («— l) a £, 



2m l+8 t 



Hence (/ — l) maj remains also below the value, which would make 

 a \ + " s ~~ 2a, , equal to 0. 



lhii now before proceeding to the comparison of the results 

 obtained here with those of the experiment. 1 shall first have to 

 discuss the question whether the disappearance of the intersection of 



= and =0 really involves the disappearance ot the com- 



plication in tin- spinodal line. and so whether the temperature 

 at which the two curves mentioned touch, is at the same time the tem- 

 perature at which the pair of heterogeneous plaitpoints occurring in 

 the spinodal line, coincide. When the points of intersection of the two 

 curves approach each other, the two heterogeneous plaitpoints will, no 

 doubt, also come nearer to each other, I>n< il need not follow from 

 this that when the points of intersection coincide, also the pair of 

 plaitpoints coincide. And a priori it is unlikely that this should be 

 the case. The existence or non-existence of points of intersection 

 depends only on properties of the two curves, without a third curve 



29 



Proceedings Royal Acad. Amsterdam Vol. XI. 



