Again, in the original molten world, it seems likely that most of the carbon present--at least, at the surface--was in the atmosphere in the gaseous form of carbon dioxide. This might be dissolved by the rain and other waters; but we know in the modern world no agency which can decompose this compound and reduce it to ordinary carbon or coal, except that of living plants, which are always carrying on this function to an enormous extent. We know that all our great beds of coal and peaty matter are composed of the remains of plants which took their carbon from the air and the waters in past times. We also know that this coaly vegetable matter may, under the influence of heat and pressure, when buried in the earth, be converted into anthracite and into graphite, and even into diamond. It is true that an eminent French chemist[15] has shown that graphite and hydrocarbons may be produced from some of the metallic compounds of carbon which may have been formed under intense heat in the interior of the earth, by the subsequent action of water on such compounds; but there is nothing to show that this can have occurred naturally, unless in very exceptional cases. Now in the Grenvillian system in Canada there is not only a vast quantity of carbon diffused through the limestones, and filling fissures in other rocks, into which it seems to have been originally introduced as liquid bitumen, but also in definite beds associated with earthy matter, and sometimes ten to twelve feet thick. The occurrence of this large amount of carbon warrants us in supposing that it represents a vast vegetable growth, either on the land or in the sea, or both. [Footnote 15: Henri Moissan, "Proceedings Royal Society," June, 1896,] In like manner, in later geological periods, beds of iron ore are generally accumulated as a consequence of the solvent action of acids produced by vegetable decay, as in the clay ironstones of the coal formation and the bog iron ores of later times. Thus the beds of magnetic iron occurring in the Upper Laurentian may be taken as evidences, not of vegetable accumulation, but of vegetable decay. May not also the great quantity of calcium phosphate mined in the Grenville series in Canada, indicate, as similar accumulations do in later formations, the presence of organisms having skeletons of bone earth? With reference to the carbon and iron ore of the Grenville series, I may quote the following from a paper published in the _Journal of the Geological Society of London_ in 1870:--