g Edmund B. Wilson 



was still everywhere separated from the entoderm by the supporting 

 lamella ^ In later stages however the filaments became perfectly con- 

 tinuous with the inner or eetodermic wall of the Oesophagus, and it was 

 owing to this faet that I failed to distinguish between the eetodermic 

 and entodermic filaments , and was led to ascribe an entodermic origin 

 to all of them . 



At Naples I have examined the development of these filaments in 

 the huds oi Alci/onium , Paralcyonium , Clavularia, Gor g onici, Kopho- 

 helemnon^ Fimiculina and Pennatula and have obtained the clearest evi- 

 dence that they are in all eases of entodermic origin. The best prepa- 

 rations are afforded by Alcyonium, which I will therefore first describe. 

 In fig. 10 we bave a longitudinal section through a bud of Alcyonium, 

 showing the first rudiment of oue of the ventral filaments at e.f. The 

 same filamentis represented more highly magnified in fig. 11. The ven- 

 tral wall of the stomodaeum is shown in longitudinal section at v.w. 

 It is composed of closely set, very long , columnar, ciliated ectoderm 

 cells, each with a small oval intensely staining nucleus. Folio wing 

 these cells towards the inner opening of the Oesophagus, we find that 

 at the beginning of the septuni they bend over the lower lip of the Oeso- 

 phagus and become continuous with the edge of the septum. The cells 

 retain their characteristic peculiarities as far as the lip , but at this 

 point undergo a decided change. In the region marked in. the cell- 

 outlines cannot be distinguished, the nuclei become pale and indistinct 

 and fiually disai)pear. Theu the characteristic entoderm nuclei suddeuly 

 appear, which differ entirely from those of the stomodaeal ectoderm. 

 They are nearly twice as large as the latter, are oiily slightly oval, 

 and stain very feebly. The outliues of the cells can only bere aud 

 there be distinguished. They are polygonal or fusiform, and as shown 

 in trans verse sections, form a thin pavenient epithelium. Altogether, 

 the contrast between the cells of the septum and of the stomodaeum is 

 unmistakeable. 



Folio wing the septum downwards we pass over a considerable 

 region in which the edge of the septum is not at all thickened and is 

 covered by ordinary entoderm cells. Just above v we find a few of 

 the elongated characteristic «gland cells«, and a little farther down, at 

 e.f., is a considerable accumulation of these cells, forming the first 

 rudiment of the filament. 



These facts render it nearly certain that the filament is simply a 



1 See Phil. Trans. 1883, PI. VIII, fig. 137, and PI. XI, fig. 159. 



