The Medusae Bougainvillea superciliaris and Ametrangia hemisphaerica are viviparous1. Segmentation is total, rarely very unequal as in Clava squamata, where the hypoblast is represented by two cells for a long time.

1 Lang thinks that DavidofF's Phialidium variable is not the Medusa usually so-called, but a younger form of his G. Raffaelei. A species, Gastroblasta timida, has been described by Keller in Z. W. Z. xxxviii. pp. 622 et seqq. It has four manubria, but fission was not observed. See on the whole subject Lang's paper in J. Z. xix. 1886.

2 I. e. in Allman's Schizocladium ramosum; see op. cit. ante, pp. 151-3, Fig. 61. Detachment has been observed in Obelia flabellata and O. (Laomedea) geniculata: Wagner, Wirbellosen des Weissen Meeres, Leipzig, 1885, p. 69. Hincks thinks it occurs in Campanularia neglecta (A. N. H. (4), x. 1872, p. 391), and Allman in Corymorpha nutans (op. cit. p. 153). Hydranths have been observed with several oral cones in Cordylophora lacustris (Price, Q. J. M. xvi. 1876), and two hydranths on the same peduncle occasionally occur in Hydractinia echinata (Wagner, op. cit. PL I. Fig. 8); whether fissiparously produced or not is unknown.

3 The medusa sometimes takes origin in the same place as does a hydranth, e. g. from the peduncle of a hydranth, the Tubularians, Dendroclava, Bougainvillea, some species of Perigonimus; from the stem of a colony, P. muscoides; from the hydrorhiza, some species of Perigonimus, the Campanularian Eucopella. As a rule it originates from the hydrocephalis of a hydranth, or from a blastostyle as in Podocoryne and most Campanularidae.

The embryo is usually from the first a solid mass of cells, the outer layer of which is ciliated and differentiated as epiblast: or there is a blasto-coele, as in Hydra and most embryoes produced from the eggs of Medusae, e. g. Eutima, Obelia, Eucope, Aequorea2. In the first case the internal cells arrange themselves in a layer: in the second the blastocoele is, sooner or later, filled with cells derived (i) by the growth and fission of the cells of the wall in general (Eueope), and the embryo has then become a planula, or of one pole, the most usual mode, or (2) by migration inwards (Obelia). The gastric cavity is formed as a central slit, or the peripheral endoderm cells devour the central, as in Clytia and Obelia (Metschnikoff), or it is in Eutima a remnant of the blastocoele. The planula leads a free existence, and then fixes itself, either by one pole or by its side. In the first case it lengthens and the mouth and tentacles appear at the free end, or it flattens into a disc, from the centre of which rises a bud, the future hydranth (Clytia, Obelia, Eucope, Eudendrium, Campamdaria, Plumularia, Sertularia): in the second, either one end becomes hydrorhiza, the other grows and becomes the hydranth (Tiara, Eutima), or the hydranth-bud appears laterally (Tur-ritopsis nutricula), or it forms a ramified and anastomosing hydrorhiza, from which buds are developed (Turritopsis (Oceania) armata). In Mitro-coma Annae the planula may become either hydrorhiza alone or.hydrorhiza plus hydranth.

There may or may not be a perisarc. Peculiarities are, - the transformation of the superficial cells completely or incompletely into protective envelopes in Hydra (p. 327), and the development in Tubularia of the planula, while within the gonangium, into an Actinula, a hydroid form with a circle of tentacles, or sometimes, as in the adult, two circles, which leads a free existence for a short time before fixation. The planula of Myriothela develops similarly within the vitelline membrane, but the Actinula possesses a set of scattered provisional capitate tentacles in addition to a certain number of permanent tentacles.

1Cf. Wagner, Wirbellosen des Weissen Meeres, p. 74; Allman, Nature, ix. p. 74. In the last instance the planula is oval and non-ciliated. On the species, which is perhaps fissiparous, see Haeckel, System, p. 636.

2According to Metschnikoff, two or three blastulae may fuse in Mitrocoma Annae; but development proceeds normally.

The freshwater Hydroidea are the genera Hydra widely distributed, the American Microhydra from the neighbourhood of Philadelphia, the Russian Polypodium from the Volga (p. 766), and Cordylophora, European, American, and Australian. A Leptomedusan, Laodice (Cosmetird) sali-narum, a modified form of the marine L. (=C.) cruciata, has been found in the variably brackish waters of the discharging canal of saltworks at Villeroy, near Cette, in S. France. Polypodium is in one stage parasitic (p. 766); a Leptomedusan Mnestra parasites lives attached by its manubrium and tentacles to the neck of the pelagic Gastropod Phyllirhoe. The marine Hydroidea are denizens of all seas; the genera are widely distributed, the species sometimes extremely restricted in their range. They live at very various depths from the surface downwards: the greatest number of forms between the lowest tide-mark and fifty fathoms. Some, however, only occur below one hundred fathoms. At very great depths they appear to be extremely rare. The colonies are as a rule of small or moderate size, but occasionally attain a great height, e. g. a Plumularian in the Pelew Islands that of a man. Deep water forms appear to be larger than those from shallow water.

A Monocaulus, seven feet long when fully extended, has been dredged in nearly 3000 fathoms in the North Pacific. The colonies are attached to fixed or floating objects, and certain genera always affect a particular habitat, e.g. Hydractinia and Podocoryne a Gastropod shell tenanted by a Hermit Crab. Some usually grow upon other Hydroids, e.g. some species of Lafoea) Amphibrachium within the canal system of the sponge Euplectella1 The Medusae are frequently carried by currents from the localities where their hydroid form grows. A few appear to swim habitually at a considerable depth below the surface. As to fossil forms a Hydractinia with calcareous skeleton is found in the Chalk, and in Tertiary strata (see note 1, p. 756, ante). There are some other allied (?) calcareous forms supposed to be Tubularian. The Silurian and Devonian Stromatoporidae perhaps belong to the same sub-order. The Campanularians are doubtfully represented by the Rhabdophora s. Grapto-lithidae, which commence in the Upper Cambrian and die out in the Upper Silurian, and by a few other Silurian and Devonian forms.

Sertularella polyzonias occurs in Miocene strata in Ayrshire: a Milleporan (Porosphaera) in the Chalk: Stylaster in Miocene strata.

1 Carter has found an undetermined hydroid in the sponge Rentera (A. N. H. (4), x. p. 50). See note 2, p. 781, post.