extends upwards of 700 miles; and Disappointment Island and Duff's Group, are connected by a coral reef of 600 miles length, over which the natives pass from one island to another. Coral reefs are divided into three great classes, namely Atolls, Barrier, and Fringing reefs. The word atoll is the name given

by the natives to the circular islands enclosing a lagoon, or still water in the centre. This is the most usual form of the coral islands, and fails not to strike the attention of every one who has crossed the Pacific. They occur of all sizes; of thirty-two examined by Capt. Beechy, the largest was thirty miles in diameter, and the smallest less than a mile, they were of various shapes and all but one, formed by living corals. This one had been raised from the water about eighty feet, but was of coral formation, and was encircled by a reef of living corals. All were slowly increasing their size, and twenty-nine of them had lagoons in the centre, which had probably existed in the others, until, in the course of time, they were filled by the labors of the zoophytes, and other substances. It was supposed by the earlier voyagers that the coral-building animals instinctively built in the form of great circles to protect themselves from the fury of the waters. So far however, from this being the case, we have seen that those massive kinds upon whose existence and increase, the reef depends, flourish best among the breakers on the outside of the reef. Another and more probable theory, is that advocated by Mr. Lyell, that they are based upon the crests of submarine craters, and this idea receives confirmation from the steep angle at which the island plunges at all sides into the surrounding ocean, and that every island yet examined in the immense region called Eastern Oceanica, consists of volcanic rocks, or coral limestones. In opposition to this opinion it is very plausibly argued by Mr. Darwin, that the form and size of some, and the number, proximity, and relative positions of others, are incompatible with this theory. Thus, Suadiva atoll is 44 geographical miles in diameter in one direction, and 34 in another; Rumsky atoll is 54 by 20 miles across; Bow atoll is 30 miles long, but only six in width. Another theory, proposed by Chamisso, accounts for the circular form of coral islands upon the well known fact, that the corals

growing more vigorously around the outside where exposed to the sea, the outer edges would grow up from the foundation before any other part; thus making a ring or cup-shaped structure; but we are not by this theory relieved from the difficulty of answering the question, upon what are these massive structures based? since it is well known that the reef-building corals cannot live at any very considerable depth, though indeed, other species have been found at a depth of 60 fathoms. Below we give a view of one of these islands, copied from Capt. Beechy. The circular form is well exhibited in this island, which is called Whitsunday, but it gives a faint idea of the singular appearance of an atoll,

being one of the smallest size. The immensity of the ocean, the fury of the breakers, contrasted with the lowness of the land, and the smothness of the bright green water within the lagoon can hardly be imagined without having been seen.

The second great class of reefs are the Barrier-reefs, these are similar in all respects to the atolls except having a high land like a castle rising out of the lagoon. The following sketch from Mr. Darwin, will give an idea of the appearance of one of these wonderful structures, being a part of the island of Bolabola in the Pacific, as seen from one of the central peaks. In this instance the whole line of reef has been converted into land, upon which trees are growing; but generally, a snow-white line of breakers, with only here and there a low islet covered with cocoa

nut trees, can be seen separating the dark heaving waters of the ocean from the light green expanse of the lagoon, the still waters

of which, within the reef, usually bathe a fringe of low alluvial soil, upon which the varied and beautiful productions of the tropical regions flourish at the foot of the abrupt and wild central peaks. In the sketch given above, the barrier-reef may be seen in the distance skirting around the island. These reefs are of all sizes from three to forty miles in diameter; and the one which encircles both ends and fronts one side of New Caledonia is upwards of 400 miles long. Externally the reef rises like an atoll with abruptness out of the profound depth of the ocean, but internally it either slopes gradually into the channel, or terminates in a perpendicular wall 200 or 300 feet in height.

There is one remarkable feature connected with the circular reefs, and that is, a deep and narrow passage almost invariably opening from the sea into the lagoon, and kept open by the efflux of the sea at low tides, and it has long been remarked in the case of the barrier reefs, that this channel or opening alway faced valleys in the included land.

The third great class are the Fringing reefs, these, so far as the coral reef itself is concerned, do not differ materially from the others, except that the encircling belt of coral is much narrower. Where the land slopes abruptly into the water the reefs are but a few yards in width, forming a mere ribband or fringe around the island, but when the slope is gradual, the width is much increased

extending sometimes as far as a mile from the land, and always to such a distance from the shore that the limiting depth of 20 or 30 fathoms is obtained, where the reef ceases. From the more flourishing growth of the outermost corals, the fringing reefs are usually highest at the outside, and the sediment washed inwards upon the reef, generally produces in the course of time, a shallow sandy channel. Such are the three great classes of coral reefs which are found scattered throughout the vast oceans, and principally in the tropical regions, but it must by no means be supposed that they are found indiscriminately united, on the contrary the atolls and barrier-reefs are never found in proximity to the fringing reefs. It has been remarked with surprise that while atolls are the most common coral structures throughout some vast portions of the ocean, such as the tropical Pacific and the Indian Oceans, they are entirely wanting, or very nearly so, in the tropical Atlantic and West Indian Seas, where the corals themselves, are exceedingly numerous. There is also another somewhat remarkable fact, that no single active volcano occurs within several hundred miles of a coral archipelago, or even a small group of atolls; and although most of the islands in the Pacific which are encircled by barrier-reefs are of volcanic origin, having remains of craters distinctly visible, yet not one of them is known to have been in eruption since the growth of the corals. In explaining by any theory the formation of coral reefs, we must consider all the phenomena presented by the three great classes as enumerated in the preceding description. To ourselves the explanation proposed by Mr. Darwin in his volume upon The structure and distribution of -Coral Reefs," is the most satisfactory, and may be briefly stated thus; islands, or a line of coast, being first skirted with fringing reefs, become atolls by a continual but gradual subsi

66

LEVEL OF SEA

dence of the land. Let us then take an island surrounded by

We

fringing reefs, and let this island with its reef, represented by the unbroken lines in the wood cut, slowly subside. As the island sinks down, the reef continually grows upward; as the island subsides the space between the inner edge of the reef and the beach becomes proportionally broader. A section of the reef and island in this state is represented by the dotted lines, A A, being the outer edges of the reef; C C, the lagoon; B B, the shores of the encircled island. This section is a real one (on the scale of .388 of an inch to the mile), through Bolabola in the Pacific. can now see why the barrier reefs are so far from the shores which they front. Supposing the island to still subside, the corals meantime growing vigorously upward, the last traces of land will finally disappear, and a perfect atoll be formed. We thus perceive why atolls so much resemble the barrier reefs in general size, form, and manner in which they are grouped together, for they are but the rude outlines of the sunken islands over which they stand. In proof of the foregoing simple and not at all improbable cause for the formation of barrier reefs, and atolls, Mr. Darwin gives some examples of actual subsidence now in progress, and also presents some evidence of the recent elevation of those islands and coasts which have fringing reefs. The sinking of the islands, or coast, for the formation of barrier reefs, or atolls, must necessarily have been very slow, and undoubtedly large archipelagos and lofty islands once existed, where now only rings of coral rock scarce break the open expanse of the sea; thus the only record left to us of the existence of vast tracts of land are the wonderful memorials of these busy architects; in each barrier reef we see evidence of land subsided, and in each atoll a monument of an island lost. Busy from the first ages of the world, when the primeval seas had but a few groups of living beings, of the lowest order of organization, the coral polype has toiled from day to day, and year to year, and is toiling now. What mighty changes have passed over our globe since that remote period in which the Geologist is first enabled to trace the existence of living beings upon the earth. How many tens of thousands of times the earth has revolved around the sun, and how many huge mountain chains of granite have been disintegrated, and their scattered frag