167 
counting for these phenomena, less liable to objection 
than any that has been hitherto proposed. It is re- 
marked by Dr Bancroft, that, though the prism and 
other transparent colourless substances shew us the 
different colours of the several rays of light, by se- 
parating them from each other, in consequence of 
their greater or less refrangibility ; yet he is per- 
suaded that the permanent colours of different bodies, 
or substances, are not produced by mere refraction ; 
but depend on other properties, which determine or 
occasion the reflection or transmission of some par- 
ticular sort or sorts of rays, and an absorption or 
disappearance of the rest : and these properties he 
conceives to be certain affinities or elective attrac- 
tions, existing in, or between the differently colour- 
ed matters, and the particular sorts or rays of light 
so absorbed or made latent *. 
431. That an affinity, or attraction, is exerted be- 
tween light and the particles of bodies, may be justly 
inferred from the great refractive power of inflamma- 
ble bodies, which, all other things being equal, must 
be supposed to attract light more powerfully than other 
substances ; and it is variation in point of strength, 
says Dr Thomson, which constitutes the characteristic 
mark of chemical affinity f. The phenomena of 
phosphorescence, and many other chemical facts, af- 
ford evidence of the same nature. Thus, many me- 
tallic oxides, as Scheele first observed, are soon 
restored to their metallic form, by the action of 
* On Perm. Colours, p. 29. 
t Syst. Chem. vol. i. p. 246. 1st edit. 
