CORALS cases, the septa may be equally divided into four groups by four shallow fossae iu the calice. A true columella may be present or absent, or there may be a pseudo-columella formed by the twisting together of the inner edges of the septa. A more or less well- de veloped zone of vesicular tissue, formed by endothecal dissepi ments, is usually found on the exterior of the visceral chamber ; and the tabulae are not complete, but occupy a more or less exten sively developed central area. The corallum may be simple or compound. All the Cyathopliyllldce, are Palaeozoic, and the principal genera are Cyathophyllum, Campophyllum, Omphyma, Pachy- l>h]]Uum, Chonophyllum, Ptychophyllum, Heliophyllum, Palceo- cyclus, Clisiophyllum, Dibunophyllum, Aspidophyllum, Rhodo- phylhim, Aulophyllum, Cyclopliyllum, Accrvularia, Strombodes, Phillipsastrcca, Smithia, Endophyllum, Spongophyllum, Syringo- phyllum, Eridophyllum, Dipliyphyllum, Lithostrotion, Lonsdaleia, Chonaxis, Axophyllum, Koninckophyllum, &c. Fam. IV. CYSTIPIIYLLID^B. The corallum in this family is pimple, or rarely compound ; the septa are rudimentary, and are generally only recognizable as so many vertical striae within the calice. The outer wall is complete, but the entire visceral chamber is rilled with small convex vesicles, sometimes arranged in iufundibuli- form layers, and formed by the dissepiments and tabulae in combi nation. A distinct septal fossula may or may not be present. The entire family is Palaeozoic, and the only undoubted genus is Cystiphyllum itself. In their internal structure, however, the sin gular operculate corals of the genera Goniophyllum, Rhizophyllum, and Calceola present striking affinities with the Cystiphyllidce, whilst a species of the genus Cystirjhyllum (0. prisinaticum) has been described by Lindstrom as being furnished with an operculum. We may, therefore, with great probability, refer the above mentioned abnormal genera to this family of the Rugosa. It should be borne in mind that the Cystiphyllidce, with or without Goniopliyllum and its allies, are such abnormal forms that it is difficult to assert positively that they belong to the Rugosa, and it remains possible that they should be separated to form a special group. As regards their geological distribution, the Rugose corals have a vast development in the seas of the Palaeozoic period, where they seem to take the place, to a large extent at any rate, of the Aporose section of the Zoantliaria, sclerodermata. In the Secondary period only one genus (Holocystis) of the Rugosa is known, and but one is known in the Tertiary rocks (Conosmilia), whilst the only two living genera which could be referred here are the Haplopliyllia of Florida and the Guynia of the Mediterranean. In the Silurian period the Rugosa are very largely developed, especially in the upper division. The principal genera of this period are Cyatlwphyllum, Zaphrentis, Cystiphyllum, Streptclasma, Acervularia, Strombodes, Omphyma, Palccocyclus, Dip hypliyllum, Amplexus, Ptychophyllum, Stauria, Syringophyllum, Goniophyllum, Rhizophyllum, and Pholidophyllum. [The Lower Silurian Colum nar ice with complete septa (Favistella) are probably Aporose corals ; Columnopora is perhaps one of the Perforata ; and Petraia and Duncanella are in a somewhat dubious position.] In deposits of Devonian age, Rugose corals are exceedingly abundant, the principal genera of this period being Cyathophyllum, Heliophyllum, Diphy- phyllum, Eridophyllum, Campophyllum, Endophyllum, Pachyphyl- lum, Phillipsastrcca, Acervularia, Smithia, Spongophyllum, Blothrophyllum, Zaphrentis, Baryphyllum, Aulacophyllum, Aniso- phyllum, Trochophyllum, Combophyllum, Hadrophyllum, Micro- cyclus, Hallia, Mctriophyllum, and Amplexus. In the Carboniferous period the Rugose corals still remain very abundantly represented by forms belonging to such genera as Cvathophyllum, Lithostrotion, Diphyphyllum, Lonsdaleia, Axophyllum, Clisiophyllum, Cyclophyl- lum, Rhodophyllum, Zaphrentis, Amplexus, Lophophyllum, Meno~ phyllum, Campophyllum, Phillipsastrcca, &c. In the Permian rocks, which so far have proved to be extremely uncoralliferous, no examples of the Rugosa are known save the genus Polyccelia, so far at any rate as Britain is concerned. In the great series of the Secondary formations no Rugosa are as yet known save two species of the single genus Holocystis of the Lower Cretaceous. In the great series of the Tertiary deposits, again, there has hitherto been discovered only a single Rugose genus, the Conosmilia of the later Tertiaries of Australia. ORDER III. ALCYONARIA (Octocoralla). The members of this order are Actinozoa in which the polypes possess eight tentacles, which are fringed on their sides with lateral pinnae, or papillae. As in the Zoantharia the mouth opens into a tubular stomach, which in turr communicates freely with the body-cavity, and the stomach is connected with the body-wall by means of a serie of vertical membranous laminoe or mesenteries. Th mesenteries, however, are only eight in number, and are lot paired, one of the tentacles corresponding with and pening into each inter-mesenteric chamber. As a general ule a corallum is secreted, though this may be wanting, ind its nature differs in different cases. In some forms Alcyonium, Xenia, &c.) the corallum is sclerodermic, and insists of variously-formed spicules of lime scattered in the oft tissues. In others (Tubipora) the corallum is external and sclerodermic, with true thecoe, but without septa or abulse, and formed of fused spicules, detached structures f the same nature existing in the soft parts. In others Heliopora) the corallum is external and sclerodermic, with rue thecse provided with septa and tabuise. In others, inally (Gorgonia, Corallium, Pennatula, &c.), there is double corallum, the one consisting of a calcareous or horny 3 clerobasis, over which the soft parts are spread, the other consisting of sclerodermic spicules scattered in the integu- nents. With the exception of the single genus Ilaimeia, which is possibly not a mature form, all the Alcyonaria are
- omposite. The tubular polypes are united by a ccenosarc,
and their body-cavities are placed in communication by means of canals which ramify through the coenosarc, and permit of a free circulation of nutrient fluids. The form of the colony varies greatly in different cases, being usually more or less branched, arborescent, or lobate, but being at other times massive, incrusting, creeping, linear, &c. None of them possess the power of independent locomotion, most being rooted to foreign objects in their adult condi tion, but some being simply sunk in the mud, and a few floating freely in the sea. Many of them exhibit the most brilliant coloration, due in many instances to the brightly- coloured spicules disseminated in their tissues. In some, lastly, it has been shown that the colony normally consists of two kinds of polypes, one sexual, the other sexless and permanently rudimentary. The Alcyonaria may be briefly discussed under the five following families : Fam. I. ALCYONID.E. The members of this family are all fixed to foreign objects, and are more or less fleshy in their consist ence, owing to the fact that the corallum consists simply of sclero dermic spicules scattered in the integuments, and there is no sclero- basis. The spicules are of various forms, but principally fusiform, and they are scattered through the ccenosarc and the soft tissues generally, but are often specially aggregated at the bases of the tentacles and along the tentacles themselves. The spicules may be present in very small numbers (as in some species of Spoggodcs), but they may be present in such numbers as to render the surface rough and prickly (as in Nepthya, Paralcyonium, and the typical species of Spoggodcs). The actinosoma may form lobate masses within which the polypes can be retracted at will (Alcyonium), or more or less branched and having the polypes non-retractile (Xenia). Others form membranous crusts, attached to foreign bodies, the polypes being sometimes retractile (Anthelia), sometimes non- retractile (Sympodium). Others form a creeping colony of slender stolons, sending up polypes at different points, the lower portion of the colony being so thickened by spicular secretions as to constitute a kind of tubular corallum into which the polypes can withdraw (Cornularia). In the genus Sarcophyton, lastly, it has recently been shown (Mosely) that the colony, like that of many of the Pcnnatulidce, is dimorphic, consisting partly of sexually perfect polypes, and partly of zooids which are destitute of generative organs and of tentacles. The zooids, however, have a stomach and mouth, and are connected with the perfect polypes by a canal- system. Fam. II. TUBIPORID.E. This forms a very small family, includ ing only the various species of the " Organ -pipe corals " ( 2 ubipora), and often included with the preceding, to which it is closely related. In the common Tubipora musica, which may be taken as the type of the family, there is a very well developed sclerodermic cor allum, with true thecae enclosing the polypes, but without septa. The corallum is bright red in colour, and is composed of tubular, cylindrical thecae, usually growing regularly side by side, but at a little distance from one another, and united at intervals by horizontal epithecal expansions, which represent external tabulae. There are no septa, nor internal tabulae ; and Dr Perceval Wright has shown that the tubes are really composed of fused spicules. The polypes are green, with eight pinnate tentacles studded with lenticular spicules. The polypes when alarmed retract
themselves within their tubes, the upper portion of the tube, &9