174. With a view to know what colours were, and what were not durable and fading, he exposed to the sun and air, for twelve days, sonic patterns of all co lours, which had been dyed in' his own house with known compositions. It appears that this time was ufficient for a trial of them ; for the durable colours are not at all, or but little injured, while the fading ones are almost entirely obliterated; so that, after twelve days exposure to the heat of the sun, and the dampness of the night air in summer, there can be no doubt as to the class in which each colour should be placed.
175. But there still remains a difficulty, which was this, that as all these colours had not been exposed to the air precisely at the same time and in the same sea son, some might have had more sun, and, consequent ly, might have suffered a greater change in the twelve days, than others exposed in more cloudy weather, or when the days were shorter. But he discovered a re medy for this inconvenience, which removed all diffi culty and doubt as to the accuracy of the experiments; he chose one of the worst colours, that is, one on which the sun had produced the greatest effect in twelve days. This colour served as a standard in the experiments; and whenever he exposed patterns to the air, he always exposed a piece of the stuff along with them ; so that he did not calculate by the number of days, but by the co lour of his standard, always keeping the pattern expo sed, till it had lost as much as that would have done by twelve days exposure in summer. As he always noted the day on which he exposed his patterns, he was led to observe, that in winter it was only necessary to leave them four or five days longer exposed to the air than in summer. By pursuing this method, lie removed all doubt as to the accuracy of his experiments. In this trial, by exposure to the air and to the rays of the sun, he had a still farther object, which was, to find the proper proof for each colour. What is called proof, is the trial of a stuff, with a view to discover whether its dye be permanent or not: a pattern is boiled with alum, tar tar, soap, vinegar, lemon juice, Ste. and its quality is judged of by the effect these substances produce. The proofs used in 1733 were so insufficient, that they did not even assist Mr Dufay in discovering more certain ones; they even destroyed some good colours, while they produced very little effect upon the bad ; so that he was obliged to fix upon several, each of which serves for a great number of colours. The following is a short description of the method he pursued in order to dis cover them.
176. After having observed the effect of air on each colour, whether good or bad, he tried the same stuff with different proofs, and stopped when he found one which produced the same effect with the air ; then no ting the weight of the ingredients, the quantity of the water, and the length of the trial, he was sure to be able to produce on a colour an effect equal to that which the air would have produced, supposing it had been dyed in the same way with his; that is, accord ing to the methods employed by the dyers of true and false colours. Having thus examined all the colours,
and all the ingredients employed in dyeing, he found out a method, which may be considered as accurate, of distinguishing the good or bad qualities of each colour, by making by means of the proof, a sort or analysis of the materials which composed it. We cannot, without injustice, refuse to acknowledge, that the means Mr Dufay employed, in the discovery of these proofs, or tests for colour, are ingenious; for the trial by air and sun cannot be made on those, where it is necessary to judge immediately, whether a stuff exposed to sale, at a fair or elsewhere, as if a true dye, be really so or not.
177. The proofs mentioned in the new regulations, made in consequence of Mr Dufay's memoirs, dischar ges,m a few minutes as much of a colour, where it is not dyed of a durable colour, as would be lost by twelve or fifteen days exposure to the air. But, as ge neral rules for such trials must be liable to many ex ceptions which neither cannot be foreseen, or, though foreseen, cannot be particularized, without the risk of causing confusion, or of affording matter for innumera ble disputes, it follows, that these rules, considered per haps as too general, are also too severe in many cases wherein light colours require salts, or quantities of salts, which shall be less active than those necessary for deep or loaded colours, which may allow a considerable portion of their colouring matter to be carried off by a proof, without sheaving a very visible alteration. It would, therefore, have been very necessary to prescribe a proof for almost every shade, which, from their infi nite variety, would be impossible. Thus the air and the sun will always be the true test ; and every colour which is not changed by them in a certain length of time, or which, by exposure, acquires what the dyers call a body, ought to he considered as a standing colour, even though it should be considerably changed by the proofs. Of this we have an example in scarlet ; as soap almost entirely discharges this colour, it has been sub mitted to the trial by alum ; and when it has been dyed with cochineal alone, without any mixture of other in gredients, it ought to assume a purple colour in a boil ing solution of alum. If scarlet be exposed to the sun, it loses some of its brightness, and becomes deeper ; but this deep shade is not similar to that imparted by alum. Proofs, then, in certain cases, cannot be substi tuted for the action of the air and sun, at least so far as respects similarity of effect.