26 Prof. J. W. Gregory on the 



Hence the specimens wliich Mr. Brook quoted as linking 

 M. palmata and ^f. cervicornis do not seem to me to give any 

 support to the belief in the specific unity of these corals. To 

 dismiss such differences in the form of the corallum as not 

 worthy of specific value appears to me inconsistent with 

 ^Ir. Brook's piactice in later pages of his monograph. Thus 

 he founded a species, vV. atteniiafa, for a form which appears 

 to be based on a series of fragments of slender branches of 

 M. cervicornis^ and he accepted Dana's M. cyclopea^ which 

 appears to be only an alciform variety of M. palmata. In 

 the case of M. attenuata it may be objected that the terminal 

 axial corallites are shorter than in M. cervicornis ; but they 

 are not shorter tlian in Agassiz's * figures of that species, 

 which show that the character is inconstant. A more serious 

 inconsistency is that Brook divided his subgenus Conocijatlius, 

 Brk., non d'Orb.t, into four sections, characterized solely by 

 the form of the corallum. The following are his diagnoses of 

 those sections : — 



A. Corallum corymbose, with or witliout confluent branches. 



If the central branches are long the habit is bushy. 

 (P. 161.) 



B. Corallum forming a subcomplanate reticulum, with 



short twigs on the upper surface. (P. 166.) 



C. Corallum cajspitose. (P. 166.) 



D. Corallum subarborescent or bushy, usually with nume- 



rous short proliferations. (P. 169.) 



If the difference between a Cfespitose | corallum and a 

 bushy corallum is of more than specific value in " Cono- 

 cj/at/iuSy" why is the well-marked difference between the 

 palmate and arborescent coralla of less than specific value in 

 Eumadrepora ? 



* L. Agassiz, " lleport on the Florida Reefs," Mem. I\Iua. Conip. Zool. 

 vol. vii. 110. 1, 1880, pi. xviii. ti^^s. 1, 4, & 8. 



t This name was preoccupied for a genu.s of corals which has living 

 Australian representatives. Another of Bmoli's subgeneric names, Odonto- 

 cyathns, is preoccupied for a deep-sea coral dredged by the ' Challenger ' 

 and described by Moseley. 



X The difference that Brook intended to suggest between a casspilosd 

 and a bushy corallum is not very easy to realize. According to Muri'ay's 

 new J^nglish iJictiouary, cjespitose means " growing in thick tufts ur 

 clumps." liut Brook places 31. Forshali, in which he describes the 

 " corallum [as] forming dense and much branclied clumps," among the 

 bushy and not among the cjespitose section ; and M. Rousseaui, in wliich 

 he describes the corallum as " consisting of tufts," is also excluded from 

 the cespitose section. Both Ogilvie and Worcesti-r's dictionaries define 

 Cfespito.se as " growing in tufts.' 



