margin of the collar constitute the so-called operculum (op.), a
structure which not only acts as a lower lip, but must be important
in separating the food-current produced by the cilia of the tentacles
from the external apertures of the collar-canals and gill-slits. The
collar-canals (fig. 3, c.p.) are a pair of ovoid organs which open from
the collar-cavity to the exterior, their external pores lying immediately
behind the base of the operculum.
While it is not improbable that the collar-pores and the proboscispores may evacuate excretory substances, there can be little doubt that their primary function is to regulate the turgidity of the segment to which they respectively belong. A pair of gill-slits (fig. 3, g.s.), which do not occur in Rhabdopleura, open immediately behind the collar-pores. It is probable that they serve to strain off the superfluous water which is introduced into the mouth during the process of feeding. An anterior median diverticulum of the pharynx (fig. 2, nch.), growing forwards in the septum between the proboscis-cavity and the collar-cavities, and supported dorsally by the median mesentery of the collar, is the representative of the so-called notochord or stomochord of Balanoglossus; and if the view that this organ is really a notochord is well founded, it may be regarded as the homologue of the anterior end of the Vertebrate notochord.
The metasome contains nearly the whole of the alimentary canal, in which pharynx (fig. 2, ph.), oesophagus (oes.), stomach (st.) and intestine (int.) may be distinguished. The remarkable position of the anus (a) on the dorsal side has already been alluded to. The metasomatic cavities are divided by dorsal (fig. 3, d.mes.) and ventral mesenteries, the latter following the outer curvature of the loop of the alimentary canal. The most conspicuous blood vessel possessed by Cephalodiscus is the dorsal vessel (d.b.v.). A ventral vessel occurs on the anterior side of the metasome and forms a loop extending down the entire length of the stalk, while a “heart” projects into the cavity of the pericardium, probably connected on the ventral side of the notochord with the ventral vessel, and on its dorsal side with the dorsal vessel. At their opposite ends the dorsal and ventral vessels are probably connected with one another by means of a splanchnic sinus surrounding the stomach. The original specimen of C. dodecalophus contained exclusively female zooids, in which a single pair of ovaries (figs. 2, 3, ov.) lie in the metasomatic cavities, and open to the exterior dorsally by short, highly pigmented oviducts (fig. 2, ovd.). In C. nigrescens and in some other species a zooid may contain a pair of ovaries, a pair of testes, or an ovary and a testis, although the males, females and hermaphrodites do not differ from one another in external characters. In C. sibogae (Celebes) the single colony known is of the male sex. The reproductive individuals have undergone an extraordinary simplification of the organs concerned with the collection and digestion of food. Thus the arms are reduced to a single pair and possess no tentacles, there is no definite operculum, and the alimentary canal is vestigial. The testes, which correspond in position with the ovaries of a female Cephalodiscus, constitute the greater part of the animal. Associated with these males are neuter zooids, which usually possess no functional reproductive organs, but have in other respects the structure of an ordinary female Cephalodiscus. It appears probable that there is a vascular connexion between these and the male individuals, which thus derive their nutriment from the neuters. The reproductive organs of Rhabdopleura have but seldom been observed. They resemble those of Cephalodiscus in structure and in position, except that in each sex the gonad occurs on the right side of the body only (Schepotieff, 1906).
The eggs of Cephalodiscus possess a large amount of yolk, and it is practically certain that there is no pelagic larval form. The embryos are hatched in an early stage, but their metamorphosis has not been observed. The early development appears to resemble that of the large-yolked species of Balanoglossus. In the bud-development, the three-segmented condition is extremely conspicuous, and a striking feature is the great relative size of the proboscis (fig. 1). A considerable part of the alimentary canal is said to be derived from the ectoderm in the buds of both Cephalodiscus and Rhabdopleura. Schepotieff (1907) states that in the young buds of the latter the central part of the alimentary canal is developed from cells which are apparently not of ectodermic origin.
The affinity of the Pterobranchia to the Enteropneusta may be regarded as definitely established. Considering the wide differences between the two groups in the size and external characters, and in the mode of life, including the mode of feeding, it is indeed surprising that in every important organ the two groups should show a fundamental morphological identity. Their relations to Phoronis are doubtful (see Phoronidea). The question of their affinity to other divisions of the animal kingdom depends principally on the views which are held with regard to the relationships of the Enteropneusta and Phoronidea respectively. The suggestion has been made by Allmann and recently upheld by Schepotieff that Rhabdopleura is related to some of the Graptolites.
Bibliography.—(1) Andersson, “Die Pterobranchier,” Wiss. Ergebn.-Schwed. Südpolar Exp. (1907) vol. v.; (2) Fowler, “Rhabdopleura,” Proc. Roy. Soc. (1893), lii. 132; Festschr. Leuckarts (1892), p. 293-; art. “Hemichorda,” Ency. Brit. (1902), suppl. vols. xxix. p. 249; Quart. Journ. Mic. Sci. (1905), xlviii. 23; (3) Harmer, “Appendix to report on Cephalodiscus,” Challenger Rep. (1887), vol. xx. pt. lxii. p. 39; “Pterobranchia,” Sibona Rep. (1905), Monogr. vol. xxvi. bis.; (4) Lankester, “Rhabdopleura,” Quart. Journ. Mic. Sci. (1884), xxiv. 622; art. “Polyzoa,” Ency. Brit., 9th ed. (1885), xix. 430, 434; “Cephalodiscus nigrescens,” Proc. Roy. Soc. (1905), B. lxxvi. 400; (5) M‘Intosh, “Report on Cephalodiscus,” Challenger Rep. (1887), vol. xx. pt. lxii.; (6) Masterman, “Cephalodiscus,” Quart. Journ. Mic. Sci. (1898), xl. 340