sure, therefore, that no malleable iron whatever can contain any such proportion of manganese. The fact is, that Bergman's mode of separating manganese from iron was defective. What he considered as manganese was chiefly, and might be in many cases altogether, oxide of iron. Many years elapsed before a good process for separating iron from manganese was discovered. 17. Bergman's experiments to ascertain the cause of the brittleness of cold-short iron need not occupy much of our attention. He extracted from it a white powder, by dissolving the cold-short iron in dilute sulphuric acid. This white powder he succeeded in reducing to the state of a white brittle metal, by fusing it with a flux and charcoal. Klaproth soon after ascertained that this metal was a phosphuret of iron, and that the white powder was a phosphate of iron: and Scheele, with his usual sagacity, hit on a method of analyzing this phosphate, and thus demonstrating its nature. Thus Bergman's experiments led to the knowledge of the fact that cold-short iron owes its brittleness to a quantity of phosphorus which it contains. It ought to be mentioned that Meyer, of Stettin, ascertained the same fact, and made it known to chemists at about the same time with Bergman. 18. The dissertation on the products of volcanoes, first published in 1777, is one of the most striking examples of the sagacity of Bergman which we possess. He takes a view of all the substances certainly known to have been thrown out of volcanoes, attempts to subject them to a chemical analysis, and compares them with the basalt, and greenstone or trap-rocks, the origin of which constituted at that time a keen matter of dispute among geologists. He shows the identity between lavas and basalt and greenstone, and therefore infers the identity of formation. This is obviously the true mode of proceeding, and, had it been adopted at an earlier period, many of those disputes respecting the nature of trap-rocks, which occupied geologists for so long a period, would never have been agitated; or, at least, would have been speedily decided. The whole dissertation is filled with valuable matter, still well entitled to the attention of geologists. His observations on _zeolites_, which he considered as unconnected with volcanic products, were very natural at the time when he wrote: though the subsequent experiments of Sir James Hall, and Mr. Gregory Watt, and, above all, an accurate attention to the scoriæ from different smelting-houses, have thrown a new light on the subject, and have shown the way in which zeolitic crystals might easily have been formed in melted lava, provided circumstances were favourable. In fact, we find abundant cavities in real lava from Vesuvius, filled with zeolitic crystals.