Mixtures of Ethane and Nitrous Oxide. 1 93 



the distance between the temperatures of the plaitpoint and 

 the critical point of contact (t? and £r), for the same mixture, 



Fig. 12. 



can be very small only. An estimation from the original figure 

 2 gives +0°*1 C. at the utmost for the distance between t? and 

 t-R, and in some parts of the plaitpoint-curve — for instance 

 near B — much less still. Now in order to fix so small a 

 difference the temperature ought to be perfectly constant for 

 a long while, which was not the case in my water-vessel. 

 There were slow changes, which of course did not prevent 

 the determination of the critical point taken as a whole. I 

 think this circumstance may be taken as a sufficient explana- 

 tion of the disagreement between theory and experiments. 



44. However, there is another cause which must have co- 

 operated in confusing the pure phenomena, viz. the influence 

 of gravitation. It was Gouy* who, in the case of pure sub- 

 stances, first pointed out the importance of this factor near 

 the critical point, where the substances become so highly 

 compressible. In the case of mixtures the influence of gravi- 

 tation is not of a quite simple nature. The result is this : 

 near the plaitpoint there is a small region in which the com- 

 plete process of the condensation is confused ; namely in this 

 manner, that with compression at a certain moment the liquid 

 surface between the phases disappears at some distance from 



* Compt. Rend. cxv. p. 720, cxvi. p. 1289 ; cf. Kuenen, ' CominunicationSj 

 etc' No. 8, p. 10. 



