212. The large specimens of native iron found in Siberia and Peru, mentioned above, § 51, are among the most curious facts in the natural history of metals. It has been doubted, however, by some, whether they really belong to natural history, or are not rather to be accounted artificial productions. If they had been found in the heart of rocks, or in the midst of metallic veins, no doubt of this sort could possibly have been entertained; but, as they lie quite on the surface, in the middle of flat countries, and at a distance from any known vein of metal, the conjecture that they may be artificial, and the remains of the iron founderies of ancient and unknown nations, is at first sight not entirely destitute of probability. This probability, however, will appear to be the less, the more carefully the specimens are examined. The metal is too perfect, and the masses too large, to have been melted in the furnaces, or to have been transported by the machinery, of a rude people. The specimen in South America weighs 300 quintals, or about 15 tons, and is soft and malleable.[109] The Siberian specimen, described by Pallas, is also very large; it is soft and malleable, and full of round cavities, containing a substance, which, on examination, has been found to be chrysolite.[110] Now, it is certainly quite impossible, that, in an artificial fusion, so much chrysolite could have come by any means to be involved in the iron; but, if the fusion was natural, and happened in a mineral vein, the iron and the chrysolite were both in their native place, and their meeting together has nothing in it that is inexplicable. [Footnote 109: Phil. Trans. 1788, p. 37. also p. 183, &c.] [Footnote 110: Kirwan's Mineralogy, vol. ii. art. Native Iron.] 213. Some circumstances in the description of the specimen in South America, such as the impressions of the feet of men and of birds on its surface, are not to be accounted for on any hypothesis, and certainly require more careful investigation. It is said, that this iron is very little subject to rust, and the analysis of a piece of it by PROUST makes it probable, that it owes this quality to its union with nickel.[111] It appears, also, that the country of Chaco, where this specimen was found, affords many others of the same kind, one of which is mentioned in the description above referred to. That country lies on the east side of the Plata, and is a plain extremely level, and of vast extent, without any appearance of mineral veins; but such veins may nevertheless exist undiscovered, in a tract subject to periodical inundations, and where the native rock is covered with alluvial earth and gravel to a great depth. The veins maybe washed away, and the more durable substances, such as those pieces of native iron, may be left behind; and, though they must be of a formation extremely ancient, according to this hypothesis, they may not have been very long on the surface.