CORALS 37$ outer margins; for the coral-polypes grow principally in a vertical direction, so that the width of the reef can be little or not at all greater than the width of its original base. If the depression of the land be still further continued, the central island will ultimately disappear altogether, and the Fio. 10. Diagrams illustrating the mode of Formation of the different kinds of Coral-reefs. A, Ideal section of a fringing-reef surrounding an is and , B, ideal section of the same, after the land has considerably subsided, and the fringing-reef has been converted into an encircling barrier-reef; C, ideal section of the same, when tlie subsidence has been so far continued as to bury the island under the ocean, and the barrier-reef has been converted into an atoll, a, sea-level; 6, coral- reef; c, land. reef will become an oval or circular ring, usually more or less incomplete, and perhaps 30, 40, or 50 miles in diameter, enclosing a central expanse of water or lagoon. It is thus seen that Fringing-reefs, Barrier-reefs, and Atolls are different stages of the same thing, the latter being produced out of the former by the progressive subsidence of the land. In order, however, that this process should be carried out, it is necessary that the rate of subsidence should not be more rapid than the rate of upward growth of the corals. If it should be so, then the reef is carried down into deep water, and becomes submerged, as is the case, for example, with the great Chagos Bank. In accordance with Mr Darwin s theory on this subject, it is found that Barrier- reefs and Atolls do not occur in the immediate vicinity of active volcanoes regions where geology teaches us that the land is either stationary, or is undergoing slow upheaval. On the other hand, the existence of Fringing- reefs is only possible where the land is either slowly rising, or is stationary ; and, as a matter of fact, Fringing-reefs are often found to be conjoined with upraised strata of post-Tertiary age. As regards their upward limits of growth, again, the coral-polypes cannot exist on levels higher than extreme low water, exposure to the sun, even for a short period, proving generally, if not invariably, fatal. The coral-polypes, therefore, can raise the reef to the level of extreme low water, but no further ; and it is to the denuding power of the ocean that the elevation of the reef above thia level is due. The breakers which fall upon the outer edge of the reef detach masses of dead coral, often of very large size, and these become gradually accumulated at particular spots, till they rise above the level of high water. The detached masses, thus heaped up, become compacted together by the finer sediment of the reef, and agglutinated by the percolation through them of water holding carbonate of lime in solution, till they become ultimately converted into a hard compact limestone. The new land produced in this way is protected from destruction by the vital activity of the living corals, which occupy a fringe at the outermost margin of the reef just below the level of extreme low water, and which by their continual growth preserve the inner parts of the reef from the waves. Another condition very essential to the welfare of the coral-polypes is an abundant supply of pure and properly aerated water. They flourish, therefore, in their highest vigour at the extreme outer edge of the reef, and on its windward side, where they are exposed to the constant beating of the surge ; and hence it is that the growth of the reef is principally carried on at these points. Mud and sediment are, on the other hand, very injurious to corals, and they rarely occur, therefore, on saudy or muddy bottoms. It is for this reason, also, as much probably as from the pernicious effect of an intermixture of fresh water, that openings in coral-reefs are always found to exist at points opposite the mouths of rivers. It has been shown, however, by experiment, that corals will flourish on a sandy bottom provided the water is free from sediment in suspension. As regards the distribution in time of the Zoantharia scleroder- mala, the distinction which obtains at the present day between the solitary and the reef-building corals is found to have subsisted in the past, so far at any rate as the Tertiary and Secondary periods are concerned. Thus the solitary and essentially deep-sea forms are represented in the Kainozoic and Mesozoic deposits by forms such as Spheiwtrochus, Flabcllum, Balanophyllia, Turbinolia, Leptocyathus, Trochocyathus, Paracyathus, Oculina, Diplohelia, Astrolielia, StephanophyUia, Stcrcopsammia, Parasmilia, Trocho- smilia, Thccosmilia, Montlivaltia, Dendrophyllia, &c. On the other hand, the reef-building and essentially shallow-water forms are represented by such genera as Madrepora, Axopora, Forties, Litharcca, Solcnastrcca, Isastrcca, Septastrcea, Dendracis, Astro ccenia, Styloccenia, &c. Taken as a whole, the Zoantharia sderodcrmatalia. ve attained their maximum of development at the present day, being largely repre sented in the Tertiary and Secondary periods, but having their place to a great extent usurped in the Palaeozoic period by the Rugose Corals. TheAporosa&re only represented with certainty in the Pala-ozoic series by the two remarkable genera Battcrsbyia and Hcterophyllia, which form an aberrant group of the Astrccidcc (Palastrcciclce], and of which the former is Devonian, whilst the latter is Carboniferous. The Silurian genus Palccocyclus was formerly regarded as belonging to the Fungidce, but it is a genuine Rugose Coral. The genus Duncanella, of the Upper Silurian, may perhaps belong to the Turbinolidce, and this may very probably be the true position of some of the corals referred to the genus Petraia, of the Silurian and Devonian. The genus Columnaria (Favistella) may perhaps also be referred to the Astrceidce. In the Permian rocks no Aporosa are known to have existed, the whole of this formation, as well as the greater portion of the Trias, being singularly destitute of remains of corals. Towards the summit of the Triassic series, however, in the St Cassian beds, we find a great development of the Aporosa, which are now represented by a number of Astrceidce, belonging to well-known Secondary types, such as Montlivaltia, Thecosmilia, Cladophyllia, Rhabdophyllia, Goniocora, Isastrcea, Thamnastrcca, Elysastrcea, Latimeandra, and Astroccenia. In the succeeding formation of the Lias the Astrceidce are repre sented by all the genera just mentioned, along with others such as Septastrcea, Stylastrcca, Cyathoccenia, Oppelismilia, and Lepido- phyllia, whilst the Turbinolidce are now represented for the first time (Thecocyathus). In the great series of the Oolites, we still find an enormous preponderance of forms belonging to the Astra idce, the principal genera of this period being Isastrcca, Thamnastrcea, Septastrcea, Clausastrcea, Convcxastrcta, Heliastrcea, Plcurosmilia, Pcplosmilia, Blastosmilia, Aplosmilia, Stylosmilia, Thccosmilia, Astroccenia, Stephanoccenia, Rliabdopliyllia, Clado phyllia, Oalamophyllia, Baryphyllia, Stylina, Goniocora, Lati meandra, Cyathophora, Montlivaltia, PMpidogyra, Pachygyra, Dendrogyra, Phytogyra, Favia, &c. The Turbinolidce are represented in the Oolites by genera such as Discocyathus, Ttocho- cyathus, and ThccocyatJius ; the Oculinidce appear under forms such as Stylophora, Euhelia, JKnallohclia, Psammohdia, kc.;
whilst the Fungidce are largely represented by species of Comoscris,