410 M. Verdet on the Optical Properties developed in 



scribed in my first memoir, adopting, however, certain modi- 

 fications, the efiect of which has been to cause either an increase 

 in the magnitude of the phsenomena to be observed, or a dimi- 

 nution in the duration of the experiment, without in the least 

 impairing the exactness of measurement. In place of the great 

 armatures made use of in my former researches, I have employed 

 much smaller ones, consisting of iron plates 8 millims. thick, 

 in the form of regular hexagons whose sides are 13 millims. long*. 

 The effect of this has been, all other conditions being the same, 

 to increase considerably both the magnetic and optical actions ; 

 and even when these two actions have not been constant through- 

 out the whole thickness of the substance placed between the ar- 

 , matures, yet such variation has not been attended by any incon- 

 venience. For when any two transparent bodies of the same 

 thickness have been placed successively in the same position 

 between the armatures of the electro-magnet, the several corre- 

 sponding layers of both have been subjected to equal magnetic 

 actions, and have therefore exercised optical actions proportional 

 to the specific optical actions of the two bodies. Hence it is 

 clear that the sums of these elementary optical actions, that is, 

 the total observed rotations, bear the same proportion to one 

 another as if the magnetic, and consequently also the optical, 

 action had been invariable throughout the whole magnetic field. 

 Now it is clear that this proportion is the only thing with the 

 determination of which we are concerned. 



The intensity of a current, and consequently also that of an 

 electro-magnet, never remains constant during a series of experi- 

 ments extending over several days, or even hours; and still 

 greater variations are produced when the liquids of the battery 

 are renewed, or the number of its elements altered. In order 

 that my determinations might be independent of the efiect of 

 these variations, recourse was had to two methods. In the first, 

 every measurement of optical action was both preceded and fol- 

 lowed by an estimation of the magnetic force midway between 

 the armatures of the electro-magnet; the mean of these two 

 values was the magnetic action with which the optical action was 

 compared. Experiment showed that for every substance this 

 relation is constant, as long as the valuations in strength of the 

 electro-magnet are confined within certain limits; — limits more- 

 over which may be much more widely apart than those bounding 

 the variations occurring in my actual experiments t- Hence, by a 



* These are the hexagonal armatures attached to all RuhmkorfF's ma- 

 chines, and which are mentioned in the first part of my researches. — Ann. 

 de Chim. et de Pkijs., ser. .3. vol. xli. p. 409. 



t Thus, on passing the current, first of twenty and then of ten of Bun- 

 sen's elements, through the wire of the electro-magnet, I obtained magnetic 

 actions represented by the numbers 100'7 and 7H3. The corresponding 



