Morphology of the Madrejjoraria. 391 



Mam'cina. For the present purpose the fragment figured 

 will serve for practically the whole poljpal system of a brain- 

 coral, however large. No directives occur, the plane of 

 fission between one poly pal system and another is always 

 entocoelic, and the mesenterial pairs usually extend as far as 

 the stomodseum, though sometimes incomplete pairs are 

 found *. 



On almost any colony belonging to the genera first men- 

 tioned as reproducing by gemmation one or more enlarged 

 polyps may frequently be found exhibiting one stage or 

 another in simple fission. These exceptional cases of fission 

 on gemmiferous colonies will be dealt with in a subsequent 

 paper. It is found that the process differs wholly from that 

 above described for Manicina. The resulting polyps are 

 altogether like ordinary gemmation-polyps in that they retain 

 the hexameral cyclical plan and possess two pairs of directives. 

 I endeavour to show that they are best regarded as repre- 

 senting a specialized form of gemmation, not true fission as 

 here understood. 



Other than these there seems to be no combination in a 

 species as regards either budding or fission, though such 

 might be reasonably expected. I have come upon no 

 instance of true fission in corals of which the normal method 

 of increase is by gemmation, and, vice versd, no example 

 of budding in polyps which enlarge by fission. Frequently 

 in fissiparous colonies fragments of the polypal system are 

 found wholly isolated from the main mass and appear as if 

 they might have arisen as buds from the wall of the parent ; 

 but in sections they exhibit all the characteristics of fission- 

 fragments — that is, they are devoid of directives, and 

 ■r present no hexameral or other cyclical regularity as regards 

 the mesenteries. 



We are now in a position to understand the true significance 

 of gemmation and fission as they occur in corals. The com- 

 plete correspondence as regards the number, character, and 

 arrangement of the tentacles, mesenteries, and septa of bud- 

 polyps with those of polyps reared directly from larvaj shows 

 that the former are to be regarded as altogether new indi- 

 viduals. The parts appear to be formed, as it were, de novo, 



* Verrill (1901, pp. 65, 68, 71) refers to cases of apparent exotliecal 

 budding on the collinal ridges of Mmtndra {Mceandrina). It would be 

 of grejTt interest to determine the actual mode of formation of tlieso as 

 concerns the polyps. According to the view presented in this paper, 

 budding would be established only if the polyps were found to possess 

 directives and the hexameral cyclical plan ;_ the absence of directives and 

 an irretrular cyclical plan would denote fission. 



29*- 



