1892.] as affected l>y Stresses in Excited Dielectrics. 65 



an equal pressure in all directions at right angles to them, superposed 

 upon such stress as would exist in a vacuum with the same value of 

 the electric force. 



(ii.) It is shown from experiments that the numerical value of 

 these additional equal tensions and pressures is, at any rate to a first 

 approximation, (K l)F 2 /87r, where F is the electric force, and K the 

 inductive capacity. 



(iii.) Such a distribution of equal tension and pressure is the result 

 of a uniform volume distribution of energy in the dielectric, irre- 

 spective of what theory is adopted as to its mode of excitation. 



(iv.) If we consider the mode of excitation to be a ^w^m'-magnetic 

 polarisation of its molecules, the numerical magnitude of these stresses 

 should be 



where X is a coefficient which depends on the molecular discreteness 

 of the medium, and is probably not very different from the value -|TT. 

 A discrete polarisation of the molecules does not account for the 

 stress, so far as this coefficient is concerned. 



(v.) The stress which would exist in a vacuum dielectric is cer- 

 tainly due in part to a volume distribution of energy, as is shown by 

 the propagation of electric waves across a vacuum. There is thus no 

 reason left for assuming any part of it to be due to a distribution of 

 energy on its two surfaces, acting directly at a distance on each other. 

 There is therefore ground for assuming a purely volume distribution 

 of energy in the vacuous space, leading to a tension F 2 /8 TT along the 

 lines of force, and a pressure F 2 /87r at right angles to them. 



(vi.) The quasi- magnetic polarisation theory rests on the notion of 

 a dielectric excited by a surface charge on the plates, and therefore 

 involves a surface distribution of energy, except in the extreme case 

 when the absolute value of K is very great ; in that case a slight 

 surface charge produces' a great polarisation effect, and in the limit 

 the polarisation may be taken as self-excited. Thus the absence of a 

 surface distribution of energy leads to Maxwell's displacement theory, 

 in which all electric currents are circuital, and the equations of electro- 

 dynamics are therefore ascertained. 



(vii.) It appears that even this limiting polarisation theory must 

 be replaced, on account of the stress-formula in (iv), by some dynam- 

 ical theory of displacement of a more continuous character. 



6. We may perhaps attempt to form a more vivid picture of the 

 interaction between asther and matter by following oat the ideas of 

 Lord Raylcigh's version of Young's theory of capillarity. We may 

 conceive the compound medium, aether and matter, to consist of a 

 very refined aethereal substratum, in which the molecular web of 



VOL. LII. p 



