Transparent Bodies by the action of Magnetism. 507 



parent substance, and. afterwards analysed beyond the screen. 

 The direction of magnetic action was first brought into parallelism 

 with the direction of the rays of light, by making the zero of the 

 vernier C coincide with the zero of the graduated scale beneath 

 it ; the transparent substance was next placed so that the faces 

 of entrance and exit were perpendicular to the same direction : 

 this was done by making the zero of the vernier on the plate O 

 coincide with the zero of the corresponding graduation; and 

 lastly, the optic phsenomenon was observed. Afterwards, the 

 electro-magnet was turned through any angle whatever ; this of 

 course caused the transparent substance to turn through an exactly 

 equal angle, but it was brought back to its primitive position by 

 an opposite and exactly equal rotation of the plate 0; the luminous 

 rays, however, preserved an invariable direction, and consequently 

 they always traversed the same thickness of transparent substance. 

 The apparatus being thus disposed, the optical phenomenon 

 was observed, care being taken to repeat each experiment twice 

 by turning the electro-magnet successively towards the right and 

 towards the left ; afterwards, the mean of these two results was 

 taken in order to eliminate the possible influence of imperfections 

 in the construction and arrangement of the several parts of the 

 apparatus. With small angular displacements of the electro- 

 magnet, these two results were not sensibly different from one 

 another; when the angular displacement attained or exceeded 45 

 degrees, the difference between the two results was greater than 

 the usual errors of observation. 



In the first place, it was easy to corroborate the fact that the 

 optical phenomenon is, in all cases, simply a rotation of the 

 plane of polarization. By using homogeneous light, and first 

 placing the analyser so as completely to extinguish one of the 

 two images of the diaphragm, I have witnessed the reappear- 

 ance of this image by the influence of magnetism ; but whatever 

 might have been the relative situations of the electro-magnet 

 and of the transparent substance, I was always able to extinguish 

 this new image by a convenient rotation of the analyser. By 

 using white light I have seen the image reappear with variable 

 colours, and by changing the position of the analyser, I have 

 always found the tints to succeed each other in the order peculiar 

 to these phenomena of rotation, whatever might have been the 

 arrangement of the apparatus. In common with my predeces- 

 sors, I have always observed that no effect was produced when 

 the magnetic action was perpendicular to the direction of the 

 rays of light. 



Thus these phenomena manifest themselves in a less compli- 

 cated manner than might have been expected ; when the direc- 

 tion of the rays of polarized light is parallel to that of magnetic 



