VOL. LV.] PHILOSOPHICAL TRANSACTIONS. 179 



the red at the greatest thickness, and the intermediate colours at intermediate 

 thicknesses. 3. The same appears from the table (p. 206) in which the thick- 

 nesses of air, water, and glass, and the colours produced by them, are set down. 

 These experiments are applied by him to transparent bodies, and the colours 

 exhibited by them ; but they are equally applicable to permanently coloured 

 bodies : and it appears from them, that denser substances ought, by their 

 greater reflective power, in like circumstances, to reflect the less refrangible 

 rays, and that substances of less density should reflect rays proportionably more 

 refrangible, and thus appear of several colours in the order of their density. 



In confirmation of this reasoning, I shall give instances of natural bodies, 

 which differ from each other in density, though circumstanced alike in other 

 respects ; and shall show that they differ in colour in the same order as they do 

 in density, the densest being red, the next in density orange, yellow, &c. In 

 such an inquiry, metallic bodies seem to deserve our first and principal attention, 

 as their specific gravities have been ascertained by well known and repeated ex- 

 periments. Without entering into a minute chemical theory of the principles of 

 metals, it is sufficient to observe that they are universally allowed to consist of 

 I . An inflammable or sulphureous matter, which is of the same kind in all the 

 metals ; 2. Of a fixed matter or calx, which appears in each of the metals to be 

 specifically different in weight, as well as in other properties. 



As the sulphureous matter, in the entire metals, acts strongly on the rays of 

 light, it is necessary to calcine, or to divide them into extremely minute parti- 

 cles, in order to examine separately the action of the calx, or fixed matter, on 

 the rays of light. In order to examine all the metals in like circumstances, by 

 reducing them into the smallest particles, and depriving them of their sulphur as 

 far as was practicable, I exposed each of them, united with a proper quantity of the 

 purest glass, without any additional ingredient, to the 

 greatest degree of fire they are capable of bearing, without 

 having all colour whatever destroyed. In this state it appears, 

 from a variety of experiments and facts, that they actually 

 do, without any exception, exhibit colours in the order of 

 their densities, as annexed. 



GOLD, 

 Which is the densest of all the metals, imparts a red colour to glass, whenever 

 it is divided into particles so minute, that it can be intimately mixed with the 

 ingredients of which the glass is made; and it seems indifl^erent in what manner 

 it is reduced to this state. Thus 



1 . From the powder obtained by rubbing gold with a pumice stone, used by 

 the goldsmiths in polishing it, mixed with nitre, borax, and potash, a beautiful 



A a2 



