( 902) 
Pv . gh ; 
eS is positive on the side of the small volumes. On the side of the 
U /J bin 
dv 
great volumes & is also positive, and so turned from the plait, 
U / bin y 
do . 
but É a) is negative there. For the- proof it would be necessary to 
av 
Pp 
examine how a cone differing little from a plane, resting with its 
apex in an horizontal plane, and the generatrices of which ascend 
towards the side of the small volumes, envelops two convex-convex 
: : dv 
parts of a surface. The points of contact give the value of il 
Ak” / bin 
If we examine besides how a cylindre the generatrices of which run 
parallel to the generatrix of the cone, as it is in the plane parallel 
to the v-axis, envelops the two parts of the convex-convex surface, 
d*v 
we find the value of ES I draw attention to the fact that for 
TL P 
the mixture water and phenol SCHREINEMAKERS has experimentally 
shown that at the temperature at which the point C exists, the 
relative situation of the two points is, as is given here as a rule of 
av 
i 
general application. On the liquid side Ee is so large negative that 
x 
d'p\ . ie 
unless ee is also very great negative, the difference between 
1X bin 
2 2 
(5) and i) is scarcely appreciable. And the first of these 
dx* ) bin U /p 
quantities being positive, this is also to be expected for the second. 
But yet thermodynamically or purely mathematically this property 
is not to be generally proved. It follows, however, if the equation 
of state is applied. If we namely draw the locus of the points of 
: dp ‘ dp é 
intersection of the two curves Dn = 0 and = 0, we find 
av da 
1 db 
v b dz ; : 
eas? a and for the locus of the points on the binodal line, where 
a 
1 db 
p is maximum, we find for very low temperatures B 
) a 
a de 
In fig. 46 PA represents the first-mentioned locus and PA’ 
4 * 
dean had 
