66 On the Theory of Electrodynamics. [June 2, 



matter is imbedded. The range of direct action between contiguous 

 parts of the aether would be very small, and that between contiguous 

 elements of matter large in comparison. There exist disturbances in 

 which the matter- web is unaffected, its free periods being too slow to 

 follow them: these are propagated with great velocity as light, or 

 electrical radiations. There are other disturbances in which the 

 matter-web is alone active; these are so slow that the aether can 

 adjust itself to an equilibrium condition at each instant; they are 

 propagated as waves of material vibration or sound waves. 



When a dielectric is excited, we find ourselves in the presence of a 

 strain of an aethereal origin somehow produced ; it would relax on 

 discharge of the system with the velocity of light. At an interface 

 where one dielectric joins another, the aethereal conditions will some- 

 how, owing to the nature of the connexion with the matter, only 

 admit of a portion of the stress being transmitted across the inter- 

 face ; and there will thus be a residual traction on the interface which 

 must, if equilibrium subsist, be supported by the matter- web, and be 

 the origin of the stress which has been verified experimentally. 

 Inside a conductor, the aether cannot sustain stress at all, so that the 

 whole aethereal stress in the dielectric is supported by the surface of 

 the matter- web of the conductor. At such interfaces the aethereal 

 part of the distribution of energy in the medium will be discon- 

 tinuous. 



A formula has been given by Maxwell* for the intensity of the 

 press ural force produced by electric undulations in the aether striking 

 against a plate of conducting matter, a force which has apparently 

 not been detected for the case of light-waves. If the notions here 

 suggested have any basis, this force may likely be non-existent. For 

 the pulsations of the aether at this surface may be so rapid as to pre- 

 vent their energy being communicated to the rnatter-web of the 

 conductor ; and the energy will then be scattered and lost instead of 

 appearing as energy o f material stress. We may take as an illustra- 

 tion a stretched cord with equidistant equal masses strung on it, for 

 which Lagrange showed that if the period of a disturbance imparted 

 at one end exceeds a certain limit, the disturbance will not be trans- 

 mitted into the cord, but will be eased off within a short distance of 

 the point of application. And also in a manner which forms a more 

 exact analogy, Sir Gr. Stokes has shown that the higher harmonics of 

 a telegraph wire vibrating in the wind have their pulsations too rapid 

 to get a grip on the air around them, and their note is therefore not 

 transmitted. 



This view would place the electrostatic and electrodynamic forces 

 on matter on a lower plane, and in the case of rapid or sudden 

 disturbance a more uncertain one, than, the electromotive phenomena. 

 * ' Electricity,' 793. 



