752 
:\m. J. LARMOR ON A DYNAMICAL THEORY OF 
of surface dS of the conductor to encroach by forward movement into the excited 
dielectric. As it proceeds, its superficial molecules somehow dissolve or loosen the 
strain of each little piece of the dielectric aether as they pass over it. Each 
fragmentary easing of strain sends a shiver through the dielectric aether, which 
however practically instantaneously readjusts itself into an equilibrium state. Thus 
the process goes on, the gradual molecular dissolution of the strain by the advance 
of the conductor shooting out minute wavelets of rearrangement of strain into the 
dielectric, which are confined to the immediate neighbourhood and are cpiite undis- 
cei’iiible directly, because on account of their great velocity of propagation the aether 
is always excessively near an equilibrium condition."^' The joressural reaction (§97) 
of these disturbances on the conductor may be taken to be the source of the mechauical 
forcive experienced by it, which does work in impelling its movement and to an equal 
extent exhausts the energy of the dielectric. 
Imagine a very thin element cZS on the surface of the conductor, thick enough, 
however, to include this layer of intense disturbance of the aether; it will be subject 
to this electric reaction of the excited dielectric acting on it on the one side, and the 
elastic traction of the material of the solid conductor actiuijf on it on the other side; 
and as its mass is very small compared with its surface, these forcives must equilibrate. 
For if this sujDerficial element is displaced outwards through a very minute 
distance ds, the following changes of energy result; the energy of the dielectric is 
altered by the subtraction of that contained in a volume dSds of it, while the elastic 
normal traction P of the conductor does work Vd^ds. These changes must coiii- 
])ensate each other by the energy principle of equilibrium (compare § 58); hence 
the normal elastic traction P is equal to the energy in the dielectric per unit volume. 
The consideration of a tangential displacement of the element leads in the same way 
to the conclusion that the tangential elastic traction, required to be exerted by its 
material backing in order to maintain its equilibrium, is null. 
Electrodijnainic Actions between Material Bodies. 
46. In order to examiue how far our energy-function of an lethereal medium 
involves an explanation of electrodynamic phenomena, wm must begin with a simple 
case of electric currents that will avoid the introduction into the field of all com- 
])lications like galvanic batteries, which could not easily be included in the energy- 
function. Let us therefore consider two charged condensers with their two pairs of 
coatings connected l^y thin wires as in the annexed diagram ; and let us suppose the 
two plates of one of the condensers to be steadily moved towards each other when 
both pairs of coatings are thus in connexion. This wall produce a steady current 
in the conducting wires, which will fiow completely round the circuit; the only 
* Cf. Sir G. G. SxoKiis, “ On the Communication of Vibrations from a vibrating body to the surround¬ 
ing gas,” ‘Phil. Ti-ans,’ 1868, p. 448; or in Lord Ratleigh, ‘ Theory of Sound,’ vol. 2, 
