GENEEAL EEMAEKS. 13 



species with similar calicles remains to be determined. This character may hereafter 

 lead to a subdivision of the genus and a separation of the species having labellate calicles 

 {M. conigera, efflorescens, cytherea, spicifera, hyacinthus, surculosa, millepora, prostrate, 

 subulata, turbinata, and convexa) as a distinct group." 



In Agassiz's 'Florida Reefs' the radial polyps of M. cervicornis are represented as 

 possessing one tentacle vifhich is longer and stouter than the others ; the arrangement of the 

 septa is only represented in the axial corallites. I find from an examination of a number of 

 specimens that the outer directive septum is sometimes, but not always, broader than the 

 inner ; both directives are broader than any of the other septa. In the case of M. prolifera 

 Agassiz's figure shows the radial as well as the axial polyps to have the tentacles arranged in 

 two cycles, alternately large and small. It appears questionable whether this figure is 

 correct : in any case it is not a good one, and apparently not drawn with the same viev to 

 minute detail which characterizes nearly all the other figures — which are, indeed, the best 

 lithographic representations of Madreporaria which have come under my notice. My 

 doubts as to the accuracy of the figure in question are further based on the following 

 considerations : — (1) I do not consider M. prolifera specifically distinct from M. cervicornis ; 

 (2) the septa of the radial corallites have the same relative importance and are subject to the 

 same variations as in M. cervicornis ; (3) there is apparently no other recorded instance of 

 the alternation of long and small tentacles in the radial polyps of any species belonging to 

 the genus. 



I conclude, from the general remarks of Klunzinger on the structure of Madrepora, that 

 he regards the presence of an elongate tentacle in the radial polyps as a usual condition; and 

 he does not mention any other arrangement. After calling attention to the fact that the outer 

 part of the wall (" Riickenwand ") is usually better developed than the inner (" Bauchwand "), 

 he refers to the association of this condition with two broad septa, which divide the corallite 

 bilaterally. He then goes on to state that, associated with this condition, we find one of the 

 tentacles corresponding to one of the principal septa broader than the others, but that this 

 only occurs in such bilateral corallites. In the descriptions of the Red-Sea species he only 

 refers twice to the occurrence of an elongate tentacle, viz. in M. corymbosa and M. cytherea. 

 In the former instance he refers to the colour of the long tentacle in a manner which appears 

 to indicate that in his opinion such a long tentacle is of general occurrence amongst the 

 species of Madrepora. Mr. Saville-Kent has kindly shown me proof copies of the plates 

 illustrating his forthcoming work on ' The Great Barrier Reef of Australia,' which contain 

 figures of the radial polyps of M. prostrata and M. hebes, and in each case an elongate 

 tentacle is shown. This, so far as I am aware, completes the list of species for which the 

 condition of the tentacles has been recorded. All the species which are known to have one 

 tentacle larger than the others, excepting M. cervicornis, belong to the group indicated by 

 Dana. The absence of M. hebes from Dana's list is probably to be accounted for by the fact 

 that he paid more attention to habit than to the form of the corallites in assigning the species 

 its position in his classification. The radial corallites of M. hebes are not, strictly speaking, 

 labellate ; but the term is used in a much wider sense in Dana's work than in the present 



