on Magnetic Rotatory Polarization in Gases. 317 



The glasses first used were very good ; but an accident 

 having caused their simultaneous breakage, I had much 

 difficulty in replacing them, and was obliged to try several 

 pairs before obtaining those that I have employed throughout 

 the work, and which are at least as good as the first ones. 

 These glasses are of St.-Gobain glass, about 0*005 metre in 

 thickness, and are finished with the greatest care, so that they 

 do not distort the images in the least degree, at any rate up to 

 the fifth passage. Even after nine passages they do not 

 depolarize the light to an appreciable extent, and the images 

 can still be reduced to equality of tint with remarkable pre- 

 cision. 



When the apparatus is not perfectly regulated, it is found 

 that the positions of the equality of tint of the various images 

 are no longer exactly the same as they ought to be. This 

 arises from the fact that the rays which reach the eye for each 

 image pass through the polariscope and the analyzer more or 

 less obliquely. This inconvenience would be obviated by ad- 

 justment ; but as the angle of the two deviations obtained by 

 reversing the direction of the electric current is always 

 measured, the absolute direction of the plane of polarization 

 of the incident ray does not enter into the determinations. It 

 is equally easy to prove that the luminous rays which pass 

 through the glasses almost normally cannot undergo any ap- 

 preciable amplification or diminution in their magnetic ro- 

 tation in consequence of the phenomena of refraction of the 

 polarized light. 



However, it was proved by a special experiment that a slight 

 rotation of the plane of polarization of the luminous rays 

 experienced by them on issuing from the polariscope was still 

 the same at the fourth image reflected ; so that the passage 

 through the glasses and the successive reflections did not, 

 with the exception of the magnetic effect, introduce any appre- 

 ciable correction into the measurements. The only important 

 correction occasioned by the glasses has relation to the mag- 

 netic rotation which they cause the plane of polarization of 

 the light to undergo. This correction was studied with the 

 greatest care, and it will be treated of hereafter. 



It would have been preferable to avoid this difficulty by 

 placing the glasses further away from the bobbins ; but this 

 arrangement would have tended to make the size of the ap- 

 paratus too large. It would also have been possible to use 

 only the four centre bobbins ; but the diminution of the 

 distance traversed by the light under magnetic influence was 

 not compensated for by the increased intensity of the current 

 of the pile circulating in a solenoid of less resistance. 



