] 898.] COKAtS OP THE SOUTH PACIFIC. 271 



o£ the calice, where a low, thin, often much-flattened columella 

 arises. 



In section the layer with living tissues is seen to be about 

 3 mm. thick. The corallum is formed of fine elements, with a 

 very close-latticed arrangement, giving it an appearance of great 

 density. Where the living tissues do not cover the corallum, it is 

 much pitted and corroded. 



Funafuti; lagoon shoals. Wakaya, Fiji ; lagoon. 



There are in the collection portions of three colonies and one 

 young colony. The latter is closely incrusting at the base and 

 differs from the older colonies in having the inner wall of the 

 calices less distinct, in most parts seemingly fused with the true 

 wall. If the wall between the calices in the Pontes is, as I 

 believe, the true theca, then these inner walls must be regarded 

 as supplementary thecae. From the comparison of the exposed 

 sections of this species with those of P. purpun'a and P. viridis, 

 I am inclined to believe that the cell-walls of these and other 

 relatively thick-walled species are formed of three elements — first 

 the fused thecse which primarily would be double, and then two 

 supplementary thecae. 



The pali vary extremely, but, from the comparison of the calices 

 in the different specimens, I think that all the septa have primi- 

 tively a prominent paliform tooth. In front of the primary septa 

 then appear the pah, thin, styliform, equally prominent projections, 

 which fuse almost at once with the edges of the septa, giving rise 

 to a crown of large and small prominences around the large central 

 fossa, in the middle of which the columella arises. Secondarily, 

 owing to physical causes, these may, I think, be enlarged on the 

 secondary septa, giving the appearance of a larger number of pah. 



From the examination of a large number of specimens, it seems 

 to me that primitively there are 6 pali in all the massive species, 

 and that all modifications are really due to causes such as I have 

 sketched above. 



5. PoEiTES UMBELLiFEEA, n. sp. (Plate XXIV. figs. If, 5.) 



Corallum massive, uneven, slightly monticulose above, often 

 table-topped with a broad central attachment. 



Calices polygonal, shallow, and almost superficial in places, about 

 1"3 mm. in diameter, or 7 in 1 cm. Cell-walls thin, with few per- 

 forations, with rough uneven summits without any definite spines, 

 but higher opposite to the septa. Within the wall arises a circle 

 of 12 thick, large, rough, thorny spines, with their summits level 

 with the top of the wall, lying on the top of a definite supple- 

 mentary wall (or theca), which is relatively more internal than 

 in P. trimurata, but less perfect, consisting in many places of 

 thickenings on the sides of the septa which have not yet fused. 

 The septa (12) then get very thin and almost smooth-walled, 

 running into the calice for about a third of its breadth, where the 

 secondaries and primaries are generally fused together and with 

 the pali, giving 6 large, rough, prominent styles with their summits 



