The freshwater Polypodium hydriforme is a parasite in its first observed stage in the ova of the Sterlet (Acipenser ruthenus). It is then a cylindrical tube or stolon which gives origin to 16 buds, each of which divides once, forming 32 in all. Each bud acquires 24 hollow tentacles. The stolon breaks up when transferred to fresh water. The free Polypodium is 2 mm. long and 4 mm. wide. It divides by longitudinal fission into 2 parts with 12 tentacles apiece, and these in turn into 2 with 6 tentacles apiece. All these forms grow, acquire 24 tentacles, and then divide again. Polypodium moves from place to place by means of its tentacles, on which it stands like stilts. Ussow, A. N. H. (5), xviii. 1886.

The hydranth is frequently detached in Tubularia, and may live some time in this condition, a new hydranth developing as a bud in its place. The two Tubularian genera Nemopsis and Acaulis are considered by Allman to be similarly detached hydranths, and the Hydroid of the former has been discovered by Brooks (Studies Biol. Lab. John's Hopkins Univ. ii. p. 468). A complete absence of hydranths during winter has been observed in Eudendrium, Cordylophora and Hydractinia, but the coenosarc persists and develops buds again at a more favourable season1. So too, when the destruction of Campanularidan colonies by parasitic Protophytes takes place, the hydrocope of certain hydranths may grow suddenly for an inch or so, and then develops a new colony at a higher level (von Lendenfeld, Z. A. vi. p. 42). The power of repairing injuries is very great, and propagation of the individual or colony by artificial section is possible, as in Hydra and Cordylophora.

A few points in the histology of the hydroid colony may be noted. The ectoderm forms an unbroken layer on the surface of the hydrophyton in the Hydractinidae and Hydrocorallina. Ectodermal ganglion cells occur on the arms, chiefly towards their bases (cf. ante, p. 329); also, but more sparingly as a rule, on the body, occasionally aggregated at its base, and even present in numbers in the coenosarc (Eudendrium ramosum). Sense cells are present, and the sense hairs may have great length as in the palpocils of Syncoryne. The ectodermal muscle cells, epithelial or sub-epithelial, are always disposed longitudinally. In the coenosarc the ectoderm is usually unilaminar, but it may be multilaminar as in Eudendrium ramosum, and its cnidoblasts are few and devoid of cnidocils. It is usually said not to contain muscle-cells,, but they have been detected in Campanu-laria and Plumularia, where they cause slow rhythmic contractions of the cavity of the tube, and they are probably present in other instances.

It secretes the perisarc, and is either in contact with it at all points or only here and there by cells with a striated structure; cf. von Lendenfeld, Z. W. Z. xxxviii. p. 534, and for the chitino-genous cells, ibid. p. 520. Where perisarc is to be absorbed the ectodermal cells enlarge; see op. cit. p. 542, and Weismann, Entstehung der Sexualzellen, etc. pp. 178 and 182. The other constituents of the ectoderm are epithelial cells, interstitial cells, and gland cells. As to the endoderm, ganglion and sense cells have been found in the hypostome of Eucopella and of other Campanularians (von Lendenfeld, Z. A. vi. p. 70). The notochord-like axial cells of the tentacles (p. 329) often become irregularly arranged in the adult, and they are in many instances united by a basal ring of similar cells, e. g. in Tubularia, Eucopella. Other elements are epithelio-muscle cells, the processes of which take as a rule a circular direction, absent in Campanularidae and Sertularidae, except in the hypostome, e. g. of Eucopella (Pothers); of glandular absorbent and excretory cells, cf. von Lendenfeld, Z. W. Z. xxxviii. pp. 522-5. Whenever muscle cells are well developed, the endoderm is thrown into prominent longitudinal rugae, at least in the contracted state of the hydranth.

These rugae are more feeble in other instances.

1 This phenomenon is perhaps not uncommon: see Hincks, op. cit. p. xliv. On Eudendrium see Weismann, Entstehung, etc. p. 102; and on Tubularia, Allman, op. cit. p. 403.

The ambulatory medusae Clavatella and Eleutheria (see note 4, p. 759) have branched tentacles. The basal or first branch ends in Clavatella in a sucker, or in-Eleutheria in a knob laden with cnidoblasts. The swimming form Cladonema has three similar adhesive processes to each tentacle, and Hincks has observed it reverse its bell and adhere by the processes in question while fishing for food with the manubrium (A. N. H. (4), x. 1872, p. 3921). The reversion and atrophy of the bell when the sexual products are ripe has also been observed in some instances, e. g. Cladonema (Allman, Gymnoblastic Hydroids, p. 359; Hincks, British Hydroid Zoophytes, i. p. xxviii.). Solid tentacles, i. e. with an endodermal axis, occur in some Leptomedusae, e. g. in Obelia. The exumbrellar ridges of Eucopella have been examined by von Lendenfeld, who finds that they consist of a meridional nerve with ganglion cells, young cnidoblast cells, and an overlying epithelium of cnidoblasts and sense cells. The otocyst in the same medusa is connected to both nerve-rings, and contains a basal ganglion.

The sexual products originate from interstitial ectoderm cells in many medusae after they are detached, or even before, e. g. Perigonimus. In some species of Obelia they migrate from their place of origin at the base of the manubrium to the radial canals (Hartlaub). The ova pass into the endoderm, wander through it, and ripen partly in the endoderm, partly in the ectoderm of the radial canals. The male cells, however, ripen exclusively in the ectoderm. The ova of the medusa of Podocoryne originate in the gonophore-bud.