THE ANATOMY OF THE MADREPORARIA. 249 



both this fact and the absence of mesenterial filaments on ten of the 

 mesenteries would seem to indicate a degeneration, of which I hope to 

 bring forward a second instance in a future paper. Between the 

 polyps runs a similar canal system to that already described by v. Koch 

 in Stylophora. The anatomy of Pocillopora, so far as mentioned, 

 appears to agree in all respects with that of Seriatopora, and the polyps 

 exhibit the same marked orientation. 



Moseley has also described the microscopic anatomy of three 

 other Madreporarian polyps (II). His observations on Fldbellum are 

 mostly incorporated with my own below, and need not, therefore, be 

 recapitulated here ; and of Stephanophyllia I hope to give a detailed 

 description in a future paper. 



Of Bathyactis, which is planoconvex in shape, the plane being the 

 basal surface, he records that on decalcification a lamina of ectoderm 

 and mesoderm separates off from the base. This fact, together with 

 its shape, suggests that the original basal ectoderm of the embryo 

 persists in this species throughout life, in its primitive position, except 

 for such part as grows up with the skeleton (the calycoblasts). 



To sum up the undoubted facts elucidated by these observers : — 



1. The adult Madreporarian polyp is built distinctly on the Actinian 

 type, except for the absence of an external body-wall in some cases 

 (Caryophyllia, Cladocora), which is then replaced physiologically by 

 the imperforate theca. 



2. The coral lum is a product of the ectoderm, and deposited outside 

 the embryo. 



This ectoderm persists in the adult as the layer of calycoblasts, to 

 which the continual growth of the corallum is attributable ; thus the 

 skeleton is morphologically external to the polyp throughout life. 



4. Between this layer and the cavity of the ccelenteron, and clothing 

 every part of the skeleton, is a layer of mesoderm and endoderm, 

 forming the internal body-Avail. 



5. Septa, when present, always lie between a pair of mesenteries 

 (entosepta), sometimes also in the spaces intermediate between pairs of 

 mesenteries (exosepta). 



6. Tentacles may be exoccelic as well as entoccelic, but exosepta may 

 be present without corresponding tentacles. 



The present classification of the Madreporaria is admittedly 

 unscientific. I have therefore laid stress on what may perhaps seem 

 the trivial point of the relations of septa and tentacles to the mesen- 



