tensive an earthquake as that of Lisbon on the 1st of November, 1755, was produced in this way? John Mitchell (Royal Society, 1760, vol. x, p. 751) drew from this memorable example the conclusion that the vapor of water intervenes in these shocks as well as in the eruptions of volcanoes. Manifest effects of a class of internal explosions, undoubtedly due to the production or sudden moving of a great quantity of superheated vapor, are exhibited at the present epoch, and are not rare. Such explosions, for instance, are exceptionally formidable in the region of Java, and the mind is naturally led to the one which has just convulsed the zone between that island and Sumatra, which has caused the disappearance of the island of Krakatoa and its mountains, has raised other mountains, and has claimed more than forty thousand victims.
At a period more remote from us, the explosive force of interior gases gave rise to very remarkable circular cavities, which have been called "craters of explosion," and are well known. Examples of them are found in Auvergne (Lake Pavin) and in the district of the Eifel, where the stratified beds have been sharply cut as if with a punch. What gases thus put in motion are capable of, as a mechanical power, could hardly have been suspected till since the explosive effects of gun-cotton, nitroglycerine, and dynamite, have been known. The effects of compressed air in the air-gun and of the powder-gases in fire-arms have been wonderfully surpassed, for we now measure explosive pressures of six thousand atmospheres and more. In the experiments in which I have had occasion to observe gases at high pressure in order to explain the action that a meteor coming with planetary speed is subjected to on the part of the atmosphere into which it plunges, I have been surprised at witnessing the great energy of gaseous masses. They engrave themselves deeply, as if with a burner, into the pieces of steel that are opposed to them, and of themselves reduce a part of it to an impalpable dust shot into the atmosphere as if it were volcanic ashes. It is no less surprising—and this observation is of much importance in explaining the problem that occupies us—to remark the tenuity of the gaseous mass that produces such results. Yet its force causes ruptures which the pressure of a weight six hundred thousand times heavier than the gas could not effect!
In short, gaseous movements under high pressure, put in operation from time to time by a simple mechanism like what Nature can and does present, will account for all the essential features of earthquakes. Much better than the hypothesis of interior collisions of solid bodies, they explain the effect of the shock, resembling the blows of a ram, their violence, their frequent succession, and their recurrence in the same regions after many centuries; they explain also the production of earthquakes in regions of dislocation, especially in those in which the disturbance is recent, and their subordination to deep fractures of the crust of the earth.