of Reflexion of Polarized Light, 



Table II. 

 Incidence =69° 37\ 



511 



In this table, with an angle of incidence greater than the 

 polarizing angle, we observe the same facts as before, viz. that Q 

 and fi are not constant but variable ; and that not in an irre- 

 gular manner, but according to some law. The value of /a is 

 nearest to the refractive index in both tables when the azimuth 

 of incident polarization is least, and diminishes as the azimuth 

 increases, being in all cases less than that of the refractive index 

 found in the usual way by refraction. 



It is impossible to believe that the law contained in equation 

 (3) can be accurate, when we find it leads to a regular deviation 

 from the results of observation. It is not that the values of jju 

 do not coincide with the refractive index, some being greater 

 and some less, in which case we might suspect some errors of 

 observation ; but the deviation is regular and real, and can only 

 arise from the inaccuracy of equation (3). 



The azimuth A' is that of the major axis of the reflected 

 elliptic polarization, and, as will be shown hereafter, this polar- 

 ization sometimes approaches to the circular. 



1. Investigation of elliptically-polarized Light. 



The instrument used by M. Jamin for the examination of the 

 elliptically-polarized light reflected from transparent bodies, 

 differs from that used in the preceding experiments by the ad^ 

 dition of a compensator placed in front of the tube containing 

 the Nicol prism analyser. This compensator is fully described 

 in M. Jamin^s paper, and serves to convert an elliptically-polar- 

 ized beam into a plane-polarized beam. Its principal planes are 

 in the plane of incidence and perpendicular to it, and the distance 

 to which the moveable prism is moved from the zero serves to 

 calculate the difference of phase between the components of the 

 elliptic beam in, and perpendicular to, the plane of incidence. 

 The plane-polarized beam which passes through it is stopped by 

 the analyser, which reads the position of the polarization of the 

 light which has become plane-polarized by the action of the 

 compensator. 



