^{page 113 note a} When lead containing silver is melted with free access of air a fusible oxide of lead (litharge) is formed on its surface. This may either be allowed to flow away, or it may be absorbed by the vessel or hearth in which the lead is melted if these are made of a porous material. In either case the whole of the lead is oxidised and removed, and pure silver remains. If impure silver is mixed with lead and treated in this way the impurities present are oxidised along with the lead, dissolve in the litharge, and are removed with it. This process is termed cupellation.

^{page 115 note a} Bone-ash consists chiefly of phosphate of lime.

^{page 115 note b} Geber, , Summa perfectionis magisterii (Venice, 1875), cap. 81Google Scholar. “Sunt autem duo in hujusperdurantia examine perfectionis corpora, sol, scilicet, et luna, propter bonam compositionem que per bonam mixtionem resultat, et illovum puram. substantiam. Narremus igitur modum illius ut tollatur cinis cribellatus aut calx ant pulvis ossium animalium combustorum ant horam omnium, commixtioaut quorundam.”

^{page 116 note a} All the fragments of Roman cupellation hearths which had previously been found consisted only of clayey marl; but in the following analysis of a slag from Thoricos, near Laurion, given in Le Laurium by A. Cordelia, p. 101, it will be seen that 2·4 per cent, of phosphoric acid is present. It is quite possible that the phosphoric acid in this slag may have been derived from the smelting of cupellation hearths along with the ore from the mines, and if so bone-ash had been employed there. Whether this slag was of Greek or Roman origin is uncertain, as although the mines of Laurion had long been worked by the Greeks, they came into the possession of the Romans in 146 B.C.

^{page 116 note b} “The Early Metallurgy of Copper, Tin, and Iron in Europe,” Archaeologia, lvi. 267.

^{page 119 note a} Strabo, iii. 146. τας δὲ τοῡ ἀρϒύρου καμίνους ποιοῡσιν ὑψηλούς, ὥστε τὴν ἐκ τῶν βώλων λιϒνὺν μετέωρον ἐζαίρεσθαι. βαρεῖα γάρ ἐστι καὶ ὁλέθριος.

The accuracy of this statement of Strabo has been sometimes questioned owing to the absence, on sites of Roman smelting works, of stones or bricks such as would be used now for the construction of similar chimnies or shafts. The structure of the chimnies of the Japanese furnaces, figs. 1 and 4, shows us, however, that these adjuncts to the hearths can be and are constructed of other and less enduring materials than those now employed in Europe. Strabo's assertion can hence be accepted without doubt.

^{page 122 note a} Eighty-five analyses of Roman coins, struck between 250 and 375 a.d., consisting chiefly of copper, but containing silver in various proportions from one per cent, upwards, are given in Die Bronzen und Kupferlegirungen der alten und ältesten Völker, by Dr. Ernst Freiherr von Bibra (Erlangen, 1869), pp. 57-68.