jEt her -par tides in a Rectilinearly Polarized Ray of Light. 193 



interference-bars for rays with vibrations perpendicular to the 

 plane of incidence in the case of glass and metal reflexion 

 must coincide for all angles of incidence ; while for rays with 

 vibrations parallel to the plane of incidence, the interference- 

 bars arising from glass and metallic reflexion must be displaced 

 in reference to one another, and the displacement must increase, 

 to a determinate limit, with an increasing incident angle *. 



By means of a heliostat the author projected solar rays in a 

 horizontal direction upon a vertical slit, which was in the focus of 

 an achromatic lens. The rays emerged from the lens almost 

 parallel, and then fell upon a plane-parallel glass plate, and, 

 after reflexion from this, upon a second parallel glass plate of 

 exactly the same thickness, at a distance of about 8 inches, 

 which was very slightly inclined towards the first ; so that the 

 rays reflected from the posterior surface of the first and from 

 the anterior surface of the second plate could interfere with 

 the rays reflected from the anterior surface of the first and 

 the posterior surface of the second, as in a similar apparatus 

 which Jaminf has used and described. A screen darkened 

 the rays reflected from both anterior or both posterior sur- 

 faces of the parallel glasses, and the rays transmitted by the 

 screen fell upon a bisulphide- of-carbon or flint-glass prism, 

 with a refractive angle of 60°, and a vertical refracting edge, 

 in which a spectrum with thick dark interference-bars was seen 

 with the naked eye. By suitably choosing the breadth of the 

 slit, and altering its distance from the lens, the Fraunhofer's 

 lines could be made to be distinctly seen in the spectrum ; and by 

 a suitable adjustment of the parallel glasses, the interfeence-bars 

 were made parallel to Fraunhofer's lines in the spectrum. 



The second plane-parallel glass was covered on the lower half 

 of the posterior surface with a metallic mirror, and in the spec- 

 trum the rays reflected from glass-glass and glass-metal could be 

 simultaneously seen, inasmuch as the upper part of the spectrum 

 had been reflected at the surface of air and the posterior surface 

 of the second glass, the lower half had been reflected at the limit 

 of metal and the same posterior glass surface. Although both 



* From Jamin's observations on elliptic polarization in reflexion from 

 metals or opake substances (Joe. cit.) with the aid of Babinet's com- 

 pensator, the question cannot be decided for which components of the 

 rays polarized parallel or perpendicular to the plane of incidence the 

 phase of the reflected ray continually alters with the angle of incidence, 

 as it remains dubious whether, by altering the compensator (in positive 

 reflexion), the component polarized in the plane of incidence must be con- 

 tinuously retarded, or the component polarized perpendicular to the plane 

 of incidence must be continuously accelerated, in order to nullify the dif- 

 ference of phase of both components. 



f Ann. de Chim. et de Pkys. (3) 1858, vol. lii. p. 163 et seq. 

 Phil. Mag. S. 4. Vol. 26. No. 174. Sept. 1863. 



