768 
MR. J. LARMOR ON A DYNAMICAL THEORY OF 
Cohesive, Chemical, and Radiant Forces. 
65. If we consider a system of these vortex atoms, each of them will be subject to 
pulsations or vibrations, some comparatively slow, under the hydrodynamic influences 
of its neighbours in its own molecule ; and each molecule will be subject to still slower 
vibrations under the influence of disturbances from the neighbouring molecules. In 
the former class we may possibly see the type of chemical forces, while the latter will 
have to represent phenomena of material cohesion and elasticity. But in addition to 
these purely hydrodynamical vibrations due to the inertia simply of the sether, there 
will be the types which will involve rotational distortion of the medium ; that is, 
there will be the electrical vibrations of the atoms owing to the permanently strained 
state of the aether surrounding them which is the manifestation of their electric 
charges; the vibrations of this type will send out radiations through the aether and 
will represent the mechanism of light and other radiant energy. The excitation of 
tliese electric vibrations will naturally be very difficult ; it will usually be the 
accompaniment of intense chemical action, involving the tearing asunder and re¬ 
arrangement of the atoms in the molecules. It is well-known that the vibrations of an 
electrostatic charge on a single rigid atom, if unsustained by some source of vibratory 
energy, would be radiated so rapidly as to be almost dead-beat, and so would be 
incompetent to j^^'oduce the persistent and sharply-marked periods which are 
characteristic of the lines of the spectrum. But this objection may be to some extent 
obviated by considering that all the vibrational energy due to any very rapid type of 
molecular disturbance must finally be transformed into energy of electric strain and 
in this form radiated away.'^' 
Voltaic Phenomena. 
66 . According to this theory a transfer of electricity can take place across a 
dielectric by rupture of the elastic structure of the medium, and only in that way ; 
and this is quite in keeping with ordinaiy notions. Further, an electrolyte is 
generally transparent to light, or if not, to some kind of non-luminous radiation, so 
that such a substance has the power of sustaining electric stress; it follows therefore 
that transfer of electricity across the electrolyte in a voltameter, between a plate and 
the polarized atoms in front of it, can only occur along lines of effective rupture 
(such as may be produced by convection of an ion) of its aethereal elastic structure. 
When two solid dielectrics are in contact along a surface, the superficial molecular 
aggregates will be within range of each other’s influence, and will exert a stress 
which is transmitted by the medium between them. The transmission will he 
partly by an intrinsic hydrostatic pressure, as in Laplace’s theory of capillarity, 
and partly by tangential elastic tractions produced by rotation of the elements of 
* I iinderstaud that a suggestion of this nature has already been made by G. F. FiTZ Geealu. 
