The Trachymedusae are marine, unless the freshwater Limnocodium, of unknown habitat, and a Medusa, lately found in Lake Tanganyika (Central Africa), belong to the order2. Some of them have been dredged at great

1The solid radial and interradial tentacles differ from one another in shape, etc. Brooks states that in Liriope scutigera the interradial tentacles are lost or retained irregularly. Hence the point is not, at least not always, as Haeckel considers it to be, of diagnostic value. See Mem. Boston Soc. Nat. Hist. iii. p. 383.

2By Allman Limnocodium was placed among Leptomedusae. There are three points about depths in the sea; e. g. Pectyllis at 1250 fathoms, Pectis at 1260, Cunarcha Aeginoides at 1675, and Aeginura myosura at 2150. The majority of the Trachomedusae are small, the Geryonidae excepted: the Narcomedusae are small or of moderate size. The two giants are the Petasid Olindias Sam-baquiensis and the Geryonid Carmaris Goltschii, both of .which attain a diameter of four inches.

The Hydroidea s. Polypomedusae, the second order of Craspedota, contains both non-colonial and colonial hydroids, fixed with the exception of Hydra and Protohydra (?). The hydroid, or in a colony the hydranth, is a purely nutritive zooid, except in Hydra, where it has sexual organs (p. 328, ante). The free sexual zooid, or Medusa, which is frequently arrested in development and consequently sessile, arises by gemmation either from the coenosarc, from a hydranth, blastostyle or Medusa.

The non-colonial hydroids are the freshwater genera Hydra and Microhydra, the marine Protohydra, Tiarella, Heterostephanns, Corymor-phidae, and Monocaulus, in all of which a hydrorhiza is absent1. In the colonial forms the hydranth is attached directly to the hydrorhiza in Hydrocorallina: by its hydrocope in some instances, e. g. Clavatella, Hydractinidae: or, as is most usual, by a more or less branched hydro-caulus2. A hydrorhiza is sometimes but feebly developed, as in Myrio-thela: but in most instances it is filiform, or reticulate and spreading, giving origin to new zooids or stems3. In Hydractinia and Podocoryne it it which makes its position doubtful, (i) It has organs at the base of the velum on its exumbrellar aspect, which appear to be auditory though they possess no otoliths. They are said by Professor Lankester to be tentaculocysts, but his description and figures do not place the matter quite beyond doubt. (2) It has floating embryoes, and the bell-cavity seems to be closed at first, a mode of origin not found in any Trachymedusan, so far as is known, certainly not in Liriope and Geryonia, a point, incontestably established by the recent researches of Brooks and Metschnikoff. As the female has never been observed, the origin of the embryoes from ova is not certain. (3) It is possible that the small columnar Hydroid, devoid of tentacles and enveloped in a tube of mud, found by Mr. A. G. Bourne on the roots of a Pontoderia growing in the tank at the gardens of the Royal Botanic Society, Regent's Park - the place where Limnocodium appeared - may belong to the life-history of the Medusa. If so, it is the solitary instance of a fixed hydroid stage among Trachymedusae met with up to the present time.

1The hydrocope of Hydra, Protohydra, and Tiarella ends in a disc; of Heterostephanus, the Corymorphidae and Monocaulus in a pointed extremity, but it is furnished in the two last-named with long extensile filamentous processes which probably anchor it in the mud or sand in which it is immersed.

2In some species of Eudendrium the main stem and origins of the branches may be fascicled, i. e. consist of a number of separate but closely apposed coenosarcal tubes contained each in their own perisarc: see Hincks, British Hydroid Zoophytes, i, and Allman, Gymnoblastic Hydroids, under the genus. Corydendrium has the same peculiarity, and the further one that from the mode of growth of the colony, 2-3 perisarcal tubes may be inclosed in a common perisarcal tube: see "Weismann, Entstehung der Sexualzellen bei den Hydromedusen, Jena, pp. 35-6. The erect growing species of the genus Lafoea have the stem composed of several tubes. Anisicola Halecioides has a second stem accompanying the hydranth-bearing stem, which itself bears no hydranths, but is connected to the primary stem at short intervals: see Jickeli, M. J. viii. p. 636.

3In Coppinia the thecae for the hydranths and blastostyles are closely packed, hiding the forms an incrusting felt-work of tubes, the chitinoid perisarc of which is thickened in the intervals, that of adjoining tubes fusing1. So too in the Hydrocorallina, where the chitinoid perisarc is replaced by a coenosteum or strong skeleton of Lime carbonate, and the hydrorhiza constitutes in-crusting or more usually erect lamellate or columnar, etc, branching structures, simulating certain corals properly so-called. Pores or depressions in this calcareous structure, either scattered or aggregated into systems and sometimes furnished with raised walls, give shelter to the dactylozooids and gastrozooids (infra, p. 758) of the colony: and in the genus Millepora, newly grown strata of tubes are cut off from older and deeper strata by calcareous platforms or tabulae. Similar tabulae cross the pores in the same genus, and in two genera of Stylasteridae.

Hydra is destitute of any natural or adventitious protective coat: Microhydra invests itself with a coating of mud, etc, held together by a secretion: Protohydra and Corymorpha nutans have a delicate cuticula. In other instances a perisarc is present, thickish, brownish in colour, and composed of lamellae; but in Tiarella its outer portion is gelatinous. It is secreted by the cells of the ectoderm, and is resorbed by them in places where budding is taking place. It usually commences in the Tubularian hydroids as a very delicate layer upon the hydrocephalis aborally to the tentacles, is thicker on the hydrocope, and thickest on the hydrocaulus and hydrorhiza2. It is confined to the two last named in the Campamdaria-like hydroids of Eutima, Octorchis and Aequorea; but in most other Cam-panularian hydroids there are perisarcal hydrothecae for the hydranths, and gonothecae (or gonangia) for the blastostyles. The edge of a hydro-theca may be thin, and thrown into plaits when the animal is retracted, or cut into moveable or fixed teeth.

In some Sertularidae it is furnished with a valve-like operculum.