282 Dr. J. Larmor. A Dynamical Theory 



the elements of volume may not fit together consistently -with the 1 

 continuity of the substance, and thus secondary strains may be 

 produced which will complicate the problem. But in fluids, in which 

 alone experiment is feasible, no such complication can occur. 



The position is aptly illustrated by the ideally simple case of a 

 perfect gas polarised in an electric field. The mechanical forci re- 

 due to the polarisation of the gas as a whole is there compensated by 

 change of pressure, which is transmitted. There is another regular 

 part of the forcive which arises from actions between neighbouring 

 molecules, so that those in the line of polarisation attract and those- 

 in lines at right angles to it repel each other, after the manner 

 of little magnets ; and this, which differs from the former by being 

 proportional to the square of the polarisation instead of its first 

 power, is not experienced as a mechanical force, because it is wholly 

 compensated on the spot where it originates by slight change in the- 

 ordinarily fortuitous distribution of velocities of the molecules of the 

 gas, by which its constitution acquires an axial character with 

 reference to the line of polarisation so that the pressure is no longer 

 quite the same in all directions. 



7. The relation is explained which exists between this organised or 

 mechanical energy and the available or free energy of thermodynamics. 

 The principle of available energy, which itself is a direct consequence 

 of the negation of perpetual motions, or rather of the negation of the 

 unlimited availability of diffuse thermal energy, is the single essential 

 foundation of that science. It is pointed out, that if we had no 

 direct perception of temperature through our senses, this negation of 

 perpetual motions would necessitate the introduction of that quantity 

 into physics, somewhat in the same way as potential is introduced 

 into electrical theory, and would yield a demonstration of its funda- 

 mental property. Instead then of making attempts, by the aid of 

 special molecular hypotheses of more or less problematical character,, 

 to obtain a purely dynamical definition of temperature and an ana- 

 lytical demonstration of Carnot's principle, it is suggested that it 

 is more philosophical to recognise that no physical scheme of matter 

 and molecular action is conceivable that would involve perpetual 

 motions (in the above sense) of matter in bulk, and to base the purer 

 theory of thermodynamics and thermochemistry directly on this 

 postulate. 



The case of homogeneous fluid media, which acquire energy of 

 polarisation of any kind when in a field of force, is considered in a 

 general manner : it is shown that such media will be in mechanical 

 equilibrium provided an extraneous traction along the normal is 

 applied over each interface between them, of intensity equal to the- 

 difference of the densities of the mechanical energy of polarisation on 

 the two sides of the interface. 



