ON OPTICAL THEORIES. 217 
Roux! had found that vapour of iodine refracted red light more strongly 
than violet, and Christiansen, in the paper quoted, announced the result 
that for a solution of the aniline dye fuchsin in alcohol the refractive 
index increases from the Fraunhofer line B to D, then sinks rapidly as 
far as G, and increases again beyond. The experimental investigation of 
the subject was continued by Kundt,? who proved that this anomalous 
dispersion was marked in all substances showing strong surface color- 
ation, and that there was an intimate relation between it and the 
absorptive power of the substance. As the result of his experiments, 
Kundt was able to lay down the rule that in going up the spectrum, 
from red to violet, below an absorption band the deviation is abnormally 
increased by the absorption, while above the band the deviation is 
abnormally decreased. Kundt has been able to see this abnormal effect 
produced by the absorption of sodium light. 
On the old theory of dispersion, as developed by Cauchy and others, 
this effect was inexplicable. Boussinesq, it is true, had explained the 
phenomena in vapour of iodine by saying that it implied that the co- 
efficient D was positive; and here, in a way, lay a germ of the truth, for 
the mutual reaction theory lends itself readily to a partial explanation of 
the whole. 
§ 5. Such an explanation was first given by Sellmeyer. He had 
been led to expect the effect from theoretical reasons in 1866, and had 
endeavoured to discover it in a fuchsin solution, but without success. 
The action between the ether and matter is a periodic one of the same 
period as the light-wave traversing the ether. Owing to the enormous 
density of the matter compared with the ether its motion will in general 
be negligeably small; but if it should happen that the period of the 
natural vibrations of the matter particles coincides with that of the 
incident disturbance this will no longer be the case. The energy of the 
light-vibration will be absorbed by the matter, and this absorption will 
tend to react on the light-disturbance, and will, it can be shown, increase 
the refractive power of the medium for disturbances of greater period 
than the critical one, and decrease it for disturbances of less period. 
The problem is much the same as that of a pendulum the point of 
support of which is undergoing a small periodic disturbance. If the 
period of the disturbance be greater than that of the natural vibration of 
the pendulum the reaction of the pendulum on its support will tend to 
quicken the motion of the latter, and vice versd. 
Sellmeyer, in the papers referred to,* published in 1872, after a most 
clear and able discussion of the difficulties of the elastic solid theories, 
adopts the hypothesis that the ponderable atoms vibrate, but with much 
smaller amplitudes than the ether particles. He then proceeds to consider 
the mechanism by which this is brought about. As with Boussinesq, the 
ether is supposed to have the same rigidity and density everywhere. The 
ether particles act directly ou the matter particles, and in consequence of 
the vibrations of the former the equilibrium positions of the latter are 
1 Ann. de Chim. 8. III. t. xli. p. 285. ' 
* Pogg. Ann. t. 142, p. 163; t. 143, pp. 149, 259; t. 144, p. 128; t. 145, pp. 17 
and 164. 
% Sellmeyer, Pogg. Ann. t. 142, p. 272. 
* Sellmeyer, ‘ Ueber die durch die Ather-Schwingungen erregten Mitschwingungen 
der K6rpertheilchen und deren Riickwirkung auf die erstern, besonders zur Erklérung 
ee Dispersion und ihrer Anomalien,’ Pogg. Ann. t. 145, pp. 399, 520; t. 147, pp. 386, 
