Transparent Bodies by the action of Magnetism. 495 



induced current by the rotation of the small bobbin ; then raising, 

 by means of a movement of the rack A, fig. 3, the transparent 

 body into the place previously occupied by the bobbin, I turned 

 the analyser so as to determine the azimuth of the tint of passage 

 if I worked with white light, or the azimuth of extinction if I 

 worked with homogeneous light, reversing the direction of the 

 current, but taking care not to break the circuit*. I deter- 

 mined anew the same azimuth. The difference of the two obser- 

 vations evidently gave the double of the rotation of the plane of 

 polarization, if the magnetic action had not perceptibly varied 

 during the experiment. In order to assure myself, I measured 

 the magnetic action immediately after the determination of the 

 second azimuth, and only regarded those experiments as good 

 where the two measurements of magnetic action did not differ in 

 their mean value by more than a hundredth of that value. It is 

 this mean value which I have set in the tables to be given further 

 on. I perceived, like observers who have preceded me, that the 

 magnetism of the electro-magnet takes a certain time to develope 

 itself and attain its maximum. It is not then necessary to com- 

 mence the experiments immediately after closing the voltaic 

 pile ; it is only after some time that the magnetism has taken a 

 value which does not vary perceptibly during a complete expe- 

 riment. 



The galvanometer ought to be placed at a great distance from 

 the electro-magnet, in order that it may not be affected by it ; I 

 was obliged to have an assistant to move the little bobbin while 

 I observed the motions of the magnetized bar. I have not had 

 recourse to any particular mechanical arrangement, experience 

 having shown me, that, notwithstanding the little irregularities 

 which movements effected by means of the hand always involve, 

 the results of many consecutive observations agreed perfectly, 

 provided the movement was sufficiently rapid. If by chance the 

 movement was too slow, or if it wa3 performed at different times, 

 I perceived it by the movement of the needle, and I recommenced 

 the experiment. 



I have experimented solely on three substances, — the heavy 

 glass of Faraday, common flint, and bisulphide of carbon ; but 

 these three substances differ so much from one another, that a 

 law which suits them equally may be regarded as general. I 

 had at my disposal two specimens of heavy glass ; the first, which 

 M. De la Hive was good enough to lend me, was a parallclopiped 

 with square base, 40 millims. long by 13 millima. on the side, 

 polished on its two bases und on two of the lateral faces; the 

 second, which belonged to the collection of the Ecole Normale 



* This precaution is useful, so as not to cause variations of intensity in 

 the current. 



