560 
in the Pe-diagrams is strongest in the part of the figure at the 
side of the component with the smaller vapour tension’), hence at 
the iodine side. Whereas the numerator of the concentration frac- 
tion is now declining, ifs denominater gets larger when we have passed 
the narrowing in the melting diagram, favourable cooperation for 
allowing the value of the concentration fraction to rapidly fall 
. 
GS ; 
again to that of ——. The point 7, will, therefore, be rapidly 
QLs 
attained. Indeed it appeared in the research of Dr. HerLDERMAN and 
myself that this point is attained at a temperature about 4° above 
the melting interval of the mixture vg—0.50 namely at rs= 
about 0.54. From the Zr side of figure 3 we can now readily read 
that the concentration fraction still continues to decrease but afterwards 
gradually rises again in consequence of the fact that 2s—7,, decreases 
more rapidly than #s—agq. Hence, the possibility of a second maxi- 
mum 7’, is created. 
We will not discuss here any possible intermediate cases where 
a degeneration of maximum and minimum to points of inflection may 
take place. If will be quite evident now, tliat, where the narrowings 
are decisive for the appearance of the minimum, the above mentioned 
configuration occurs the more decidedly when the compound is less 
dissociated. | 
5. We will now call attention to the difference in behaviour 
between a compound as explained by Fig. 2 and Fig. 3, that is to 
say between a dissociating compound and one miscible with its 
dissociation products in the solid condition. If such a compound 
without formation of mixed erystals is heated to fusion at a constant 
pressure the equilibrium will be quasi-unary *), the compound has 
a sharp melting point. With the compound which forms mixed 
erystals such is not the case,,for such a compound has a melting 
interval. In connexion therewith the equilibria at their own vapour 
tension, as read off from the P7-diagram are also more complicated. 
First of all let us remember that in the point /” of figure 2 
IP Qs 
8 
PL ==es and that consequently in equation (2) becomes 7’ — 
3 d1 Vis 
*) The expression quasi unary, quasi-binary etc seems very appropriate 
for characterising the condition in which a system behaves as if it possesses 
one variable (or two, three etc.) less than indicated by the phase rule. (As to these 
equilibria see BAKHUIs RoozeBoom, Heterogene Gleichgewichte I pg. 34 and 
following). 
