Morphology of the Madreporaria. 153 



size, it will tend to separate from its parent, and in so doing 

 it will appear as if an enlarged polyp were undergoing fission 

 into equal halves, whereas, strictly speaking, it is the 

 components of the bud-polyp separating from the parent 

 body. 



In such a specialized form of gemmation, the constituent 

 mesenteries of the bud and parent polyp may be so arranged 

 that division actually involves half of each polyp — that is, 

 lialf is new formation, and half is old. Such is manifestly 

 what happens in Porite.t and Madrejyora, where the plane of 

 separation is along the primary directive axis ; but in Clado- 

 coi-a, Siderastrcea, &c. the actual distribution of the new and 

 old mesenteries in the two fission-polyps has not been 

 determined. 



Fiir. 7. 



Transverse section through a bioral polyp of Cladocora arbuscula, below 

 the stomodseal region. The polyp to the left is as yet incompletely 

 developed, and may represent a bud which has developed indepen- 

 dently on the disk of another polyp, or a stage in fission beyond that 

 shown in fig. 3. 



In no other way than as discal gemmation it seems to 

 me can one account for the occurrence of two oral apertures, 

 two stomod^ea, and two Hexaclinian systems of mesenteries 

 ■within a single system of tentacles and a single column-wall. 

 In discal gemmation the division of the stonioda^um has not 

 the same significance as in true fissiparity. It is rather a 

 separation of two distinct stcmoda^a, one belonging to the 

 parent and one to the bud; wlureas, in true fission, it is the 

 division of an enlarged stonio(Jc€um into halves, and neither 

 represents a distinct individual. 



Some of the enlarged bioral polyps of Cladocora arbuscula 

 seem to afford evidence that discal gemmation may occur 



