264 ME, J. S. GABDINEB ON THE PEEFOEATE [Apr. 5, 



channel of Eotuma. Great spreading masses are formed which 

 die in the centre and become somewhat hollowed out, but continue 

 to grow at the sides. The mode in which the colony grows, 

 whether explanate, pulvinate, or globular, is, I think, due to local 

 conditions as to depth below low tide and current, and also to the 

 character of the rock on which the embryo first fixed itself. In 

 my collection there are three specimens of A. listeri, one of which 

 is typically pulvinate, one shows approximation to the globular 

 type, while the third, a young colony about 6 cm. in diameter, is 

 distinctly globular. The great variety shown between the upper 

 and the inider sides of the species I have named A. tabulata seems 

 to show that there is little value in the naming of species of this 

 genus from the skeleton alone. 



1. ASTE^OPOEA LiSTEEi Bernard (6). 



This species seems to be an extremely variable one, but the three 

 specimens in the collection closely correspond to types in the 

 British Museum. 



Funafuti ; lagoon. 



2. AsTE^EOPOEA TABULATA, n. sp. (Plate XXIII. fig. 4.) 



Corallum showing the pulvinate type of gi'owth. Corallites 

 slightly protuberant, hemispherical, generally about 3 mm. high, 

 often coalescing at the sides, but the valleys between usually distinct, 

 with here and there young corallites. The calices are from 1*8- 

 2-2 mm. in diameter 'and from 3-4 mm. apart ; the primary and 

 secondary septa are of nearly equal size, scarcely visible above, but 

 below can be traced as 12 very thin laminate narrow plates with 

 smooth edges, not meeting at "the centre. A few of the tertiary 

 septa are sometimes visible. About 7 mm. below the opening of 

 the calice somewhat thick tabulte, often arched in the centre, occm' ; 

 of these there are about 11 in 1 cm., but the septa are very 

 distinctly continuous through them and the cell is not filled up at 

 all with stereoplasm. The ccenenchyma is extremely echinulate, 

 ending on the surface with somewhat flattened low spinulous pro- 

 jections, which on the sides of the corallites tend to form very 

 regular striations. In sections the interlacing of the costal 

 elements from neighbouring cells is very distinctly visible. The 

 colour of the living colony is green. 



Funafuti ; lagoon. Eotuma ; boat-channel. 

 I have referred to the same species another specimen from 

 Funafuti, which is apparently the under part of a colony, the top 

 of which has broken off and rolled over; the greater part of it 

 has been killed by incrustiug nullipores and the corallites on its 

 surface do not generally project. The calices vary greatly in size, 

 and generally have the primary septa distinct and projecting 

 nearly to the centre of the cell ; the secondary septa are small. 

 The ccenenchyma is very echinulate, and the section shows the same 

 arrangement of the tabulae and of the costal elements as in the 

 types above. 



