424 M. H. Becquerel's Experimental Investigations 



Almost the same results are obtained as with the glasses 

 placed inside the first coil, especially with the white light. 

 For example, if the rotations are compared with that which is 

 deduced from the fourth image, we shall find that the rota- 

 tions for a single passage of the white light are expressed by 

 the following numbers : — 





4th 



m 3rd 



t 2nd 



1st 



defiant gas 



image. 

 . 1 



image. 

 1-054 



image. 



1-083 



image. 



1-066 



St.-Gobain glass 



. 1 



1-046 



1-058 



1-061 



We see, then, that the divergences obtained between num- 

 bers which should be constant are partly due to the same cause 

 — that is, to the difference in the colour of the various images. 



If this is the case, the successive images with the white 

 light ought to become more and more yellow, and the rota- 

 tions should diminish from the first to the fourth. For the 

 green light, on the contrary, the first image ought to be more 

 yellow than the others, and the corresponding rotation more 

 feeble. We find, in point of fact, that this deduction is con- 

 firmed by experiment. The absorption of the luminous rays 

 by the coloured screens renders observations on the third and 

 fourth images very difficult. It is necessary to increase the 

 pressure of the oxygen from the blowpipe : the temperature 

 of the line rises ; but as the brilliancy increases it becomes 

 more and more variable for very slight differences in tempe- 

 rature. 



The want of regularity in the value of numbers deduced 

 from the different images apprises us of these variations ; 

 nevertheless by taking the average of the numbers obtained, 

 we see that we have results comparable with those deduced 

 from observations made for the same colours with the carbon 

 bisulphide. 



II. Dispersion of the Planes of Polarization of Luminous Rays 

 having different Wave-lengths. 



The study of the phenomena of magnetic rotatory polariza- 

 tion in bodies comprises two parts: — 



1st. The relative measurements of the rotations of the planes 

 of luminous rays having different wave-lengths and traversing 

 the same substance. 



2nd. The comparison of the magnetic rotatory powers of 

 various bodies for rays having the same wave-length. 



I have demonstrated in the course of this memoir that the 

 want of homogeneity in the luminous source compelled me to 

 begin with the first of these studies — that is to say, to deter- 

 mine the relation of the rotations of the planes of polarization 

 of rays of different colours. 



