[305] Lib. xxxv. c. 26. and lib. xxxvii. c. 9. The _lapis obsidianus_, which Obsidius first found in Ethiopia, and made known, is undoubtedly the same as that vulcanic glass which is sometimes called Icelandic agate, _pumex vitreus_, and by the Spaniards, who brought it from America and California, named _galinace_. [306] Historiæ Augustæ Scriptores, in vita Gallieni, cap. 12. [307] Ib. in Vopisc. vita Saturnini, c. 8. [308] Strabo, Amst. 1707, fol. lib. xvi. p. 1099.--Some consider the glass earth here mentioned as a mineral alkali that was really found in Egypt, and which served to make glass; but, as the author speaks expressly of coloured glass, I do not think that the above salt, without which no glass was then made, is what is meant; but rather a metallic oxide, such perhaps as ochre or manganese. [309] Sen. Op. Lipsii, p. 579. [310] Hist. Nat. lib. xxxvii. c. 12. A passage in Diodorus Siculus, lib. ii. c. 52, alludes, in my opinion, to this method of colouring by cementation. [311] Magia Naturalis. Franc. 1591, 8vo, p. 275. [312] Kunkel’s Ars Vitraria. Nur. 1743, 4to, pp. 98, 101. [313] Comment. Soc. Scient. Gotting. ii. p. 41. [314] Montamy von den Farben zuni Porzellan- und Email-malen. Leipsic, 1767, 8vo, p. 82. Fontanieu, p. 16. [315] [The extensive use of this substance in colouring glass and porcelain has rendered its best and most œconomical preparation a subject of interest both to the chemist and the manufacturer. Although the determination of its true chemical composition has presented obstacles almost insuperable, still many important points with regard to its manufacture have been elucidated. It has been found that the tin salt used in precipitating it must contain both the binoxide and protoxide of tin in certain proportions, and it has been also discovered that the degree of dilution both of the gold and tin solutions exerts a very perceptible influence on the beauty of the preparation. Capaun has examined this latter point with great attention, by testing all the different products as to their power of colouring glass. The first point to be attained is the preparation of a solution of sesquioxide of tin; and for this purpose Bolley proposes to employ the double compound of bichloride of tin with sal-ammoniac (pink salt). This salt is not altered by exposure to the atmosphere, and contains a fixed and known quantity of bichloride of tin, and when boiled with metallic tin it takes up so much as will form the protochloride; as the exact quantity of the bichloride is known, it is very easy to use exactly such a quantity of tin as will serve to form the sesquichloride. 100 parts of the pink salt require for this purpose 10·7 parts of metallic tin.