320 M. H. Becquerel's Experimental Investigations 



to employ. The corrections due to this class of phenomena 

 with the piles of glasses were thus very complex. 



This multiplicity of experimental measures and the un- 

 certainty as to the real degree of amplification in each case, 

 take away all the advantage of the large rotations which may 

 be obtained, and do not permit a greater relative approximation. 

 Therefore, in order not to multiply to excess the very delicate 

 precautions with which the measurements were already sur- 

 rounded, I relinquished the idea of making use of piles of 

 glasses. 



On the Distribution of Magnetic Intensities in the Apparatus. — 

 Magnetic Rotation of a Column of liquid Carbon Bisulphide. 



It is well known that the magnetic rotations of the plane of 

 polarization of light are proportional to the magnetic intensity; 

 the numbers given below will have real significance only 

 when the corresponding magnetic intensity is accurately 

 determined. 



This can be estimated in absolute measure ; but the deter- 

 mination offers many difficulties ; and it is much simpler and 

 more accurate to measure the magnetic rotation of a well- 

 defined body under the same conditions of magnetic intensity 

 as those used for the experiments, and to compare the numbers 

 obtained with the magnetic rotation of this body taken as 

 unity. 



In my previous researches on magnetic rotatory polarization 

 I had already taken as unity the rotation of the plane of 

 polarization of the yellow rays of the flame of sodium tra- 

 versing a column of liquid carbon bisulphide at a temperature 

 of 0°, with a thickness equal to that of the body studied, and 

 subjected to the same magnetic intensity. In the present 

 experiments I have kept to the same unit. 



It thus became necessary to measure the magnetic rotation 

 obtained with the yellow light of sodium passing through 

 a column of carbon bisulphide equal to the length of the 

 distance traversed by the luminous rays in the apparatus above 

 described. Now, at different points of the line traversed by 

 the luminous rays the magnetic intensity is not the same : in 

 the inside of the large solenoid formed by the six electro- 

 magnetic bobbins it is nearly constant; but it decreases very 

 rapidly outside these bobbins, and is no longer appreciable at 

 about 0*20 metre from their extremities. 



This exterior action is very weak compared with the mag- 

 netic influence of the interior of the solenoid ; nevertheless, on 

 account of the very peculiar conditions of the experiments, it 



