GASEOUS AND LIQUID STATES UNDER VARIOUS CONDITIONS. 
53 
The temperature being kept steady, the liquid carbonic acid appeared at first with 
the usual concave surface, and on augmenting the pressure its volume for some time 
steadily increased without any marked change in its appearance (fig. 3, b). By further 
increase of pressure, the surface of separation became a faint line without curvature 
(fig. 3, c), and on continuing the pressure it finally disappeared, the whole mass 
assuming the homogeneous state (fig. 3, cl). 
Fig. 3. 
The position of the surface of separation in the tube before it disappeared depended 
upon the temperature at which the observations were made. At 14° the liquid 
occupied about two-thirds of the entire space just before the surface was effaced. 
It is difficult to fix with precision the exact pressure at which for a given tempera¬ 
ture the last trace of the surface of separation disappears, but a definite point can be 
obtained if, after having completely effaced the surface by pressure, we diminish the 
pressure till a cloud appears, when the liquid surface in a faint form is immediately 
restored (fig. 3, e ). The appearance of this cloud is remarkable. It occupies several 
millimetres of the tube, and as it subsides the plane surface of separation appears, 
not in the middle of the space occupied before by the cloud, but about one-third from 
the lower end. 
For a mixture composed of 1 volume of nitrogen and 3‘43 volumes carbonic acid, 
or containing 22 ‘5 per cent, nitrogen, the critical point of temperature was found to 
be 1 4°‘0, and the corresponding pressure 98 atmospheres. This point was determined 
by gradually lowering the temperature till it was just possible to obtain a small trace 
of liquid by the application of pressure. With this mixture a number of experiments 
were made at lower temperatures than the critical point (14°), in order to fix the 
pressures at which, for the same temperature, the liquid first appeared, and was after¬ 
wards effaced. As it was difficult to fix the latter point directly with precision, 
I found it better in the first place completely to efface the liquid, and afterwards to 
