SURFACES OF DISCONTINUITY 695 



mercury surface has no charge and according to Gibbs (even 

 apart from the fact that the charges in mercury and electrolyte 

 must compensate one another) the electrolyte layer has no 

 excess or defect of ions, it follows that the solution should be 

 higher than the mercury by about 0.2 volt. Of course we can- 

 not be sure that our assumption for the KCl is correct, which 

 only makes matters still more ambiguous; for it is clear that 

 the situation renders doubtful the whole basis of the various 

 methods hitherto employed to measure an absolute elec- 

 trode P.D. 



An explanation for this behavior has been offered on thermo- 

 dynamic grounds as follows. We have seen that in the natural 

 state positive ions (cations), such as hydrogen ions, will be in 

 defect in the electrolyte layer of the discontinuous region, while 

 negative ions (anions) would preponderate. It is assumed 

 therefore that in this state there are present anions which are 

 capillary-active, in the sense defined in the earlier part of this 

 article; i.e., they tend to lower the surface tension and are 

 "specifically adsorbed" at the surface of the solution on that 

 account. Now, in so far as this has any meaning, it apparently 

 assumes that the negative charge of these adsorbed anions will 

 be to some extent neutralized by the positive charge on the 

 cations in the electrolyte layer. The corresponding positive 

 charge on the mercury will exert the usual depressing effect on 

 the surface tension represented in the Lippmann equation by 

 — q 8V. But in addition to this, these anions will exert a still 

 further depressing effect represented by an additional term of 

 the Gibbs type — TSix (not an equivalent term). Whether this 

 "combination" effect can be derived from a really theoretical 

 treatment we shall consider presently. Of course it is part of 

 the assumed state of affairs that the cations are capillary- 

 inactive and are therefore not "specifically" adsorbed, their 

 presence in the layer is determined by the external electrical 

 influence. Without the specific adsorption of the anions it is 

 assumed that we would have the "normal" parabolic curve; 

 with the adsorption we have an additional depression and the 

 curve begins lower down than the normal. As the E.M.F. rises 

 the electric field drives the active anions out of, and brings 



