Farre's observations were made on certain transparent Ctenostomes (Bowerbankia and Farrella). He states that the parietal muscles "were distinctly seen to contract whenever the protrusion of the animal took place, and to become relaxed again upon its retiring into its cell." Their contraction may indent the outline of the ectocyst, or may cause the separation of the endocyst from the ectocyst. The endocyst is then drawn into longitudinal lines at the origin and insertion of these fibres. It is further suggested that some part is played in the process by the muscular walls of the alimentary canal, which is a good deal bent in the retracted condition. The effort to straighten itself is believed to have some share in forcing out the polypide. The flexible, membranous character of the "aperture" (see p. [524]) in Membranipora (Fig. 256, A) is said by Nitsche[[561]] to be an arrangement for the protrusion of the polypides; the parietal muscles passing from the lateral walls of the zooecium to the upper membranous wall, which is accordingly depressed by their contraction.
Although it is hardly possible to doubt the accuracy of Farre's observations, which have, moreover, been confirmed by Hincks, it is by no means certain that this is the whole explanation in all cases. Oka,[[562]] for instance, states that protrusion of the polypide in Phylactolaemata can be effected in a branch whose body-wall has been cut open. Pergens[[563]] believes that the diaphragm (Fig. 234, d) acts as a pump, introducing water from the tentacle-sheath into the body-cavity, into which it is said by him to open, and so forcing out the polypide. It is probable that many of the forms which have a stiff, unyielding ectocyst possess special arrangements for introducing water in some way into the space bounded by the ectocyst,[[564]] and so forcing out the polypide. Such, for instance, may be the median pore which occurs beneath the orifice in Microporella (Fig. 241, A, mp), and in certain other cases.
Reproduction of Phylactolaemata.—Sexual reproduction takes place in Cristatella from June to August. The spermatozoa are ordinarily produced on the funiculus. The ovaries usually occur on the inner side of the common wall of the colony, not far below the orifice of a polypide. Each ovary matures a single egg, which develops in situ, the free larva leaving the colony by the orifice of one of the degenerated polypides.
A second method of reproduction takes place by means of the statoblasts, which are developed on the funiculus (Fig. 249). According to Verworn,[[565]] each statoblast arises from a single cell of the funiculus; and on this view, the statoblast is, as supposed by the earlier observers, a special kind of winter-egg. According to more recent researches,[[566]] the funiculus consists of a central axis, formed from the ectoderm, and of an outer sheath of mesoderm-cells; the statoblast is developed from the two kinds of cells of which the funiculus is composed, and is consequently comparable in its mode of origin to an ordinary bud. Its special peculiarities are: its origin as an internal bud, its possession of a chitinous shell, and the fact that it is destined to leave the parent colony, and to develop, after a period of rest, into a new colony. Germination takes place by the formation of a polypide-bud inside the statoblast, which finally splits along its equator into two halves. The contents emerge as a young colony which possesses at least one fully-formed polypide.
Remarkable structures known as "hibernacula" occur in the fresh-water Ctenostomes, Paludicella and Victorella. These bodies are in the former (Fig. 250, B) specially modified external buds, which persist through the winter when the rest of the colony dies down. At the close of winter the shell splits into two halves, exactly as takes place in the statoblasts, and a young colony emerges. It is possible that the statoblasts may have been evolved from a hibernaculum, which was at first produced externally, but has become modified in such a way as to acquire an internal mode of origin.[[567]]
The simplest known statoblast is that of Fredericella (Fig. 251, A), which differs from that of other Phylactolaemata in having no ring of air-cells. In Plumatella, the statoblast (Fig. 251, B) has a broad equatorial ring of air-cells, which enable it to float at the surface of the water on the decay of the parent tubes. In some species, certain statoblasts which are produced in the adherent parts of the colony remain attached to the substratum. These "sessile statoblasts" may have no trace of the ring of air-cells; but the fact that many sessile statoblasts have rudiments of this structure suggests that they are a secondary modification of the floating statoblast. In Lophopus (Fig. 251, C) the ring of air-cells is very broad, and is pointed at each end; while in Cristatella (Fig. 251, D) and in Pectinatella the statoblast is circular, and possesses an armature of hooked spines. That of Cristatella, measures about .75 mm. in its greatest length.
Fig. 250.—Paludicella ehrenbergi van Beneden, × about 3. A, Part of a colony with expanded polypides; B, remains of part of a colony which has produced hibernacula or winter-buds (h); z, zooecium. (From Kraepelin.)
Kraepelin has suggested that the above order of increasing complexity of the statoblasts corresponds with the order in which the genera to which they respectively belong would be placed, on the assumption that the Phylactolaemata have been derived from the Ctenostomata. Thus, in Fredericella, the form of the lophophore is circular, as in the Gymnolaemata. The number of the tentacles is comparatively small (20-24). The arborescent form of the colony resembles that of many Ctenostomes, and the zooecia are more or less cut off from one another by incomplete septa.
In Plumatella, the lophophore has become horse-shoe-shaped, and the tentacles are more numerous (38-60). In general form and in the arrangement of the septa this genus resembles Fredericella, with which it may easily be confused.