314 M. H. BecquerePs Experimental Investigations 



coloured screens were always placed in front of the collimating- 

 lens. 



Polariscope. — The polariscope was formed by a very fine 

 Nicol's prism, 45*5 millim. in thickness by 45 millim. in 

 width. The prism was cut in half lengthwise, and the two 

 halves brought together so that their principal planes made an 

 angle of about 5°. The field of this prism was a circle of 

 about 0*022 metre in diameter. Such a large opening was 

 considered to be necessary — first of all to allow a sufficient 

 quantity of light to pass through, and secondly that the 

 images should retain, after the successive reflections, an ap- 

 parent diameter suitable for observation. 



The polariscope, the collimator, and the source of light were 

 firmly mounted on the same piece of copper, which was capable 

 of receiving a horizontal movement, and which could also turn 

 through a slight angle round a vertical axis, so as to cause the 

 axis of the optical system to assume a slight inclination to the 

 axis of the tube (PI. VII. fig. 2). 



Mirrors. — Suppose, for the sake of simplifying the descrip- 

 tion, that the glasses G, Gi which close the tube had been 

 removed (see PI. VIII. fig. 2). The luminous rays, rendered 

 parallel by the collimator, after having passed through the 

 polariscope, graze the edge of a vertical mirror M l5 pass through 

 the tube, and encounter at the other extremity a vertical mir- 

 ror M, from which they are reflected ; they then fall upon the 

 mirror M 1; and in this way are reflected successively on the 

 two mirrors until the last ray reflected grazes the edge of the 

 mirror M. The number of the successive reflections depends 

 on the mutual inclination of the mirrors and the direction of 

 the incident ray. We have seen how it is possible to deter- 

 mine the latter. As to the mirrors, they are mounted in cop- 

 per frames, which are fixed by three spring screws to vertical 

 wooden supports firmly fastened to the wooden block which 

 sustains the apparatus. By means of the three screws the 

 mirrors can be moved in any required direction (Plate VII. 

 figs. 2 & 4). 



If the light were polarized only in a vertical plane, the plane 

 of the successive reflections being horizontal, it would follow 

 that the plane of polarization of the reflected images would 

 remain absolutely vertical. 



With this apparatus the light of each half of the image is 

 polarized in two planes equally inclined to the vertical (that 

 is to say, to the plane of the reflections) ; each half will thus 

 experience at each reflection an equal slight rotation, and the 

 bisecting plane of the planes of polarization should remain 

 about the same for all the images. Consequently, the posi- 



