Mr. G. C. Bourne informs me that in species are in the young condition very closely alike, and sometimes indistinguishable;' and 'in the very early stages the young of even widely different species of the same genus are almost absolutely alike' (Moseley), facts especially noted in the genus Flabellum. Tabulae occur in Pocillopo-ridae and a few Poritidae.

As to the zooids they are, in the majority of instances properly examined, Hexactinian in structure. The tentacles do not appear to be perforated by an apical pore. They may correspond to the intraseptal chambers alone, e. g. in Rhodopsanimia parallela, or to the interseptal as well, the more usual arrangement. They are simple in form and often armed with batteries of aggregated nematocysts. The mesenteries may be only twelve in number, all complete (Stylop/wra, Madrepora), or, when the calycles are large or the animal solitary, much more numerous and some complete, others incomplete. They are disposed in the typical Hexactinian mode. The septa are contained only in the intraseptal chambers, or in the interseptal as well. Septal stomata have not been described nor Rotteken's sphincter muscle. Sexual organs may be present on all the mesenteries, or confined to those of a certain order, or, as in Madrepora Durvillei, to two specially elongated mesenteries. The sexes appear generally to be separate even in the colonies.

In the coral last named a remarkable dimorphism occurs between the zooids in respect of the mesenteries1.

Seriatopora differs in several respects from typical Madreporaria. Its calycles are elongated and disposed lengthwise on the branches of the coral, and are traversed by a longitudinal plate. There are twelve tentacles in two cycles, and twelve mesenteries, two of which, towards the basal or ventral end of the calycle, are of great length, and alone have mesenterial filaments and genital organs. The chambers to which they belong are greatly prolonged and received into deep pits in the calycle. There are three complete septa and two rudimentary on each side. The form and arrangement of the mesenteries and septa bring about a marked difference between the two ends of the zooids or calycles. The colony is unisexual. Pocillopora closely resembles Seriatopora.

The zooids in a colony are either isolated, e. g. Cladocora, or where there is a coenosarc, they are connected by a superficial canal system as in Stylophora, Seriatopora and Pocillipora, by a deep system as well in Madreporaria Perforata, or by their gastric cavities, as in Coeloria, Maeandrina, etc, corals in which they are disposed in linear series.

Fungia the synapticulae or plates connecting the septa commence as points on the opposing surfaces of two adjoining septa, which grow, meet and fuse, perforating the intervening mesentery in the process. The edge of the flattened theca grows similarly; an extrathecal space underlies it, divided into radial chambers by the aboral continuations of the mesenteries and connected to the intrathecal space peripherally as well as by canals passing through the theca itself.

1Fowler states (Q. J. M. xxv. p. 588) that there are acontia (?) and peristomial cinclides in Flabellum patagonicum. Mr. G. C. Bourne tells me that he has observed the protrusion of much convoluted mesenteries from peristomial cinclides in Maeandrina.

Asexual reproduction has been observed in few of the non-colonial Actiniaria. Part of the limbus may separate off in some Hexactiniae and develope into a new zooid: such a process is known as scissiparity (p. 241). Fission occurs commonly in Ammonia sulcata ( = Anthea Cereus), and may be the cause of the occurrence of twin individuals in Actiniloba dianthus, etc. The colonial forms increase in various ways which have been classified by von Koch under the two heads of internal and external gemmation. In the former, the young calycle originates with the parental, and is derived partly or entirely from it: in the latter, it is external to it altogether. Internal gemmation occurs (I) as fission-budding, in which the original calycle is constricted into two parts, e. g. in Mussa; (2) as septal budding, seen in the Silurian Stauria and its allies, where the so-called primary septa become in part the walls of the young calycles; (3) tabtdar budding, in which the young calyx is produced from the parental, but is completed by the development of a pocket-shaped floor or wall.

As to external gemmation, the bud is either formed (4) from the wall of the parental theca, and adheres closely to it as in Favosites, which is probably, however, an Alcyonarian; (5) in coenosarcal budding from the tube of the coenosarc, e. g. Madrepora, or of the coenenchyma (?), e. g. Rhodopsammia; or (6) in stoloniferous budding, from a stolon, as occurs in some extinct operculate Corals1 . Individuals are sometimes found with a new but smaller calycle symmetrically formed within an old calycle with which it corresponds completely. This is termed 'rejuvenescence' by von Koch. In Fungia there is a small fixed stock simple (or branched ?) from which the apical zooid-portion is detached, a new zooid arising by gemmation from the pedicle. The process may be repeated 3-4 times. The detached zooid is 11/6 in. in diameter and shows a scar of detachment which disappears during subsequent growth (Moseley)2.