The power of resisting vegetable acids, alkalies, and soap, and, above all, the action of air and light, constitutes the durability of a colour. But this property has a very unequal standard, according to the nature of the colour and the species of the stuff. There is no obscurity in the action of water, alkalies, acids, and soap: for a solution is effected by means of these agents, or a small portion of acid or alkali unites to the combination, which forms the colour. But this is not the case with the action of light and air. Till lately, however, it was not known in what this action consisted. Of the two principles which compose atmospheric air, it is only the _oxygen_ gas which acts on the colouring particles. It combines with them, and thus impairs their colour or makes them fade. But its action is soon chiefly exerted on the hydrogen which enters into their composition, and it thereby forms water. This effect may be compared to a feeble combustion. Hence the carbon, which enters into the composition of the colouring particles, becomes predominant, and the colour usually passes to yellow, dun, or brown, or other appearances. Light promotes this decomposition of the colouring particles, which frequently takes place only with its concurrence, and thus it contributes to the destruction of the colour. Heat also favours the same result, but less efficaciously so, unless it have a certain intensity. It is concluded, therefore, that colours are more or less fixed in the air, according to the greater or less tendency which the colouring particles have to undergo this change[6]. Hence the utility of _mordants_ in rendering _fugitive_ colours _fast_. _To prove the colours of Dyed Stuffs, &c._ The natural proofs of a dye's being effectual, are exposure to the air, to the sun, or to rain. If the colour be not changed by such exposure after twelve or fourteen days, it may be considered as fixed. These proofs are not, however, adapted to every colour: for some resist the action of air, light, and rain, yet are nevertheless injured by certain acids. There are also colours which do not resist the natural proofs and yet remain unchanged by acids. Colours may be arranged in this respect in three classes: the first class is tried with _alum_, the second with _soap_, the third with _tartar_. For the proof with _alum_, half an ounce of this salt must be dissolved in a pint of water in an earthen pipkin, and into this liquor is to be put half a quarter of an ounce of the dyed thread or stuff, the whole being boiled about five minutes; it is then to be washed clean with water. Thus are tried _crimson_, _scarlet_, _flesh-colour_, _violet_, _ponceau_, _peach-blossom_, different shades of _blue_, and other colours bordering on these.