806 
MR. J. LARMOR ON A DYNAMICAL THP]ORY OF 
electroniotive jiheiiomeiia in electrostatics and electrodynamics. It assigns correctly 
the magnetic rotatory action on light to a subsidiary term of definite type in 
the enei’gy function of a material medium ; while to avoid a magnetic translatory action 
of such amount as would be detectable, it is compelled to assign a high value to the 
coefficient of inertia of the free aether. In unravelling the detailed relations of 
rether to matter it is not very successful, any more than other theories ; but it 
suggests a simple and precise basis of connexion, in that form of the vortex-atom 
theory of matter to which it leads ; and even should the present mode of representa¬ 
tion of the phenomena become on further development in this direction definitely 
untenable, it may still be of use within its limited range as illustrating wider views of 
possibilities in that field. The theory also leads to the correct expressions for the 
ponderomotive class of electrostatic and electrodynamic phenomena, or rather it is not 
in disagreement with them ; for here again knowledge of the details of the relation 
between the rether and the matter is defective, and thus for example the laAv of the 
attraction between permanent magnets is left unexplained. It supplies also a more 
definite view of the essentially elastic origin of all electrodynamic action than has 
perhaps hitherto been obtained, especially in cases of induction by motion across a 
steady magnetic field. 
[Added August 13, 1894.] 
Introduction of Free Electrons. 
114. The conclusion to which we are led in § 107 is that a sim])]e vortex-atom 
theory is not in a position to attempt to explain the law of the forcive between 
permanent magnets, if only for the reason that on such a theory no explanation of 
the inertia of matter has yet been developed. This difficulty is, however, not peculiar 
to the present special view of the electric field ; any representation of a magnetic 
molecule, which assigns to it a purely cyclic motional constitution, is subject to an 
equal or greater difficulty in explaining why it is that the law of the forcive between 
magnets is the same as between currents, and not just the reverse. 
What is required in order to obtain a decisive positive result is, that the assump¬ 
tion of a purely cyclic character for the motions associated with permanent magnets 
shall be avoided by giving the elasticity of the medium some kind of grip on them. 
The movements of rotation and vibration of the simple vortices which constitute a 
vortex-aggregate are not competent to secure this, however sudden they may be, for 
in the irrotatioiial fluid motion the constraint of the rotational elasticity has only to 
reduce a labile condition of the medium into a stable one ; thus there is no sensible 
play of elastic energy introduced, such as would be required to explain induction in a 
steady magnetic field. 
