322 M. H. Becquerel's Experimental Investigations 



in the third decimal place ; beyond that the difference is 

 marked. I have assumed that the first formula expresses 

 tolerably approximately the progress of the phenomenon; and 

 I have reduced the rotations observed to what they would be 

 at a temperature of 0°. The numbers obtained differ by only a 

 few minutes of arc from those which would result from the ap- 

 plication of the formula of dilatation of the carbon bisulphide. 

 The following table contains the numbers obtained for the 

 rotation of carbon bisulphide in the interior of the different 

 bobbins. The multiplicity of the measurements in each case 

 enables us, as an average, to count upon an accuracy of some 

 tenths of a minute; the hundredths of a minute, resulting from 

 the calculation of the averages, have been recorded. 



Magnetic Rotations of Carbon Bisulphide (yellow light D). 





Average tem- 

 perature of 

 each series. 



-r, , , . Rotations due 

 Rotations re- , , 



duced to the t0 ™f ^ asses 



which close 

 temperature the . 50 . metre 



Of0 - tube. 



1st bobbin 



15-6 



716-21 

 746-39 

 770-04 

 767-83 

 747-97 

 702-41 



1-88 



2-51 

 260 

 2-60 

 2-51 



1-88 



2nd „ 



17-2 

 20-6 

 226 

 23-5 

 23-6 



3rd „ 



4th „ 



5th „ 



6th „ 







Total 



4450-85 



13-98 









The glasses which close the tube of 0*50 metre are of crown 

 glass, and are about 1*5 millim. thick. The magnetic rotation 

 which they cause the plane of polarization of the light to 

 undergo is very feeble. It was ascertained directly, the tube 

 being empty of liquid; and the rotations are recorded in the 

 last column of the preceding table. The sum of these correc- 

 tions is I3 /# 98, and it must be subtracted from the total 

 amount of the rotations observed. 



It is equally important to subtract from each measurement 

 the rotation due to magnetic influence on the column of air 

 traversed by the luminous rays. In accordance with what 

 will hereafter be shown, this rotation at 20° and at a pressure 

 of 760 millim. amounts to about /# 58, say 3 /, 48 to be sub- 

 tracted from the sum of the rotations. 



Placing the O50-metre tube on the outside of the first bob- 

 bin under the same magnetic conditions and at a temperature 



