436 MEMOIRS OF THE NATIONAL ACADEMY OF SCIENCES. 



A goniilial groove at oacli end of tlie stomodfeuiii is, with certain exceptions, present in all 

 Hexactiniw; and a single groove occurs in the Zoanthea' and Ceriantheaj. In the Zoantheie the 

 organ is ventral or posterior, while, according to Carlgren (l.S!»3, p. 243), it is dorsal or anterior 

 in the Ceriantheie. A ventral groove, first termed by Professor Hickson (1883) the Siphonoglyphe, 

 is likewise found in nearly all Alcyonaria. It is, therefore, a little remarkable to find that such a 

 typically Anthozoan organ has never been established for the Madreporaria, and it is absent from 

 each of the twenty-six species here .studied". Its non-development is probably indicative of 

 the more primitive character of coral polyps generally comi)ared with most Actiniaria. 



The suggestion may be ottered that the grooves, alread}^ described as occurring all the way 

 round the stomodaunn in some species of Madreporaria, are to be regarded as the morphological 

 and physiological equivalents of the two axial grcjoves in the Hexactinian polyps. Instead of 

 a groove occurring only between each pair of directives, one is found between all the complete 

 mesenteries. The same histological differences are found in each case, though not so pronounced 

 in corals. No experiments have been made to determine whether the grooves in the ]\Iadre- 

 poraria have any special function in directing the inhalent and exhalent currents, and with such 

 small oral apertures experiments of this character would be difficult to conduct. 



In living polyps of Cladocora arhimcula, Solt^iastrcea hyadeJi, and others, the lateral por- 

 tions of the lips and .stomodwal walls have at times been observed to come into close contact, 

 leaving a small aperture at each (>xtremity of the mouth, through which currents of water enter 

 or leave the gastric cavity. A similar approximation has also been recorded ))y ditt'ercnt 

 observers as occurring among the Aetinise, but is there associated with the presence of gonidial 

 grooves. In the Zoanthidw, provided with only one gonidial groove, only one terminal apei-ture 

 remains when the lips are approximated. 



The inner stomodwal extremit\' may become reflected upwardly and outwardly, so that in 

 transverse sections the stomodtval walls are cut through twice; or, if they are much folded in 

 addition, they may appear several times in succession in the same section. The appearance of 

 the reflection in longitudinal section is shown on PI. VII, ffg. 50, and in transverse section on 

 PI. VI, flg. 51. The stomodwum terminates internally at practically the same level all the way 

 round, or the two axial extremities, with the directives attach(^d. may extend a little below the 

 lateral walls, but nothing comparable with the "Languettes" of Actinians has been oliserved. 



Upon complete retraction of the polyp, the distal parts of the polypal tissues — upper column 

 waH and disk — mostly come to rest upon the skeletal projections — septa, pali, columella; in 

 consequence of which the stomodaHjm becomes flattened and more or less irregularly folded. 

 As a result it is often with difficulty that transverse sections of the stomodiBum, exhibiting the 

 relations of the mesenteries, can be ol)tained, especially as the organ is comparatively short. 



On retraction the stomoda-al walls as a whole are sometimes thrown into a few deep vertical 

 folds, which assume a symmetrical ffgure. This is especially the case in Porites (figs. 28, 30); the 

 folds may be four or six in number, and approach so as to touch one another in the middle, 

 practically obliterating the lumen. The stomodiual foldings of a bud of Cladocora likewise 

 assumed a regular arrangement (PI. VIII, fig. 60); Fowler (1888) also describes and figures a 

 similar appearance in a transverse section of the stomodanim of Scriafopora ><iihhtfa. No doubt 

 it is a consequence of the strong contraction of the circular endodermal muscle. 



The histological details of the stomodffial wall are practically alike in all Madrepoi'arian 

 polyps, and agree closely with those of the Actiniaria. The ectoderm is always a broad, strongly 

 ciliated layer, comprised largely of supporting cells, the nuclei of which are closely arranged, 

 and give rise in sections to a characteristic, brightly-staining zone. The ciliation is luiiform 

 throughout, and is nearly always persistent in preserved luaterial. As a rule nematocysts of 

 two or three kinds occur, while both cknir and granular gland cells are numerous, 'i'he latter 

 arc particularly abundant in Chidm-ora (fig. 52/-). In some cases, e. g., PlnjIIaiujia, distinct 

 ectodermal nervous and muscular elements can be made out near the mesoglceal surface, but are 



"Saville Kent refers to a siphonoglyp in a Barrier Reef Fuiigia, and Bourne adds one to his diajiraininatic- fiirure 

 of a coral, on p. 62 of his article Anthozoa (1900). 



