Let us now consider volcanoes as geological agents, and see what they do. A volcanic eruption may be described in a general way as follows: Its advent is heralded by earthquakes affecting the mountain and the whole country round; loud underground explosions are heard, resembling the fire of distant artillery. The vibrations are chiefly transmitted through the ground; the mountain seems convulsed by internal throes, due, no doubt, to the efforts of the imprisoned steam and liquid rock to find an opening. These signs are accompanied by the drying up of wells and disappearance of springs, since the water finds its way down new cracks in the rocks, caused by the frequent shocks and quiverings. When at last an opening has been made, the eruption begins,--generally with one tremendous burst that shakes the whole mountain down to its foundations. After this, frequent explosions follow with great rapidity and increasing violence, generally from the crater. These are indicated by the globular masses of steam which are to be seen rising up in a tall column like that which issues from the funnel of a locomotive. But sometimes the whole mountain seems to be more or less engaged in giving out steam, and thus to be partly enveloped in it. This is illustrated by our engraving from an instantaneous photograph of Vesuvius in eruption in the year 1872. The steam and other gases, in their violent ascent, hurl up into the air a great deal of solid rock from the sides of the central opening, after first blowing out the stones which previously stopped up the orifice. Blocks of stone falling down meet with others coming up; and so a tremendous pounding action takes place, the result of which is that great quantities of volcanic dust and ashes are produced, generally of extreme fineness. Winds and ocean currents transport these light materials for long distances. The observations made during the famous and fruitful voyage of H. M. S. "Challenger" showed that fine volcanic dust is carried by wind and marine currents to almost all parts of the oceans. The darkness so frequently mentioned in accounts of eruptions--sometimes at a very great distance from the volcano--is entirely caused by clouds of volcanic dust hiding the light of the sun. Perhaps the best example of this is the case of the eruption of Krakatoa (in the Strait of Sunda, between Sumatra and Java) in 1883. Its explosions were heard in all directions for two thousand miles, and a perceptible layer of volcanic dust fell at all places within one thousand miles; while the finest dust and vapour, shot up fifteen or twenty miles high, were spread all over the globe, causing, while still suspended in the atmosphere, the peculiar red sunsets noticed in all parts of the world for some months after the eruption.