ANTHOZOA
463
coral reefs between 1838 and 1842, but it was not long before it was attacked by other observers. In 1851 Louis Agassiz produced evidence to show that the reefs off the south coast of Florida were not formed during subsidence, and in 1863 Karl Semper showed that in the Pelew islands there is abundant evidence of recent upheaval in a region where both atolls and barrier-reefs exist. Latterly, many instances of recently upraised coral formations have been described by Wharton, Guppy, Gardiner, and others, and Alexander Agassiz and Murray have brought forward a mass of evidence tending to shake the subsidence theory to its foundations. Murray has pointed out that the deep-sea soundings of the Tuscarora and Challenger have proved the existence of a large number of submarine elevations rising out of a depth of 2000 fathoms or more to within a few hundred fathoms of the surface. The existence of such banks was unknown to Darwin, and removes his objections to Chamisso’s theory. For although they may at first be too far below the surface for reef-building corals, they afford a habitat for numerous echinoderms, molluscs, Crustacea, and deepsea corals, whose skeletons accumulate on their summits, and they further receive a constant rain of the calcareous and silicious skeletons of minute organisms which teem in the waters above. By these agencies the banks are gradually raised to the lowest depth at which reef-building corals can flourish, and once these establish themselves they will grow more rapidly on the periphery of the bank, j because they are more favourably situated as regards food-supply. Thus the reef will rise to the surface as an atoll, and the nearer it approaches the surface the more will the corals on the exterior faces be favoured, and the more will those in the centre of the reef decrease, for Fig. 20.—Diagram showing the formation of an atoll during subsidence. experiment has shown that the minute pelagic organisms (After Darwin.) The lower part of the figure represents a barrier reef on which corals feed are far less abundant in a lagoon surrounding a central peak. A, A, Outer edges of the barrier reef at tlie than in the sea outside. Eventually, as the margin of the sea-level; the cocoa-nut trees indicate dry land formed on the edges of the reef; L, L, lagoon channel; A', A', outer edges of the atoll formed by upgrowth of the coral during the subsidence of the peak; I/, lagoon of the reef rises to the surface and material is accumulated upon atoll. The vertical scale is considerably exaggerated as compared with the it to form islets or continuous land, the coral growth in horizontal scale. the lagoon will be feeble, and the solvent action of seaand none emerging above it. But on the supposition that water and the scour of the tide will tend to deepen the the atolls and encircling reefs were formed round land lagoon. Thus the considerable depth of some lagoons, which was undergoing a slow movement of subsidence, amounting to 40 or 50 fathoms, may be accounted for. their structure could easily be explained. Take the case The observations of Guppy in the Solomon Islands have of an island consisting of a single high peak. At first gone far to confirm Murray’s conclusions, since he found the coral growth would form a fringing reef clinging to in the islands of Ugi, Santa Anna, and Treasury and its shores. As the island slowly subsided into the ocean Stirling islands unmistakable evidences of a nucleus of the upward growth of coral would keep the outer rim of volcanic rock, covered with soft earthy bedded deposits the reef level with or within a few fathoms of the surface, several hundred feet thick. These deposits are highly so that, as subsidence proceeded, the distance between the fossiliferous in parts, and contain the remains of pteropods, outer rim of the reef and the sinking land would continu- lamellibranchs, and echinoderms, imbedded in a foraminially increase, with the result that a barrier-reef would be ferous deposit mixed with volcanic debris, like the deepformed separated by a wide channel from the central peak. sea muds brought up by the Challenger. The flanks of As corals and other organisms with calcareous skeletons these elevated beds are covered with coralline limestone live in the channel, their remains, as well as the accumu- rocks varying from 100 to 16 feet in thickness. One of lation of coral and other debris thrown over the outer the islands, Santa Anna, has the form of an upraised atoll, edge of the reef, would maintain the channel at a shallower with a mass of coral limestone 80 feet in vertical thickness, depth than that of the ocean outside. Finally, if the resting on a friable and sparingly argillaceous rock resemsubsidence continued, the central peak would disappear bling a deep-sea deposit. A. Agassiz, in a number of beneath the surface, and an atoll would be left consisting important researches on the Florida reefs, the Bahamas, of a raised margin of reef surrounding a central lagoon, the Bermudas, the Fiji Islands, and the Great Barrierand any pause during the movement of subsidence wou.d reef of Australia, has further shown that many of the result in the formation of raised islets or a strip of dry peculiar features of these coral formations cannot be land along the margin of the reef. Darwin’s theory was explained on the theory of subsidence, but are rather published in 1842, and found almost universal acceptance, attributable to the natural growth of corals on banks both because of its simplicity and its applicability to formed by prevailing currents, or on extensive shore platevery known type of coral-reef formation, inc.uding such forms or submarine flats formed by the erosion of predifficult cases as the Great Chagos Bank, a huge submerged existing land surfaces. In face of this accumulated evidence, it must be admitted atoll in the Indian Ocean. Darwin’s theory was adopted and strengthened by Dana, that the subsidence theory of Darwin is inapplicable to a who had made extensive observations among the Pacific ! large number of coral reefs and islands, but it is hardly
natural growth of corals and the action of the waves. He pointed out that the larger and more massive species of corals flourish best on the outer sides of a reef, whilst the more interior corals are killed or stunted in growth by the accumulation of coral and other debris. Thus the outer edge of a submerged reef is the first to reach the surface, and a ring of land being formed by materials piled up by the waves, an atoll with a central lagoon is produced. Chamisso’s theory necessarily assumed the existence of a great number of submerged banks reaching nearly, but not quite, to the surface of the sea in the Pacific and Indian oceans/ and the difficulty of accounting for the existence of so many of these led Darwin to reject his views and bring forward an explanation which may be called the theory of subsidence. Starting from the wellknown premiss that reef-building species of corals do not flourish in a greater depth of water than 20 fathoms, Darwin argued that all coral islands must have a rocky base, and that it was inconceivable that, in such large tracts of sea as occur in the Pacific and Indian oceans, there should be a vast number of submarine peaks or banks all rising to within 20 or 30 fathoms of the surface
