Magneto-optic Phenomenon. 



313 



Equatorial Magnetization per nnit volume = 1400 C.G.S. 



units. 



Mirror T. J = 77° 23'-5 ; 

 „ II. J=76°30'-5; 



H = 26°34'. v ,, r , , 

 H = 26°44'. Yellowll g nt - 



Mirror. 



Angle of 

 Incidence. 



Amplitude of 



the Magnetic 



Component of 



Light. 



1 Obs. Am pi. 

 Calc. Ampl. 



x 10 ~ 3 



Phase of the Mag- 

 netic Component 

 of Light. 



Observed 



—Calculated 



Phase. 







Calc. 

 XA. 



Obs. 

 X10- 3 - 



Calc. 



Obs. 

















180° 



180° 



II. 



86 



0-226 



0-284 



1-26 



-54 165 



+29 26 



+83 42-5 



II. 



82 30 



0-358 



0530 



1-48 



-62 22 



4-24 22 



+86 44 



II. 



76 30-5 



0-493 



0-715 



1-45 



-69 51-5 



+ 14 49 



+84 40-5 



II. 



71 52 



0-548 



0-815 



1-48 



-73 445 



+ 10 3 



+83 47-5 



I. 



61 30 



0-598 



0-820 



1-37 



-81 36-5 



+ 1 49 



+83 35-5 



II. 



51 22 



0-545 



0-760 



1-39 



-85 55 



- 1 



+86 55 



II. 



36 10 



0-426 



0-630 



1-48 



-90 15 



- 5 51 



+84 49 



II. 



24 16-5 



0-266 



0-430 



1-62 



-91 56 







II. , 



12 



0-152 



0'260 



1-69 



-93 2-5 







II. 



6 



0074 



0-125 



1-69 



-93 15 







Since in the factor A, Hall's constant h for iron is positive, 

 and since this depends upon the choice of the system of 

 coordinates, and the positive directions of the magnetic force 

 N in these observations and in the theory of magnetic re- 

 flexion are opposite to each other, the signs of the calculated 

 and observed amplitudes do not agree with each other, be- 

 cause they are both referred to positive magnetization. The 

 sign of the amplitude of a wave-motion can, however, be 

 changed by making the phase 180° large or smaller. In the 

 rest of the paper the phases have been so changed in order to 

 give the amplitudes the proper signs. The differences given 

 in the last column will then be 180° greater or smaller. 



The observations showed that yLt. = /x^; m^=m v This rela- 

 tion follows also from theory. As we can only write down 

 proportional values for the theoretical amplitudes, the observed 

 and calculated amplitudes agree with each other within the 

 limits of experimental error. But there is a constant differ- 

 ence between the observed and calculated phases. The devia- 

 tions from the mean value lie within the limits of error. 



