"The foregoing considerations make it tolerably clear that volcanoes are fed from deposits of water contained in ancient rocks which have become greatly heated through the blanketing effects of the strata which have been laid down upon them. The gas which is the only invariable element of volcanic eruptions is steam; moreover, it is the steam of sea-water, as is proven by analysis of the ejections. It breaks its way to the surface only on those parts of the earth which are near to where the deposition of strata is lifting the temperature of water contained in rocks by preventing, in fact, the escape of the earth's heat." Another very common theory is that of chemical action, or the heat produced by the oxidation of various substances inside the earth, such, for example, as iron pyrites, a compound of iron and sulphur. When Sir Humphrey Davy discovered metallic sodium and it was found that this material, when thrown on water, possessed the power of liberating intense heat, the discovery was welcomed by geologists as affording a possible explanation of the cause of volcanoes and earthquakes. It may be said generally concerning chemical action as the source of the earth's interior heat, that the chief objection against it is the fact that such heat is liberated too slowly to result in the production of a very high temperature. This objection does not exist in the case of such substances as metallic sodium, since here the heat is rapidly developed and is sufficient in amount to fuse the substances produced. But in the lava produced in such great quantities as it is in volcanic districts there must be liberated at the same time large quantities of gaseous hydrogen. Now, although hydrogen is, as we have already seen, sometimes given off with the gases that escape from volcanic craters, yet the quantity which escapes is so small that this theory of volcanic activity has been practically abandoned. Quite recently, however, among the various chemical substances that are produced under the extremely high temperatures of the electric furnace have been found, or formed, a number of curious substances such as _calcium carbide_, _calcium silicide_, _barium silicide_, etc., that possess the property of becoming highly heated on coming in contact with water. Now it is an interesting fact that the hydrogen and other gases which are given off by the action of water on these substances are absorbed in large quantities by the materials themselves, so that the objection of the absence of hydrogen and similar gases in the craters of the volcanoes would not be quite as objectionable as in the case of such substances.