Coelomate Metazoa fixed to some foreign object usually by a peduncle■, and provided with a bivalved calcareous shell. This shell is inequivalve, but each valve is equilateral (see p. 124). Its valves are either free from one another; or one, the dorsal, is hijiged on the other, the ventral. They are lined by tzuo mantle folds or extensions of the body-walls, the free margins of which are generally beset with chitinoid setae. There is a lophophore which typically surrounds the mouth and bears ciliated tentacles or ch'ri, but is in some cases produced into spirally coiled 'arms.' The digestive tract is ciliated and furnished with glands, the liver so-called. The anus is either absent, or when present, is either posterior or lateral. There are one or two pairs of nephridia. The sexes are either separate or united (?). There is a metamorphosis. Exclusively marine.

A difference of size between the two valves of the shell is well-marked in all Brachiopoda with the exception of the Lingulidae where it is but slight. The larger valve is termed ventral, or superior; the smaller, dorsal or inferior1. The outline of the valves varies in shape and is rounded, oval, or triangular, but the shell is usually narrowed towards the attachment of the peduncle or the line of the hinge, which may be pointed, curved, or straight. The shell may be convex, flattened, or concave, and the two valves may not be similar in this respect. They may be smooth, ribbed, or spinosel. One valve may have depressions or sinuses to which correspond elevations or juga on the other. In the hinged Brachiopoda or Testicardines the dorsal valve is furnished with a projecting cardinal process to which are attached the divaricator muscles. At the base of the cardinal process is a right and left socket into which fit the two teeth of the ventral valve. The latter is prolonged posteriorly into a more or less prominent beak which is generally deeply curved and grooved for the passage of the peduncle.

The groove is usually converted into a foramen by a 'deltidium,' which consists of two calcareous pieces remaining separate ('d. discretum'), e. g. in Terebratella; fusing into either a single flat plate ('d. sectans '), e.g. in Terebratirta, or one so curved as to embrace the orifice for the peduncle ('d. amplectens'), e. g. in Rhynchonella. In the Spiriferidae, etc. the groove of the beak is frequently closed by a simple calcareous plate or 'pseudo-deltidium' which grows towards the hinge line. The aperture for the peduncle consequently becomes closed, and the shell must then have been set free. The peduncular aperture is wanting in Productns, etc, and the shell is found either free, e. g. Trimerellidae, Strophonemidae, many Prodnctidae, or attached to some foreign object by the beak of the ventral valve, by its surface, e. g. Crania, Thecidinm, or by spines2. Discina among Ecardines has the ventral valve perforated centrically or excentrically for the peduncle.

1 The terms 'dorsal' and 'ventral' are applied by Von Buch and Quenstedt to the larger and smaller valves respectively, i. e. conversely to what is usual in English Manuals. The larger valve appears to have been called ventral because the chief nerve ganglion corresponds to it. In the living animal it lies uppermost except where, as in Crania, it is fixed to some foreign object. .

As to the structure of the shell. A thin external periostracum or cuticula covers its external surface and is traceable into a groove at the margin of the mantle (at least in Lingula pyramidatd) as well as over the peduncle. The bulk of the shell in Lingulidae and Discinidae consists of horny and structureless calcareous layers alternating one with another. A thin layer of horny matter extends to the mantle edge, and is said by Beyer to become continuous in Lingula pyramidata with the supporting substance of the mantle. In Crania the shell substance is almost entirely calcareous, but it is pierced by vertical canals which in the dorsal valve commence with wide apertures, bifurcate three times, and then divide near the surface into a number of minute branches, but in the ventral valve scarcely bifurcate, and are connected with an irregular network of canali-cules. In the Testicar dines a fine structureless calcareous layer underlies the cuticula3; it is followed by a layer of calcareous prisms united more or less distinctly into lamellae between which are delicate organic membranes. The axes of the prisms are oblique to the surface of the valves. It is rare for the systems of prisms to cross one another or to fuse.

A second prismatic layer may be found in old specimens internal to the first. The substance of the valves is sometimes perforated, but not in Rhynchonellidae, by vertical canals occasionally branched, which widen at their outer extremities beneath the cuticula, e. g. in Terebratida, Terebratella, Waldheimia, Spirifer. The canals in question contain hollow or solid processes (van Bemmelen) of the mantle l. The superficial contour lines parallel to the margins of the valves probably represent periods of little or no growth. The calcareous matter is in the form of Calcite except in Lingula, where it consists chiefly of Calcium Phosphate.

1A living spinose Rhynchonella (R. Doderleinit) has been recently dredged off the coast of Japan. See Davidson, A. N. H. (5), xvii. 1886, who states that spines • formerly prevailed among Palaeozoic Productidae, Orthidae, etc, and the Oolitic Spiriferidae and Rhynchonellidae^ and that 'no spinose Brachiopoda are known from the Cretaceous or Tertiary period.'

2 The presence and extent of a flattened triangular 'Area' horizontally and vertically striated; and situated on the dorsal, i. e. inferior aspect of the beak of the ventral valve, and sometimes extending on to the dorsal valve too, is to be noted as a point of classificatory importance.

3 Shipley states that in Argiope the calcareous matter of this layer increases and that the outer ends of the prisms become square and are packed together like bricks.