400 On the Experimental Relations [1857. 



untouched. The Arts do not seem to furnish any process 

 which can instruct us as to this condition, for all the operations 

 of polishing, burnishing, &c. applied to gold, silver and other 

 metals, are just as much fitted to produce the required state 

 under one view as under the other. 



To return to gold : it is clear that that metal, reduced to 

 small dimensions by mere mechanical means, can appear of 

 two colours by transmitted light, whatever the cause of the 

 difference may be. The occurrence of these two states may pre- 

 pare one's mind for the other differences with respect to colour, 

 and the action of the metallic particles on light, which have yet 

 to be described. 



Many leaves of gold, when examined by a lens and transmitted 

 light, present the appearance of red parts ; these parts are 

 small, and often in curved lines, as if a fine hair had been there 

 during the beating. At first I thought the gold was absolutely 

 red in these parts, but am inclined to believe that in the greatest 

 number of cases the tint is subjective, being the result of the 

 contrast between the white light transmitted through bruised 

 parts, and the green light of the neighbouring continuous 

 parts. Nevertheless, some of these places, when seen in the 

 microscope, appeared to have a red colour of their own, that is, 

 to transmit a true red light. As I believe that gold in a certain 

 state of division can transmit a ruby light, I am not prepared 

 to say that gold-leaf may not, in some cases, where the effect 

 of pressure in a particular direction has been removed, do the 

 same. 



Many of the prepared films of gold were so thin as to have 

 their reflective power considerably reduced, and that in parts 

 which, under the microscope and in other ways, appeared to be 

 quite continuous : this agrees with the transmission of all the 

 rays already mentioned, but it seems to imply that a certain 

 thickness is necessary for full reflexion ; therefore, that more 

 than one particle in depth is concerned in the act, and that the 

 division of gold into separate particles by processes to be de- 

 scribed, may bring them within or under the degree necessary 

 for ordinary reflexion. 



As particles of pure gold will be found hereafter to adhere 

 by contact, so the process of beating may be considered as 

 one which tends to weld gold together in all directions, and 



