( 708 ) 
for 7,;. Then we find namely 
1 Shy Sn. v — dt a 
dp h ia 7 ef de, 
dn <a ! : dv, 
Virgin (w,— 2.) ue 
dr dv : ; dy, dv 
For SR and <2 we may write (32) and (=) because the 
pL 
vy U, Uo) pT da, 
phases 2 and 3 then have equal p and 7. Now the point Q, is a 
liquid phase, and so a point of the upper sheet of the surface of 
0, Se 
saturation. In general the value of (GE) for such a-point is not 
x 
great at low temperatures. But yet it is larger on the whole than 
5 Op : 8 : 
the value of ar) on the vapour sheet, even for sections for which 
Kij 
xv is smaller. At least for temperatures which lie pretty far from 
Tj, so that there are therefore two possibilities chiefly dependent on 
: ; ON 
the temperature: either the value of se the point Q, may be 
x 
OT 
of x for the point Q', may either run back or proceed. 
Over the full width of the three-phase-curve on the right of Q, 
the line p,,, leaves the upper sheet of the surface of saturation with 
rising temperature. This is also still the case for points on the left 
of Q,; but a point will soon occur where the three-phase-curve 
passes to the lower sheet. So this point must lie on the “contour 
apparent” with regard to the 7’, r-surface; or in other words: it 
must be a critical point of contact. Then too the three-phase-curve 
still passes to smaller value of w. And only afterwards a point can 
occur where w has minimum value, but this only on the lower sheet. 
And if the temperature of @ is comparatively low, the vapour 
branch of the three-phase-curve will certainly run again to the right 
with falling temperature. Accordingly I have drawn the vapour 
branch in fig. 39 in this way, though there too the circum- 
stance may occur that w runs back. Besides, the circumstance occurs 
there that « shows minimum and maximum value for the liquid 
phases. The condition for 2, whether maximum or minimum is 
Op Op dp 
(ar tr (5) = qT? if we denote the phase where z runs 
back, by 1, 
0 
larger than (=) for the point Q',, or smaller — and so the value 
x 
128 
