908 



MISS U. M. HARRISOX AND MISS M. POOLE OX [Dec. 14, 



psammia) fig. 7 a ; another individual with an incomplete fourth 

 cycle {Coenojysammia) fig. 7 6 ; while a third individual which has 

 lately become separated off has not even the tyjaical six systems of 

 cycles, fig. 7 c, but four complete cycles of four systems. In the 

 lai-ger colony is an indivi^lual which clearly proves that new zooids 

 are formed by tissiparity and not by gennnation (Lobopsaimnia), 

 fig. 7 d ; for here the lip of the calice has been drawn out into 

 an oval, a strong ridge has grown across the centre, and the 

 two individuals are incompletely separated ; the septal systems 

 are incomplete in both, but four systems can be distinguished 

 in one and two in the other. A comparison of this figure 

 (Hg. 7c?)Avith fig. 7 c throws a light on the interpretation of 

 the septal arrangement of the latter ; the ridge that has grown 

 across the double calice divides it into two im equal parts. 

 The larger part contains four nearly complete systems of four 

 cycles, and it is here possible to trace rudiments of the two 

 remaining systems, which will ultimately complete the typical 

 zoantharian six systems ; but in the other part, where only two 

 systems of four cycles are discernible, it is possible that the six 

 systems will never be complete, and that this has actually occurred 

 in the individual represented in fig. 7 c. The process of fissiparity 

 has not gone far enough in the double-caliced individual to decide 

 that each zooid will ultimately have its full complement of septal 

 systems, and it has gone too far to be certain that the dividing- 

 ridge has grown, not across the middle but rather across one end 

 of the calice ; nevertheless, considering the arrangement of the 

 septa of the zooid represented in fig. 7 c, it seems probable that 

 this has been the case, and that this represents four cycles of four 

 systems rather than three cycles of eight ; for it is easier to believe 

 • that an individual which has been formed by a process comparable 

 to simple binary fission will have a shortage rather than an excess 

 of the normal complement of characters. 



From the above discussion it is clear that this species from the 

 Mergui Archipelago breaks down Milne-Edwards and Haime's 

 original generic characters for the three genera Dendrophi/Uia, 

 Coeno])Scimmia, and Lobopscmiinia ; and the same is true of Dendro- 

 phyllia rohusta {Lohopsmnmia robusta Bourne). Of the three 

 colonies in the Burmese collection, one does not differ in any 

 particular fi-om the original Ceylon specimen, but in both the 

 others it is quite obvious that new individuals are not formed by 

 fissiparity. One of these colonies is represented in fig. 6, but 

 whether the youngest zooid has been formed by gemmation or by 

 the process described and figured by von Koch (20. pi. iii. fig. 21) as 

 " Theilknospung," it is not possible to decide from the material 

 available ; but the fa,ct that it has not been a process of fissiparity 

 breaks down an importa.nt generic character that separated Lobo- 

 psamonia from Dendt^ophyllia and Coenopsmnmia. 



The remaining colonial form appears to be BendropJiyUia gracilis 

 (M.-Edw. k H.) and offers no special feature that bears on the 

 present argument. 



