438 On the Experimental Relations [1857. 



its result with that of the light which had passed through 

 the gold. 



The gold-leaf plates, deprived of green colour by heating in 

 oil, were found with the glass in such good annealed condition, 

 as not to affect the ray ; but when they were moved, until the 

 oblique colourless gold came into the course of the ray, it was 

 depolarized ; a red image appeared ; direct rotation of the 

 analyser reduced this a little in intensity and then changed the 

 colour to blue. The reduction was not much, and both in 

 that and the first appearance of the red image, there is a 

 difference between the heated and the unheated gold: pro- 

 bably the green tint of the latter, which would tend to extin- 

 guish the red and produce a minimum, may be sufficient to 

 account for the effect. Gold which had been re-greened by 

 agate pressure acted in like manner on the polarized ray, but 

 the experiments were imperfect. 



A glass plate having gold-leaf on one part of it, had a 

 second glass plate put over it and gummed at the edges. In 

 the sulphide of carbon, therefore, it represented in one part a 

 plate of air, and in another a compound plate of air and gold ; 

 both acted in the same direction, but the air and gold much 

 more than the air. Gold on glass in this medium, or gold in 

 air, or glass in air, all gave results in the same direction, i. e. 

 required direct rotation of the analyser to compensate for 

 them. 



I proceeded to examine the other forms of gold ; and first, the 

 deposits on glass obtained by electric deflagration. These af- 

 fected the ray of polarized light exactly in the manner of gold- 

 leaf, and that even at the distant parts of the deposit. It was 

 most striking to contrast the thinnest and faintest portion of 

 such a film with the neighbouring parts of the glass from 

 which it had been wiped off. It must be remembered that 

 such a preparation is a layer of separate particles ; that these 

 particles are not like those of starch or of crystals, for they 

 have no action whilst in a plane perpendicular to the polarized 

 ray ; nor have they a better action for being in a thick layer, 

 as in the central parts of the deposit. The particles seem to 

 form the equivalent of a continuous plate of transparent sub- 

 stance ; and as in such a plate it is the two surfaces which act, 

 so there appears to be the equivalent of these two surfaces 



