of the Rujosa to the Uviiiy Zoanthece. 391 



modern coral, and in any natural taxononjic arrangement 

 must call for its sharp separation as the representative of a 

 distinct group. Though agreeing as concerns the protoseptal 

 stage, the njctaseptal sequence shows Lophophyllum and 

 modern cyclical corals to be widely divergent. 



Fig. 8. 



Upper surface of a polished specimen, showiugthe alternating small septa, 

 which occur only towards the upper region of the c dice. The prin- 

 cipal septa, witli the exception of the main and counter axial septa, 

 are now practically equal in size and radially arranged, presenting 

 little evidence of their successive origin in bilateral pairs. In such a 

 tigui'e it is impossible to determine the ordinal value of the septa, 

 but the total number present agrees with that in the previous figure. 

 The dark median lines are not seen by reflected light. 



The Palaeozoic corals sufficiently resemble those of modern 

 days to warrant the assumption that the general relationships 

 of the polyp to the corallum were the same as those now fully 

 understood, and therefore, knowing their septal sequence, we 

 can assume what must have been the mesenterial. With 

 such a metaseptal sequence as that above demonstrated, both 

 by tlie serial sections and by the external ridges and grooves, 

 modern coral morphology permits the assertion that in the 

 polyps of Lophophyllum new mesenteries were successively 

 developed within four of the primary exocoelic chambers, and 

 at only one region in each, immediately next a primary pair 

 of mesenteries. Though no known living coral presents such 

 a mesenterial sequence, yet it bears the closest resemblance 

 to what is found to be characteristic of Zoanthid polyps, 

 except that here the mesenterial increase takes place in only 

 two exocoelic chambers, whereas in Lophophyllum the process 

 was carried on in four chambers. The researches of E. van 



