cause as peculiar as its effects. Moreover, this cause must be one affecting not merely or only the peculiarities of the animal which builds up the coral, but some cause affecting also the solid rocks and crust of the earth. The coral animals must build on some foundation. They must begin by attaching themselves to something solid. Every coral reef, therefore, whatever be its form—every line of barrier-reef however long—every ring however small or however wide, must indicate some corresponding arrangement of subjacent rock. What cause can have arranged the rocky foundations of the coral in such curious shapes? Extreme cases of any peculiar phenomenon are always those which most attract attention, and sometimes they are the cases which most readily suggest an explanation. Ring-shaped islands of such moderate dimensions that the whole of them can be taken in by the eye, supply such cases. There are atoll-islands where ships can enter, through some break in the ring, into the inner circle. They find themselves in a perfect harbor, in a sheltered lake which no wave can ever enter, yet deep enough and wide enough to hold all the navies of the world. Round about on every side there are the dazzling beaches which are composed of coral sand, and crowning these there is the peaceful cocoanut-palm, and a lower jungle of dense tropical vegetation. On landing and exploring the woods and shores, nothing can be seen but coral. The whole island is a ring of this purely marine product; with the exception of an occasional fragment of pumice-stone, which having been floated over the sea from some distant volcanic eruption, like that of Krakatoa, here disintegrates and furnishes clay, the most essential element of a soil. But reason tells us that there must be something else underground, however deeply buried. When the corals first began to grow, they must have found some rock to build upon, and the shape of these walls must be the shape which was thus determined. One suggestion is obvious. Elsewhere all over the globe there is only one physical cause which determines rocky matter into such ring-like forms as these, and which determines also an included space of depth more or less profound. This physical cause is the eruptive action of volcanic force. When anchored in the central lagoon of a coral atoll, are we not simply anchored in the crater of an extinct volcano—its walls represented by the corals which have grown upon it, its crater represented by the harbor in which our ship is lying? The vegetation is not difficult to account for. The coral grows until it reaches the surface. It is known to flourish best in the foaming breakers. These, although confronted and in the main resisted by the wondrous tubes and cells, are able here and there in violent storms to break off the weaker or overhanging portions of the coral and dash them in fragments upon the top of the reef. Often the waves are loaded with battering-rams in the shape of immense quantities of drift-timber. These bring with them innumerable seeds and hard nuts able to retain their vitality while traversing leagues of
