524 MEMOIRS OF tup: NATIONAL ACADEMY OF SCIENCES. 



As already mentioned in tleseril)ino- tiie external eharacters, the ectoderm cells may contain 

 zooxanthella;, mainly restricted to the oral pole, but at times occurring sparsely throughout. 



At the aljoral extremity of all the larvw examined the ectoderm undergoes an important 

 alteration: nervous elements become developed to such a degree as to suggest that the region 

 represents a special sensory organ. The general features of the diHerentiation are much the 

 same in each species (PI. XXV, fig. 165). From the narrow mesogloea a number of delicate 

 fibrils extend parallel with one another and at right angles to the layer, and unite in a reticulum 

 which in sections seems largely made up of the cut ends of nerve fibrils. The nerve layer may 

 be very })road, and on the outer side is continued into the ectoderm cells. The latter are usually 

 more elongated, and more compactly arranged; the mucous cells are greatly diminished in inun- 

 bers, and the nematocysts and supporting cells have undergone a corresponding increase. 



The special nervous development is not restricted to the actual aboral pole, Init extends 

 some distance up the wall, gradually becoming weakei- and weaker, until ultimately, a little 

 below the middle of the larva, it is scarcely distinguishable. 



I have described the occurrence of a similar sense organ at the aboral pole of the larva of 

 Lebni/nia coraUi<jen>< (1899), and Professor McMurrich (1S'.»1) has found the same in the larva 

 of RKodacth sand i- Thorn a'. It is suggested that the organ is in some way associated wnth 

 the forward position of this end of the larva in swinmiing, and disappears when the larva 

 settles by this extreiuity. jVppellOf (1900) has found a less marked ectodermal modification at 

 the aboral pole of tiie larva of Aetaila equinu^ liut in this species no special nerve layer is 

 developed. The layer is clearer than elsewhere, and the cells are long and extraordinarily fine, 

 and some even seem to terminate in two or more fine fibrils, while on the outside a group of 

 longer, less mobile cilia occurs. Appellof observed no corresponding difl'erentiation in the aboral 

 ectoderm of Urticin/i. 



2£oKth and SUmiodxum. — When the larvsv are first extruded an oral aperture as a rule is 

 indeterminable, though a few hours afterwards a small circular opening can be made out, and 

 later the wall around may be partly depressed. Transverse and longitudinal sections through the 

 oral pole of freshly extruded individuals also indicate that for a time the mouth and stonuxheum 

 are not functionally active, and the ectoderm at the entrance to the interior often ajjpears without 

 any 1)reak. In sections through the stomoda-al tube an extremely narrow lumen occurs, })ut 

 the condition of the canal does not suggest that ciliary activity has been estal)lished, any more 

 than the nearly solid interioi- of many of the larva' would permit of the circulation of a nutrient 

 fluid. The storaodteal ciliation is not always distinguishable, yet when fully active the cilia 

 here are the strongest in the whole polyp. The deeper parts of the ectodermal epithelium at 

 this stage contain the develoi)ing stages of many nematocysts, and the nuclear zone so character- 

 istic of the .stomodieal ectoderm in adult polyps is not yet strongly differentiated. 



The ectoderm never stops short all the way round at the actual inner termination of the 

 stomoda?al tube, l)ut is partly reflected along the endodermal surface, and thence becomes 

 continuous with the mesenterial filaments, passing down the free edge of whatever mesenteries 

 are wholly complete (PI. XVIII, fig. 127). In endeavoring to establish the homology of the 

 mesenterial filaments, nuich significance has been attached' to this reflected ectoderm, and to the 

 apparent passage of the stomodieal ectoderm on to the mesenteries (p. -ITT). 



Endodenn. — Much variation exists as to the condition of the interior of the larva? when 

 the latter are newly hatched. In some instances it is filled with a highly vacuolated tissue, so that 

 the larva is a nearly solid mass of spheroidal cells; other larvte are hollow toward the middle, 

 but provided with a In-oad endodermal lining. A comparison of the figures on Pis. XV and 

 XXV will give an idea of the ditterent internal conditions which have been encountered. 



The vacuolated tissue filling the coelenteric cavity appears as if made up of distinct spheroidal 

 or polygonal cells, each with a definite boundary, and having a nucleus applied to the wall. 

 Each cell is occupied almost completely by a large vacuole, but around the walls arc granules of 

 different sizes which do not stain. The appearance of the tissue is the same throughout, in what- 

 ever direction the sections mav be made. Zooxanthella- are numerous and may be uniformly 



