of the Plane of Polarization of Light. 321 



Since the rotations are very small in amount, it was neces- 

 sary, in order to obtain results in any measure reliable, to 

 multiply the observations. As a rule, four observations were 

 taken for each angle of incidence, for each of the two positions 

 of the plane of polarization, viz. in the plane of incidence and 

 at right angles to it. In each of the four observations the 

 polarizer was adjusted afresh, and ten readings taken with 

 each direction of the current. In the following a rotation is 

 called positive, as before, when it has the same direction as 

 the projection of the amperian molecular currents upon a 

 plane at right angles to the ray of light. 



First Series of Experiments. 



A circular iron mirror, of the dimensions given above, was 

 employed. The iron pole-pieces were adjusted at a distance 

 of 29 millim. from each other, and maintained at this distance 

 apart by means of a block of wood inserted between them. 

 The mirror was placed sideways upon the magnet-poles, so as 

 to touch each through a distance of 3 millim. The magnet 

 was thus closed through the steel mirror. The reflecting 

 surface was placed Vertical. The angle of incidence J is 

 the angle between the incident rays and the normal to the 

 mirror. 



The following Table gives the results obtained. The sign 

 = denotes that the plane of polarization (the long diagonal of 

 the nicol) is parallel, and the sign J_ that it is at right angles 

 to the plane of incidence. 



The amount of rotation is given in minutes. Since each 

 number is the result of eighty separate adjustments, the tenths 

 of minutes are given, although the error may amount to several 

 tenths of a minute. 



J. = -J- 



19°-0 -4-8 +2-7 



29-9 -4-5 +7-3 



39-5 -6'6 +7-7 



50-1 -7-7 +6-9 



61-3 -8-0 +7-5 



65-0 -9-4 +8-7 



70-0 -7-1 +8-1 



75-0 -6-0 +6-8 



80-3 - +2-6 



82-0 -4-3 -2-3 



85-2 -3-9 -1-9 



These experiments completely confirm Kerr's results — that 



