MEMOIRS OF THE NATIONAL ACADEMY OF SCIENCES. 537 



4. Madi'epoi'iiria with a basal pseudotheca and no " Randplatte," FJahelhim. 



None of these characters, however, can be reg'arded as of importance in the foundation of the 

 principal subdivisions of the group. Investigations on the Actiniaria, as well as those here given 

 on the Madreporaria, indicate the exact value to be assigned such details as the presence or 

 absence of directive mesenteries, while questions of sj'mmetry, unaccompanied by developmental 

 history, have very little significance. For example, the apical polyps of Madrepora exhibit 

 externally the most pei'fect radial symmeti'y, while the radial polyps from which they are derived 

 arc markedly bilateral. The presence or absence of directives is a secondary, not primary, char- 

 acter, already shown to be dependent upon the mode of asexual reproduction of the species, and 

 has therefore no fundamental significance. The marked bilateral arrangement of the parts in 

 Mitdrcpora , Pocllloponi, and SerlatojMira are a retention of larval characteristics. 



The quotations from Bourne accuratel}' represent the opinion of zoophytologists with regard 

 to the classificatorv value of the Madreporarian polyp, and little progress along such lines has 

 since been made, while much attention has been concentrated on the skeleton. 



Undoubtedly the mesenteries are the organs of greatest taxonomic importance among the 

 Anthozoa; for the tentacles and most other outgrowths which may occur are arranged in strict 

 accordance with them, and in the Madreporaria the arrangement of the septa follows most 

 directly upon that of the mesenteries. From a truly morphological standpoint all other polypal 

 structures are of subordinate value. It may, therefore, be safely accepted that so far as any 

 classification among the Madreporaria can be founded upon differences in the mesenteric system 

 it will be fundamental, and of course the same remark applies to the septal system, as this is 

 determined by the former. 



Reviewing the arrangement of the mesenteries so far disclosed within the Madreporaria a 

 perfect uniformity' occurs as far as the protocnemic stage, or stage with-only six pairs of mesen- 

 teries. It seems doubtful, however, whether anj' species of living coral invariablj' retains this 

 primarj' condition in all its mature polyps. No such group of Actinians is now known since 

 Faurot (1895) discovered four or six pairs of rudimentaiy metacnemes in Edwardsia. By far 

 the majorit}' of the adult polyps of Porites and Madrepora never get beyond the protocnemic 

 stage, but occasional!}' such examples occur. According to Moseley and Fowler, the adult poh'ps 

 of Pocillopora and Seriatopora have only six pairs of mesenteries, but Verrill and Quelch 

 mention that occasionally' twenty-four septa are present, which would imply the occurrence of 

 twelve pairs of mesenteries in the polyp. 



The protocnemic stage being probably alike in all modern corals, it is clear that any diver- 

 gences in the mesenterial plan must be looked for in the subsequent development, that is, in the 

 metacnemic succession. 



Two altogether difi'cj'ent t}'pes of metacnemic sequence and adult arrangement are now 

 known — the one in which the metacnemes appear in vinilateral (isocnemic) pairs all round the polyp, 

 and in the adult present a cycli(^ disposition, represented by the majoritj- of corals; and the other 

 with a bilateral origin and arrangement of the mesenteries throughout, as j'et definitely ascertained 

 only for the genera Purites and Madrepora. The two tj'pes have been shown to be somewhat 

 comparable with the metacnemic sequence and resulting arrangement in the Hexactinia; and 

 Cerianthese among the Actiniaria, and I propose to make of them two Madreporarian groups 

 of nearly equivalent value as follows: 



Entoenemaria. — Madreporaria in which the mesenteries always arise in bilateral pairs, and 

 beyond the protocnemic stage the increase takes place within one or both of the directive 

 entocoeles. 



Cydocnemaria. — -Madreporaria in which the mesenteries beyond the protocnemic stage arise 

 in isocnemic unilateral pairs within the primary exocceles. The mesenteries in the adult are 

 usually arranged in two or more alternating cj'cles. 



So far as our knowledge of the anatomy and development of coral polj'ps goes, the second 

 group will include the majority of recent forms, and fossil genera in which a regular multicyclic 

 disposition of the septa can l)e established, while the first will comprise P<>rlte»., Madrepora, and 

 probably certain fossil corals exhibiting a bilateral arrangement of the septa. The Entoenemaria 



