Double Refraction of Gold Colloids. 361 



Law. Consequently the spherical form presumed by this law 

 might be proved by those investigations. Westgren *, for 

 instance, writes concerning these solutions : " Sie besitzen 

 nicht nur den Vorteil, spharische Teilchen zu haben. . . .V 



As regards the structure of the particles we have been 

 confined to assumptions. If the gold particles have a form 

 at least approximately spherical, one must attribute to them 

 a micro-crystalline structure. Further, one expects the 

 microcrystals to be constituted by a space lattice belonging 

 to the regular system as in the case of ordinary gold \. 

 One can imagine the microcrystals in a particle to be rather 

 small, but on the other hand all transitional states between 

 this structure and particles consisting of a single regular 

 crystal are possible. It is probable that these alternatives 

 do not differ perceptibly with regard to light absorption and 

 hydrodynamical action. Starting from the current con- 

 ception of the particles of the colloidal gold solutions which 

 I have outlined above, it would be necessary to use 

 Havelock's % theory, assuming an ieolotropic distribution of 

 the particles in order to account for the magnetic double 

 refraction. The particles would arrange themselves in a sort 

 of space lattice of rhombic symmetry. Havelock himself has 

 not given any reason to show why the packing o£ these rather 

 scattered obstacles is disturbed from the "cubical arrange- 

 ment of the natural state. " One might possibly imagine an 

 influence of the Brownian movement. Supposing the 

 particles charged, the magnetic field will exercise a deviating 

 action on the particles, whose vector of velocity has a com- 

 ponent perpendicular to the lines of force. On the contrary, 

 the field does not influence the particles that have a longi- 

 tudinal movement of translation. Certainly the packing is 

 disturbed to some extent by the field, but the symmetry of 

 the distribution will hardly be changed. Further, the 

 hypothesis that the particles are changed might be rather 



* Arkiv fur kemi, mineralogioch geologi utg. av K. Svenska Vetenskaps- 

 <ikademien y B. vii. No. 6, p. 8. 



t In a preliminary note (Nachricht K. Ges. Wiss. Gottingen, 1918, 

 p. 98), P. Scherrer mentions some experiments that support the latter 

 suggestion. He has not, however, given any data about the mode of 

 preparation of the solutions. Debye-Scherrer's method permits also of 

 a determination of the size of the microcrystals — of the crystalline 

 degree of dispersion — by measuring the " Halbwertsbreite " of the density 

 belonging to a maximum. Indeed, as the photographic curve o( 

 density for X-ray light is not yet accurately investigated, the results 

 are open to discussion and we must wait for further communications. 



t Proc. Roy. Soc. vol. lxxvii. p. 170 (190G), vol. lxxx. p. 28 (1908). 



