99 



I. and II. The columella is a plate, one-third to one-half the 

 length of the calicle, its upper edge intermediate in height be- 

 tween the upper edges of the septa of the two series ; it is 

 joined by processes from the 12 larger septa, about i'^ mm. 

 below its edge. 



The theca has the appearance of having been formed by 

 thickenings of the septal edges. It has no definite 

 costae, but the whole of the exterior is covered by 

 a series of minute, linearly extended granules, forming lines, 

 which commonly converge a little round the base of the coral- 

 lum. They start somewhat irregularly as elongated paired gran- 

 ules, each pair forming a V, on the exsert portions of the septa, 

 but below, on the widest portion of the corallum, 8 or 9 Hnes 

 correspond to two neighbouring septa. 



Height of the specimens by the greatest and least diameters 

 of their calicles in mms., 13 x 1 1 x 6, 1 1-5 x 9-5 x 5-5, 9-5 x 8"5 

 X 5, 11x8x3-5, 10 X 8" 5 X 5, io'5 X 8 X4-5.. 



Locality. — Near Kowie, Lat. 33° 45' 20" S., Long. 26° 44' 20" 

 E. Depth, 44 — 43 fms. Date, November 19th, 1898. (Seven 

 specimens.) 



The species most closely approaches to 5. granulosum, Ed. 

 and H., from the Eocene. It differs from the four recent species 

 described by various authors, in the fact that it has no costae 

 and no wings on either side. Two specimens have the columella 

 slightly notched in the centre, and one has two pairs of thick- 

 enings on the sides of the same, opposite the tertiary septa on 

 either side of the large central pair. 



Genus TROCHOCYATHUS, Edw. and H., Duncan. 



The genus appears to be fairly well defined, but I do not see 

 any reason in the characters that Duncan gives for the retention 

 of Thecocyathus as a subgenus, since they are due to purely 

 environmental causes. The hollow root-like growths from the 

 " interseptal chambers," described by Pourtales in Th. cylin- 

 draceiis and laevigatus {III. Cat. Mus. Comp. Zool., No. IV., 

 1871, pp. 13, 14), may provide a character of value, but this de- 

 scription must be re-read in view of the now known ectodermic 

 origin of the corallum. Similar roots are also found in T. 

 cmcticulatus below, a close ally. If they have really any con- 

 nection with the interseptal chambers, they can only be formed 

 ill the youngest stage of growth of the polyp, unless it be as- 

 sumed that the polyp redissolves its theca in places so as to push 

 them out. I have, however, been quite unable to see any such 

 connection, and if there be none, the innermost wall alone can 

 be regarded as a thecal formation, and all the external walls 

 must be epithecal. 



