West-Indian Species of Madrepora. 31 



8. No. 92. 6. 8. 214. Thursday Island. Saville Kent 

 collection. 



This is the last specimen in Brook's list, and it is that 

 which most closely resembles M. cervicomis. It agrees with 

 that form in (1) the radial corallites being nariform below 

 and tubo-nariform in the distal parts, (2) in the thin walls of 

 the terminal radial corallite, and (3) in the striate or echinu- 

 late ornamentation of the walls of the corallites. 



These three characters are common to most of the arbo- 

 rescent Madrepores ; the first and third characters occur in 

 nearly all, as, e. g., in M. intermedia. The most impDrtant 

 point of resemblance between this specimen and M. cervi- 

 comis is the length of the terminal corallite. This structure 

 is shown, however, only on one branch of the Thursday Island 

 specimen. 



There are not wanting differences between this coral an I 

 Brook's description of M. cervicomis. Thus he states that 

 the primary septa are subequal, whereas in this specimen they 

 are very unequal, while the one terminal corallite, though of 

 the same length as in M. cervicomis, is narrower, bsing 

 3 millim. instead of 4—5 millim. in diameter. 



Why this specimen, with its long narrow terminal corallite, 

 was regarded as the same species and variety as the Port 

 Darwin specimens, with their short, broad, thick-walled, 

 terminal corallites, is not obvious. 



Hence I am driven to the conclusion that the evidence of 

 the eight Madrepores which Brook catalogued as Indo-Pacific 

 representatives of M. palmata, cervicomis, and prolifera is 

 insufficient to prove the occurrence of those species in the 

 Indo-Pacific Ocean. 



SUMMAKY OF CONCLUSIONS. 



1. M. palmata, Lam., may be conveniently kept distinct 



from M. cervicomis, since (a) the two forms live under 

 identical conditions, their differences are not due to 

 environment, and (b) the evidence of the intermediate 

 forms is inconclusive. 



2. M. palmata, Lam., should not be treated as the typical 



form of M. muricata, L., from which Linnseus ex- 

 cluded it. 



3. If the name M. muricata be retained, which seems un- 



desirable, it should be used for an Indo-Pacific species. 



4. The evidence of the range of M. palmata, M. cervicomis, 



and 21. prolifera into the Indo-Pacific is inadequate. 



