ACTINOZOA 131 of its own, which is either confluent with that of the others, or is united with them by calcification of the con necting substance of the common body. This intermediate skeletal layer is then termed ccenenchyma. The Octocoralla (excepting Tubipora) give rise to no thecce and their dependencies, the skeleton of each polype, and of the superficial portion of the polyparium, being always composed of loose and independent spicula. But in many, as the Gorgonida;, Pennatulidce (and in the Antipathidue among the Hexacoralla), the central part of the common stem of the compound organism becomes hardened, either by conversion into a mere horny axis (which may be more or less impregnated with calcareous salts) without spicula; or the cornification may be accompanied by a massive development of spicula, either continuously or at intervals; or the main feature of the skeleton may, from the first, be the development of spicula, which become soldered together by a subcrystalline intermediate deposit, as in the red coral of commerce (Corallium rubrum). It has seemed advisable to say thus much concerning the hard parts of the Actinozoa in this place, but the details of the structure and development of the skeleton of the Coralligena will be discussed under CORALS and CORAL REEFS. The Tabulata, or Millepores, and the Rugosa, an extinct and almost exclusively Palaeozoic group of stone-coral form ing animals, are usually referred to the Coralligena. Judg ing by the figures given by Agassiz 1 of living Millepores, the polypes which cover its surface are undoubtedly much more similar to coryinform Ilydrozoa than they are to any Actinozoon. But it is to be observed, firstly, that we have no sufficient knowledge of the intimate structure of the polypes thus figured; and, secondly, that the figures show not the least indication of the external reproductive organs which are so conspicuous in the Ilydrozoa, and which surely must have been present in some one or other of the Millepores examined, were they really Ilydrozoa. As re gards the Rugosa, the presence of septa is a strong argument against their belonging to any group but the Actinozoa, though it is not to be forgotten that a tendency to the development of septiform prominence is visible in the walls of the gastric passages of certain calcareous sponges. Phenomena analogous to the "alternation of generations," which is so common among the Ilydrozoa, are unknown among the great majority of the Actinozoa. But Semper 2 has recently described a process of sexual multiplication in two species of Fungice, which he ranks under this head. The Fungice bud out from a branched stem, and then become detached and free, as is the habit of the genus. To make the parallel with the production of a Medusa from a Scyphistoma complete, however, the stem should be nourished by an asexual polype of a different character from the forms of Fungice, which are produced by gemmation. And this does not appear to be the case. Dimorphism has been observed by Kolliker to occur extensively among the Pennatididce. Each polypary pre sents at least two different sets of zob ids, some being fully developed, and provided with sexual organs, while the others have neither tentacles nor generative organs, and exhibit some other peculiarities. 3 These abortive zooids are either scattered irregularly among the others (e.g., Sarcophyton, Veretilluni), or may occupy a definite position (e.g. , Virgularia). (2.) The Ctenophora. These are all freely swimming, 1 Contributions to the Natural History of Vie United States. Vol. iii. Plate xv. 8 Ueber Generalions-.Wcchsel bei Steinkomllen. Leipzig, 1872. 3 AWumdlungen der Senkenbergischen Naturforschenden Gesdl- schqft, bd. vii. viii. actively locomotive, marine animals, which do not multiply by gemmation, nor form compound organisms such as the polyparies of the Coralligena. Like the latter they are composed of a cellular ectoderm and endoderm, between which a mesoderm, containing stellate connective tissue corpuscles and muscular fibres, is interposed. But, in most parts of the organism, the mesoderm acquires a great thick ness and a gelatinous consistency; so that the body of one of these animals differs in this respect from that of au Actinia in the same way as the body of a Cyanoea differs from that of a Hydra. The bilateral symmetry, which is obscure in most of the Coralligena, becomes obvious in the Qtenophora, in which the parts are disposed symmetri cally on each side of a vertical plane passing through the longitudinal axis of the body. The oral aperture is situated at one end of this axis (or its oral pole), while at the opposite extremity (or aboral pole) there is very generally situated a sac containing solid mineral particles the litJiocyst. The oral aperture leads into a visceral tube, which undoubtedly performs the functions of a stomach. Never theless, as in the Coralligena, it is open at its aboral end, and its cavity is thus placed in direct communication with a chamber, whence canals are given off which penetrate the gelatinous mesoderm. Of these canals, one continues the direction of the axis of the body, and usually ends by two apertures at the aboral pole. The others take a direction in a plane more or less at right angles with the axis ; and after branching out, terminate in longitudinal canals, which lie beneath the series of locomotive paddles, or come into relation with the tentacles when such organs are developed. In addition to these, two canals frequently extend along the sides of the stomach towards the oral pole. The paddle-like locomotive plates are disposed in eight longi tudinal series (ctenophores) on the outer surface of the body. They are thick at the base ; thin and, as it were, frayed out into separate filaments, at their free edges; and each plate is set transversely to the long axis of the series of which it forms a part. The ovaria and testes are developed in the side walls of the longitudinal canals. It is clear, therefore, that these canals answer to the intermesenteric spaces of an Actinia; that the common cavity into which they and the stomach open answers to the common cavity of the body of the Actinia; that the apertures at the aboral pole answer to the terminal aperture of Cereanthus; and that the wide interspaces between the longitudinal canals repre sent the mesoderm of the Actinian mesenteries immensely thickened. In their development the Ctenophora resemble the Coralligena in all essential respects, though they differ from them in some details. Thus the process of yelk division goes on at a different rate in the two moieties of the egg, so that the vitellus becomes divided into one set of small and another set of large cells, whereof the latter become overlaid by the former, and give rise to a large- celled hypoblast, enclosed within a small-celled epiblast. But in the manner in which the body cavity is formed, and the visceral tube (which becomes the stomach) is developed, the Ctenophora resemble the Actinia?. The paddles make their appearance at four points of the circumference of the body, in the form of elevations beset with short cilia; but each of these divides into two, and thus the eight defini tive series are constituted. There is a general agreement among anatomists respect ing the structure of the Ctenophora thus far; but the question whether they possess a nervous system and sensory organs or not, is, as in the case of the Coralligena, one upon which there exists great diversity of opinion. Grant originally described a nervous ganglionated ring, whence
longitudinal cords proceed in Cydippe (Plwrobrachia);
