

Mixtures of Ethane and Nitrous Oxide. 191 



point-curve rises, and the border-curves lie on its left side. 

 R lies beyond P. In this region lie the mixtures the com- 

 position of which is between 02 and 0*5 C 2 H 6 . 



In consequence of the complexity of the curves as com- 

 pared to fig. 1, the region of r.c. II. is a great deal smaller 

 than it would have been according to tig. 1 (BA instead 

 of CiC 2 ). It is easy to see that this was the only thing which 

 could happen : r.c. II. could not disappear entirely. 



39. These rather strange conclusions are entirely made 

 clear by figs. 3, 4, 5. These represent exactly the same 

 phenomena, only in a different manner. If P lies above R 

 we have r.c. I. ; if it lies below R, we have r.c. II. At first, 

 after the dividing of the plait, the plaitpoints P x and P 2 lie 

 on opposite sides of R x and R 2 in the two plaits. This corre- 

 sponds to the existence of r.c. I. below A, and r.c. II. above 

 A in fig. 2. While with rise of temperature the plait P 2 R 2 

 shrinks together, R 2 approaches the maximum-line. P 2 

 approaches R 2 at the same time, and at the moment that the 

 maximum reaches the end of the plait, P 2 and R 2 coincide. 

 At that moment there is no retrograde condensation. The 

 mixture ^ = 0'2 behaves at its critical point like a pure sub- 

 stance. P 2 , however, now continues to move upwards, and 

 henceforth lies a little above R 2 . That explains the existence 

 of r.c. I. between B and C 2 . If there had not been a maxi- 

 mum, r.c. II. would have existed all the way from A to C 2 . 



40. The critical phenomena near the point where the plait 

 divides into its two branches (£ = 25'8, p = b7'7 atm.) are of a 

 rather complex nature. This will be understood by the con- 

 templation of fig. 11, where the two branches of the con nodal 

 curve at the moment of parting have been drawn on a much 

 larger scale. 



Let us consider the condensation of the mixtures x x and x 2 

 separately. It will be seen (vid. § 1) that the mixture x\ in 1 

 begins to show a liquid phase which increases, decreases, and 

 disappears in 2 (r.c. I.). Then, in 3, the liquid appears again, 

 and in 4 the whole mixture will be liquid (normal condensa- 

 tion). Mixture x 2 liquefies in 1 : the liquefaction is at an end 

 in 2 (normal condensation) : in 3 a vapour phase appears 

 which disappears in 4 (r.c. II.). 



reasons for believing that the plaitpoint-curve is continuous during its 

 whole course, so that neither A nor B would be singular points. The 

 conclusions drawn in §§ 40-42 and figures 11 and 12 entirely depend upon 

 the existence of a singular point in A. There is, however, no connexion 

 between those and the rest of this paper. Professor van der Waals will 

 publish his results ere long. 



