and Gases near their Critical Temperatures. 149 



in the liquid. Continuing the heating, the meniscus (c) in the 

 capillary is seen below that (C) in the external tube (fig. 4). 

 It then gradually loses its concavity, becomes successively 

 plane and less defined, and frequently presents a more or less 

 convex appearance before finally vanishing. Meanwhile the 

 liquid in the external tube expands and undergoes the same 

 changes, the defined surface appearing more or less convex, 

 and then becoming black and ill-defined. It first ceases to be 

 visible in the capillary tube. The curve fig. 5, Plate III., shows 

 the changes in the height of liquid sulphurous anhydride in a 

 capillary tube of 0*4358 millim. diameter in an external tube 

 3'2 millims. wide. The abscissae are in degrees centigrade, the 

 ordinates in millimetres. The measurements near the critical 

 temperature could only be made with difficulty, and are con- 

 sequently less reliable than those made at a lower temperature. 

 Measurements of the depression of the liquid, or even of the 

 diameter of the capillary tube in which the depression is ob- 

 served, possess but little value, other conditions (e. g. rate of 

 heating, &c), incapable of exact expression, exercising too great 

 an influence upon it. When the tube is rapidly cooled a local 

 cloud of very fine particles suddenly makes its appearance, in 

 the middle of which the liquid first appears as a fine horizontal 

 line (fig. 4). In the capillary tube this cloud extends to a 

 considerably greater distance, both upwards and downwards, 

 than in the external tube. Probably this is connected with a 

 surface-action, to be described subsequently. The mean 

 position of the cloud is nearly that at which the tube-con- 

 tents disappeared on heating. The formation of this cloud 

 reminds one forcibly of the sudden crystallization of a su- 

 persaturated solution. The volume which the liquid then 

 first occcupies is less than that which is possessed before its 

 disappearance, and is considerably less if the black ill-defined 

 surface visible on heating be taken into consideration. A 

 persistent tendency on the part of the ether and sulphurous 

 anhydride gases to condense at a lower temperature (par- 

 ticularly when somewhat rapidly cooled) than that at which 

 they were formed is very noticeable *. Rapidly cooled 

 (especially in narrow external tubes) the liquid, when it first 

 becomes visible on condensation, is either level in the capillary 

 and external tubes, or higher in those tubes in which the de- 

 pression on rapid heating has been the greatest. When 



* Although it is premature to draw any definite conclusion from this, it 

 may be of interest to recall the researches of Magnus (Pogg, Ann, xxxviii. 

 592), and Eegnault (Memoires de V Academie 1802, xxvi. 71-5-749, 335- 

 6G4) on the vapour-pressures of liquid mixtures, which seem to bear a pos- 

 sible relation to it. In this connexion see also Cailletet's paper (Phil. 

 Mag. March 1880, p. 235). 



