REPORT ON CORALS — HYDROCORALLIN^E. 77 



iu Anthozoans, than does the very curious simulation which exists in its coenosteum. 

 The resemblance is, however, in both instances merely superficial, and of no genetic 

 significance. 



The branched and fringed processes of endoderm described as embracing the embryos 

 in Errina and Stylaster appear to correspond with the similarly brauched structures 

 in Cordylophora lacustris, described and figured by Allman and F. E. Schulze. 1 

 I have described them as outgrowths of the spadix, but possibly the cup-shaped 

 endodermal structure supporting the ova should not be so designated. 



The endoderm of the Stylasteridse is always coloured, and seems most frequently 

 to assume various shades of red or violet coloration, but in Astylus subviridis it is 

 green. The coenosteum itself is in some species coloured, especially, it would appear, in 

 Distichopora, but no doubt in many instances the coloration ascribed to the calcareous 

 structures is in reality due to endoderm dried up within the interstices of the corallum. 



In a former paper I conjectured that possibly shallow water Stylasteridas might bear 

 free gonophores, and perhaps medusiform ones, and that the occlusion of the gonophores 

 within calcareous structures, and their adelocodonic condition, was due to the fact that 

 the forms examined lived in the deep sea. This suggestion was in accordance with the 

 observations of Allman, who has found fixed sporosacs in all deep-sea Hydroids 

 examined by him. 2 I find, however, from specimens sent me by Count de Pourtales, 

 that ampullae are especially well developed on the shallow water Stylaster roseus ; those 

 in the female stocks being very large and prominent. There can, therefore, be little 

 doubt that these structures occur throughout the family. 



In all the Stylasteridae in which the gastropores have styles, the gastrozooids must 

 be protrusible in the expanded condition to a very slight extent. And the fact that in 

 some genera the gastrozooids lose their tentacles seems to bear out this supposition. 

 No doubt in active life the dactylozooids extend like long and filiform tentacles and 

 catch and convey food to the gastrozooid, which nourishes them in return by means 

 of its basal canals and the general circulation. It is to be noted that in those genera 

 in which the gastrozooids have no tentacles, tentacles are wanting in the entire stock. 



The nariform and tubular projections of Errina are no doubt contrivances for 

 extending the reach of the dactylozooids, whilst at the same time protecting them. 

 In Acanthopora the bases of the dactylozooids are pushed out to a remarkable dis- 

 tance from the gastropore mouths, and subsidiary dactylozooids of a smaller kind 

 seem to be necessary to ensure the conveyance of food to the gastrozooid. G. 0. Sars, 

 who is the only naturalist who has observed a Stylasterid alive, never saw the zooids 

 raise themselves above the level of the mouths of their cyclo-systems. 



1 F. E. Schulze, Uber den Bau unci die Entwicklung von Cordylophora lacustris, Leipzig, W. Engeluiann, 

 1877, p. 34, plates iii., iv. 



- Allman, Gymnoblastic or Tubularian Hydroids, vol. ii. p. 155 ; Nature, Oct. 28, 1875, p. 556. 



