440 Proceedings of Royal Society of Edinburgh. [sess. 
points on Yan der Waals’ thermodynamical surface, and to the 
points P in figs. 1 and 2. The tops of the loops M correspond to 
the tops of the border-curves M in the pv figures. C is the point 
where a loop has a tangent parallel to the p axis. Above the 
temperature belonging to that point no condensation takes place, 
but the critical phenomenon belongs to the temperature of the 
plait-point which is always lower. The circumstance that the 
plait-point curve rises agrees with the fact that P is found to 
be to the left of M. 
It is impossible to say anything about the exact course the plait- 
point curve will follow outside the field of observation, except that 
the critical temperature of the second substance seems to be higher 
than for ethane, and the vapour-pressures lower. The diagram, as 
far as we can judge, agrees entirely with the supposition of butane 
being the second substance, but this point can only be decided by 
preparing butane with a view to determine its constants, and by 
mixing it with pure ethane as prepared by electrolysis. If there 
was a simple law connecting the critical constants of mixtures with 
those of the constituents, we might calculate the constants for the 
second substance. But such is not the case. Pawlewski’s law that 
the critical temperature is proportional to the composition expressed 
in weight units is very inaccurate, the deviations being sometimes, 
considerable in both directions. 
Table III . — Mixture a. (Fig. 1.)* 
t. 
P- 
V. 
Vliq. 
14-95 
25-73 
268-5 
14-95 
26*90 
249-7 
... 
14-95 
28-26 
230-3 
14-95 
29-27 
215-2 
l 
— 
14-95 
29-45 
210-7 
l 
— 
14-95 
29-66 
207-3 
l 
— 
14-95 
29-78 
204-4 
l 
— 
14-95 
30 32 
190- 
l 
-- 
14-95 
30-95 
172-2 
l 
4*4 
14-95 
30-96 
170-8 
l 
4*4 
14-95 
31-76 
135-2 
l 
10-4 
14-95 
32-32 
91-3 
l 
19-7 
14-95 
32-57 
66-1 
l 
25-8 
14-95 
32-73 
48-4 
l 
29-9 
14-95 
32-91 
34-3 
e. 
c. 34 *3 
14-95 
40 
33-6 
33'6 
14*95 
50 
32-6 
32*6 
:: In the tables l indicates that the snbstance is partly liquefied. 
