ANIMALS. 37 



than usual, and their lower part, on the celluliferous surface, is strongly angulated 

 through the prominency of the dividing ridge, and the close proximity of the tubercles ; 

 while higher up they possess the usual characters. Similar differences obtain with 

 the meshes, the form of which depends chiefly on the distance of the stems from each 

 other, being circular in one specimen, oval in another, oblong in this, and linear in that. 

 The fossil represented by Goldfuss in the ' Petrefacta Germanise,' pi. xxxvi, fig. 2 3, c, 

 is a rare condition for this species. I have been fortunate, however, in obtaining a 

 specimen resembling it, showing the tubercles with an aperture at the apex, as repre- 

 sented in Plate II, fig. 1 1 . Never having been able to find any internal casts of 

 these appendages, so frequent as regards the cellules, I am inclined to think that they 

 remained closed until a late period of their growth. Mr. Lonsdale, referring to 

 Goldfuss's figure, simply designates them " abraded vesicles."'^ Why may they not 

 be considered gemmuliferous vesicles, which have become ruptured through the 

 discharge of their contents ? M'Coy's Fenestella carinata, and some other species, 

 display similar open cellules.^ 



This Coral sometimes attains a large size : the frond of my largest specimen 

 measures eight inches in width. The under side or non-celluliferous face is occasionally 

 seen with root-like processes, apparently hollow, from i to ^ of an inch in length, 

 striking off from the stems (vide Plate II, figs. 18 and 19). From what is displayed 

 in a specimen before me, I am inclined to think that in many cases all the fronds of 

 one Coral possess these processes, and that they served, not only as supports to the 

 fronds, but as stays to keep them separated from each other. Specimens are some- 

 times so completely folded, that it is difficult to abandon the idea of its being the outer 

 surface of the frond, which is celluliferous. Is not the specimen represented by Geinitz, 

 in his ' Versteinerungen,' pi. vii, fig. 1 5, in this condition ? 



Mr. Lonsdale, in showing that Fenestella has no relation to Gorgonia, the genus in 

 which Goldfuss and others have placed the Permian Corals, states, that " beneath the 

 thick series of obliquely-overlying cylindrical cells, limited to one side of the Coral, 

 there is no axis formed of concentric bands, yet distinct from the crust," as in Gorgonia, 

 " but a parallel layer of vertical, capillary tubes, the walls of which differ not 

 apparently from those of the cells. In the young state this layer constitutes the outer 

 surface of the non- cellular side."^ In Fenestella retiformis the vertical capillary tubes 

 are sometimes finely displayed, when they are seen to be straight, and crowded with 

 minute foramina, as shown in the magnified representation in Plate II, fig. 17. 

 According to Mr. Lonsdale, the intermediate layer of vertical capillary tubes is totally 

 wanting in Retepora {Elasmopord), proving that the species under consideration does 

 not belong to this genus, as was formerly supposed. 



' Geology of Russia, &c., vol. i, Appendix A, p. 630. 



- Sj'nopsis of the Carboniferous Fossils of Ireland, pi. xxviii, fig. 12 ; pi. xxix, figs. 1, 2. 



^ Geology of Russia, vol. i. Appendix A, p. 627. 



