392 Dr. J. E. Duerden on the 



as in the growth of a polyp from the larva. It is other- 

 wise, however, in fission-corals. Only the early larval polyp 

 follows the hexameral cyclic plan, and possesses two pairs of 

 directives. No wholly new individual polyps seem ever to 

 arise by fission ; growth consists in an increase in number 

 and size of organs in intimate association with those of the 

 larval polyp, and fission is simply a separation, either partial 

 or complete, of parts of this growth. It is not a separation 

 of parts arising independently; in the early stages of fission 

 the constituent organs of the original larval polyp can be 

 traced as constituents of separate stomoda^al systems. Fission 

 does not result in the production of new individuals ; it merely 

 divides partially or completely an enlarged growth, and each 

 fragment continues its growth on the same plan until fission 

 may again step in. Even where the products of fission 

 become completely or almost completely separated from one 

 another, as in Favia, Isophyllia, Mussa, and Lopholielia, 

 the isolated fragments are not new polyps in the sense in 

 which they are in gemmiferous species; they merely repre- 

 sent a part pinched off an enlarged growth. 



A new oral aperture and stomodajum with the associated 

 mesenteries cannot be considered in themselves as constituting 

 a new individual polyp. Only when parts arise de 7iovo, or, 

 at any rate, completely reproduce the main essentials of the 

 sexually produced polyp, can they be regarded as additional 

 members of a colony. 



It follows, therefore, that the entire polypal system of a 

 fissiparous coral, however large, is not made up of individual 

 polyps, but is an enlarged complex growth of the primary 

 larval polyp in which new oral apertures have been formed to 

 meet physiological needs. Morphologically, an enlarged 

 fissiparous coral admits of comparison only with a single 

 individual of a gemmiferous colony; only two pairs of direc- 

 tives occur in each. The former is irequently meandering in 

 character, or by frequent transverse growths may become 

 broken up and assume more the character of separate polyps. 

 A fissiparous coral, whatever its extent or complexity, has 

 not the morphological value of a colony in the sense of being 

 constituted of distinct individuals. 



Summari/. 

 a. The bud-polyps of gemmiferous corals arise as new indi- 

 viduals. In the course of their development they pass 

 through the same stages as larval polyps, and ultimately 

 possess all the distinctive characteristics — cyclical hexameral 

 plan and directive mesenteries — of sexually produced polyps.. 



