•280 
MR. J. LARMOR ON A DYNAMICAL THEORY OF 
Vo — will still hold for the dissociated portion, if u and v denote the actual 
velocities of the ions when free, not the abnormally small efiective velocities as 
determined by Whetham. Thus the total diffusion would now consist of this part 
belonging to the ionized portion, with coefficient independent of the degree of ioniza¬ 
tion, together with the actual diffusion of the non-ionized portion. On the same 
Iiypothesis the potential difference between the fluids would depend, as might have 
been foreseen, only on the actual concentrations of the ions in the two solutions, the 
amount of non-ionized substance being immaterial except in so far as it gives rise to 
an ordinary contact difference (§ 56): but ft may not be computed from the abnormal 
ionic velocities by the ordinary formula unless the degree of ionization is indejrendent 
of the concentration. 
Critique of von Helmholtz’s Theory of Electric Stresses: Electrostriction 
not due to Mechanical Forcive. 
64. A theory of electrostatic stress in dielectric media, based on the method of 
energy, and avoiding molecular theory, has been origmated by Korteweg,* 
formulated in general terms by von Helmholtz,! and further developed by Lorbeeg, 
Kirchhoee,;]; Hertz§ and others : it is desirable to examine the relation in which it 
stands to the views here set forth. The investigation of von Helmholtz postulates 
a dielectric medium which is effectively continuous, not molecular; also a potential 
function, that namely of the distribution of uncompensated polarity which represents 
the electric state of the medium, satisfying a characteristic equation, that of the 
Faraday-Maxwell theory. The energy per unit volume is expressed in terms of 
this potential, in such form that the variation of the integral which represents the 
energy for the whole volume leads, on integration hy quarts, to this characteristic 
equation as one of the conditions of internal equilibrium ; the integral is then asserted 
to be ill the normal form, which would mean, in our order of ideas, that it represents 
the actual distribution of the energy in the medium as well as its total amount. Its 
variation with sign changed, owing to change of material configuration, should then 
give the extraneous forcive that must be applied in order to maintain mechanical 
equilibrium ; the varicition with respect to the electrical configuration being null, so 
that electric internal equilibrium is provided for, by the characteristic equation 
already satisfied. The variation without change of sign should thus give the mechanical 
forcive of electric origin that acts on the medium. But the data do not even on 
these assumptions suffice to lead to a definite stress-system for the material; a 
certain geometrical stress-system is merely assumed which yields on the element of 
* D. J. KoEirwEG, ‘ Wied. Auu.,’ 9, 1880. 
t H. VON Helmholtz, ‘Wied. Aun.,’ 13, 1882: ‘Abliandluno-en,’ I., p. 798. 
t G. Kikcuhoff, ‘ Wied. Ann.,’ 24, 25, 1885: ‘ Abhaudluugen,’ ‘ Naclitrag,’ p. 91. 
§ H. Hektz, ‘ Wied. Ann.,’ 41, 1890 : “Papers on Electric Waves,” Euglisli edition, pp. 259-268. 
