ON COLOURS IN METAL GLASSES, ETC. 259 



requires that practically the whole of the silver must have been in suspension in the 

 form of small spheres.* 



Once more, CAREY LEAt prepared suspensions of silver in water by precipitating 

 the silver from the nitrate by means of a mixture of ferrous sulphate and sodic citrate. 

 He describes how, after careful washing, the silver frequently " dissolved," forming a 

 liquid which varied from red to yellow^ and was generally blood red ; he adds : 



" On one occasion the substance was obtained in a crystalline form. Some crude red solution had been 

 set aside in a corked vial. Some weeks after the solution had become decoloured with crystalline deposit 

 on the bottom, The bottle was carefully broken ; the deposit, examined by a lens, consisted of short 

 black needles and thin prisms." 



If, then, the diffused particles of silver when aggregated and precipitated had 

 become crystalline, they must before have been in the form of nascent crystals, and 

 for gold and for all the substances examined by VOGELSANG, such nascent crystals 

 were spherical. 



7. Blue Reflection from the Stained Face of Silver Glass. 



When clear glass is flashed with silver glass, or when a clear glass is so stained on 

 one face with silver that the volume proportion /A of silver does not gradually diminish 

 to zero as we proceed inwards from the stained face, but that the stained region ends 

 abruptly, a blue reflection from the interface can be observed if the glass is held with 

 the stained face away from the eye. No blue reflection can be seen from the 

 air-glass interface when the stain is held towards the eye. STOKES observed this 

 blue reflection, and stated that the interface presented the appearance of being coated 

 with a fine blue powder. || 



We proceed to examine whether the presence of small spheres of silver, which has 

 been shown to account for the colour of the light transmitted by silver glass, will also 

 account for this blue reflection. Consider, then, plane polarised light travelling in a 

 medium of refractive index v' and directly incident on the surface, z = 0, of an 

 absorbing medium whose optical constants are n' and K', where N' = n' (I IK'). Then 

 we may take as the electric and magnetic vectors for 



* BARUS and SCHNEIDER (foe. cit., p. 297) make the following comment on their experiment: 



"KuNDT has found for normal metallic silver a refractive index of about 0'27. It would, therefore, be 

 expected that the presence of the silver would diminish the refractive index of the water. It is by no 

 means denied that it might be possible to explain the normal refractive indices of the above table in 

 accordance with MAXWELL'S Theory of Light." 



The investigation in the text attempts to give such an explanation. 



t CAREY LEA, ' Amer. Journal of Science,' 1889, and 'Phil. Mag.,' 1891. 



I Cf. above, p. 243, especially second footnote. 



Cf. above, 1. 



|| STOKES, 'Collected Works,' vol. III., p. 316. 



2 L 2 



