MEMOIRS OF THE NATIONAL ACADE^NIY OF SCIENCES. 479 



mesentery i>^ full}' formed (tig. 57 c). If the filaments of the incomplete mesenteries originate 

 from a reflected tract of stomoda^al ectoderm, it might reasonably be expected that they would be 

 l)est developed in the uppermost region of the polyp, whereas, as a matter of fact, they are here 

 absent or only incipient, even when fully developed below. 



All the evidence seems to favor the view that in the Madreporai'ia the mesenterial filaments 

 first appear independently of any connection with the ectodermal lining of the stomodajum, but 

 that in the case of the complete mesenteries such a continuity is early established, while with 

 incomplete mesenteries the separation is permanent. 



When describing the adult mesenteries, it is shown that the first part of the filament differs 

 in form and structure from that below, and histologically is indistinguishal)le from the stomodajal 

 ectoderm, with which it is in direct continuity. It is manifest, therefore, that between the 

 actual stomoda?al ternnnation and the conunencement of the true filament there is a tract which 

 partakes more of the nature of the stomoda?al ectoderm than of the filamental tissue. It forms 

 the connecting link, as it were, between two tissues which may be considered as wholh- distinct 

 both phylogenetically and ontogenetically. Such, it may be conceived, is the significance of the 

 "reflected ectoderm" as seen in corals. It is the stomodwal ectoderm passing along the mesenter\' 

 to establish structural continuity with the upwardly growing fllament. 



The strongh" ciliated charai'ter of the stomodseal ectoderm marks out the layer as specially 

 concerned in the circulation phenomena of the polyp, and the same must be affirmed to a less degree 

 of the upper part of the mesenterial fllament. For the proper carrying out of this it is manifest 

 that a close histological continuity should be maintained between the two structures, and it can 

 be conceived that in estal)lishing this the stomoda?al ectoderm passes some distance down or 

 along the mesenterial edge to meet the upgi-owing filament proper. 



The mesenterial fllaments on the incomplete mesenteries are the homologue of those on the 

 complete filament, but the latter are in more or less direct histological contiiuiity with the 

 stomodfeal ectoderm by a downgrowth from the latter, while the others are free throughout. 



In his recent paper (1900, p. 73), Appellof contends for the ectodermal origin of the whole 

 Actinian filament, his fig. 2.5 showing a distinct passage of the stomoda?al ectoderm down the free 

 edge of the mesentery. The occurrence of an interval of endoderm between the stomodteal 

 ectoderm and early filament, such as McMurrich (1891) and I (1899) have found in other larvse, 

 Appellof would explain as the result of a more or less accidental severance of continuity upon 

 retraction of the larva". Such a suggestion would scarcely be applicable to the conditions 

 already described in the larvaj of Favia, for the first indications of the filamental tissue occur at 

 very difl'erent levels on the various mesenteries, in some instances at a considerable distance 

 below the stomoda3al termination. On the view presented above, the appearances which Appellof 

 desci'ibes in Urtichm are not dii*ectly concerned with the formation of the filament; the downward 

 growth of the ectoderm from the stomoda'al wall is merely the means of establishing continuity 

 with the true filament which will appear independently below. 



BASAL DISK, SKELETOTROPHIC OR SKELETOGENIC TISSUES. 



Under these terms will ))e included the three Coelenterate layers — ectoderm, mesogla?a, and 

 endoderm — which everywhere cover the surface of the skeleton in the living portion of a coral. 

 G. von Koch (1SS2) has fully demonstrated that the ectodenn alone is the true skeletogenic laj'er, 

 and is actually adherent to the corallum; l>ut it is convenient to study along with it the associated 

 niesoglcea and endoderm. The early stages in the growth of the young polyji after fixation 

 (PI. XIX) reveal that the basal disk onl}' is concerned in the formation of the skeleton, and 

 therefore all the subse^juent foldings, invaginations, and evaginatioiis of tiie skeletotrophic 

 tissues are but so many extensions of this region of the polyp, produced pari jkikku with the 

 deposition of the calcareous particles. The theca, septa, costie, columella, and all the teeth, 

 spines, etc., connected with them, represent so many foldings of the basal disk, for all take their 

 origin from the same continuous layer, and their surface remains covered by it so long as they 

 belong to the living parts of the polyp. 



