to the high degree of compressibility of the substance in the critical 
state; l.c. p. 13 (Proc. p. 637). 
At the time it was not possible to form a true judgment about 
the influence of admixtures and the required corrections. While other 
corrections, which had probably to be applied and which might 
have the same influence, were fully discussed, about the correction 
for admixtures, nothing could be said but that (Le. p. 6, Proe. p. 
631) measurements were being made at the laboratory, which would 
spread the desired light on the influence of small admixtures. 
Since that time have been published those measurements by myself 
on mixtures of carbon dioxide and hydrogen (Comms. N°. 45, Jan. 
"99 and N°. 47, Febr. ’99) and those by Kersom on carbon dioxide 
and oxygen (Comm. N°. 88, Jan. ’04). In the series of “Contributions 
to the knowledge of var DER WAA1S’ y-surface” occur several calculations 
of Krgsom (Comms. N°. 75, Dec. OL and N°. 79, April 02) and of 
myself (Comm. N°. 81, Oct. 02 and Suppl. N°. 6, May and June ’03). 
These calculations in which the law of corresponcing states has been 
applied according to KAMERLINGH OnNES for substances with admix- 
tures, reduce all the deviations from the properties of the pure 
ve 1 (dT x 
substances to the knowledge of the two quantities a = ~ *) 
zl, 
1 dps . Zn i 7 
En Hs and of the empirical equation of state. 
Pk an en 
I have availed myself of the obtained results to investigate what 
differences of density will be observed in a tube of CAGNIARD DE LA 
Tour, containing carbon dioxide mixed with a small molecular 
composition of oxygen, if in the manner indicated by Trtcuner *) 
floats are placed into it to determine the density. I found it confirmed 
that the nature of the deviations which would be observed in TrrcunEr’s 
experiments in consequence of small admixtures (if pressure and 
temperature are in equilibrium), corresponds entirely to that of the 
deviations observed. 
It seems to me an important result that, on the strength of the 
knowledge of the behaviour of the mixtures of carbon dioxide and 
oxygen, we can calculate that even very small quantities of oxygen 
in carbon dioxide (a few 0.001 mol.) are sufficient to produce the 
1) Drupe’s Ann., 13, 595, 1904. The explanation of Trrcuner’s experiments 
covers that of Gauirzine’s experiments, where the density was determined at dif- 
ferent heights by an optical method. In tubes filled with carbon dioxide, Gouy 
(C. R. 116 p. 1289, June 1893) has observed a slow displacement of the meniscus 
a little below the critical temperature, and has ascribed this phenomenon, and 
rightly I hold, to impurities. 
