[196] Nicephori Schol. in Synesium, in Synesii Op. Par. 1612, fol. p. 419.
[197] In the collection of antiquities at St. Denis, an ancient mirror was shown, which was said to have belonged to Virgil. It was oval, and before Mabillon let it fall, was fourteen inches in length and twelve in breadth, and weighed thirty pounds. It is transparent, and of a brownish-yellow colour. According to experiments made on purpose, it was found to consist of artificial glass, mixed with a considerable portion of lead; and as it had been preserved in the above collection from the earliest periods, the practice of adding lead to glass must be very old. But whether this mirror was covered at the back, and how it was covered, though these are the most important points, I find nowhere mentioned. In the collection of the Grand Duke of Tuscany there is a piece of the same kind, said also to have been the mirror of Virgil. See Le Veil, Kunst auf Glas zu malen, Nurnb. 1779, 4to, p. 23, and Hist. de l’Acad. des Sciences, 1737, p. 412.
[198] Villaret, Hist. de France. Par. 1763, xi. p. 142.
[199] In Miscellanea Berolinensia, i. p. 263; but nothing further is said respecting the art, than that it was daily used in the glass-houses. Had I an opportunity, I should make experiments of every kind in order to discover a method of forming plane mirrors also in the like manner.
[200] Wecker, in his book De Secretis, lib. x. p. 572, seems to say, that one must lay the saturated tin leaf so carefully on the glass plate, that no air can settle between them. According to Garzoni, the tin leaf is spread out on a smooth stone table, and after it has been rubbed over with quicksilver, the glass is placed above it.
[201] Tome iii. p. 87, art. Glace.
[202] Dict. Géog. de la France. Amst. 1762, fol. v. pp. 415, 672.
[203] A furnace for casting large glass plates, before it is fit to be set at work, cost, it is said, 3500l. It seldom lasts above three years, and even in that time it must be repaired every six months. It takes six months to rebuild it, and three months to repair it. The melting-pots are as big as large hogsheads, and contain above 200 cwt. of metal. If one of them burst in the furnace, the loss of the matter and time amounts to 250l.—Trans.