THE ELECTRIC AND LUMINIFEROUS MEDIUM. 
247 
P rismatic Deviation hy Opaque Media. 
34. The tact that light preserves its period shows (Part I., § 90) that the circum¬ 
stances of its propagation across opaque media are determined simply by a complex 
index of refraction, of which the imaginary part represents the absorption. Measures 
of deviation by opaque prisms, such as those made by Kundt, yield directly the value 
of this complex index, by simple consideration of the geometrical continuity of the 
traces of the waves along the interfaces, without the necessity of the intervention of 
any dynamical theory whatever and therefore free of all ambiguity of interpretation. 
The thickness of the portion of the prism that is traversed does not affect the devia¬ 
tion of the light; so it may be taken as null, and we have only to consider refraction 
into the prism, and then out of it at a plane inclined at an angle differing by a, the 
angle of the prism, without changing the point of incidence. Let the axis of x be in 
the first face of the prism, towards the edge, that of y normal to it and that of s 
parallel to the edge. Then for the incident, reflected, and refracted waves, of period 
2 ttIp, the vibration-vectors are proportional respectively to 
exp t [lx dr ~ ~ pO’ d" U ~ PO 
where 
70 t 0 ^—0 o 70 I /o Tr f —0 0 
r + = c r + ^ = Iv c 
c being the velocity in free space, and K' the complex value of the square of the index 
of refraction. It will now be convenient to refer the second refraction to correspond¬ 
ing axes f, 7), ^ related to the second face of the prism ; thus 
X — ^ cos a — y sin a, y = ^ sin « -|- y cos «, : = 
The vectors of the ’ icident, reflected, and emergent beams are then proportional to 
A' exp t (\^ r /xy — pt), B exp i [\^ — jxy — pt), B' exp i -}- yy — pt), 
where 
\ = I cos a -p ni sin a, jx — — I sin a -{- ni cos a, and /x'-' = 
When the refracting angle a is small, this gives approximately jjp ~ nd — 2 hn a, 
tlae primary exciting wave. The analogous medium for sound, filled with fixed attuned resonators, is 
absorbent solely on account of the secondary radiation of the resonators: consequently if they were all 
alike and regularly distributed and they occupied, a very large number of wave-lengths, there would be 
no absorption and the medium would be trausparent to sound, unless it has the same period as the 
resonators when it could not penetrate into the medium at all. The coiTelative absorption of light 
would thus be a process special to it, arising from ionisation and molecular impact. Unless the 
absorption in iodine vapour is accomplished by ionisation so that it goe.s mainly into heat, there must be 
scattered light accompanying it and representing part of it. 
