three conditions we have just been considering in the lower parts of the dislocated regions: cavities, water, and a high temperature, all constituting an agency capable of producing considerable dynamic effects at any moment. Suppose a barrel of powder exploded in a cavity situated a hundred metres underground. At the surface we would hear a rumbling explosion, and feel a vertical shock within a limited space, and an undulatory thrill over a wider circle. The phenomenon will be much like that of an earthquake, except that the essential element of repetition will be wanting; for all will be over with the first shock. But, in the majority of earthquakes, the shocks come in succession, as if the cause of them were a regenerating one.
Many ways may be conceived in which these enormous tensions may end in reiterated shocks, according to the hypothesis on which we place ourselves. Thus the water in a cavity having in time reached the temperature of explosion, suddenly displaces some of the walls of its prison. Hence, a first shock, followed by an expansion into the cracks and adjoining cavities, which have lower temperature and tension. Then, the pressure in the original focus of explosion having fallen off, the walls which had given way return upon themselves to their former position, to give way again when the primitive reservoir has regained its lost tension. This flow from cavity to cavity, which, instead of being continuous, is made by bursts and starts, may continue to be reproduced time after time till the principal reservoir is exhausted. But the mechanism is not destroyed then. During the period of calm, following the seismic period, it can be charged again. Something analogous to this takes place in volcanic eruptions, which are separated by the lapse of time necessary to recharge their apparatus by a slow alimentation. Furthermore, reservoirs of water may be suddenly displaced under the effect of contractions of the crust, and may thus be brought into contact with masses having a high temperature.
If we suppose a sea of melted matter to exist beneath the crust of the earth, we should have analogous effects whenever hydrated rocks, broken off from the shell, fall into the ignited masses.
The theory of the agency of vapor is also supported by the rumblings and subterranean thunders which sometimes continue for months and even for years without being attended by shocks, and for which it is difficult to imagine any other causes than sudden condensation, or a flow of gaseous matter at very high tension through a narrow orifice. The vapor, having escaped from its prison, must in many cases resume the liquid state very quickly, on account of the enormous expansion it undergoes. It also has to traverse miles of relatively cold rocks, more or less water-charged and full of cracks. It may thus contribute to the production of thermal springs. Examples are on record, too, of earthquakes, remote from any volcanic point, in which both hot water and gaseous matter have been seen to issue from crevasses.
