Of these various modes, (1) and (2) are often termed simply 'fission,' and (3) 'calycular gemmation,' very common in the extinct Rugosa. Fission may be quite imperfect; e. g. Dana states that the long lines of the Maeandrine corals are due to the lengthening out of the peristome and the formation of a series of mouths, the elongated and many-mouthed disc being fringed by a line of tentacles on either side. (4) Is termed by von Koch 'Zwischenknospung,' and does not occur in any living corah (5) Is a very usual form, and is commonly spoken of as 'parietal gemmation.' As to (6) it is, strictly speaking, only a mode of (5), inasmuch as every stolon contains coenosarcal tubes. There can be no doubt that the stolons of extinct Anthozoa contained similar extensions of the coelenteron. According to Martin Duncan, the epithecal roots of the living Rhizotrochus are hollow and open into the calycle.

2The Fungia stock is probably never branched, but two or three stocks may grow close upon one another. Stutchbury states that the pedicle of the stock becomes bare of soft tissues when the terminal part is detached. The mode in which the second zooid buds from the stock is not known. Semper mentions in his paper (Z. W. Z. xxii.) several interesting facts relative to the asexual multiplication of Madrepores. Blastotrochus nutrix is simple, properly speaking, but from the bare sides of the theca originate buds which drop off. The pedicle of the bud left in situ forms

Development is only known, and that imperfectly, in some Hexacti-nians, in Cereanthus, in the Madreporian Caryophyllia and Astroides1. It may take place externally to the parent as in Cereactis (= Actinia) aurantiaca, but more generally within it, and then it is probable that the ovum is fertilised in the ovary. An invaginate Gastrula is found in Actinia (sp. ?) and Cereanthus, and probably in Caryophyllia. In other instances (Adamsia Rondeletii - Actinia (Sagartia) parasitica; Astroides) the endo-derm probably originates by delamination from a central mass of cells as in Alcyonaria. The ciliated larva is usually free-swimming and elongate, sometimes with a long posterior bunch of cilia; but in some cases (Cereactis aurantiaca and occasionally in Actinia equina - A. mesembryanthemum) it quits the parent as a young Actinian. It usually assumes the Actinian condition while in the free-swimming stage. A single tentacle, corresponding to one end of the mouth, may appear first of all as in Actinia equina; or there may be two, one at each end of the mouth, as in Heliactis (Sagartia) bellis, but the number may be greater, e. g. four in Cereanthus. The twelve primary mesenteries in Hexactiniae are probably developed in the following order: - (i) a pair at each end of the mouth, but one before the other, with retractor muscles turned away from one another; (2) a second pair, the future directive mesenteries, with retractor muscles similarly disposed in the intraseptal spaces of (1); (3) a pair on the right and left sides of the mouth with muscles also similarly disposed.

The members of these two lateral pairs, however, become coordinated with the members of (1), to form two right and left pairs with retractor muscles turned towards one another. The remaining mesenteries arise in pairs in the interseptal spaces 2.

The Actiniaria are found in all seas and at depths not greater than 2900 fathoms: and the deep sea yields a distinct fauna in which a tendency to disappearance of the tentacles and to variations in structure and arrangement of the mesenteries is observable (R. Hertwig). Some of the Madre-poraria descend to great depths, e. g. Bathyactis symmetrica from 70-2900 another bud, but if it drops off a scar remains, laying bare the coelenteric cavity of the coral. Buds may develope from the sides of the scar. He has also shown that a pedunculate and fixed Flabellum, his Fl. variabile, may detach the upper part of its body. The part set free is the Fl. Stokesii or Fl. Oweni (according to age) of Milne Edwards: the attached part, the Fl. spinosum, and the original parent form the Fl. aculeatum of the same author. Semper mentions that in the simple Fungid genus Diaseris the coral has several mouths; that the lobes of the corallum may break away to form new zooids, new lobes being developed after such a separation.

He observed a Fungia when reversed, i. e. turned topsy-turvy, forming a number of zooid mouths around the edges of its base.

Semper regards the budding of the Ftmgza-stock and the pedicle in Blastotrochus as instances of Alternation of Generations, the stock and the pedicle being asexual. Nothing, however, is definitively known as to the ultimate fate of the structures in question. fathoms. With one or two exceptions the deep sea forms are widely-distributed and most of them occur also in shallow water. Leptopenas, however, is an instance of a coral limited to the deep sea (1500-2250 fathoms). The majority form the well-known coral reefs. These structures, which are absent from the west coasts of America and Africa, are for the most part confined within the limits of 300 North and South of the equator, where the mean temperature of the sea does not fall below 68° R, growing in water of, as a rule, not greater depth than twenty fathoms and for the most part less, different genera and species flourishing at different depths. The coral banks may follow the coast line of the land either closely as 'fringing reefs'; at a distance, sometimes great, as 'barrier reefs'; or they inclose spaces of water, the 'lagoons' and then constitute'atolls.' They are formed in the first instance on sub-marine banks and elevations mostly of volcanic origin, combined with deposits of mud and calcareous matter; and their subsequent extension depends on various factors, 'the temperature, solvent power, currents, tides, and waves of the sea;' on 'the amount and direction of the supply of pelagic food, the up-building of calcareous deposits to the zone of reef-builders, the outward vigorous growth of the coral-masses, and their death and decay' forming a fresh extension of ground' and the solution of their skeletons in the inner parts of the reefs' (Geikie). Neither subsidence nor upheaval of the land are necessary factors, though both may occur, especially the latter.