152 Dr. J. E. Duerden on the 



Madrepora, Pontes, Cladocora, Sfeplianoccenia, &c., is there- 

 fore not comparable with fission as manifested in normally 

 fissiparous genera, like Municina, Favia, Afceondruiny &c. 

 Fission in tiie former gives rise to new individual polyps, 

 just as truly as in the more usual method of budding; while 

 in tiie latter fission does not produce new polyps, but merely 

 a comj)]ex multioral condition of the primary larval polyp. 

 In gemmiferous genera tlie process is not a division into 

 two of a simple hexanierous polyp, as in the first division of 

 Manicnw, &c., but the separation of the constituents of two 

 individual polyps which have grown together, one much later 

 than the other. Evidently the term fission cannot be applied 

 in the same sense to the two processes, for morphologically 

 the results are altogether different. The term is best retained 

 for the conditions met with in Manicina, Favia, and their 

 allies. 



In what manner, then, should the phenomenon of appa- 

 rently simple fission in gemmiferous colonies be conceived? 

 Though all the stages in proof thereof are not yet forthcoming, 

 a consideration of the facts presented leads to the conviction 

 that it is best understood as a specialized form of budding, and 

 the entire process may be regarded as fissiparous gemmation. 



In most species of gemmiferous corals the buds arise on 

 the polypal wall, within fairly definite limits, which are 

 characteristic of the species. Tlius, in the branching Madre- 

 pora, buds appear mainly on the cociiosarc, a little below tiie 

 apical polyps; in Forites, Stvphanoccenia, and SoJenastraa 

 they are intercalary in position — tiiat is, arise at the point of 

 union of two or more adult polyps ; in Cladocora buds are 

 developed toward the upper part of the column-wall of indi- 

 vidual polyps; while in Oculina they arise in a spiral 

 manner, close together at the apex of the branches, and become 

 wider apart below. 



Though the above distribution of the buds seems fairly 

 constant for the particular species studied, there is no reason 

 to suppose that it is invariably followed ; rather, I conceive 

 that gemmation may occasionally take place at almost any 

 part of the free polypal wall, from the disk as well as from 

 the ccluran-wall ; and, if within tiie disk, then also around 

 the oral aperture. In this last case the one mouth and 

 storccda^um would be common to the bud and the parent j 

 the mesenteries of the one would intermingle with those of 

 the ether ; two additional pairs of directives would be deve- 

 loped, as in all other buds; and the mesenteries as a whole 

 would have the same ordinal and cyclic value as in buds 

 arising on the column- wall. When the bud reaches its full 



