1886.] MADREPORARIAN CORAL. 133 



appearance, doubtless exaggerated by the action of the alcohol in 

 which the coral is preserved. 



The tentacles are in four cycles ; the innermost cycle are of the 

 largest size, and correspond in position to the primary and secondary 

 septa, and are therefore twelve in number. The second cycle 

 correspond to the tertiary septa, and are also twelve in number. 

 The third cycle correspond to the quaternaries, and are twenty-four 

 in number. These differences between the three cycles are very slight, 

 consisting only of small differences in size. The fourth cycle, placed 

 considerably nearer the circumference of the coral, are very much 

 shorter and more stumpy than the other three cycles ; they correspond 

 to the quinary septa, and, as will be seen below, are evaginations of 

 intermesenterial sfiaces [exocoelic. Fowler, (4)], not of the intra- 

 mesenterial chambers [entocoelic, Fowler (4)] as are the others ; they 

 are forty-eight in number ; so that there are altogether ninety-six 

 tentacles. 



The mesenterial chambers are forty-eight in number ; there are 

 two cycles of mesenteries, larger and smaller, forming larger and 

 smaller chambers ; the larger chambers contain septa of the first, 

 second, and third order; the smaller chambers septa of the fourth 

 order. From these chambers the tentacles of the first three cycles are 

 evaginated ; the fifth cycle of septa, on the other hand, are placed in 

 the spaces between the mesenterial chambers, i. e. exosepta, Fowler ; 

 and the fourth cycle of tentacles are in the same way evaginated from 

 these exoccelic spaces. The whole arrangement will best be under- 

 stood by reference to the diagram (Plate XII.). 



In all the chambers, with the exception of those in which are 

 placed the primary septa situated at the long ends of oval mouth- 

 opening, the longitudinal muscles are placed, facing one another ; 

 but in the chambers above mentioned, the muscles are placed on the 

 outer faces of the mesenteries ; these thei'efore are the directive 

 mesenteries. So that in the arrangement of the muscles on the 

 mesenteries, Stephanotrochus exactly corresponds to the Hexactinian 

 type, as do all other Madreporarian corals that have been hitherto 

 examined. 



The tentacles are all of one shape, they are simple conical evagina- 

 tions of the coelenteron ; they end in a knob separated from the rest 

 of the tentacle by a narrow neck ; the knob contains no pigment, and 

 is of a white colour ; the main part of the tentacle is madder-coloured 

 like the rest of the polype. 



A section of the knob and part of the conical part of the tentacle is 

 figured (Plate XIII. fig. 6) ; the knob is packed full of nematocysts, 

 and in fact is a single immense battery of thread-cells. The ectoderm 

 consists of a series of very long and very narrow cells, which are 

 nucleated ; there are also gland-cells present, but the elements are 

 difficult to distinguish ; the endoderm cells of the knob are much 

 longer than those lining the other parts of the ccelomic cavity of the 

 tentacle. I was unable to distinguish any muscular elements in the 

 knob ; it is perhaps in consequence of this that the outline of the 

 knob appears straight, since it apparently has not been contracted 

 by the action of the spirit, as has been the rest of the tentacle. 



