( 896 ) 
d, 
on the left of 5 = 0 and which has been drawn in fig. 44, is 
U) yv 
not to be expected here. A third point of intersection of these lines, 
: dp 
also drawn in fig. 44, on the right of é = 0 and for smaller 
Ak v 
Fig. 44. 
d 
volume than (3) = 0, is of no importance for the splitting up of 
LV] x 
the spinodal eurve — because no points of this line can be present 
there. There the p- and q-lines can, namely, not touch. The p-lines 
run there almost parallel to the w-axis, and the g-lines, on the other 
hand, almost parallel to the v-axis. It follows from all this that if 
T;, has a minimum value, the splitting up of the spinodal line will 
take place into what we might call, a righthand branch and a lefthand 
branch — at least in the case in which the component with the 
greatest value of 5 possesses the smallest value of 7. But this will 
also be so in the opposite case. Then, however, we must first make 
the observation that fig. 43 must be modified, or rather that we have 
then almost entirely the case of fig. 40 back. Then we have to choose 
a region from the general fig. 1, which begins just before the point, 
in which 7; has minimum value and further greatly extends to the 
right. Of course we have also to-choose the values n, ¢, and ¢, in 
~ 
ee a 
