A. K YerriU—Bermiidian and West Indian Heef Corals. Ill 



have been made until isol, when Lamarck (Syst. Anim., pp. 369-:}75) 

 clivided it into eight genera.* Unfortunately he restricted the name 

 Madrepora to the group that inchided 31. muricata and M. porites 

 Pallas. The latter was made the type of Porites, by Link, 1807. 



The next restriction of the name was by Oken (Lehrb., 1815), who 

 established a number of additional generic subdivisions and restricted 

 Madrepora (which he spelled 3fatre2)ora\) to four species, one of 

 which, 31. ramea, became the type of Bendrophyllia Bv., 1830 ; the 

 others Avere earlier (1816) placed in Oculina by Lamarck. One of 

 these {31. oculata Linne), which is the long-known and officinal "white 

 coral " of the Mediterranean, the " 3Iadvepora vulgaris " of Tourne- 

 fort, may well be taken as the true type of 3Iadrepora, not only on 

 <account of Oken's restriction, but also because of the rule, advocated 

 and followed by many naturalists of the Linnaean period, that the 

 type of a genus should be the most common or officinal and well- 

 known species, if such were included. Certainly 31. oculata would 

 answer Avell to this requirement, and so would 31. prolifera. 



Moreover, in following the principle of elimination, this was one of 

 the very last of the determinable Linnaean species to receive a special 

 generic name (1849). 31. prolifera, the second species of Oken, and 

 the type of Lophohelia PI and IL, is now made congeneric Avith 

 3[. oculata. 



Therefore, it appears that ocidata should be taken as the true type 

 of the restricted genus Madrepora, if the Lamarckian nomenclature 

 must, in this case, be abandoned, as argued by VaughanJ and other 

 recent writers. 



■" These genera are as follows : — Cyclolites, p. 3G9 ; Fimgia, p. 369 ; Cdiyo- 

 phyllia, p. 370 ; Madrepora, p. 371 ; Anfrea, p. 371 ; Meandrina, p. 372 ; Pavona, 

 p. 372 ; Ayaricia, p. 373. 



f That Oken, in using Matreporn, did not intend it as a new name, but only as 

 a corrected spelling of Madrepora, is proved by the fact that in citing the Lin- 

 naean names of species tinder various genera, he invariably quotes them as 

 " Matre2Jora" or "Mat." of Linne. The generic divisions of Madrepora pro- 

 posed by Oken are as follows : — Asfrea, p. 65 = Astrea (pars) Lam., 1801 ; Acro- 

 pora, p. 66; Turbinaria, p. 67; Favia, p. 67; Pectinia, p. Q8 = Meandrina 

 Lam., 1801 ; Undaria, p. 69 = Agaricia, Lam., 1801 ; Mycedium, p. 69 ; Mcean- 

 dra, p. 70; Matrepora, p. 71 (includes 4 species, viz. — M. ramea, M. prolifera, 

 M. virginea, M. oculata) ; Gakixea, p. 72 (with 4 species) ; Mussa, p. 73 (2 species) ; 

 Fungia, p. 74, 2 STp. z= Fungia Lam., 1801. 



Probably Lamarck's Systeme Anim. sans Vert., 1801, was not known to Oken, 

 for he makes no reference to it. The coincidences in some of the names were 

 probably due to the influence of the older specific and polynomial names. 

 Neither does he refer to Link's work of 1807. 



t Samml. Geol. Reichs-Mus., ii, p. 68, 1901. 



