538 ox FUWGID COBALS mOM THE SOUTH PACIFIC. [Juue21, 



calices in the same valley are continuous betw'een the fossae, but 

 not over the collines. 



There is no true theca, but the calices are separated from one 

 another, except those in the same valley, by a false wall formed 

 by synapticula nnd trabeculse. 



The septa arise at tirst almost perpendicularly from the fossa, 

 bat towards the exterior of the calice extend more gradually 

 outwards, generally having one very distinct row of synapticula 

 at a short distance within the false wall. They vary greatly in 

 number, but commonly at least thirty can be counted near the 

 edge of the calice, some running singly to the axial fossa, but the 

 majority fusing up iiTegularly, only about twelve ending freely. 

 Their upper edges are covered with very small, bluut, subequal 

 spines, themselves minutely spinulose at the sides and summit. 



The axial fossa is closed in below by trabeculse, which often 

 above are prolonged into a few blunt spines, which may simulate 

 a columella. 



Funafuti ; lagoon shoals. 



7. PsAMiiocoBA SAViGNiENSis, n. sp. (Plate XLV. fig. 4.) 



Corallum at first incrustiug, but later forming low, convex masses, 

 which often die in the centre but continue to grow at the periphery. 

 The margin may be free for a few mm., when the base is seen 

 to be formed by a thin, solid, imperforate wall, probably an 

 epitheca. 



The surface is marked by very irregularly-shaped calices, which 

 either lie quite separately, or in short series with 2-7 distinct 

 centres. Increase is by gemmation over the septo-costse at the 

 margin, or anywhere within the calice. The axial fossae in separate 

 valleys are usually from 3'5-5 mm. apart, in the same valley rather 

 less. In depth the calices vary up to about .'i"5 mm. 



The septa vary greatly in number, but usually towards the 

 exterior of the calice about 36 can be counted, of which perhaps 

 half reach the axial fossa, the remainder coalescing with them 

 and seldom ending freely; no arrangement of cycles can be 

 distinguished. In the deeper calices they are quite thin, but in 

 the serial calices generally wider than the interseptal loculi. 

 Their free edges are covered with low, rough, subequal, blunt 

 spines, which are often somewhat flattened on the thicker septa 

 at right angles to their plane. The collines are rounded, 1-1-5 mm. 

 in breadth, the lower covered with spines, which are directly 

 continuous between the septa of neighbouring calices, but the 

 higher somewhat trabecular in appearance and irregularly spined. 



In section the septa can be seen to be much perforated, especially 

 near the axial fossfe, and appear to be directly continuous between 

 neighbouring calices. There is no true theca, but between the 

 deepest calices a thin, much perforated wall has been formed by 

 the fusion of synapticula and trabeculse from the septa ; a well- 

 marked row of synapticula on either side in addition is a common 

 feature, but it is never fused with the central row. 



