Double Refraction of Gold Colloids. 359 



finally attaining a constant value, the double refraction 

 characteristic of the sol. 



The Amplitude effect. — Besides the double refraction all 

 the sols show dichroism under the influence of the magnetic 

 field, the absorption of the component of the vibration that 

 is parallel to the lines of force not being the same as the 

 absorption of the component at right angles to the former. 



The amplitude effect is rather great and increases with the 

 intensity. In fig. 9 (PI. XIII.) (Table IX.) the expression 



( — — 1 J 1000 as ordinate is plotted to the intensity H (Gauss) 



as abscissae, where a u and a ± are the amplitude of the 

 electric component parallel or perpendicular to the lines of 

 force. The concentration of the sol u as c = 2'5 x 10 ~ 4 normal 

 and the degree of dispersity 2r = 48 /jl/ju. 



The change of the amplitude effect with the wave-length 

 is clear from fig. 10 (PL XIII.) (Table X.). The curve A 

 corresponds to a sol 2r — ±9 fifi (cone. c = 2*5 X 10~ 4 normal), 

 B to another 2r=140/zyu, (cone. c = 20 x 10" 4 normal). The 



wave-lengths (/ul/jl) are abscissae and the expression { — — 1 ) 



x 1000 ordinates. In the case of the high disperse sol the 

 effect has a minimum about \= 500 yu^. The low disperse 

 sols show effects of opposite signs, the component parallel 

 to the lines of force being most absorbed. 'A range in red 

 forms an exception, the effect changing sign with increasing 

 wave-length. The amplitude effect is a rather good test of 

 homogeneity of low disperse sols ; sols that are not quite 

 homogeneous exhibit curves of a different type. 



The time of relaxation. — One of the reasons why the 

 Majorana effect was at first considered " magneto-optical in 

 the direct sense " *, is the extreme rapidity with which the 

 double refraction appears when the field is established. 



Investigating gold sols of varying degree of dispersity, 

 one finds that the time of relaxation increases with the size 

 of the particles; the time cannot be measured for the high 

 disperse sols ; in the case of the low disperse ones it amounts 

 to seconds. Hence a theory assuming a change of frequency 

 of the electrons is out of the question. 



The form and structure of the particles. — Before discussing 

 the nature of the effect I will o-ive a short account o( the 



* This expression is due to Voigt, who added terms of the second 

 order to the ordinary electrical equations in order to account for the 

 action of the magnetic field. Cp. Accademia dci Lincei, JRendieontt\ 



vol. xi. i. p. 505 (1902). 



