( 826 ) 
Ò 
line p,,,, and that in such a way that (35) is greater after the 
x 
0 
intersection than before the intersection. If (4) should be equal to 
zero before the intersection, this quantity is again positive after the inter- 
section. If it should be negative for a section of still smaller value of 
x, it is negative but smaller in absolute value or perhaps even positive 
after the intersection. If the intersection should take place in the critical 
0 
point of contact, and so (5) be —o, so that the whole of the 
zr 
upper branch has been entirely completed when the break appears, 
then — and the properties of the following sections depend on the 
conclusion at which we arrive — the following direction either 
: ij 
coincides with the preceding one, so that sE) remains — — oo, or 
XY 
it is negative. Now the former supposition is excluded, for it requires 
that in the same point v,,=0 and v,,=0. But for following 
sections the point of intersection with the three-phase-line must then 
lie on the lower branch of the sheet 2,1 and on the upper branch 
0 
of the sheet 3,1. This continues till the value of = =o. in the 
d 73) 
point of intersection. So there are sections for which a line normal 
to the 7,z-plane can intersect the surface of saturation 4 times. 
If fig. 42 I have drawn the quantity 7—7;, as function of # on 
a large scale for the plaitpoint line, for v,, =O (the point of contact 
line for the equilibrium 2,1), for »,, =O (the point of contact line 
for the equilibrium 3,1), and for the vapour phase of the three- 
phase-curve. Only for the value of « contained between the points 
in which v,, =O and v,,=0 intersects the dotted curve (vapour 
phase), the discussed phenomenon occurs. 
Where the curves v,,=0O and v,, =O intersect the value of x 
is the same, but the value of the pressure is different for the value 
of 7'—T',, belonging to the point of intersection. For v,,=0, p is 
larger than for v,,—= 0. And this is in perfect harmony with what 
is known of the course of the isokars, which I shall set forth at 
bine ee 
some length. As the line = = 0, because we are dealing with tem- 
v 
peratures above 7}, is closed in a critical point, there is a locus 
for the points of inflection of the isobars which passes through this 
critical point, and speaking roughly, about coincides with the critical 
points of the mixtures with smaller value of x. If we now think 
