1896 - 97 .] Prof. J. P. Kuenen on Ethane and Butane. 437 
ethane and butane is an extremely difficult matter. I know of no 
method except fractional distillation. The quantity left after all 
my experiments were finished was too small for the purpose. 
We will now consider the diagrams (figs. 1-3) more closely, assum- 
ing the mixtures to contain ethane and butane. 
The reason why it seemed worth while to communicate the 
numerical results and the curves which are based upon them is, 
that the mixtures form a type which has not been investigated 
before. Andrews’ curves for carbonic acid really belong to a 
mixture of CO., and a trace of permanent gas, say air, i.e., of a 
substance of much lower critical temperature than carbonic acid. 
The features of his isothermals inside the border curve are well 
known. They start almost horizontally from the point where the 
condensation begins ( b.e .), hut are really slightly curved, especially 
near the end (e.c.), the curves turning their concave side upwards. 
The mixtures of ethane and butane, on the other hand, are mix- 
tures of one substance with a small quantity of a second substance 
of much higher critical temperature and lower vapour-pressures. 
The isothermals inside the border-curves have a correspondingly 
different shape. The curves turn their concave side downwards, 
the curvature being relatively high near the beginning, and dimin- 
ishing towards the end. If air had been present in the mixtures 
as well as butane, it would have shown itself in a change in curva- 
ture in the curves near e, as in Andrews’ curves. 
The values for p and v at the end of the condensation may be 
read with great accuracy, as the volume is small and consequently 
stirring is very effective : concordant values for those points are 
easily obtained. This is not the case as regards the points b.e., 
especially not at low temperatures : the column occupied by the 
substance is very large in those cases, and it is difficult to obtain 
perfect homogeneity of the mixture ; even then, however, the 
point where condensation commences is not easily determined (at 
least not at low temperatures) owing to the difficulty of observing 
the minute quantity of liquid which is formed in the beginning. 
Some of the b.e. points (in the tables XII. and XIII. printed in 
italics) have therefore been obtained by interpolation as the point 
of intersection of the two parts of the isothermals inside and out- 
side the border-curve. This intersection always takes place at a 
