A small group, the Gastreadae of Haeckel, of doubtful affinities may be mentioned here. They are marine sponge-like organisms with a body-wall composed of an ectodermal syncytium (? mesoglea + ectoderm) and a ciliated collared endoderm, supported by a skeleton of sponge-spicules, Radiolarian shells, etc, according to habitat. Pores are absent. There are two sub-groups: (1) Physemaria, body-wall thin, solid, a single osculum, with four genera, two solitary, Haliphysema, Gastrophysema, and two colonial (?), Dendrophysema, Clathrophysema; (2) Caementaria, body-wall thick, traversed by gastral tubes lined wholly or in part by ciliated endoderm, mostly deep sea, with four genera, Caementascus with a single osculum, Caementoncus, Caementissa and Caementura with several. Ova have been seen in several instances: filiform spermatozoa and an invaginate gastrula in Gastrophysema dithalamium.

The group requires a careful re-examination. Haliphysema Tumanowiczii, which was originally described as a sponge by Bowerbank, has been proved to be one of the Protozoan Reticularia by Kent, Lankester and Mobius, as it was surmised to be by Carter. Moreover, it is identical with Carter's Squamulina scopula = Gastrophysema scopula according to Norman, which Carter supposed originally to be a Reticularian. Haeckel, however, states that he has himself never seen the Squamulina in question, and classes it among Physemaria on the strength of his own views relative to Carter's observations. A fate similar to that of H. Tu-manowiczii has betaken Wagnerella borealis from the White Sea, described by Merejkowsky as a sponge, but now held to be one of the Protozoan Heliozoa.

Neue Gastraeaden, Haeckel, SB. Jen. Ges. 1883, pp. 84-9. Physemaria, Id., J. Z. xi. 1877. Haliphysema Tumanowiczii, Saville Kent, A. N. H. (5), ii. 1878; Ray Lankester, Q. J. M. xix 1879; Mobius, Beitrage zur Meeresfauna der Insel Mauritius, etc., Berlin, 1880, pp. 72-5, Pis. i, ii, I. On Haliphysema and allied forms, Norman, A. N. H. (5), i. 1878.

The Porifera are classified as follows: I. Calcarea = Calcispongiae: skeleton composed of calcareous spicules.

1. Homocoela: gastric cavity lined in all parts by collared cells; three families, (i) Asconidae=Ascones: gastric cavity flask-shaped and body-wall plain, Leucosolenia =Ascetta, Ascandra, &C.; (ii) Homodermidae: body-wall produced into radial tubes similar to those of Syconinae (infra), Homoderma Sycandra; (iii) Leucopsidae: a large amount of mesoglaea imbedding 'mouthless Ascon-persons,' a large pseudo-gaster and pseudostome, Leucopsis pedunculata.

2. Heterocoela: endoderm differentiated into two forms of cells, pavement epithelium cells lining a central portion of the gastric cavity, and collared flagellate cells restricted to limited regions either radial tubes or ampullae; four families, (i) Syconidae= Sycones, body-wall produced into radial tubes or cones with sensory cells round the orifices of the inhalent canals, with three sub-families (pp. 793-4), Syconinae, Sycon = Sycetta, Sycandra, Sycortis, Sycaltis in part, Uteinae, e.g. Ute, Grantessa, etc, Grantinae, e. g. Grantia; (ii) Sylleibidae, with a complicated inhalent canal-system, and sac-like ampullae either placed radially to main axis of the sponge and connected to central cavity by a complicated exhalent canal-system, Vosmaera=Leucetta in part, or placed radially to the wide exhalent canals, the layer of ampullae being thrown into folds, Polejna=Leucilla; (iii) Leuconidae: a ramified inhalent and exhalent canal-system, and spherical ampullae, Leucetta in part, Leuconia (=Leucetta, Leu-caltis, Leucandra in part), Pericharax (=Leucandra in part); (iv) Teichonidae, external surface differentiated into two planes, one pore-bearing, the other with oscula, Teichonella, Eilhardia; the internal organisation of the latter does not differ from that of Leuconidae (Polejaeff)1.

Extinct group, Pharetrones of Zittel, probably a sub-group of Leuconidae.

II. PoriferaNon-Calcarea=Fibrospongiae\ skeleton rarely absent, composed of either siliceous spicules or spongin fibres; the spicules are either isolated, or united by silica or spongin; gastric system belonging to either type (3) or (4); see p. .794.

1. .Hyalospongiae=Hexactinellidae: skeleton wholly siliceous, spicules triaxile, either isolated or united by silica into a trellis-work; canal-system known only in Euplectella, belonging to type (3), see p. 794; marine, and for the most part deep-sea forms; extend from the Silurian to the present epoch; two sub-orders, (i) Dicty-onina: spicules united by the tips of their arms; skeleton a trellis-work with square or irregular meshes; flesh-spicules present or absent, e. g. Farrea, Aphrocalistes, Dactylocalyx, and a large number of fossil forms, including the Ventriculitidae; (ii) Lyssakina: spicules united only by protoplasm, or a small quantity of silic- flesh-spicules usually numerous and of very various forms, e. g. Hyalonema, Euplectella, Holtenia, Pheronema, and fossil forms including the Receptaculitidae.

2.Spiculispongiae: skeleton very rarely absent; spicules generally independent, united either by interlocking processes or into bundles by organic material, with five sub-orders: (i) Lithistina: body strong and massive, a central cloaca, or scattered oscula; cloaca frequently replaced by vertical tubes; spicules tetraxile or branched regularly, often covered entirely or at their extremities with knobs, or much divided, for the most part firmly interlocked; monaxile needles frequent; flesh-spicules often present; extend from the Silurian to the present epoch, the majority extinct; (ii) Tetractina: spicules to a great extent tetraxile; large monaxiles common; both forms frequently disposed radially; stellates and globules almost always present; Geodidae, e. g. Geodia, Pachymatisma, etc.; Ancorinidae, e. g. Stel-letta, Tetilla, Craniella; Plakinidae; Corticidae; (iii) Oligosilicina: skeleton, if present, composed of isolated stellates, canal-system belonging to type (3) or (4); Chondrosidae; Halisarcidae (=Myxospongiae), no skeleton, Halisarca, Oscarella; (iv) Pseudotetraxonia: body with a radial structure, a cortex usually well-defined, spicules for the most part monaxiles; stellates may also be present, canal-system belonging to type (4); Tethyadae, e. g. Tethya; (v) Clavulina, firm in consistence; a cortex common; a radial character sometimes visible in skeleton; spicules frequently knobbed; canal-system belonging to type (4), occasionally to (3); Polymas-tidae, e. g. Polymastia, Rinalda; Suberitidae, e. g. Poterion, Suberites; ? Clionidae.

3. Cornacuspongiae: skeleton consists either of principally monaxile spicules, which are united by more or less spongin, or of spongin fibres with or without foreign inclosures; inhabitants of the sea, of brackish or fresh water; extend from the Carboniferous to the present epoch, with two sub-orders: (i) Halichondrina: skeleton for the most part spicular; spongin often almost if not quite absent; Hali-chondridae, e. g. Reniera, Halichondria; Spongillidae, spicules smooth or spinose, no spongin, asexual gemmae, cosmopolitan and freshwater, e. g. Spongilla, Meyenia, etc, see pp. 249-53; Desmacidonidae; Ectyonidae, spiculose and hispid spongin fibres; (ii) Ceratina,=Keratosa,= Ceratospongiae: skeleton of spongin fibres, proper spicules never present (but see note 2, p. 798); foreign inclosures, e.g. sand, spicules, etc, often taken up, and present both in mesoglaea and spongin fibres; canal-system belonging to type (3) or (4); Spongelidae, Spongelia, Velinea, Psammoclema, Psam-mopemma; Spongidae, e. g.