ON COLOURS IN METAL GLASSES, ETC. 257 



calculated suggests that the colouring agent of the yellow silver glass consists primarily 

 of diffused spheres of silver. Since discrete silver molecules would produce an 

 absorption maximum at X = '360, not more than a comparatively small amount of 

 silver can be present in the molecularly subdivided condition. The conclusion that 

 silver glass owes its colour to diffused spheres of silver will be verified in the following 

 section. 



The absorption spectra of some colloidal solutions of silver, prepared by BREDIG'S 

 method,* have been measured by EHRENHAFT. The continuous curve shown in fig. 3, 

 representing the calculated absorptions of a diffusion of silver spheres in water, is of 

 the same form as that which, according to EHRENHAFT'S measurements, represents 

 the absorption of visible light by a colloidal solution of silver.! Using ultra-violet 

 light, he further found that a brown colloidal solution of silver, examined before 

 coagulation had seriously affected its colour, showed an absorption band which began 

 a,t X = '503 and attained a maximum at X = '380, while the fluid was again quite 

 transparent at X = '335. Except for the fact that the maximum ordinate of the 

 calculated curve for silver spheres in water is at X = '389 instead of at X = '380, the 

 above observations admirably describe the continuous curve shown in fig. 3. Since 

 the dotted curve given in that figure shows a maximum at X = '360, and the absorp- 

 tion band does not begin until X = '450, about, the colour of the " colloidal " solution 

 is not that which would be exhibited by a suspension of discrete silver molecules, i.e., 

 by a true solution. We conclude, therefore, that the silver in a " colloidal " solution 

 is present in the form of small spheres ; discrete molecules may, however, also be 

 present, and, as indicated above in the case of gold, prepared by BREDIG'S method, 

 probably also crystallites, the number and size of which will increase with the age of 

 the solution. 



That the silver in a colloidal solution is in the form of small spheres is further 

 shown by an experiment of BARUS and SCHNEIDER j who measured the refractive 

 index of such a fluid. Their results are given in the following table, in which n 

 represents the measured refractive index : 



the spectrum formed by the light which has traversed the glass is not quite pure, so that that image of 

 the slit which should be illuminated only by light of wave-length, say, A = -433, is also, owing to 

 reflections from dust particles, &c., illuminated by light of other wave-lengths which has experienced a 

 less absorption. 



* BREDIG, ' Zeitschr. f. Electrochemie,' 4, pp. 514, 547. 



t EHRENHAFT, loc. cit., p. 506. 



J BARUS and SCHNEIDER on "The Nature of Colloidal Solutions," 'Zeitschr. f. Phys. Chem.,' VIII., 

 p. 278. 



Tabelle 5, loc. cit., p. 296. 



VOL. CCV. A. 2 L 



