The Peridinidae are rarely naked (Gymnodinium, all species(?), Hemi-diniutn). The cuticle is seldom structureless (Glenodinium), but is usually broken up into a number of plates arranged as a rule in a definite manner on each side of the transverse furrow1. They vary however in number size, etc, in different genera. An apical set crowns the anterior pole of the body, i. e. that end which is usually directed forwards in locomotion. They are generally prolonged into a conical tube open at its extremity. An ant-apical set crowns similarly the opposite pole; whilst two circles of plates, a prae- and post-equatorial lie respectively in front of and behind the equatorial furrow, which is usually placed near the centre of the body, rarely anteriorly. When the furrow is absent the arrangement of the plates marks the equatorial line. The furrows are lined by a delicate cuticle, and in Ceratiutn the longitudinal usually expands into a sub-triangular area. In the same genus the posterior pole is produced into a horn, sometimes recurved; and one or more horns may proceed from the right and left post-equatorial plates. They contain prolongations of the body. The aperture for the longitudinal flagellum is a slit. The second family, the Dinophysidae, differs from the foregoing in several respects.

The body is elongate, and sometimes remarkably so (Amphisolenia). The transverse furrow which is circular, is generally approximated to the anterior pole, or almost terminal, and the longitudinal is very short2. The aperture for the longitudinal flagellum is round and somewhat tubular. The cuticle is wanting in Amphidinium alone; it constitutes a bivalved shell, and its valves are right and left. The margins of the furrows tend to become produced into lamellae. The lamellae of the transverse furrow are directed forwards so as to form two cones one within the other; the anterior is the cephalic, the posterior the nuchal. The right margin of the longitudinal furrow is but slightly produced; its left greatly, and in some instances into broad lamellae, which may be supported by three, rarely more, rib-like thickenings. Cuticular lamellae may also be produced along the junction of the valves of the shell posteriorly. The Polydinidae, or third family of Dinifera, has but one genus, Polykrikos. It is barrellike in shape, compressed dorso-ventrally, and devoid of cuticle. The body is encircled by eight somewhat spiral transverse furrows, each lodging a transverse flagellum, and all connected by a longitudinal furrow.

The longitudinal flagellum is typically single, and situated a little in front of the posterior end of the body. A second terminal and posterior flagellum is sometimes present (Biitschli).

1The structureless cuticle of Glenodinium is said to resolve itself into a number of plates under certain conditions.

2In the Peridinidans Oxytoxum and Ceratocorys a similar approximation is found, unless Stein's view is right, and the furrow is posterior in these two genera.

The cuticle is brittle, colourless and hyaline, and at first structureless. It generally developes delicate ridges, inclosing in Dinophysidae areolae, with thicker ridges in the Peridinidae along the line of union of the plates. Growth of the plates in the last-named family, and of the valves of Dinophysidae, is supposed to take place at the edges which are often transversely striated 1. In Dinophysidae each areola is perforated by a pore, and pores are probably present also in Peridinidae. The protoplasm has a clear layer of exoplasm in some naked forms, but probably not in other instances. It may be reddish in hue when chromatophores are absent. These last-named structures are as a rule numerous, small and placed peripherally. A single and reticulate chromatophore liable to break up occurs in Ceratium; a single central one with out-runners in Pyrophacus, etc, and two thin plate-like bodies in Extiviaella. Their colour varies from brown to various tints of green, according as diatomin or chlorophyl predominate in them. Unmixed chlorophyl occurs in the marine Proto-peridinium viride. Colourless genera occur, e. g. Polykrikos, colourless species in other genera, as well as colourless examples of normally coloured species.

Starch-granules, said to be lamellate in Ceratium, are present even in colourless individuals. Oil drops, yellow or reddish, are found especially in individuals about to pass into a resting condition, or actually in that state. A drop of large size and tinted red with haematochrome, the so-called eye-spot or stigma, is not infrequently present in freshwater species beneath the longitudinal furrow. A black mass of pigment, with an apposed transparent refractile body, supposed to be an eye, has been observed in a marine species of Gymnodinium by Pouchet. The nucleus is globular, ovate or band-like; it has a reticulate structure, and is single except in Polykrikos, which has four, one behind the other. Each of these four nuclei is said to have 3-6 small paranuclei apposed to it, as in some Infusoria. The protoplasm is sometimes vacuolate; and one or two large vacuoles are common, whether contractile or not is uncertain1. Polykrikos has nematocysts similar in structure to those of Coelenterata, but supposed by some authorities to be taken in with food. Some and probably all coloured Dinoflagellates are holophytic.

Solid foreign bodies have been observed in naked forms, and in two instances even their actual ingestion 2.

1 The old plates, according to Joseph's account of Peridinium stygium, may remain overlying a new cuticle (Z. A. ii. p. 116). A lamination of the old cuticle has been noted in an old Ceratium tripos.

Reproduction takes place by binary fission, longitudinal in Exuviaella and Dinophysidae, oblique in most Peridinidae, but transverse in Hemi-dinium and Polykrikos. The process rarely occurs except in Ceratium and Polykrikos, while the animal is in a motile condition. It usually loses its flagella, the protoplasm contracts from the cuticle and very generally forms a cyst of a cellulose-like substance (?) within the cuticle, which is detached by the swelling of the cyst before or after fission has been accomplished. The cyst is in some instances gelatinous. In Peridinium membranous semilunar cysts may be produced in which four, or in a marine species eight, individuals are produced fissiparously. These cysts appear to be attached temporarily by one horn3. The chromatophores collect together and generally undergo resolution, the organism acquiring a prevailing red tint. Fission is sometimes incomplete, resulting in double individuals. It is uncertain whether conjugation ever takes place. No such significance can be assigned to the chains of two or more individuals observed in Ceratium and in one or two other instances; they are probably formed, at least in the genus named, by fission (Bergh).