92 Extension of Maxwell's Electro-magnetic Theory of Light. 



p? x is found to be equal to the specific inductive capacity of the 

 medium, as previously determined. 



For a medium which might be compressed without altering the 

 period of vibration of the constituent molecules, Gladstone and 

 Dale's law, in the modified form /r lac density, would follow. 



Double refraction, in case of a uniaxial crystal, is explained on the 

 assumption that the molecules are arranged with their axes parallel 

 to a certain direction. Electrical disturbances perpendicular to this 

 direction will produce a molecular rotation, whilst those parallel to 

 this direction will produce an inter-atomic separation. The doubly 

 refracting nature of a dielectric when subjected to electric strain is 

 thus explained ; and it is pointed out that Lord Kelvin was led to 

 postulate a crystalline structure similar to the above to account for 

 the pyro-electric properties of tourmaline, ,&c. 



For infinitely quick vibrations the refractive index of the above 

 medium will be equal to unity, a result possibly explaining the 

 action of material bodies on Rontgen radiations. 



Assuming a metallic or quasi-metallic substance to have a struc- 

 ture essentially similar to that described above, with the addition 

 that a viscous term is included in the equation for the atomic vibra- 

 tion, the refractive index of a metal is found in the form of a 

 complex quantity, the imaginary part of which is essentially positive. 

 The ordinary laws of metallic reflection, as deduced by Cauchy and 

 others, will therefore hold. It is shown that for those metals in 

 which the real part of the square of the refractive index is a large 

 negative quantity, the velocity of propagation of light will be in- 

 versely proportional to the molecular viscosity. Since Mr. Tomlin- 

 aon has shown that for those metals which he had examined the 

 order of magnitude of the specific electrical resistances was the same 

 as that of the molecular viscosities, a connection is established 

 between the velocity of light and the* electrical conductivity of a 

 metal, agreeing with that obtained experimentally by Kundt. 



The initial assumptions in the above investigation are similar to 

 those made by Helmholtz in his papers on the "Electro-magnetic 

 Theory of Dispersion."* Some doubt has been expressed as to 

 whether Helmholtz's developments are in consonance with Maxwell's 

 theory .f In the present case the principle of Least Action is not 

 used. The dispersion formula obtained differs from that of Helm- 

 holtz, but bears a general resemblance to that obtained by Rieff in 

 his modification of Helmholtz's theory. J A more definite physical 

 significance is, however, given to the various constants introduced. 



* < Wied. Ann.,' vol. 48, pp. 389405, 723725. 



f O. Henviside, ; Electrician,' vol. 37, August 7, 1896. 



J ' Wied. Ann.,' vol. 55, pp. 8295. 



