SURFACES OF DISCONTINUITY 699 



other by the passage of the mass 8m of an ion, viz. (^t + ^V)8m. 

 It appears that in the transfer the division of the energy into 

 two parts nSm and /375w(or V8e) is of no practical importance. 

 The writer need not discuss the point fully here, since the reader 

 can find in a paper by E. A. Guggenheim (Jour, of Phys. Chem., 

 33, 842, (1929)) some very interesting remarks on it which 

 will repay careful consideration. We shall consequently 

 replace the expression /x + fiV by M, referring to it, as is 

 suggested by Guggenheim in his paper, by the name "electro- 

 chemical potential" of the ionic component. This quantity M 

 has equal values in equilibrium conditions for an ion on opposite 

 sides of an interface if the ion is an actual component of both 

 phases. It has of course the same value in the region of dis- 

 continuity if it is an actual component of this region. If it 

 exists at the interface and in one bulk phase only, the electro- 

 chemical potential has the same value in each, a value which 

 cannot be greater than the value in the bulk phase in which it 

 is only a possible component. 



Now it appears on reading parts of the literature that some 

 authors take it for granted that if the electric potential of a 

 phase is altered by the amount 57, then the quantity ^ + /3F or 

 M must alter by the amount /3 8V. This is no doubt based on 

 an implicit assumption that /x does not change in the meanwhile; 

 but this view seems to the writer to be too narrow and based on 

 the artificial splitting of the real chemical potential of the ion, 

 its electrochemical potential as we call it, into a "purely 

 chemical" and a "purely electric" part, which can vary 

 independently of one another. Even on this physically non- 

 significant analysis, one cannot guarantee that a change in V 

 will not alter the concentration of the ion and therefore the n 

 of the ion. The truth is, that the only chemical potential of an 

 ion of which we have any direct cognizance is the quantity we 

 have denoted by M, and we actually would have preferred still 

 to use the symbol m for it, but for the possibility of confusion 

 with the terminology of other writers. In fact, the electric po- 

 tentials of the phases are now to be reckoned among the thermo- 

 dynamic variables of the system, and the electrochemical 

 potentials of the ions (although to be quite exact the term 



