GENERAL AND COMPARATIVE STRUCTURE. 51 



h. Without spiniform coral lites, and having numerous interstitial tubes 

 intercalated uniformly amongst the large ones. Ex.., M. ramosa, 

 D'Orb. (with M. ri/gosa, E. and H., and Af. Dalii, E. and H.), 

 M. ONea/li, James, J/, nodi/losa, Nich. 



c. With spiniform corallites, and with numerous interstitial tubes inter- 



calated among the large corallites. Ex., Af. mamnnilata, D'Orb., 

 M. calycula, James, M. Jamesi, Nich., M. U/richii, Nich., M. 

 frondosa, D'Orb., Af. Whiieavesii, Nich. 



d. With spiniform corallites, but with very few interstitial tubes (these 



latter sometimes almost obsolete). Ex., AI. tumida, Phill., AT. 

 gracilis, James, AI. Dawsoni, Nich. 



e. Spiniform corallites few or occasionally wanting, but with well- 



marked clusters of small tubes. Ex., AI. subpulchella, Nich., 

 AI. Selwynii, Nich. 



The above rough grouping of a number of species of Mo7i- 

 ticulipora in accordance with the nature and disposition of 

 the coralHtes forming the colony, will show that this feature 

 cannot be relied upon as singly sufficient for a basis of natural 

 classification ; since, while it certainly brings together many 

 nearly allied types, it widely separates others which in their 

 main structural peculiarities are unquestionably closely related. 

 At the same time, if due weight be attached to other important 

 structural characters, such as the condition of the walls of the 

 corallites, and the nature of the tabulae, we may safely avail 

 ourselves of the variations in the specialisation of the coral- 

 lites as a guide to a sound classification of the species of this 

 difficult genus. 



VI. Nattire of the TabulcE. — The next point in the construc- 

 tion of the Monticuliporoids which needs notice is the form of 

 the tabulce, and the arrangement of these structures in the dif- 

 ferent corallites of the colony. So far as I have seen, tabulce 

 are never wholly absent in any Monticuliporoid, though they 

 are sometimes {e.g., in M. irregitlai^is, Ulrich, and M. clava- 

 coidea, James) reduced to a minimum, and may be entirely want- 

 ing in more or less extensive portions of the corallum. In the 

 great majority of the Monticuliporoids, also, the tabulse are 

 "complete," and are either horizontal or slightly curved. In 

 no case known to me do the tabulse assume what is properly 

 called the "incomplete" type of these structures (such as 



