( 643 ) 
phase-pressure higher than the vapourpressures of each of the com- 
ponents. Only of H,O + SO, (Baxnuis RoozeBoom), and of some 
systems more (S + Xylol and Toluol, CO, + H,O) we know with 
certainty, that the threephase-pressure is between the vapourpressures 
of the two components. 
There exists, however, still another, important difference. Wheras in 
the case of type I (fig. 1) the eritical curve QC, presents alternately 
ry 
dT C . 
positive and negative —, and in that of type II (fig. 7) — is of 
dp dp 
course positive in the beginning in M/' (indeed, the point Q just 
appears in M') — in the case of type III the initial course of MQ may 
be as well to the left as to the right. 
a ee 
Is this course to the /eft, that is to say is = negative, then — 
ap 
just as in fig. 1 between C, and the maximum and between the 
minimum and the lowest temperatures — increase of pressure will 
lower again the critical temperature in the case of these upper eritical 
points, and the plait on the y-surface in its v, v-projection will again have 
turned the plaitpoint Q to the side of the smal/ volumes. (This is 
equally the case with C,H, + isoamylalcohol, where no threephase- 
equilibrium could be stated, but where the plaitpoint Q, which has 
become here identical with P, has removed strongly to the «x-axis, 
just as in the case of mixtures of C,H, and the /ower alcohols). We 
find this e.g. with C,H, + CH,OH. 
But when the initial course of MQ is to the right, as in the case 
of C,H,, + CH,OH, C,H,, + CH,OH, and of phenol and water, then 
increase of pressure will raise the point Q. and the mentioned plait 
will now have turned the plaitpoint Q for the first time to the side 
of the large volumes. 
The question, whether the longitudinal plait, as in the case of phenol 
and water, will present still a second plaitpoint at very small volumes, 
consequently at very high pressures -— in other terms, whether the 
coexistent liquid phases, after diverging initially, will reapproach 
afterwards in composition, has not yet been answered theoretically 
with certainty. It however appears to me, that where in the 
case of C,H, + CH,OH the plait has turned the plaitpoint Q to 
the side of the small volumes, whereas C,H,, and C,H,,, equally 
with CH,OH, have turned this point to the side of the large 
volumes, there must exist a continuous transition between the two 
kinds of longitudinal plaits, and that also the latter (as long as it 
has not yet detached itself from the liquid curve of the transversal 
plait, that is to say below the upper critical point) must be regarded 
