586 RICE ART. L 



Bircumshaw (Phil. Mag., 6, 510, (1928)) has found that the 

 surface tension of mercury in contact with such gases exhibits 

 some anomahes with lapse of time which have not yet been 

 explained. Finally, reference should be made here to the ex- 

 cellent work of H. Cassel and his collaborators on the adsorption 

 of gases and vapors on mercury and water surfaces (Z. Elektro- 

 chem. 37, 642 (1931); Z. physik. Chem., Aht. A, 155, 321 (1931); 

 Trans. Faraday Soc, 28, 177 (1932); Kolloid-Z., 61, 18 (1932)). 



XII. The Thermal and Mechanical Relations Pertaining to the 

 Extension of a Surface of Discontinuity 



SS. Need for Unambiguous Specification of the Quantities Which 

 Are Chosen as Independent Variables 



In this subsection Gibbs makes use of the results obtained in 

 the previous subsection of his work, to which we have already 

 referred at the beginning of the part of the commentary just 

 concluded. The results are in equations [578], [580] and [581]. 

 When there is one component in two homogeneous phases and a 

 surface of discontinuity, o- is a function of t and n (the one 

 potential involved). The transformation effected at the 

 bottom of page 265 still leaves it a function of two variables t 

 and p' — p". If the surface is plane there is only one variable, 

 t, involved; this is obvious in any case since with only one 

 component in two phases, say vapor and liquid, p is a function 

 of tf and of course o- is also. 



Equation [580], which refers to two components in two homo- 

 geneous phases, and equation [581] are framed as if cr were 

 again a function of two variables, and yet a is originally regarded 

 as a function of three, viz., t and the potentials ni and m2 of 

 each component. The reason is clear. Equation [579] shows 

 that there are really three variables involved, t and the two 

 pressures; but since the surface is regarded as practically plane, 

 the difference between the two pressures is ignored. Actually, 

 since the surface is plane and p' = p", this gives us an equation 

 between two functions of t, ni, /X2 and thus /Lt2 is a function of t and 

 Hi and is not an independent variable; so o- is really a function 

 of t and Ml or t and p. It would be a great assistance to students 



