Transparent Bodies by the action of Magnetism. 501 



M. Wiedemann has demonstrated that the rotatory power 

 developed by the direct action of electric currents is proportional 

 to the intensity of these currents. He has made no experiments 

 on the influence exercised by the relative situation of the electric 

 currents and of the transparent bodies ; but if my results are 

 compared with those of M. Wiedemann, if further, the general 

 identity of the properties of magnets and of systems of closed 

 currents be taken into account, it will appear evident that the 

 rotatory power developed by a system of closed currents in an 

 infinitely thin plate of a transparent substance ought to be pro- 

 portional to the action which the system would exercise on a 

 molecule of magnetic fluid. 



I find a complete contradiction to the law announced by M. 

 Bertin, according to which the rotation of the plane of polariza- 

 tion due to the influence of a single magnetic pole will decrease 

 in geometric progression when the distance from the transparent 

 substance to the pole increases in arithmetical progression. The 

 explanation of this disagreement is not difficult to give. M. Bertin 

 considers as pole, the terminal surface of soft iron of one of the 

 branches of Ruhrnkorff's electro-magnet. Now this surface 

 could not be regarded as pole, at least if to this expression be 

 attributed its precise signification; it is a system of magnetic 

 centres distributed over a great surface, and whose action cannot- 

 be assimilated to that of a single centre. We cannot then look 

 for the elementary law which makes the rotation of the plane of 

 polarization depend on the distance from the transparent sub- 

 stance to this polar surface ; we can only find an empirical for- 

 mula, which ought to change when the electro-magnet is changed, 

 or even when we simpl y change the terminal armatures of the 

 same electro-magnet. But if the law which I have established 

 in this memoir be true, the empirical formula which represents 

 the decrease of the rotations at different distances from the polar 

 surface ought also to represent the decrease of the magnetic 

 action, since these two quantities are always proportional; con- 

 sequently, in the apparatus of M. Bertin, the magnetic actions 

 ought to decrease in geometric progression when the distances 

 from the polar surface increase in arithmetical progression. 



Direct experiments have completely confirmed this conclusion. 

 The apparatus which I used, and which belongs to the Ecolc 

 Normale superieure, was that which M. Bertin also used in his 

 researches. I took away one of the branches of the electro- 

 magnet, and at the extremity of the branch that I retained, I 

 screwed instead of the large armature of my experiments, the 

 small hexagonal armature which M. Ruhmkorff generally uses 

 in his apparatus, and which M. Bertin himself used. 1 deter- 

 mined by ; the method before indicated, the magnitude of the 



