42(5 MEMOIRS OF THE NATIONAL ACADEMY OF SCIENCES. 



maturity thev iniijrato to the periijhfiv. iind arrange themselves in a vertical direction, jiarallel 

 with the other cellular constituents. 



A weak longitudinal ectodermal musculature seems to })e always present, the cut ends of 

 the tibrils lieing displayed in transverse sections, and most pronounced toward the proximal 

 extremity. In some species — e. g., Cladocora, Jlfufrrponi, and probably others — a distinct 

 nerve layer also occurs, situated some distance from the mesogkea (PI. II, tig. 10 /ir. /.). The 

 ectodermal gland cells and supporting cells are practically the same as in the cohunii wall, 

 but the former are less numerous. 



The tentacular mesogloja is always a very thin layer, usually smooth on both sui-faces, while 

 the endoderm is comparatively l>road with irregular internal limitations. The endoderni is 

 generally richly supplied with zooxanthelhe. but the algw are absent from PJiyllangia^ Axti-angia^ 

 and certain of the tentacles of Madrepora. In all instances a weak circular musculature is 

 developed, but the mesogl(ea is rarely folded to afford it additional support, as in the larger 

 tentacles of Actinians. The lumen is preserved, even in fully retracted tentacles. 



The tentacles of several genera present so many peculiarities of form and arrangement as to 

 call for special description. 



TENTACLES OF MADREPORA AND PORITES. 



The tentacles of the polyps of Madrepora and J'liriteK are exceptional among the genera 

 studied in that they are, with certain exceptions, oidy six or twelve in number, and in the 

 adult usually exhibit constant variations in size. The tentacles of the apical polyps of 

 Madrepora will be rirst described (tig. 1, a-, h). In the most typical instances only six tentacles 

 occur, all eipial in size, and communicating with the entocLelie ch.imbL'rs. They are widest at 

 their origin in the margin of the disk, where a consideral)le interval separates one from another, 

 and terminate either acutely or in a rounded manner. The surface is smooth throughout, no 

 urticating spots being yisil)Ic. 



Polyps with such a tentacular system are found at the apex of the long established branches of 

 colonies of both the palmate and arborescent types of growth. On polyps at the ends of short, 

 rapidly growing branches, rudiments of other tentacles also occur, alternating with the members 

 of th(> tirst order. In regions of vigorous growth, as at the margin of palmate colonies, it is 

 found that certain of the ordinary polyps, bearing the full complement of twelve tentacles, may 

 become larger and assume an apical character, an<l among these the separation of the tentacles 

 into an inner and an outer cycle can t)e recognized. Sometimes, oidy two or four of the six 

 memljers of the outer cycle will be present, always nuich smaller than the entotentacles. All 

 stages in the diminution in numl)er and size of the outer exocu-lic tentacles are, however, 

 represented, according as the polyp has recently assumed or long maintained the apical position; 

 at the same time, the six members of the inner cycle become larger and more equal. 



From all the variations observed, it is clear that on any I'olyp taking on the axial condition 

 the six exoccelic tentacles, ]iresent on all the radial polyps and smaller from the beginning, tend 

 to completely disappear, and oidy the six entocujlic members ultimately remain, becoming at the 

 same time larger and e([ual. Like the other regions of the axial polyp, the tentacles are perfectly 

 colorless, owing to the absence of zooxanthelliv, and are not often secMi fully exj)anded. 



Among the fully developed radial poh'ps of Madrep>ora twelve tentacles occur; rarely, the 

 numlier may be incresCsed to sixteen, eighteen, or as many as twenty-four. The usual forms and 

 arrangement are given on PI. I. tig. 1 {d.-n.). The separation, as regards distance from the center 

 of disk, into two alternating cycdes of six each, is not clearly detined, ])ut the members of one 

 series are always larger than those of the other. The anterior or abaxial tentacle, adjacent to 

 the nariform apex of the corallite. is longer and stouter than any of the others, and colorless, 

 except toward its origin. It may be nearly twice as long as the others, and stands out very 

 prominently; even in partly retracted polyps, when the t<'nta(des are arranged vertically, it 

 easily overtops the rest(c.). The opposite or axial tentacle — that is, the one adjacent to the stem — 

 is the next in size, but diti'ers very little, sometimes not at all, from the four large lateral 



