THE ELECTRIC AND LUMIHIFEROUS MEDIUM. 
787 
cause sensible absorption in a verj^ small fraction of a wave-length, the analytical 
formulae might be entirely altered. 
Sir George Stokes* has however supported the view that besides the effects due 
to simple absorption, metals probably also show reflexion phenomena involving change 
of phase, such as were originally discovered by Airy for the diamond, and were 
afterwards found in other highly refractive substances. These effects, which were 
extended by Jamin to ordinary media, have been eliminated by Lord Kayletgh for 
the case of water by cleansing of the surface, by which means the sharpness of the 
optical transition would be improved. The phenomena for the case of diamond were 
long ago classed by OREENt as a result of gradual transition ; and this might be 
expected to be more marked between hard substances whose optical properties are 
very different. On this view we may not be driven to try the hypothesis of extreme 
absorption in the interficial layer, wdiich is unsatisfactory for the same reasons as 
apply to Kirchhofe’s doctrine of extraneous forces ; the quality above mentioned, 
for which Sir George Stokes proposes the name of the adamantine property, being 
sufficient. 
Reflexion hy Partially Opaque Media. 
94. The ordinary formulse for reflexion at the surface of an absorbing medium may 
now be derived from the analytical functions which express the averaged dynamical 
constitution of the mther for the case of its vibrations in ponderable bodies. Tf the 
general argument is correct, it is to be expected that these formulae would be verified 
for reflexion at the surfaces of such media as are not too highly absorbent in com¬ 
parison with the length of the wave. There are in fact two extreme cases ; first the 
reflexion of electromagnetic waves of sensible length from metallic surfaces, where the 
reflexion is complete and there is no absorption at all; and second the reflexion of 
waves from perfectly transparent media, where the reflexion is incomplete because 
part of the energy goes on in the transmitted wave. The reflexion of light from 
metals may conceivably be more nearly akin to the first of these limiting cases thau 
to the second ; but for media more transparent than metals wm should expect closer 
agreement with the ordinary theory, now to be developed. 
95. The general variational equation of the motion is 
|(8T-8W-8'dF)c?T=0, 
leading to 
* In a note appended to a paper by Sir J. Conroy, “ Some experiments on Metallic Reflexion,” ‘ Roy. 
Soc. Proc.,’ Feb., 1§93. 
t Gr. Green, “ Supplement to a Memoir on tlie Reflexion and Refraction of Light,” ‘ Trans. Camb. 
Phil. Soc.,’ May, 1839. 
5 H 2 
