Mirrors seem for a long time to have been made of a mixture of copper and tin, as is expressly said by Pliny[153], who adds, that the best were constructed at Brundisium. This mixture, which was known to Aristotle, produces a white metal, which, on account of its colour, may have been extremely proper for the purpose, and even at present the same mixture, according to the careful experiments made by Mr. Mudge, an Englishman[154], produces the best metal for specula. It appears that the ancients had not determined the proportion very accurately; for Pliny assures us twice that in his time mirrors of silver were preferred. It is indeed not easy to ascertain the quantity of each metal that ought to be taken, and the most advantageous degree of heat; upon which a great deal depends. One of the principal difficulties is to cast the metal without blisters or air-holes, and without causing any part of the tin to oxidize, which occasions knots and cracks, and prevents it from receiving a fine polish. A passage of Lucian[155], which no one as yet has been able to clear up, alludes certainly, in my opinion, to these faults. A mixture of copper and tin is so brittle, that it is very liable to crack; and a mirror formed of it, if not preserved with great care, soon becomes so dim, that it cannot be used till it has been previously cleaned and polished. For this reason a sponge with pounded pumice-stone was generally suspended, from the ancient mirrors, and they were kept likewise in a case or box, as may be seen by the greater part of those still extant. Mirrors of silver were less subject to this inconvenience, and I am inclined to think that the latter on this account made the former be disused, as we are informed by Pliny. As ancient mirrors of metal are still to be found in collections of antiquities, it might be of some importance to the arts if chemical experiments were made on their composition. Those who have hitherto given us any account of them have contented themselves with describing their external figure and shape. Count Caylus[156] is the only person, as far as I know, who caused any chemical experiments to be undertaken on this subject. They were made on a mirror found near Naples, by M. Roux, who asserts that the composition was a mixture of copper and regulus of antimony, with a little lead. Antimony however was not known to the ancients. If that metal was really a component part, the mirror must have been the work of more modern times, or it must be allowed that the artist had metal combined with antimony without knowing it; but the latter is not probable. The experiments made by Roux do not seem to me to have proved in a satisfactory manner the presence of regulus of antimony; moreover, no certain information can be derived from them, for the antiquity of the mirror was not ascertained; nor was it known whether it ought to be reckoned amongst the best or the worst of the period when it was made.