The formation of anisospores in Collozoum and Sphaerozoum contrasts with the above (i) in the grouping of 2-3 nuclei together, and the more refractile and homogeneous character of the protoplasm directly inclosing each group; (2) in the multiplication of the nuclei in each group and an increase of size in the corresponding mass of protoplasm; (3) in the differentiation of a nuclear chromatin network; and (4) the formation of a group of oil-drops in each mass, whilst the central oil-globule dwindles, the oil-drops increasing in number pari passu with the nuclei, but being ultimately dispersed in granules. The ripe anisospore is bean-shaped; it may or may not contain a minute crystalloid; whether it has a single or double flagellum is uncertain. It is of two sizes; a large, the macrospore, with a fine chromatin network in the nucleus, and a small, the microspore, one half less in all dimensions, with a coarse chromatin network. Both forms of spores occur in the same central capsule. It has been suggested that they are sexually differentiated, but copulation has not been observed.

If the suggestion is true and if the isospores are regarded as asexual, then there is Alternation of Generations. In Collosphaera and its allies a corresponding state is evidenced by the transient grouping of nuclei, which however become scattered and then differentiated into macrosporal and microsporal nuclei (supra), the two kinds however in different central capsules. Crystalloids are also always present. The changes undergone during sporulation by the extracapsular region of the body are - the disappearance of the vacuoles, the dwindling and softening of the jelly, the aggregation of the capsules in the centre of the colony, the retraction of the pseudopodia and sarcoplegma, and the sinking of the colony in the water. The capsular membrane appears to be resolved and the spores escape. Then the remains of the protoplasm break up into smaller and smaller portions which shrink into brown globules and disperse. The spores of Acantharia are small and pyriform; with several, probably three, flagella, two at one spot, and one vis-a-vis to it: a crystalloid may or may not be present1 (Brandt).

Haeckel to the vacuoles of the extracapsular region. Brandt denies the existence of a special membrane to the central jelly-spheres of Collosphaera, etc, said to be present by Hertwig and Haeckel: see pp. 59-60 of his monograph cited p. 874, note 2.

1The flagella of the spores are exceedingly difficult to see. It must remain doubtful for the present whether or no there is any instance of a single flagellum only. The life-history of the

The nuclei of the spores in Thalassicolla nucleatea, (and probably other species of the genus) are said to be derived from the nucleoli of the single nucleus (Hertwig).

The simple Radiolaria are mostly of small size, under 1/25 in. Thalas-solampe maxima (Colloidea) reaches 1/2 in. circa, but the Phaeodarian family Coelographida contains the giants of the class, several species reaching fin. and Coelothamnus maximus 11/4 in. The dimensions of the colonies vary much. The globular Myxosphaera coerulea may have a diameter of 1/5 in.; the elongated colony of Collozoum inerme is nearly 1 in. long, whilst C. pelagicium has not rarely a length of 4 in., and has been observed over 10 in., with a thickness of 2/25 in. Most Radiolaria are phosphorescent. The light centres round the intracapsular oil-globules, and is emitted readily in response to physical or chemical stimuli. With the exception of the Phaeodaria, most Radiolaria contain 'yellow cells' or symbiotic algae; see p. 243. They are either extracapsular, or in Acantharia intracapsular. They maybe present or absent in. a given species, and when present their numbers are inconstant. The extracapsular cells possess a distinct cell-membrane of cellulose, a nucleus, yellow pigment granules and granules of amylum, i. e. starch, or of an amylum-like substance. They multiply by fission within the cell-membrane, each cell giving origin to four which then escape.

If removed from their host artificially, or naturally by its death, the cell-membranes swell up and gelatinise; the cells themselves become amoeboid and multiply by binary fission. They may escape from their gelatinous coats, and form them anew repeatedly; but if brought into a plentiful supply of water, they wander out and become biflagellate. They are destroyed however during the sporulation of some colonial forms with assimilation-plasma (note 1, p. 878), e.g. Collozoum inerme. The intracapsular yellow cells of the Acantharia occur most commonly in the Acanthometra, rarely in the Acanthophracta. They are sometimes found outside the capsule, and appear to be devoid of a cell-membrane. Both forms of cells are supposed to contribute oxygen and starch to their host. Radiolarians appear to be capable of nourishing themselves also in the ordinary way by their pseudopodia, digestion taking place either superficially or sometimes even within the calymna when there is no shell. Living diatoms have been observed within a Collozoum and its allies; Myxosphaera coertdea and young Collozoa are infested parasitically by a species of Hyperia (Amphipod). Thalassicolla sanguinolenta frequently takes up coccoliths and coccospheres, etc, colonial Spumellaria is divided by Brandt into (1) the spore-stage, (2) the young vegetative stage, (3) the young reproductive stage when extracapsular bodies or their analogues are formed, (4) the old vegetative stage, and (5) the old reproductive or fructificative stage when sporulation takes place.

It may be noted that Brandt believes the peculiar nucleus of Acantho?netrida, described by Hertwig (Dk. Jen. Ges. ii. pp. 148-53), to be an internal bud, which is probably extruded, and grows into an Acanthomeira: see" p. 209 of his monograph cited p. 883. in such quantity that it becomes deformed, the soft jelly being drawn out into arms by the weight. It was in this state supposed to be a distinct genus, Myxobrachia.