But, as I have said, we do not clearly know why great reservoirs of melted rock occur from time to time deep under our feet. We may perhaps one day find the clue from the fact that nearly all, if not all, volcanoes occur near to the water's edge, either on the coast of the great oceans or of some enormous inland sea or lake. But at present all we can say is, that in certain parts of the globe there must be from time to time great masses of rock heated till they are white-hot, and having white-hot water mingled with them. These great masses need not, however, be liquid, for we know that under enormous pressure white-hot metals remain solid, and water instead of flashing into steam is kept liquid, pressing with tremendous force upon whatever keeps it confined.
But now suppose that for some reason the mass of solid rock and ground above one of these heated spots should crack and become weak, or that the pressure from below should become so great as to be more powerful than the weight above, then the white-hot rock and water quivering and panting to expand, would upheave and burst the walls of their prison. Cannot you picture to yourselves how when this happened the rock would swell into a liquid state, and how the water would force its way upwards into cracks and fissures expanding into steam as it went. Then would be heard strange rumbling noises underground, as all these heavily oppressed white-hot substances upheaved and rent the crust above them. And after a time the country round, or the ground at the bottom of the sea, would quake and tremble, till by and by a way out would be found, and the water flashing into vapour would break and fling up the masses of rock immediately above the passage it had made for itself, and following after these would come the molten rock pouring out at the new opening.
Such outbursts as these have been seen at sea many times near volcanic islands. In 1811 a new island called Sabrina was thrown up off St. Michael's in the Azores, and after remaining a short time was washed away by the waves. In the same way Graham's Island appeared off the coast of Sicily in 1831, and as late as 1885 Mr. Shipley saw a magnificent eruption in the Pacific near the Tonga Islands when an island about three miles long was formed.