(ii) The highly extensile tentacle-like and apparently solid zooids of the Campanularian Ophiodes, attached singly to the hydrocope or in numbers to the hydrorhiza. They are capitate, and the terminal knob supports numerous cnidoblasts. There appear to be very similar structures seated on the hydrorhiza of Oplorhiza and Lafoeina and below the hydrothecae of Halecium Gorgonidae 2. (iii) The structures known as nematophores, sarco-thecae, guard-polypes or macho-polyps which are confined to the Campanularian family Plitmidaridae and are in close relation to the thecae of the ordinary hydranths. They are tentaculoid with a solid endodermal axis, longitudinal ectodermic muscle cells, and are capitate. The knob in one form contains cnidoblasts, sense cells, supporting cells, sub-epithelial ganglion cells, and radial muscles; see PI. xiv, Fig. 8, and p. 330. In some species the cnidoblasts are replaced in the guard polypes situated in front of the hydranths by cells containing adhesive globules similar to the structures so called in Ctenophora. And in the genus Aglaophenia certain guard-polypes of this last-named kind placed behind the hydranths, possess a basal ectodermic thickening with cnidoblasts, ganglion cells, etc, protrusible from an aperture of the theca separate to that by which the macho-polypes themselves are protruded.

The macho-polypes are usually disposed in a median and two lateral rows 3. (3) The Gastrozooid, a term applied to the digestive zooid of Hydrocorallina. It is relatively short, and large, with a wide mouth, and in the Stylasteridae is sometimes devoid of tentacles. The base of the gastric cavity gives off many communicating tubes to the coenosarc. (4) A spine occurring in the genus Podocoryne alone1. (5) The claspers of Myriothela, simple filiform structures, with a sucker-like extremity, which bear each one an ovum during its development.

Weismann considers (Entstehung der Sexualzellen, etc. pp. 245-6) that the blastostyle is correlated with the degeneration of the Medusa (pp. 760,768,post). A Medusa-bearing blastostyle occurs in Podocoryne alone among Tubulariae, but the Campanularian Medusa is invariably developed from one.

1The spiral zooid is confined to the edge of the colony bordering the aperture of the Gastropod shell on which the colony is planted, or of any accidental hole in it. The zooids are found in the colonies of both sexes, not in the male only, as Grobben thought. See Weismann, Entstehung der Sexualzellen, etc. pp. 63-65.

2See note, p. 6, in Allman's Plumularidae, Challenger Reports, vii.

3Von Lendenfeld states that the guard animal with cnidoblasts is protruded when the colony is disturbed; the one with adhesive cells, on the contrary, retracted; and that in the third variety above described the part or head with adhesive cells is protruded for holding fast the prey, the basal thickening, on the contrary, if the colony is roughly disturbed, the two parts being retracted under

In many Plnmidaridae a ramulus, hydrocladium or theca-bearing portion of a branch, becomes modified to form a phylactocarp for the gonothecae which here occupy the position of hydrothecae. This phylactocarp, to which numerous machopolypes are attached, varies in structure and in the degree to which the ramulus is modified. It may form a cage completely open, nearly closed, or altogether closed, e.g. in Aglaopheniapluma, where it receives the special name of corbida2.

The coenosarc connecting the hydranths and other members of the colony is nearly invariably a simple tube. But in Tubularia indivisa, Corymorpha and Monocaulus, the endoderm cells of the stem proliferate, leaving only tubular passages: and in some Plumtdaridae the main stem and principal branches are hollow, with a number of peripherally placed endodermal tubes, communicating laterally with one another, all inclosed in a common thin ectoderm and perisarc3. See also note 2, p. 755.

The sexual zooid or Medusa has the typical structure detailed on pp. 247-8, and Fig. 11 4. The Medusae of the Hydroidea constitute Haeckel's first sub-legion of Craspedota, the Leptolinae, in which there are either ocelli or ectodermic otocysts. There are two groups, the Anthomedusae and Leptomedusae. The former possess ocelli and never otocysts: their generative products ripen in the walls of the manubrium, and their radial canals usually number four, very rarely six to eight, and they are derived from Tubularian Hydroids. The latter may have ocelli, but generally reversed conditions. He also observed that the third variety has at first cnidoblasts in its terminal knob, but that during growth the nematocysts are lost, their cells undergo degeneration, and certain supporting cells develope adhesive globules. Metschnikoff has observed the ectoderm of the macho-polyps in two species of Plumularia ingest not only carmine suspended in the water, but the dead hydranths of the colony: see Q. J. M. xxiv. p. 94 possess vesicular otocysts; their generative products ripen on the sub-umbrella in the course of the radial canals, which are in number from 4, 6, 8, 16, 32, 80, to several hundreds, and they are derived from Campanularian Hydroids. In both groups the Medusa may after its detachment grow in size, develope additional radial canals, tentacles, etc, and it may become sexual before its full development is attained.

1Spines also occur in Hydractinia; but, judging from Allman's figure, their structure is unlike that of the spines of Podocoryne.

2 It is impossible to explain the structure of the phylactocarp without figures. For details, see Allman, Plumularidae, Challenger Reports, vii. pp. 10-15; and for the development of the corbula, Id. Gymnoblastic Hydroids, pp. 59-60, and Fig. 30. In Cladocarpus, the phylactocarp is a superadded bifurcating branch, and in Pleurocarpa it consists of phylactocarpal appendages in the shape of ribs taking the place of the proximal hydrocladia of a branch.

3 See Allman, Gymnoblastic Hydroids, pp. 124-6; Id. Plumularidae, Challenger Reports, vii. p. 4; Agassiz, Natural History of United States, iv, p. 267; Hamann, J. Z. xv. p. 30.

4 The Medusae Eleutheria dichotoma and Clavatella prolifera have so feebly developed a bell that they are incapable of swimming, and only creep about. See Allman, op. cit. ante, pp. 31, 212, ^84, and PI. XVIII; Hincks, op. cit. i. p. 71; Haeckel, System der Medusen, pp. 105-7, and Hartlaub, Z. A. ix. 1886. The Medusae in question appear to be distinct, not, as Haeckel considers them, one and the same species: see Allman and Hincks.