260 On the Magnetic Rotation of Polarized Light. 



Creosote. 



C. D. E. F. G. 



Observed rotation . 0573 0*758 1*000 1*241 1*723 



Calculated, formula I. 0*617 0*780 1*000 1*210 1*603 



„ „ II. 0*627 0*789 1*000 1*200 1*565 



„ „ III. 0*976 0*993 1*000 1*017 1*041 



To examine the direction of the action, we must see what the 

 relative direction of the currents and magnetism are in the 

 equations, as I have not taken the signs with respect to any 

 system. 



Let the positive direction of the current be the direction in 

 which the positive electricity moves, and the positive direc- 

 tion of the magnetic lines of force be the direction in which 

 the north pole tends to move ; then we easily find that our 

 equations are on the right-handed screw system, the right- 

 handed screw being such that, if we turn it in the direction of 

 the hands of a watch with its face towards us, it will move 

 away from us. According to this system, Mr. Hall has found 

 that the value of c is positive for gold and some other dia- 

 magnetic substances, and negative for iron. Hence a mag- 

 netic force in the positive direction will cause the ray to be 

 rotated in the positive direction in diamagnetic substances, 

 and in the negative direction in magnetic ones, which is 

 exactly what has been observed. 



To compare the numerical amount of the revolution with 

 observation, we can take the constants as observed by Mr. 

 Hall for gold, and thus find at least whether it is of the proper 

 order of magnitude. 



From more recent observations than those published, Mr. 

 Hall finds that in the field of his magnet he can cause the 

 lateral electromotive force to be at least as great as 20V0 °^ 

 the force along the strip. According to the system of units 

 used in this paper, the new electromotive force will be (in the 

 case of conduction, the current passing along Y and the mag- 

 netism being in the direction of z) 



A // =-cc i y=-47r / a / cc 1 B / ; 

 but Mr. Hall finds 



^ = -2000, nearly. 



A" ? 



Hence, using the C.G.S. system, in which ^ = 2000 nearly, 

 we shall have 



CCi= — , nearlv, for gold. 



47T 



