B.—CHEMISTRY. 79 
by Kenrick, Thorwaldson, and McTaggart, any specific chemical 
activity or ionisation of the oil or gas would seem improbable. Any 
theory which attempts a general treatment of the problem must be 
prepared to deal with cases such as these. 
Many measurements have been made of the potential differences 
between solids and liquids, or between pairs of immiscible (or partly 
miscible) liquids, using electrometric methods. Thus Haber and 
Klemensiewicz determined the potential difference at a glass-water 
solution interface, and found the glass to act like a hy drogen electrode. 
Their results have been recently confirmed by W. S. Hughes. Tt will 
be at once obvious that these results are not in agreement with those 
obtained by cataphoretic and electroendosmotic methods. A somewhat 
similar type of discordance has keen observed in the electrometric 
measurements of the potential difference between solid paraffin and an 
aqueous solution made by G. Borelius, and of the P.D.’s between pairs 
of liquids made by R. Beutner, E. Baur, and others. Freundlich and 
Gyemant have drawn attention to the fact that in all such electrometric 
measurements, where in the process of the measurement an electric 
current must pass from one phase to the other, we measure the total 
or ‘thermodynamic’ potential difference between the phases in bulk, 
whereas in determinations by the methods of electroendosmose and cata- 
phoresis, we measure only a portion of this total potential difference. 
These ‘ electro-kinetic’’ P.D.’s, although of fundamental importance in 
relation to the stability of suspensoid (lyophobic) systems, need not, 
and in general will not, coincide in value with the total (thermodynamic) 
potential differences. It will be recollected that I drew attention to a 
quite analogous difference in discussing the measurements of the 
potential differences at gas-water interfaces made by Kenrick and by 
McTaggart. 
We may illustrate this point by considering the P.D. between two 
immiscible phases, L, and L,, in equilibrium with each other, and 
each of which contains dissolved in it the electrolyte KA. Ife denote 
the positive potential of L, above L,, and F the quantity of electricity 
associated with an ionic gram equivalent, then by a virtual variation 
of the equilibrium system it follows that 
(tux): — (Ux)o= Fe = (Ya)o— (Ha): 
where the subscripts refer to the cation or anion and tc the phases 
L, or L,, and the #’s denote the chemical potentials per gram equivalent 
(partial equivalent free energies) of the ionic constituents in the bulk of 
the two phases. Whatever may be the ‘electro-adsorption ’ or ion 
adsorption of K and A at the interface I\,—L,, it is clear that ¢ depends 
only on the ‘ bulk ’ values of the respective chemical potentials, which 
likewise determine the surface concentrations. If the phases L, and 
-L, be not in equilibrium, then velocity or diffusional terms will enter 
into the equations, and the potential difference will be partly or wholly a 
“diffusional potential.’ These relationships were clearly established 
many years ago by R. Luther. 
In discussing the ‘ stabilities ’ of hydrocarbon oil emulsions, it must 
‘not be forgotten that I was dealing with very dilute suspensions of oil 
H 
