VAUGIIAN : MADKEPORARIA. 71 



fig. 8 e-f represent the usual condition very well. I wish to call 

 attention to four possible dilFerences between the Manga Reva specimen 

 and A. cancdiculata : — 1. the somewhat thicker branches ; 2. its slightly 

 smaller apical corallites ; 3. the absence of radial corallites with slit 

 upper walls ; 4. the occurrence of a few calicos that are slightly dilated 

 at the mouth. Of these differences, No. 3 can probably be entirely 

 discarded, as the usual form of the calices corresponds to that for 

 A. canaliculata. The other differences are in degree. I therefore 

 think that the Manga Reva is A. canaliculata, but with so little material 

 a positive identification is hazardous. 



Locality : — Outer edge, Motus reef flats, Manga Reva. 

 Porites paschalensis, sp. nov. 



Plates 9, 10. 



Corallum attached by a large base, increasing in diameter as it grows upward, 

 upper surface flat, with rather wide shallow furrows and some circumscribed 

 depressions. Height of largest specimen 15 cm., greatest distance across the top, 

 22 cm. Plate 9 gives two views, natural size, of a smaller specimen, and shows 

 the mode of growth. 



Calices polygonal, varying greatly in size ; on the more elevated portions of the 

 upper surface they may be fully 2 mm. in diameter, while in neighboring depres- 

 sions the diameter may be scarcely 1 mm. ; on the sides 2 mm. is the maximum 

 diameter. The variation in depth is considerable, on the upper surface relatively 

 deep, between .5 and .75 mm. ; on the sides, shallow. 



The walls between neighboring calices are thin, elevated, with numerous per- 

 forations between the synapticula joining the component trabeculae ; the upper 

 edge is spinulose, and with spinules more numerous than the septa, the spinules 

 irregularly granulated. On the sides of the corallum, within the wall and at a 

 lower level is a fairly regular ring of thick synapticula, above each of which is 

 a frosted granulation. The trabeculae forming this ring are what Bernard calls 

 the " wall trabeculae." ^ There is intervening between the mural trabeculae and 

 the pali usually a ring of elevated, pointed spines, Bernard's "septal granules," 

 which equal in height the mural granules and exceed that of the pali. The pali 

 arc distinct low points. The columella is a compressed tubercle, situated in a 

 depression. 



The dorsal, solitary, directive and the lateral pairs bear pali as is usual in the 

 genus. The members of the ventral triplet are not intimately united, although 

 their ends are joined by the palar ring. The middle septum of the triplet may 

 bear a palus, but usually in this group the septal granules simulate pali and seem 

 to form part of the palar ring. 



The preceding description of the wall and septa is based upon the 

 calices on the sides of the corallum ; on the top, the wall ridge is decidedly 



1 Cat. Madrep. Corals, Brit. Mus. 5. The Genus Porites, p. 14, 273, 1905. 



