228 



CEYLON PEARL OYSTER REPORT. 



' 4 



3 



JD 1 



Fig. IV. Diagram of a specimen of Hdempmmmia michelini in which the calice has not yet divided by 

 transverse fission. The endosepta are cross-shaded, the exosepta, the porous theca, and the columella 

 are black. The mesenteries are indicated in the upper half of the figure but not in the lower. D, D, 

 directive mesenterial pairs. 1, 2, 3, endosepta of the first, second, and third cycles. 4, 4, fourth- 

 cycle endosepta, each enclosed in a chamber formed by the forked peripheral ends of an exoseptum. 

 In several cases the inner ends of the exosepta may be seen to unite in front of a tertiary endoseptum 

 and to be prolonged beyond the jtoint of union into the columella. 5, an endoseptum of the fifth 

 cycle formed in a bifurcation of an exoseptum adjoining one of the directive endosepta. 



Pourtales and Duerden's law in the case of Eupsammid corals. It should be 

 noted, however, that the quinary mesenterial pairs do not arise according to the 

 sequence observed by Duerden in Siderastrcea, but are contiguous to the sulcar 

 (dorsal) and sulcular (ventral) primary septa. 



My second series of sections, through a polyp which had been formed by transverse 

 division from the original single parent polyp, gave some interesting results. A 

 tracing of a section taken a little above the level of the stomodeeum is given in 

 fig. 22. The arrangement of the septa and mesenteries is remarkably regular. 

 There are apparently six systems and three cycles of endosepta, with an additional 

 quaternary septum in the chamber on the right side of the directive septum, that 

 is to say, there is just half the number of endosepta that there was in the elongated 

 and undivided specimen. The endosepta may be classified according to size and 

 relations as primaries, secondaries, and tertiaries, but they cannot have made their 

 appearance in this order, as they were derived from the pre-existing septa of the 

 parent polyp. The number of parental mesenteries has evidently been halved in the 

 process of division. No new septa have been added, but the quaternaries of the 

 parent have become the tertiaries of the offspring. 



