Formerly when much that is regular and constant in nature was considered as mere aberration and accident, the colours arising from refraction were but little attended to, and were looked upon as an appearance attributable to particular local circumstances. 286. But after it had been assumed that this appearance of colour accompanies refraction at all times, it was natural that it should be considered as intimately and exclusively connected with that phenomenon; the belief obtaining that the measure of the coloured appearance was in proportion to the measure of the refraction, and that they must advance _pari passu_ with each other. 287. If, again, philosophers ascribed the phenomenon of a stronger or weaker refraction, not indeed wholly, but in some degree, to the different density of the medium, (as purer atmospheric air, air charged with vapours, water, glass, according to their increasing density, increase the so-called refraction, or displacement of the object;) so they could hardly doubt that the appearance of colour must increase in the same proportion; and hence took it for granted, in combining different mediums which were to counteract refraction, that as long as refraction existed, the appearance of colour must take place, and that as soon as the colour disappeared, the refraction also must cease. 288. Afterwards it was, however, discovered that this relation which was assumed to correspond, was, in fact, dissimilar; that two mediums can refract an object with equal power, and yet produce very dissimilar coloured borders. 289. It was found that, in addition to the physical principle to which refraction was ascribed, a chemical one was also to be taken into the account. We propose to pursue this subject hereafter, in the chemical division of our inquiry, and we shall have to describe the particulars of this important discovery in our history of the doctrine of colours. What follows may suffice for the present. 290. In mediums of similar or nearly similar refracting power, we find the remarkable circumstance that a greater and lesser appearance of colour can be produced by a chemical treatment; the greater effect is owing, namely, to acids, the lesser to alkalis. If metallic oxydes are introduced into a common mass of glass, the coloured appearance through such glasses becomes greatly increased without any perceptible change of refracting power. That the lesser effect, again, is produced by alkalis, may be easily supposed. 291. Those kinds of glass which were first employed after the discovery, are called flint and crown glass; the first produces the stronger, the second the fainter appearance of colour. 292. We shall make use of both these denominations as technical terms in our present statement, and assume that the refractive power of both is the same, but that flint-glass produces the coloured appearance more strongly by one-third than the crown-glass. The diagram (Plate 3, fig. 2,) may serve in illustration. 293.