( 880 ) 
system hexane and aniline we are saved from this error by the 
circumstance that its critical end-point can be realised, but it is of 
course very well possible that in other cases the appearance of the 
crystalline phase prevents this realisation. Only the determination of 
the melting-point line can give certainty in such cases. 
How the facts mentioned here, and indeed so many others could 
be made to harmonize with the “theory” which was advocated many 
years ago by Spnryers’), and which has been upheld again of late 
by DorrzarEK *), according to which the partial pressures of the 
components are always proportional to the liquid concentration, is 
not to be understood. Even if we restrict ourselves to mixtures of 
normal substances, the abundance of phenomena is so great that it 
can only be accounted for by a theory which takes account of the 
great difference which can exist in the three quantities which together 
determine the state of such a binary mixture: the ratio of the mole- 
b a, 
cular size n= —, the ratio of the molecular attraction 4 = —, and 
b, a, 
a,. 
the attraction inter se of the molecules /= —*_. 
V da, 
CORCLUSLONS: 
We have demonstrated in this communication : 
1. That the expectation pronounced by van per Waars that the 
line of the three-phase pressures 4, -+ L, + G would tonch the 
plaitpoint line in the critical end-point is in contradiction with the 
experiment, but that also further theoretical consideration proves this 
expectation to be unfounded, and gives two other rules which are 
both confirmed by the experiment instead of VAN DER Waars’ thesis. 
2. That in general the rule given by vaN per Waars that splitting 
up of a plaitpoint line only occurs in systems from the middle 
region of the diagram of isobars, is corroborated by experience. But 
that nevertheless also cases seem to occur in which systems from 
the lefthand region of the diagram of isobars (descending 7), with 
ascending 4 without the minimum 7%, being reached) present splitting 
up of the plaitpoint line. Also for other above mentioned reasons a 
further investigation of these mixtures remains very desirable. 
3. That in opposition to the theories of OstwaLp, SPEYERS, Doun- 
ZALEK, and others, but in agreement with the theory of var DER WAALS 
mixtures of normal substances can possess p‚v-lines with a point of 
inflection, and even with more than one, when we approach the 
region of non-miscibility. 
1) Journ. of Phys. Chem. 2 p. 347, 362 Journ. Amer. Ch. Soc. 21 p. 312, 725. 
2) Zeitsehr. f. Phys. Ch. 62721571, 190. 
