2 IV. B. Cart 111 el 



upon tliis were carried out the experiments which form the basis 

 of the following paper. 



As already pointed out by Wood\ the great difficulty in deter- 

 mining the dispersion of strongly absorbing substances by inter- 

 ferential methods is that the ray which passes through the sub- 

 stance is so reduced in intensity that it is not capable of causing 

 interference when it meets the undiminished light of the other 

 ray. This may be easily conceived to be the case with fuchsin, 

 when we consider that a film of fuchsin a wave-length thick 

 transmits only five parts in one hundred million of the incident 

 light within the absorption band. Experiments were, therefore, 

 made with the object of reducing, if possible, the intensity of the 

 light in one of the paths of the interferometer, without producing 

 any change in its optical length, in order that the fuchsin might 

 . be placed in the more intense beam. However, none of the 

 various plans tried were adopted. It was then decided to use a 

 form of interferometer in which the light does not return upon 

 itself, and for two reasons : First, because in traversing the film 

 twice the diminution in intensity is squared while the retardation 

 is only doubled; and, second, because the enormous reflection 

 from the surface of the film obscures the fringes in the ordinary 

 form of interferometer, but in the type used the reflected light 

 does not reach the observer's eye at all. 



After a number of trials to determine the best adjustment, it 

 was found possible, by using this type of interferometer and mak- 

 ing the films sufficiently thin, to obtain distinct fringes through- 

 out the spectrum. An unsymmetrical arrangement was first tried, 

 in which the partly silvered plates were very lightly silvered, or 

 not silvered at all, and the fringes observed at a direction at right 

 angles to that at which the light entered the instrument. In this 

 way fringes were obtained from beams of unequal intensity, 

 though the method was finally abandoned for the following, 

 which is more satisfactory. 



The partly silvered mirrors of the interferometer are silvered 

 so as to reflect and transmit equally, in order that the light in the 

 two paths may be of equal intensity, and then the fuchsin film is 



^P/ii/. 3Iag., vol. 1, p. 43. 1901. 



102 



