496 MEMOIKS OF THE NATIONAL ACADEMY OF SCIENCES. 



conditions. 1 recognize veg-etative growth only by hiddhuj and by fission. Tiip differences 

 manifested within esich division are mainly such as are dependent upon the position and method 

 according to which the process takes place, and these do not in any way modify the essential 

 distinctions between the two types. 



When studying the mesenteries of adult polyps, two great divisions were determinable. 

 In one section, including the genera OrlnceUa^ Solenastrsea, Ocidina, Cladocora, Astrangia, 

 PhyUanqia, and S/demsf,raia, the mesenteries of all the polyps in a colony were found to be 

 arranged according to the regular, hexameral, cyclic plan, with two pairs of directives; while in 

 the other section, embracing the genera Favia, Dichocfenla, hopliylUa, Manicina, Msmndrhia, 

 and ColpojjhyUia, the mesenteries have lost their hexameral cyclic regularity, including the 

 directives, and little more than a distinction into complete and incomplete pairs can be established. 

 It was further found that the tirst-mentioned group comprises genera whose asexual growth is by 

 gemmation, while tissiparity is characteristic of the latter. In whatever position the buds are 

 produced, whether on the disk, upper part of column wall, intercalary, marginal, apical, 

 ccenosarcal, or stolonic. matters not; the polyps retain a hexameral disposition of the organs. 

 Also, whether the products of fission assume an individuality, or remain as constituents of a 

 complicated system, makes little difference as regards the irregularity of the arrangement of the 

 mesenteries, tentacles, and septa. 



This fundamental difference in the adult polyps of the two groups seems to be determined 

 by the fact that in gemmation the polyp as a whole is formed practically as a new individual, 

 whereas, in ffssiparity, some parts at least of the essential organs of the new polj'p are obtained 

 fully formed from a parent polyp. In the one case the polyp as a whole is free to develop 

 according to a deffnite plan characteristic of the species, while in the other new organs are to be 

 added and adapted to parts already formed, and tissiparity may again take place before any 

 second regularity has been established. Growth in the one is altogether new, and in the other 

 it is patchwork — some regions new, some regions old. 



it has not been possible to determine whether in every case of gemmation the mesenteries are 

 formed wholly independent of those of the parent. In some instances they certainly are. and in 

 others it seems very prol)able. In very young buds the mesenteries are already found to be 

 wholly cut off' from those of adjacent polyps, and the bud is free to develop as symmetrically 

 as any sexually-produced polyp. 



Either one or the other method of growth is in the main characteristic of any species; 

 sometimes a case of simple fissiparity may be found in a species where gemmation prevails," as in 

 Mndrepora and Porites, but the converse has never been found— that is, the production of buds 

 where ffssiparity is the rule. 



Intermediate stages are not wanting which seem to indicate how the passage from the one mode 

 of colonial growth to the other has l)een brought about. In corals like CJadocora and Ocidina 

 the buds usually arise toward the upper extremity of the column wall, and it is easy to understand 

 how gemmation may overstep, as it were, the usual boundary and occur on the discal wall. Such 

 apparently happens, for occasionally polyps of C. arhuscidd and 0. diffusa are found in which 

 two oral apertures are inclosed within one system of tentacles, and a common coliunn wall and 

 theca occur. In such cases the two polyps may be ci^ual, or one may be larger than the 

 other. ^licroscopic (>xamination of these shows that the normal hexameral regularity of one of 

 the polyps, along with the presence of two pairs of directives, has in no way been disturbed, and 

 the other polyp is either perfectly hexameral, or evidently on the way to become so. Such double 

 polyps can certainly not be regarded as fission products, at any rate not according to the plan 

 followed where ffssiparity prevails. They seem best undei'stood as discal buds, or as examples 

 of ffssiparous gemmation (see foot-note). 



It is but one step from discal budding to oral fission, or perhaps the conception may 1)6 



"The Ofca^ional instance-s of simple fit'sion in corals reproducing by gemmation have since been found to be a 

 modified form of budding, which I have termed " Fissiparous Oennnation " ; the products are altogether different 

 from those in onlinary fis:?ijiarouH growth, being cyclical, liexamerous polyps, with two pairs of diret'tives. This 

 disco\'ery greatly strengthens the seiiaration l:)etween the two groujjs of corals. " Morphology of the Madre- 

 poraria. — IV. Fissiparous Gemmation." Ann. Mag. Nat. Ilist. (Injiress. ) 



