Tabulate Corals with Existing Species. 363 



Porites, and has some additional special characters. The 

 transverse septa are usually quite numerous and thin, usually 

 irregular, but with an evident tendency to coincide in height 

 in all the chambers of the same polyp-cell, though much 

 broken up and forced out of the transverse plane by the pre- 

 sence of the large irregular columella. In one species of 

 Goniopora I have occasionally seen cells with a deeply infun- 

 dibuliform septum completely closing the cavity below, thus 

 recalling the septa of Roemeria. 



The three genera Goniopora^ Alveopora, and Porites agree 

 closely in the characters of their polyps ; the first, however, has 

 24 tentacles, while the others usually have but 12, although 

 there are often a few larger polyps with 24 tentacles, scattered 

 among the smaller ones, in both the latter genera. It seems 

 necessary, therefore, to place these genera and the others that 

 are evidently closely allied to each of them in one family, Pori- 

 tidcB. It will also be understood, from what has already been 

 said, that it is impossible to assign any characters sufficient for 

 separating the Favositince^ even as a family, from the Poritidm, 

 It is very doubtful whether the group can be maintained even 

 as a subfamily ; for Alveopora and Goniopora combine the 

 characters of both groups. The family PoRiTiD^*, thus ex- 

 tended, might, perhaps, be provisionally divided into three 

 subfamilies : — PORITIN^, for Porites and the closely allied 

 genera ; Alveoporin^, to include Alveopora, Goniopora^ 

 Litharcea, and, if considered distinct, Koninckia and Favositi- 

 pora ; Favositin^e, to euihYSiCeFavosites,Fmmonsia,Micheliniaj 

 and the other closely allied genera. It is probable, however, 

 that even such a slight separation of Alveopora and Favosites 

 is greater than the differences actually observed will warrant. 



Admitting these necessary changes in the classification f, it 

 follows that the Madreporaria perforata or Madreporacea, 

 Avhich is generally regarded as the highest division, or sub- 

 order, of the true corals, was abundantly represented even in 



* The genus Montipora, for which Edwards and Haime constituted 

 their second subfamily of Poritidce {Montiporinfe), belongs properly to 

 the Madreporiclce, as explained elsewhere by the writer (Trans. Conn. 

 Acad. vol. i. p. 501), and where it was also placed by Prof. Dana. 



t The opinion that the Fca^ositince belong to the Madreporacea was 

 advanced by the writer in 1870 (Trans. Conn. Acad. i. p. 518). Mr. Kent, 

 in the article referred to, published simultaneously with mine, expressed 

 the same opinion and used independently nearly the same arguments. 

 He also uses the argument with reference to the impossibility that a 

 coral with radiating septa could be formed by hydroid polyps, as I had 

 also done both in the paper referred to and in that of 18G7. This coin- 

 cidence of opinion, an-ived at through studies pursued in different ways 

 and approached from different directions, could not fail to be gratifying 

 both to the writer and to Mr. Kent. 



25* 



