58 Mr. J. Larmor. Theory of Electrodynamics, [June 2, 



the existence of a quasi- magnetic polarisation of the elements of the 

 medium; that would lead, on the interface between two media, to an 



T7- _ -. T7- _ -I 



uncompensated sheet of poles of density -- F x -- - F 2 , subject 



47T 47T 



to a mean force (Fi + F,,) ; so that, as K^ = K 2 F 2 , the traction would 



"U12 / -i \ 



be (F!* F 2 2 )/87r, or, in the example chosen above, M_ l. The 



8 7T ^ Jt\. j 



discrepancy between these values might, perhaps, be amenable to 

 experiment; but I find, on trial, that the difficulty of obtaining a 

 clean unelectrified surface is not easily overcome. 



The observation of Faraday, that short filaments of silk or other 

 dielectric material suspended in a fluid dielectric set themselves 

 along the lines of force when it is excited, is also evidence of actual 

 internal polarisation related to the lines of force. 



For the case of a fluid, the Faraday-Maxwell stress is made up of a 

 hydrostatic pressure, KF 2 /8 TT, which is consistent with simple fluidity, 

 together with a tension KF 2 /4?r along the lines of force, which 

 requires for its maintenance qualities other than those of isotropic 

 mechanical fluidity. 



3. The polarisation theory, in the form of Mossotti and Helmholtz, 

 which locates part of the electrification in a displacement existing in 

 the elements of the dielectric, and part of it in an absolute electric 

 charge situated on the plates of the condenser the cause of that dis- 

 placement, is the representation of a wider theory which supposes 

 the electrostatic energy to be in part distributed through the di- 

 electric as a volume density of energy, and in part over the plates as 

 a surface density. If experiment show that the latter part is null, 

 we are precluded from imagining any superficial change on the 

 plates which has a separate existence, and is not merely the aspect 

 at one end of the displacement across the volume of the dielectric. 

 We shall find reason to conclude that there is no superficial part in 

 the distribution of energy ; this would carry the result that the ex- 

 citation of a condenser consists in producing a displacement across 

 the dielectric which just neutralises the charge conducted to the 

 plates ; it would also carry the result that all currents, whether in 

 conductors or in dielectrics, must flow in complete circuits, and 

 would therefore confirm the Maxwell theory of electrodynamics. 



The conclusion that the location of all the electrostatic energy in 

 the dielectrics involves that all currents flow in complete circuits 

 seems of importance sufficient to justify a few remarks on the nature 

 of the evidence on which it is based. The only precise notion or 

 illustration of the nature of the dielectric polarisation which has yet 

 been advanced is that of Poisson, which has been at various times 

 usod and developed by Mossotti, Faraday, Thomson, and Helmholtz. 

 It might be held merely on general grounds that it gives a correct 



