244 



Sir J. Conroy on 



[Jan. 9, 



were determined directly, but less accurately than by Dr. Wiede- 

 mann's arrangement, which, however, involves a good deal of cal- 

 culation. 



The analysing portion of Dr. Wiedemann's instrument appears to 

 differ merely by the addition of a small direct vision spectroscope from 

 the Elliptic Analyser of Professor Stokes, described in the "Report of 

 the British Association for 1851," Part II, p. 14. 



The experiments were made with sun-light and with lamp-light ; 

 with the former,, the angle of incidence and the azimuth of the ana- 

 lysing nicol were altered till the dark band in the spectrum was most 

 intense at certain definite positions, as measured by the scale of the 

 spectroscope ; with the latter, till the light which had passed through 

 a piece of red glass was reduced to a minimum. 



Numerous measurements were made of both these angles with a 

 plate of gold in air, water, carbon bisulphide, and carbon tetrachloride ; 

 and of copper, in air, w^ater, and carbon tetrachloride ; but the results 

 were not very satisfactory. In addition to the difficulty of deter- 

 mining accurately the zero of the nicols, and of placing the neutral 

 axis of the quarter undulation plate in the plane of incidence, I found 

 that very different values were obtained for the principal incidence, 

 according as one or other of the neutral axes of the quarter undulation 

 plate I was using was in the plane of incidence. 



In all cases, however, the principal incidence which, as is well 

 known, is less for the more refrangible rays, diminishes, and the prin- 

 cipal azimuth increases with the increase of the refractive index of 

 the medium in contact with the metallic surface ; arid further, the 

 diminution in the value of the principal incidence appears to be nearly 

 in proportion to the increase of the refractive index of the surround- 

 ing medium. 



The decrease of the principal incidence, with an increase in the 

 refractive index of the surrounding medium, is exactly what might be 

 expected to take place if the principal incidence for a metal were the 

 same as the angle of polarization of a transparent substance; that is, 

 the angle whose tangent is equal to the refractive index. 



If such is the case, the metals must all have very high refractive 

 indices; but some experiments of Quincke's ("Pogg. Ann.," vol. cxix, 

 p. 379, and vol. cxx, p. 602) appear to show that their refractive indices 

 are less than 1. 



The following are some results I obtained with a gold plate (formed 

 by soldering a slip of thin sheet gold to a brass plate), in air, with 

 lamp -light, a deep red glass being interposed; the position of the 

 quarter undulation plate in which the ray polarized perpendicularly to 

 the plane of incidence was retarded relatively to the other, being 

 called A, and that in which the retarded ray was the one polarized 

 in the plane of incidence, B. The signs of the azimuth of the plane of 



