Transparent Bodies by the Action of Magnetism. 413 



by examiuing the passage-tiut from distilled water, through a 

 thickness of the coloured substance under investigation, equa 

 to that afterwards employed in acting on the light. The lumi- 

 nous pencil which fell upon the analyzer had thus the same 

 composition as in the case of the coloured substance. The 

 passage-tint in both experiments corresponded therefore to the 

 plane of polarization of the same colour of the spectrum, and the 

 relation of the deviations of this tint gave the relation between 

 the optical actions. 



Finally, the particular nature of the substances which I ex- 

 amined, rendered another correction necessary. In order that 

 my experiments might be confined to perfectly definite bodies 

 and such as may be obtained in a state of perfect puritj^, I have 

 only examined liquids* ; and these of course had to be introduced 

 into cases closed at their ends with transparent jjlates. If care 

 be taken to choose plates of glass free from Haws, and to close 

 the cases in such a manner that the pressure upon the plates is 

 sufficiently small and uniformly distributed upon their whole 

 circumference, then the plates, in their ordinary state, have no 

 action upon polarized light ; but under the influence of mag- 

 netism they exercise a sensible action, which assists the action 

 of the liquid, and which must therefore be taken into account. 

 For this purpose, before employing a case in my experiments, I 

 examined it, successively, empty and full of distilled water, so as 

 to determine the relation between the rotation produced by the 

 action of the plates and the total rotation produced by the simul- 

 taneous action of the plates and the water. It is easy to under- 

 stand how the knowledge of this relation enables us to correct all 

 the observations made with the same case. The above determi- 

 nation was, moreover, repeated as often as I had occasion to alter 

 the relative positions of the case and the armatures of the electro- 

 magnet. In my experiments, twice the rotation of the plane of 

 polarization due to the action of the glass plates varied from 30' 

 to 1°, according to the nature of the plates and their distances 

 from the armaturesf. 



* Tiie only solid bodies suitid)Ie for exiierimenting ou are glasses, — 

 bodies whose composition is too ill defined for much interest to attach to 

 th(;ir action. The singly-refracting crystals which might he employed arc 

 80 difficult to obtain in pure and flawless pieces of any thickness, that I 

 have made no use of them in this part of my investigation. 



t It may perhajjs excite sur|)rise that 1 have sought to measure with 

 any degree of accuracy such feeble rotations as these. In order to show 

 that this method admits of exactness, I shall give the elements of a deter- 

 mination made with plates enclosing a case 4-1 millims. in thickness. The 

 distance between the arnuitures of the electro-magnet was 51 millims., and 

 tiie current was produced l)y twenty of Bunscn's elements. The current 

 having been made to i)ass in succession in two opposite directions round 



