MAGNETISM AND LIGHT 323 



where C is a constant for the substance for the particular wave 

 length used. It is known as Verdet's constant, though it might 

 more appropriately be named after Faraday. If V", V 6 , be the 

 magnetic potentials at the beginning and end of the path con- 

 sidered we may put R in minutes = C (V 6 V a ), a concise form of 

 expression, though the integral form is the one practically useful. 

 Verdet worked before the value of absolute units was realised, 

 and he was content to express the constant of a substance in terms 

 of the constant for water. Taking the rotation in water of light 

 at the E line in the spectrum as 1, he found the rotations of the 

 light of the various lines as follows. Underneath are the rotation, 

 which would have been observed if they had been inversely as the 

 squares of the wave length. 



WATER 



Lines in spectrum C D E F G 



Observed rotations 0'63 0'79 I'OO 1'213 1'605 

 Calculated from 1/X 2 0'64 0'80 I'OO 1'08 1'50 



For carbon bisulphide, creosote, and various other liquids the 

 dispersion is about as near to that calculated from l/\ 2 as in the 

 case of water, departing as notably from it as the wave length 

 diminishes. 



If the substance examined absorbs the light of one particular 

 wave length there are peculiarities in the rotation, which were first 

 predicted from theory, and which will be more appropriately 

 described when we discuss the theory now accepted. 



Effect of rise of temperature. Bichat was the first to 

 make a careful examination of the effect of change of temperature 

 of the Verdet constant, and he found that in general the constant 

 lessens as the temperature rises, though his numerical values have 

 not been confirmed by subsequent workers. 



Absolute values of Verdet 's constant. The first 

 determination of the absolute value of a Verdet constant was 

 made by Gordon* for sodium light in carbon bisulphide. 



A few years later another determination of the same constant 

 was made by Lord Rayleigh. f He used a coil with a known 

 number of turns surrounding a tube of known length containing 

 the bisulphide. The current was measured by a potentiometer 

 method, assuming the E.M.F. of a Clark cell. If the tube had ex- 

 tended a great distance beyond the coil at each end the difference 

 of magnetic potential between the ends would have been 4x X 

 number of turns of coil X current. But as the ends were at points 

 where the field was still appreciable a correction was needed. 



To measure the rotation Lord Rayleigh used a special " half- 



* Gordon, El. and Mag,, vol. ii. chap, xlviii. 



f Phil. Traitx. 176, p. 343 (1885) ; or Scientific Papers, vol. ii. p. 360. 



