194 On the Oscillations in a Rectilinearly Polarized Ray of Light. 



interfering rays had different intensity in the lower spectrum, 

 the interference-bars could be perfectly well seen, and appeared 

 displaced towards the interference-bars of the upper spectrum, 

 when looked at through a NicoPs prism the principal section of 

 which was parallel to the plane of reflexion of the glasses, that is, 

 when only rays polarized parallel to the plane of reflexion reached 

 the eye. If the NicoPs prism was turned 90°, the interference- 

 bars in both spectra coincided, in which case luminous rays, 

 polarized perpendicular to the plane of reflexion, could alone 

 reach the eye. In the latter case the intensity of light was, of 

 course, feebler. In a position of NicoPs prism in which its 

 principal section is almost parallel to the plane of reflexion, a 

 different position of the interference-bars in both spectra is seen, 

 so long as the intensity of the light polarized in the plane of 

 incidence predominates ; and for the same reason the phenomenon 

 is seen without a NicoPs prism, just as with a NicoPs prism whose 

 principal section is parallel to the plane of reflexion. 



The author used various glass plates of from 4 millims. to 

 8*5 millims. in thickness, whose refractive index varied between 

 1*45 and 1*5. For metallic coating, mercury (ordinary mirror- 

 surface), silver, gold, and platinum were used. 



The angle of incidence $ at which the rays met the anterior 

 surface of the plane-parallel glasses was varied from 20° to 60° or 

 70° ; yet not under all angles of incidence could the interference- 

 bars in the lower spectrum (glass-metal) be seen for rays pola- 

 rized perpendicular to the plane of incidence, on account of the 

 too small intensity of the light of the rays reflected at the limit 

 of air and glass, as compared with that reflected at the surface of 

 glass and metal. The displacement of the interference-bars 

 towards each other in both spectra, for rays polarized parallel to 

 the plane of incidence, increased in all metals with increasing 

 angle Q ; yet when <p is > 60° the phenomenon cannot be well 

 seen, as too much light is lost by reflexion at the anterior glass 

 surfaces. The phenomenon is best seen with an angle of inci- 

 dence <p of 45° or 60°, in which case the displacement of the 

 interference-bars amounted to about 0*3 or 0*4 of the distance 

 of the fringe. 



With an incident angle $ = 20° and less, the intensity of the 

 rays reflected at the limit of glass and air, and of the rays 

 reflected from glass and metal, is so different, that the author 

 could not discern with certainty a displacement of the interfe- 

 rence-bars in both spectra towards one another; yet from his 

 experiments it may be concluded that if it exists it is less than 

 0*1 of the breadth of a fringe. 



It may here be observed that this apparatus, by means of a 

 compensator of Soleil and Duboscq's construction, permitted a 



