520 MEMOIRS OF THE NATIONAL ACADEMY OF SCIENCES. 



LABV/E AND POSTLAEVAL DEVELOPMENT. 



In all corals yet observed, the segmentation of the ovum and early stages in the development 

 of the larva take place within the internal cavity of the polyp, and are therefore not easily 

 accessible for study. The extrusion of a few eggs and semen, which sometimes occurs, appears 

 to be fortuitous in character, and is not to be regarded as the normal method of sexual reproduc- 

 tion. It is a little remarkable, that in the many sections of adult polyps which have been pre- 

 pared, none of the intermediate stages between the egg and the fully developed larva have been 

 secured, though fertile polyps, and others containing free advanced larvae, are by no means rare. 



Probably by keeping under observation, for a lengthened period, colonies which are known to 

 be fertile, it would be possible to secure the earlier embryonic stages; polyps which are charged 

 with advanced larva? often contain in addition nearly ripe eggs, from which another batch of larva? 

 might soon be expected. 



From the colonies of Manicina areolata, which H. V. Wilson had under observation for 

 a period of several months, eggs, semen, and larvae in very different stages were extruded. The 

 first specimens of coral poured forth eggs and semen on the loth and 17th of March, while on the 

 20th the polyps had given birth to larva' a little more advanced than the planula. After this date 

 only larva' were ejected, their stage of development becoming much more advanced as the 

 Bahama season progressed. 



Among Actinians also the larva? are usually ejected in an advanced stage of development; 

 but unfertilized eggs and semen are extruded from a few forms — Urtieina crassicornis, JL tridium 

 ma/rginat/un, Sagartia parasitica, and Cerianthus memhranaceus, so that different investigators 

 have found it possible to follow the segmentation and early embryonic stages of these. 



Manicina areolata is the only coral in which the segmentation and formation of the germinal 

 layers has hitherto been followed. But the early stages of development, as far as the formation 

 of the skeleton, seem so completely alike in the Madreporaria and Actiniaria that the fuller 

 details obtainable within the latter group may be considered to hold for the former. Appellof 

 (1900), in his admirable paper. "Studien fiber Actinien-Entwicklung." has followed very com- 

 pletely the early stages in the development of the Actinian Urtieina crassicornis, and at the same 

 time reviews the results of other workers throughout the Zoantharia in the light of his own. 



The following account will be restricted to a description of the larva' either upon extrusion 

 or shortly after, and as far as their subsequent development has been traced. During the 

 course of the present investigations numerous larva' of the following species of corals have 

 been obtained: Manicina <n;<>l<tfii. Fa/oia fragum, Siderastrsea radians, Porites clavaria, 

 Tsophyllia dipsacea, and Agaricia agaricites; not always, however, under circumstances in which 

 their future history could be followed. In some cases it has been possible to trace their 

 growth through various stages, but in others only the larva? themselves are available for study. 

 Some of the larva' were preserved directly in corrosive acetic, and others in forniol. The former 

 method gives the best results. During preservation mucus was often extruded from the 

 unicellular ectodermal glands, resulting in the adherence of foreign particles to the larval wall; 

 often also the larvae collapsed when transferred to the preserving medium. 



Where the superficial tissues of viviparous polyps are partly transparent, the larva' can 

 be seen moving about within the gastro-coelomic cavity, coming into view above and then 

 disappearing below. Very often they enter the tentacles, and may remain there for some time, 

 so that when tin? polyps retract they give rise to small protuberances of the surface. Whether 

 the motion within the polvpal cavity is clue to the larva's own ciliary activity, or is brought 

 about as a result of the general circulation of the internal nutrient fluid, is not readily ascertained; 

 certain larva' are provided with cilia and able to swim about immediately on extrusion, while 

 others remain motionless for a short time, showing that cilia are not yet active. 



The actual extrusion of the larva' seems always to take place suddenly, not with the slow 

 convulsive movements more usual in parturition in other groups. In Manicina <n'i<iltit<i the 

 larva' were ejected through the mouth in batches, a dozen or so at a time, by a peculiar jerking 

 motion of the adult: but in Porites clavai <ia, Wavia fragum., and Siderastrsea radians they 



