110 BULLETIN OF THE 



slightly, whereas it now begins to develop rapidly, leads to the presump. 

 tion that it has ceased to migrate, and has come to a state of comparative 

 rest, relative to the surrounding ectodermal cells. 



One of the first indications of further development is seen in the 

 arrangement of the cells of the inner layer, which is such that their 

 nuclei come to lie near the surface of a hemisphere whose convex side is 

 turned toward the ccenoccel. The beginning of this process is seen in 

 Figure 10 (Plate II.), and, further progressed, in Figure 11. Figure 14 

 exhibits a still later stage in the development of the polypide. In the 

 lower portion of the two-layered sac of this figure a separation of the 

 cells (lu. gm.) has begun. This is the first indication of the atrium. 



In all cases there exists at this stage a condition of the ectoderm like 

 that shown in Figures 5 and" 14. The absence of ectodermal cells directly 

 over the bud may be accounted for by supposing that they have come to 

 lie upon, and form part of, the neck of the polypide. While it would be 

 impossible to deny that they might migrate through the cells composing 

 the neck of the polypide, and thus come to form the nervous elements, 

 a careful study of the successive stages figured will not show the slight- 

 est evidence of any such migration, nor is it a priori probable, from what 

 is known of the action of epithelium the world over, that such a migra- 

 tion would occur. 1 



According to the description of Nitsche ('75, p. 353), there is a 

 lumen in the bud of Alcyonella (where the ectoderm is much less meta- 

 morphosed than in Cristatella), which is always in direct communication 

 with the outer world, the bud having been formed by a typical invagi- 

 nation. Braem ('88, pp. 506, 507), however, states that he has never 

 seen in Alcyonella this communication of the bud cavity with the outer 

 world. In the much more obscure process of polypide development 

 in Gymnolsemata the lumen first appears after the cells of that mass 

 from which the bud is to arise have arranged themselves in two con- 

 centric layers. In Endoprocta, according to Nitsche ('75, p. 374) and 

 Seeliger ('89, p. 179), the lumen arises by a virtual or actual invagi- 



1 Since writing the above paragraph I have cut some sections of Plumatella 

 in which this process is much clearer, owing to the absence of secreted bodies in 

 the ectoderm. Instead of a few ectodermal cells dropping down upon the upper 

 part of the neck of the polypide, as is the case in Cristatella, there is a cup-shaped 

 invagination of the ectoderm, which is quite deep, and thus gives rise to an elongated 

 " neck." That none of these ectodermal cells go to form any part of the polypide 

 proper is certain in Plumatella. But it is also true, that ectodermal cells are thus 

 incorporated into the neck of the polypide, and probably into the stolon which pro- 

 ceeds from it. 



