SURFACES OF DISCONTINUITY 561 



behavior of mercurous sulphate, saHcylic acid and picric acid 

 at a mercury-water interface, but the experiments only gave 

 qualitative results from our point of view, as a quantitative 

 estimate of F could not be made. Later, Patrick and Bachman 

 (Journ. Phys. Chem., 30, 134, (1926)) found that the cation 

 is more readily adsorbed than the anion of a mercurous salt at a 

 mercury-water interface. 



Frumkin in Zeit. physik. Chem., 116, 498, (1925) described a 

 method for testing the law which differed considerably in the 

 experimental procedure from those previously used. He worked 

 with lauric acid, chosen because of its relatively slight solu- 

 bility in water, and managed to produce a saturated layer of 

 the acid on the water whose concentration he could measure, 

 obtaining an adsorption of 5.2 X 10"'^" moles per sq. cm. Using 

 the ((T, c) curve in the neighborhood of saturation he calculated 

 r to be 5.7 X lO"'-" moles per sq. cm. He made control experi- 

 ments to test the accuracy of his measurements and concluded 

 that the error in the calculated value was not more than 10 per 

 cent, and that about the same Uncertainty affected the observed 

 amount. If this is so, Frumkm's measurements constitute one 

 of the most satisfactory tests yet made. 



Reference should also be made to some experiments made by 

 Bancelin (J. chim. phys., 22, 518, (1925)) on the adsorption 

 of dyestuffs (at very low bulk concentration) both at liquid-air 

 and liquid-mercury interfaces. Rather remarkably, Bancelin 

 obtained fair agreement between calculated and observed 

 values for these solutes. 



Historically, the next important contribution is that of 

 Schofield (Phil. Mag., 1, 641, (1926)), who observed the adsorp- 

 tion by mercury of its own ions from solution. However, in this 

 work we are concerned with somewhat wider issues than those 

 raised by the Gibbs capillary adsorption equation. Questions 

 concerning the electric potentials at the surface enter into the 

 discussion, and we shall postpone deahng with these until we 

 treat electrocapillarity towards the end of this article. 



23. The Experiments of McBain and His Collaborators 



The most extensive and exact experimental test of Gibbs' 

 equation carried out up to the present is that of McBain 



