( 955 ) 
after equilibrium had been reached in our experiment on the 
density of the vapour and the tap &, had been closed we 
altered the temperature of the cryostat slightly until the liquid 
phase first disappeared and then reappeared in the appendix. 
We found that a temperature change of '/,, to */s, of a degree 
was sufficient to cause the liquid phase to disappear completely. The 
absolute values of the temperatures are accurate to about */,,th of 
a degree except in one case (— 210° C) in which, owing to 
unfavourable circumstances, the accuracy attained was '/,th of a degree. 
§ 6. Results. For gig the density of the liquid oxygen, for @ yap 
the density of the saturated vapour with which it is in equilibrium 
1 
at the same temperature, and for Dz = > (@uig + Qrap) the ordinate 
el 
of the diameter we obtained the following values: 
Qlig Qvap Dz (obs.) Dz (cal) OC 
SENOS, 1D 46 0.0001 0.63738 0.6373 () 
— 182.0 1.4415 6.0051 0.5733 0.5730 + 0.0008 
— 154.51 0.9758 0.0585 0.5072 0.5107 — 0.0035 
~~ 140.2 0.8742 0.0805 04775 0.4783 — 0.0010 
oe shh, OT TSL 0.1320 0.4550 0.4550 0.0000 
= 193-0 0.6779 0.2022 0.4400 0.4400 0 
4001 HEI a 02701 — 04366: 04335 7 SE OOR 
The calculated values of the diameter are taken from the equation. 
Decca = 0.1608 — 0.002265 1 
The results are plotted on Plate III. 
By putting ¢ equal to the critical temperature — 118.°8C. a value 
0; = 0.4299 is found for the critical density. This value compared 
with the value of Qj, at — 210.°C., is in good agreement with the 
law of the third of the density. 
From 
ba == — 0,002265 
the absolute value of the slope of the diameter D= az + ba T, 
taking 7, = 273.1 — 118.8 the reduced slope is found to be 
_Trba 
Qk 
eee | i a 
ba 
The deviation at —154°.5 C., the density of the vapour being 
1) E. Maraias. Remarques sur le Théoréme des états corresnondants. Ann, de 
Toulouse t. V. 1891. 
