CTENOPHORA somewhat similar condition, nothing so highly specialized as the mesenchyme of Ctenophora occurs in any other Coelenterate. The nematocysts being nearly absent from their group, their chief function is carried out by adhesive lasso-cells. The Ctenophora are classified as follows :— Subclass i. Tentaculata, Order 1. Cydippidea, Hormiphora. „ 2. Lobata, Deiopea. ,, 3. Ckstoidea, Cestus. ,, ii. Nuda, ,, Bcroe. The Tentaculata, as the name implies, may be recognized by the presence of tentacles of some sort. The Cydippidea are generally spherical or ovoid, with two long retrusible pinnate tentacles : the meridianal and paragastric canals end blindly. An example of these has already been briefly described. The Lobata are of the same general type as the first Order, except for the presence of four circumoral auricles (processes of the subtransverse costs) and of a pair of sagittal outgrowths or lobes, on to which the subsagittal costae are continued. Small accessory tentacles lie in grooves, but there is no tentacular pouch ; the meridianal vessels anastomose in the lobes. In the Cestoidea the body is compressed in the transverse plane, elongated in the sagittal plane, so as to become ribandlike : the subtransverse costae are greatly reduced, the subsagittal
i
Fiq. 3.—Schematic drawing of cestus. {After Chun.) Subs, subsagittal cost®; Subt, much reduced subtentacular costae; Subt, branch of the subtentacular canal which runs along the centre of the riband; Pg, continuation of the paragastric canal at right angles to its original direction along the lower edge of the riband. At the right hand end the last two are seen to unite with the subsagittal canal. costae extend along the aboral edge of the riband. The subsagittal canals lie immediately below their costae aborally, but continuations of the subtransverse canals round down the middle of the riband, and at its end unite, not only with the subsagittal, but also with the paragastric canals which run along the oral edge of the riband. The tentacular bases and pouches are present, but there is no main tentacle as in Cydippidea ; fine accessory tentacles lie in four grooves along the oral edge. The subclass Nuda have no tentacles of any kind ; they are conical or ovoid, with a capacious stomodseum like the cavity of a thimble. There is a ccelenteric network formed by anastomoses of the meridianal and paragastric canals all over the body. The embryology of Callianira has been worked out by Metschnikoff. Segmentation is complete and unequal, producing macromeres and micromeres marked by differences in the size and in yolk-contents. The mioromeres give rise to the ectoderm ; each of the sixteen macromeres, after Fio. 4.—Schematic drawbudding off a small mesoblast cell, passes ing of beroe. {After on as endoderm. A gastrula is established Chun.) by a mixed process of embole and epibole. The mesoblast cells travel to the aboral pole of the embryo, and there form a cross-shaped mass, the arms of which lie in the sagittal and transverse planes (perradii). There can be but little question of the propriety of including Ctenophora among the Coelentera. The undi-
301
vided coelenteron (gastro-vascular system) which constitutes the sole cavity of the body, the largely radial symmetry, the presence of endodermal generative organs on the coelenteric canals, the subepithelial nerve-plexus, the mesogloealike matrix of the body—all these features indicate affinity to other Coelentera, but, as has been stated in the article under that title, the relation is by no means close. In the ninth edition of this work (see Hydrozoa) some stress was laid on Haeckel’s discovery of Ctenaria as a possible linkform between Hydromedusae and Ctenophora, but this view has now been generally abandoned. At what period the Ctenophora branched off' from the line of descent, which culminated in the Hydromedusae and Scyphozoa of today, is not clear, but it is practically certain that they did so before the point of divergence of these two groups from one another. The peculiar sense-organ, the specialization of the cilia into paddles with the corresponding modifications of the coelenteron, the anatomy and position of the tentacles, and, above all, the character and mode of formation of the mesenchyme, separate them widely from other Coelentera. The last-named character, however, combined with the discovery of two remarkable organisms, Coeloplana and Ctenoplana, has suggested affinity to the fiat-worms termed Turbellaria. Ctenoplana, the best known of these, has recently been redescribed by Willey {Quart. Journ. Micr. Sci. xxxix., 1896). It is flattened along the axis which unites sense-organ and mouth, so as to give it a dorsal (aboral) surface, and a ventral (oral) surface on which it frequently creeps. Its costae are very short, and retrusible ; its two tentacles are pinnate and are also retrusible. Two crescentic rows of ciliated papillae lie in the transverse plane on each side of the sense-organ. The coelenteron exhibits six lobes, two of which Willey identifies with the stomodaeum of other Ctenophora; the other four give rise to a system of anastomosing canals suck as are found in Berde and Polyclad Turbellaria. An aboral vessel embraces the sense-organ, but has no external opening. Ctenoplana is obviously a Ctenophoran flattened, and of a creeping habit. Coeloplana is of similar form and habit, with two Ctenophoran tentacles: it has no costae, but is uniformly ciliated. These two forms at least indicate a possible stepping-stone from Ctenophora to Turbellaria, that is to say, from Diploblastic to Triploblastic Metazoa. By themselves they would present no very weighty argument for this line of descent from two-layered to threelayered forms, but the coincidences which occur in the development of Ctenophora and Turbellaria,—the methods of segmentation and gastrulation, of the separation of the mesoblast cells, and of mesenchyme formation,— together with the marked similarity of the adult mesenchyme in the two groups, have led many to accept this pedigree. In Iiis Monograph on the Polyclad Turbellaria of the Bay of Naples, Lang regards a Turbellarian, so to say, as a Ctenophora, in which the sensory pole has rotated forwards in the sagittal plane through 90° as regards the original oral-aboral axis, a rotation which actually occurs in the development of Thysanozoon (Muller’s larva); and he sees, in the eight lappets of the preoral ciliated ring of such a larva, the rudiments of the costal plates. According to his view, a simple early Turbellarian larva, such as that of Stylochus, most nearly represents for us to-day that ancestor from which Ctenophora and Turbellaria are alike derived. For details of this brilliant theory, the reader is referred to the original monograph. A list of the chief works relating to the group may be found at the end of Bourne’s chapter on Ctenophora, in part ii. of Lankester’s Treatise on Zoology (1900). (g. h. FO.)
