( 876 ) 
of normal components which differ greatly in volatility highly im. 
portant deviations from the rule of van ’r Horr’) may occur. For this 
it is required but also sufficient, that the value of = 
TD 
descends only comparatively little below unity (in these cases e.g. 
to about 0.85). This lower value of a,,, however, is always accom- 
panied by a region of non-miscibility. In connection with conside- 
rations on systems of an entirely different nature one of us will 
shortly return to this question,’ and ‘discuss it} more {fully ; we 
shall therefore not dwell any longer on this point. We shall only 
draw one more conclusion’ from what has been said. If already 
e wy 
an unusually small value of / is required to prevent (in 
v 
. 1 
from becoming equal to zero for mixtures of a very diverging degree of 
TU, 
volatility, this value will of course have to be taken still smaller, if fijn 
wv 
is put larger than 1, and so if it is assumed that the concentration 
of the so much less volatile component is yet greater in the vapour 
than in the liquid, as the existence of a maximum would require. 
Hence we see a corroboration in what has just been said, of the 
indication obtained from the boilingpoint lines that at lower tempe- 
rature a point «,—.2, does not occur in the figure either. For 
the same reason it seems less probable that the system sulphuric 
acid -+ decane will possess a maximum vapour pressure, though 
it belongs to the type of the splitting up of the plaitpoint line. *) 
From this it would follow that vaN peR Waats’s rule which states 
that the splitting up of the plaitpoint line always involves the occur- 
rence of a minimum 7% is too narrow, because a point of splitting 
up of this line can also occur in systems from the lefthand region of 
the diagram of isobars, provided they do not move too far away from 
the region where liquid and vapour concentration become equal. 
Yet an examination of all the known data shows that by far in 
the majority of the cases splitting up of the longitudinal plait will 
only oceur for systems with a minimum T’.. We have collected all 
the experimental data in Table VI. Those concerning the plaitpoint 
1) Of course we do not mean here deviations for concentrated solutions, for they 
occur in every system. But we mean that van ‘r Horr’s rule need not hold as 
limiting law for extreme dilution either for all mixtures which differ greatly in 
volatility. Moreover the condition must be fulfilled that the two substances readily 
mix in the liquid state, in other words that we are very far from a region of 
non-miscibility. 
2) TIMMERMANS and Konnsramm, These Proc. Sept. 1909, N° 15 of the table. 
