of the Antliozoa Tahulata. 3 



the place of the dissepiments (as, for example, in some Ptycho- 

 phjUa), just in the same way as the dissepiments may occu])y 

 the place of the tabula. This identity of the tabulge and 

 dissepiments is perhaps in no forms so evident as in the Cya- 

 thojyhylla, in which there are frequent passages between both 

 these sclerencliymatous secretions, whereby it is demonstrable 

 that they are immediate continuations or transformations of 

 one another. On tlie other hand, there are Heliolitidas in 

 which a longitudinal section shows dissepimental tissue of quite 

 a Cystiphyllidean type partially superseding the usual regular 

 tabulae. A compound Cystijohyllum (such as C. cylindricum, 

 Lonsd.), where the individual corallites are often very narrow 

 and are each traversed by crowded and regular horizontal dis- 

 sepiments, quite resembles a " tabulate " coral in its longitu- 

 dinal section, and cannot be distinguished from one so far aJthis 

 particular point is concerned. The fact seems to be that some 

 corals which, like Syringopora and Columnaria, have been 

 placed amongst the Tabulata on account of their " floors," are 

 rather to be regarded as Eugose corals. It is also very diffi- 

 cult in longitudinal sections to see any great difference between 

 a MicheUma or Emmonsia and a Cystipliyllum^ all alike having 

 the visceral chamber filled up with abundant vesicular dissepi"^ 

 ments.^ Besides, there are several recent corals of quite remote 

 zoological affinities, such as TuUpora^ which are provided with 

 tabulae, thus resembling Syringophyllum and Syringopora. 

 Duncan has also shown how LopJwhelia is provided with tabula? 

 (Madrepor. of the ' Porcupine' Exp. p. 323). Amongst Meso- 

 zoic genera, Clausastrcea and CyathojiJwra, according to De 

 Fromentel (Intr. Pol. Foss. pp. 278, 280), have tabula? so 

 tongly developed as to lead him to place them in the 

 Tabulata. I am of opinion, therefore, that there is no difference 

 of kmd between dissepimental tissue and tabula, both belono-- 

 ing to the same sort of endotheca. The Rugose corals there- 

 fore, and some other forms, are just as much tabulated as the 

 Tabulata, and the latter are just as much dissepimental as the 

 former, there being in this respect a complete agreement be- 

 tween the two groups. 



There are, moreover, other animals which in their hardened 

 tissues possess tabulae, or have the cavity formerly occupied 

 by their body divided into compartments by transverse floors 

 placed at tolerably regular intervals ; and these have therefore 

 been regarded as Tabulate corals, though I think there is no 

 longer any reason for retaining them amongst the Anthozoa. 

 This is the case with Millepora, and probably also with Axo- 

 pora. In a former paper ("Anthozoa Perforata of Gotland," 

 p. 3) I endeavoured to show that the polypary of Millepora 



1* 



