272 MB. J. S. GAKDINEE ON THE PEEFOUiTB [Apr. 5, 



little below that of the wall. In the centre of the cell the 

 columella is similar in appearance to the pali, and almost as pro- 

 minent, springing from the fusion of a number of strands of 

 corallum arisiug from a regular, deep ring joining the edges of 

 the septa and pali. 



In section the corallum is seen to be formed of rather coarser 

 and more open elements than in P. trimnrata, with the usual 

 palisade arrangement. It is also distinctly less heavy. 



Funafuti ; lagoon shoals. 



There are two specimens, a chip from the summit of a massive 

 block and a part of the edge of a table-topped colony. The former 

 shows very well the triple nature of the wall, but has the pali and 

 columella less distinct. 



The species is closely allied to P. trimurata, but is at once 

 distinguished by its smaller calices, more open corallum, and 

 regular pali. 



6. PoEiTES PAEViSTELLATA Quelch. (Plate XXIV. fig. 1 g.) 



Porites 2)cirvistellata Quelch, Challenger Report on Reef-Corals, 

 p. 184, pi. xi. figs. 8-8 a. 



I have obtained two specimens of this species, which very 

 closely correspond to the type in the British Museum. The 

 calices generally are about 1 mm. broad, and the same in depth, 

 but near the edges of the colony tend to become somewhat larger, 

 thinner-walled, and almost superficial. 



Eotuma ; boat channel. 



7. PoEiTES ARENOSA Esper. (Plate XXIV. figs. 1 h, 6.) 



Madrepora arenosa Esper, Pflauz. t. i., Suppl. p. 80, Madr., 

 tab. Ixv. (1797). 



Porites arenacea Lamarck, Hist, des Anim. sans Vert. t. ii. 

 p. 270 (1816). 



Porites arenosa Milne-Edwards & Haime, Cor. iii. p. 180. 



In the collection there are 13 specimens, which closely resemble 

 the published descriptions of this species, but all of which bear 

 considerable resemblances to P. lutea also. In addition I have 

 examined a large number of named (?) and other specimens in the 

 British Museum without being able to find any point which I 

 should consider of specific difference between them. I leave, 

 therefore, the additional characters of the two so-called species, 

 given by Klunziuger, Quelch, and others, with the remark that a 

 series shows very great variability in the arrangement of all the 

 parts within the calices, and secondly that no character taken 

 from the form of the colonies of the two species can be of any 

 specific importance. The description would then be as follows : — 



Corallum primarily incrustiug at the base, with a great tendency 

 to thicken, so as to form moderately thick flat masses with generally 

 level summits, but occasionally a few low, rounded elevations. 

 The top, on the colony reaching a certain size, invariably dies in 



