Double Refraction of Gold Colloids. 365 



or if tc u k 2 , and k 3 are the principal susceptibilities of the 

 substance and k the susceptibility of the surrounding- 

 medium, H the field in the body, components along the 

 principal axes of crystal H l5 H 2 , and H 3 : 



w= — ij[(«i — ^H^oi-f- (fc 2 — a: )H 2 H 02 



4-(fc 3 —/e )H. 3 K 03 ]dv. 



For the position of equilibrium of the crystal we have 

 8w = 0. The equilibrium is stable when iv is a minimum. 

 In the case of weakly magnetizable substances we may be 

 able approximately to substitute for the intensity in the 

 crystal the intensity of the external field H : 



Suppose the crystal left free to turn about its centre of 

 gravity. 



The expression is a minimum when the external field 

 coincides with the direction of the largest of the quantities 

 Ki — k . If /q — /e <0, the crystal is oriented in such a 

 position that the direction corresponding to the absolutely 

 less of the expressions k 1 — k coincides with the lines of 

 force (H ). 



Let us compare these facts with the result recorded above. 

 T have shown that the ellipsoidal particles of a colloidal gold 

 solution have their major axes situated perpendicularly to 

 the lines of force. 



Is it possible to account for the behaviour of the particles 

 by assuming an isotropic microcrystalline substance ? 

 Obviously this will hardly do. Isotropic elongated ellipsoids 

 of rotation are placed longitudinally, independently of the 

 sign of the susceptibility. 



The particles of the gold sols exhibit a tendency for their 

 major axis to be situated perpendicularly to the lines of 

 force ; hence they must consist of a magnetically seolotropie 

 medium. 



The susceptibility of ordinary gold has been determined by 

 J. Konigsberger *, who found k— — 3*0 X 10~ G . Assuming 

 also the particles to be diamagnetic, the smallest axis, or one 

 of the smallest, will thus correspond to the absolute minimum 

 of the expressions k x — k$. 



The crystalline form of ordinary gold has been determined 

 by many investigators f . Either the melted metal is allowed 

 to solidify, or a solution is reduced by ferrous sulphate — for 



* Wied. Ann. lxvi. p. 698 (1898). 



f Cp. e. g. G. Pose, Pogg. Ann. lxxiii. p. 8 (1848). 



