AND GLTPHASTR^A, DUNCAN (1887). 209 



on the exterior of the theca, and this is the case even when the 

 septa of the third cycle are developed at the summit of the calices, 

 as in S. sexradiata. It also often happens that even when 12 septa 

 are fairly well developed in the interior of the corallites there are 

 only six clearly shown in the calices, the other septa of the second 

 or third cycles being only faintly indicated by slight projections 

 at the margins of the calices. 



The calices in the mature corallum are generally shallow (in 

 extreme instances nearly level with the surface of the corallum, in 

 other cases from 1 to 3 mm. in depth). The floor is, in mature 

 corallites, complete and usually gently elevated at the centre, so as 

 to form a slight dome ; it is radially divided by the septal upper 

 edges, which project slightly above it. In some instances the floor 

 conceals the central union of the septa ; in others their junction is 

 distinctly shown. The septa within the calices of mature corallites 

 are much more regular than in the lower and central portions ; they 

 are for the most part 6 or 12 in number, but they may range from 

 4 to 24. In certain cases two opposite septa extend quite across 

 the corallum, and are more prominent than the others ; but this 

 feature is by no means general, and in some specimens is hardly 

 noticeable (PL IX. figs. 2, 3, 3 a, 3 6, 6). 



When the corallites have not reached their full limit of growth, 

 or, perhaps, in some cases owing to abnormal conditions, the calices 

 are much deeper, and the floors are incomplete and limited to the 

 central portions of the corallites. In weathered specimens also the 

 basal floor of the calices may disappear, and they are then open for 

 a considerable depth. 



In the interior of some calices, more particularly in those which 

 are less filled up with stereoplasm, there are delicate linear grooves 

 within the wall, extending vertically from the summit margin, one 

 on each side of the principal septa (PI. IX. fig. 17). Sometimes the 

 line is not continuous, but instead of it there are closely arranged 

 minute holes. These structures are probably connected with the 

 insertion of the muscles of the mesenteries* of the living polyp. In 

 the interior of the calice of the existing Flabellum patagonicJmm, 

 Moseley, there are similar rows of small holes on either side of the 

 septa ; these do not penetrate through the theca. Prof. Lindstromf 

 has described apparently similar rows of dots or small holes in the 

 recent Schizocyathus fissilis^ Pourtales. 



By the junction of the inner margins of the septa with each other 

 the internal cavity of the corallites is divided into a series of closed 

 longitudinal chambers or interlocular spaces, and each of these is 

 independently partitioned off at irregular intervals by the transverse 



* I am not aware that these structures have been previously notified in fossil 

 corals; they would evidently afford greater hold to the mesenterial muscles, 

 and explain the statement of Prof. Moseley, that these are attached with such 

 firmness that when the corallum is broken away small pieces of it hang tena- 

 ciously to the muscular shreds. Report of H.M.S. * Challenger,' vol. ii. p. 163. 



t "Actinology of the Atlantic Ocean," p. 15. Kongl. Svenska Vetensk.-Akad, 

 Handl. Bd. xiv. no. 6, 1877. 



