The Interference of Natural Light on a Magnetic Field , 3 



electromagnetic rotation, the making of the current circuit should 

 cause the reappearance of the interference fringes. Such was the 

 phenomenon observed by Sohncke. 



The writer has recently been using a form of apparatus 1 which 

 suggested the further verification of this work in a somewhat 

 different manner. The apparatus used consisted of the more 

 general type of the Michelson interferometer, in the path of each 

 ray of which was placed a tube of carbon bisulphide surrounded 

 by a coil and supported independently of the coil. The coils were 

 connected so as to produce a rotation in the case of plane pol- 

 arized light- in opposite directions in the two tubes. The source 

 used was sodium light. If natural light' is capable of rotation 

 in an electromagnetic field, varying the current should cause the 

 interference bands to pass through alternate maxima and minima 

 of intensity. The positions of the minima should then correspond 

 to the times at which the current would produce in the two tubes 

 a rotation of an odd multiple of 90 . 



Plane polarized light in passing through one coil was found to 

 be rotated 90 by a current of 13.06 amperes. In the case of the 

 other this was found to be 13.23 amperes. The mean of these 

 two values should give the current necessary in both coils if 

 acting in opposite directions to produce a rotation of 180 . 



Using the apparatus described, a series of readings were taken 

 as follows: A rheostat in the circuit was adjusted, increasing 

 the current until the quantity passing through the coils was just 

 sufficient to cause the disappearance of the bands ; at this point 

 tbe ammeter reading was taken by an assistant. The current 

 was then increased until the bands reappeared. The motion of 

 the lever arm of the rheostat was then reversed and the current 

 decreased until they again disappeared ; at this point a second 

 reading was taken. The mean of these two readings should give 

 the value of the current for which the fringes were entirely invis- 

 ible. The point of disappearance of the bands it was found could 

 most easily be ascertained by the following device. The com- 

 pensating plate of the interferometer was mounted on a brass 



1 The apparatus used was that granted to Messrs. Morley and Miller by 

 the A. A. A. S. See Phys. Rev. 7, p. 283, 1898. 



147 



