782 
MR. J. LARMOR ON A DYNAMICAL THEORY OP 
range of the chemical forces of another molecule; and it implies that the encounters 
of the molecules that are contemplated in the kinetic theory of gases are not of so 
intimate a character* as the encounters in a solid or liquid mass; in the latter case 
there is perhaps not sufficient space for free repulsion, and the molecules become so to 
speak jammed together. In the theory of exchanges of radiation, a gas would thus 
act simply as a medium for the transfer of radiations from one surface to another 
without itself adding to or subtracting from them. 
It follows from the second law of Thermodynamics that the heat-equivalent of the 
radiation of a given substance rises with the temperature, and this may be extended 
to each separate period in the radiation ; this is however a theorem of averages not 
directly applicable to single molecules. 
It seems a noteworthy consequence of the foregoing that the kinetic theory of 
gases is valid without taking any account of radiation. Without some tangible mode 
of presentation such as the mechanism of radiation here put forward, there would be 
a strong temptation to assume that the interchange of energy in that theory must 
take place not only between the different free types of vibration of the molecule 
{i.e. hydrodynamical vibrations of the vortices), but that also there is even in the 
steady state continual interchange with the aether. According to the present views 
such interchange would involve dissociation in the molecules; and there exist in fact 
observations relative to the action of ultra-violet radiations in producing discharge 
of electricity across a gas and consequent luminosity in it, a j^henomenon which very 
probably depends on dissociation. Whether the ideas here indicated turn out to be 
tenable or not, they at all events may serve to somewhat widen our range of 
conceptions. 
87. The result that the electric vibrations of a molecule depend on its configuration 
and the relative motion of its parts, not directly on its motion of translation through 
the aether, seems also to be of importance in connexion with the fundamental fact 
that the periods of the radiations corresponding to the spectral lines of any substance 
are precisely the same whatever be its temperature. The lines may broaden out 
owing to frequency of collisions due to increase of density or rise of temperature of 
the substance, but their mean period does not change. If we consider a system of 
ordinary hydrodynamical isolated vortex-atoms, a rise of temperature is represented 
by increase of the energy, and that involves an expansion of each ring and a 
diminution of its velocity of translation; such an expansion of the ring would in 
turn alter the periods of its electric vibrations. The question arises, how far the 
action of the atomic chai’ge will modify or get rid of these two fundamental 
objections to a vortex-atom theory of gases. Independently of this, it seems quite 
reasonable to hold that in the case of atoms paired together into molecules by their 
electrical and chemical forces, the size and configuration of the rings will be 
* [The difficulty of chemical combination of dry gases confirms this conclusion; as also for example 
the fact that molecular impacts do not explode a mixture like hydrogen and chlorine.] 
