We must now see what can be made out of the somewhat rambling account of the origin and manufacture of glass given by Pliny at the end of his thirty-sixth book (cap. 44-47). Pliny regarded glass as a Syrian invention. For many centuries, he tells us, the sole source of the principal constituent was a small tract of sand thrown up by the sea at a spot on the Phœnician coast near the town of Ptolemais, where the river Belus[46] flows into the Mediterranean. With this sand the natives mixed the _nitrum_, imported oversea in cakes,[47] and thereby for the first time formed glass. According to Pliny, these Phœnicians were astute and ingenious craftsmen, and they, in time, took to adding to their glass-pots the ‘_magnes lapis_, which, it is asserted, draws to it the melted glass like iron.’ This is a statement most characteristic of Pliny. The _magnes lapis_—magnetic iron-ore or loadstone—is the last substance in the world any one would think of adding to glass. But we know that the ancients knew of two kinds of black stone, for one of which they used the masculine form _magnes_—this was the loadstone—for the other the female form _magnesia_;[48] and this _magnesia_, at any rate at a somewhat later period, can be undoubtedly identified with the black oxide of manganese (_MnO_{2}_), a substance known of old as the ‘soap of glass,’ from its power of removing the green colour derived from iron. Now we have seen that pure white glass, ‘cleansed’ probably by this method, had only comparatively lately been introduced into Italy, and some confused account of the new discovery had probably reached Pliny’s ears. ‘In the same way,’ he continues, ‘they took to adding to the fused mass shining pebbles, then shells and sandy concretions (_fossiles arenæ_).’ In these ‘fossils’ we may, perhaps, recognise the source from which was obtained the lime, an essential constituent of glass. Passing over some obscure references to the nitre of Ophir and the copper of Cyprus, Pliny goes on to say that the whole is melted ‘like bronze,’ in closely grouped furnaces, and that a blackish mass of fatty aspect is obtained. This we must regard as a preliminary frit, for we are told that the mass is melted again in the glass-house, where the requisite colouring matter is added to it. ‘So the work was carried on of old in the famous glass-works of Sidon.... At times the glass was shaped by blowing, or again it was abraded by the wheel, or carved in the manner of silver.... Such was the ancient way of making glass. At the present day in Italy also, by the mouth of the river Vulturnus, for a space of six miles between Cumæ and Liternum, a white and most soft sand is collected, which is pounded both in mortar and mill; it is then mixed with three parts of nitrum,[49] by weight or by measure, and after melting is transferred to other furnaces. In these the substance, now known as _ammonitrum_, is melted and then cast into cakes. These cakes