123 



When a polyp extends down the outside of its cup, the 

 mesentenes and niter-mesenterial spaces continue into its edge- 

 zone over the top of the corallum, dividing it up into a series 

 of spaces lying over the costae. The tentacles first appear 

 over the top edges of the primary septa, and continue to in- 

 crease with the growth of the other septa, until they have 

 attained their adult number. Thin-threaded nematocysts in 

 batteries cover all except in Paracyathiis parvulus, which has 

 the whole epithelium of the tentacles evenly set with nemato- 

 cysts. 



The Stomodoeum in all forms shows ridges over the 

 attachments of the mesenteries, the filaments of which merge 

 into them, and evidently belong likewise to the ectoderm. No 

 definite groove is present, the whole epithelium being usually 

 ciliated. The lower ends of the filaments commonly form 

 coiled masses, the acontia ; that they are ever capable of 

 being ripped off the mesenteries and shot out is doubtful, but 

 Icops can and do frequently project into the stomodoeum. The 

 mesenterial nematocysts, when present, are always of the thick- 

 threaded type found in Coenopsammia {Willeys Zoo. 

 Results, p. 357 et seq.). They perhaps kill or paralyse the prey 

 when it is taken into the stomodoeum, but their size and 

 enormous number suggest that they must have, or must have 

 had, some other function. 



The body layers in all parts of the polyps have so 

 nearly the same structure as in F labelluni rubruni 

 that their description in that form, applies to all. They 

 only vary in their minute anatomy in the presence or absence 

 and distribution of the gland cells and nematocysts. Small 

 thread-cells of the thin and thick-threaded tentacular and 

 filamentar types are found anywhere in the ectoderms of the 

 edge-zone, tentacles, peristome, stomodoeum, and filaments, 

 though the large ones of each kind are restricted to their proper 

 positions. But nematocysts also occur in the endoderm towards 

 the tops of the septa. In CaryophyUia epithecata they are not 

 so degenerate as in other forms, and clearly belong to the usual 

 filamentar type. They are evidently not now functional in any 

 way, though their arrangement and regularity is such that one 

 is led to consider that they must have been so at one time. In 

 this connection one may recall their abundance in the young 

 post-larval stages of Flabellum rubrum. Indeed, it seems not 

 unlikely that the method of development described in that 

 form may be in reality the primitive one ; the stomodoeum then 

 would be of quite secondary formation (p. 151, iialics). 



Of course, I am well aware that the view set forth above is 

 absolutely opposed to nil preconceived ideas of the origin of the 

 stomodoeum and the phylogeny of the Actiniaria. It, however, 

 serves to explain their anatomy — endodermal nematocysts, 

 ectodermal mesenterial filaments, massed filaments and true 



