342 Mr. H. J. Carter on Stromatopora dartingtoniensis. 



tain the basal or vermicular coenosarc of Millepora, so in 

 Stromatopora it becomes necessary to make a similar distinc- 

 tion ; but the " main canals " here assuming for the most 

 part a stellate form, in which the branches radiate from 

 fixed points in the lamina? successively, instead of meandering 

 about irregularly as in Millepora, I propose henceforth to 

 designate each group by the name of " astrorhiza," as above 

 mentioned (PL XVIII. fig. 1, ccc). At the same time it 

 should be remembered that, although the astrorhiza is so 

 strongly developed in Stromatopora dartingtoniensis , there are 

 other species in which it is so little differentiated from the 

 general ccenosarcal structure that none but an experienced eye 

 can detect its presence. 



What, then, is the condition of the tabula? in Millepora 

 alcicomis (which so far may be considered the nearest 

 living representative of Stromatopora) ? Here they present 

 themselves in the way above mentioned, in the tubular spaces 

 extending from the axial structure of the branch through its 

 ccenenchymal tissue to the circumference. I have said 

 " spaces ;" for the coenosarc; from its extreme tenuity, 

 almost wholly disappears on desiccation, when the tube 

 itself is left as a mere excavation without wall in the 

 midst of the ccenenchymal tissue ; hence the coenosarc of 

 the latter is continuous with that which originally filled these 

 tubular spaces, and but for the presence of the tabulae, which 

 act like so many diaphragms in dividing the space into 

 separate compartments, the coenosarc in them would be 

 directly continuous throughout. In structure, composition, 

 and size the tabula? (often, however, provided with a stelli- 

 form prolongation outwardly, as in Tubipora musica) are 

 thin imperforate plates of calcspar, about l-1800th inch thick, 

 and, on an average, about l-120th inch apart. Here 7 again, 

 it is necessary to remember that the ccenosarcal cavities, which 

 are vermiculate, open into the " tubular space " throughout, 

 and therefore that the coenosarc which originally filled the 

 tubular spaces was continuous with the coenosarc which filled 

 the ccenosarcal cavities or vermicular channels of the ccenen- 

 chyma. 



Thus we are prepared to follow out my description of the 

 tabulation in the larger branches of the astrorhiza in Stro- 

 matopora, which I have no doubt will be received with much 

 hesitation ; but if I can show that, under certain circum- 

 stances, these tabula? are left in the branches of the astrorhiza 

 as they appear in the tubular spaces of Millepora, while all 

 the rest of the calcspar has been removed, and that they are 

 also to be seen in the midst of the calcspar filling the branches 



