104 REPORT ON PHYSICAL OPTICS. 



reflexion at which this light is restored to plane-polarized light, in 

 any azimuth of the plane of the second reflexion with regard to the 

 first, may be represented by the variable radii of an ellipse ; while 

 these angles are equal in all azimuths in the case of light circu- 

 larly-polarized. 



Sir David Brewster seems to have been led to employ the term 

 " elliptic polarization " in this manner, in his desire to avoid as 

 much as possible all reference to theory. The laws which he has 

 obtained, however, belong to elliptically-polarized light, in the 

 sense in which the term was introduced by Fresnel. It appears, 

 in fact, from the theory of the composition of vibrations, as laid 

 down by this author, that the vibration resulting from the union 

 of two rectilinear and rectangular vibrations will be in general 

 elliptic; so that two oppositely-polarized pencils compound in 

 general a pencil elliptically-polarized the ellipse becoming a right 

 line, when the difference of phase of the two portions is an integer 

 multiple of 180. When, therefore, by the effect of reflexion, two 

 such pencils are made to differ 90 in phase as Sir David Brewster 

 has shown to be the case when a ray polarized in the azimuth of 45 

 is incident at the maximum polarizing angle of the metal a 

 second reflexion, in the same plane and at the same angle, will raise 

 the difference to 180, and the resulting light will be plane-polarized. 

 In other parts of his memoir, however, Sir David Brewster seems 

 to acknowledge that theory, for he speaks of elliptic polarization 

 as produced by the interference of two unequal portions of oppo- 

 sitely-polarized light, and even calculates their difference of phase 

 for any incidence. 



The identity of the light produced by metallic reflexion with 

 the elliptically-polarized light of the wave-theory seems to be 

 placed beyond all doubt by an observation of Professor Airy. 

 When Newton's rings are formed between glass and metal the 

 incident light being polarized, and the angle of incidence exceed- 

 ing the polarizing angle of the glass it is found that the rings 

 dilate, as the azimuth of the plane of polarization with respect to 

 the plane of reflexion is increased ; the dilatation being a maxi- 

 mum when these two planes become perpendicular. In order to 

 account for this fact, Professor Airy has shown that if the vibra- 

 tions of the incident pencil be resolved into two, one in the plane 

 of incidence, and the other in the perpendicular plane, it is neces- 

 sary to assume that their phases are unequally changed by reflexion; 



