CORALS FROM MURRAY, COCOS-KEELING, AND FANNING ISLANDS. IOQ 



reproduction in F. pallida is not by subequal but by very unequal fission. There- 

 fore, so far as I may base an opinion on the material available for comparison, 

 Favia dana is a valid species. 



Favia matthaii, new species. 

 Plate 39, figures 2, la, lb. 



The following is a description of the type of this species: 



Corallum with domed upper surface; 10.5 cm. long, 8.5 mm. wide, 4 cm. thick. 

 Corallites rounded, subcircular, elliptical, or oval. Greater diameter from 9 to 1 2 mm. ; 

 lesser diameter from 7 to 9 mm. Calices rather shallow; their edges slightly raised; ltiter- 

 corallite areas from 2 to 5.5 mm. across. Costa; correspond to all septa, but they are low, 

 thin, and distant, as much as I mm. apart, their edges dentate or beaded. Those from 

 adjoining calices often do not meet, but disappear on the surface of the vesicular exotheca. 



Septa about 36 in a fully developed calice, of which 16 meet the columella, about 12 are 

 rudimentary, and about 8 are well developed but not long enough to reach the columella. 

 Within the calice they are thin, but the outer ends of the larger septa are thickened in the 

 wall, and exsert up to 1.5 mm. The exsert part is dentate according to a peculiar pattern. 

 There is a prominent tooth over or nearly over the edge of the wall, and both '"side and 

 outside it are one or two prominent, spreading teeth. These teeth may be forked and the 

 tips of the forks frosted. Below these larger teeth there is within the calice one or two 

 smaller, pointed, thin, distant teeth, and still farther within the calice a circle of simple or 

 double, wider, prominent teeth stand above the outer margin of the axial fossa. 1 he part 

 of the margin next the fossa slopes steeply or is perpendicular; there may or may not be 

 dentations on it. Septal faces granulate. 



Columella laxly developed, relatively small. 



Endotheca and exotheca greatly developed, very vesicular. 



Asexual reproduction by unequal fission; there is no instance of the division crossing 

 the columella. 



Type: No. 38391, U. S. National Museum. 



Locality— Indian Ocean, Dr. W. L. Abbott, collector. Dr. Abbott visited the 

 Seychelles, Aldabra, Gloriosa, and Assumption Islands, all in the western Indian 

 Ocean, but the locality label does not state from which island he obtained the 

 corals he collected. 



This species groups with Favia pallida (Dana), but the septa are very diHerent, 

 as both the descriptions and figures show. 



Genus FAVITES Link. 



1807. Favites Link, Besch. Nat. Samml., Rostock, 3d pt., p. 162. 



1901. Favites Vaughan, Samml. Geol. Reichs Mus. Leiden, 2d. ser., vol. 2, p. 21. 



1902. Favites Verrill, Trans. Conn. Acad. Arts and Sci., vol. 11, p. 92. 



Type species: Madrepora abdita Ellis and Solander = Madrepora favosa Esper 

 plate 45 A, figure 2. . 



In my publication referred to in the synonymy I proposed using t antes Link 

 instead of Prionastrea Milne Edwards and Haime, type species Astrea abdita (Lam.), 

 and Verrill followed my suggestion. 



Favites abdita (Ellis and Solander). 

 Plate 4 o, figure 1, Dana's type of Astraa robusta; figure 2, Dana's type of A.flexuosa; figure 3, specimen identified 

 by Dana as A. fusco-viridis; figures 4, 5. specimens from Murray Island. Also plate 16, figure 31. ot 

 Dr. Mayer's article. 



1846. Astrea flexuosa Dana, U. S. Expl. Exped., Zooph., p. 227, plate 11, figs. 6, 6a-6c. 

 1846 Astrea fusco-viridis Dana U. S. Expl. Exped., Zooph., p. 228, plate II, figs. 7, 7 a-7c. 

 1846. Astra-a robusta Dana, U. S. Expl. Exped., Zooph , p. 248, plate 13, figs 10, loa-iod 

 1914. Favia abdita Matthai, Trans. Linn. Soc. London, 2d ser., Zool., vol. 17, P- 9h plate 9, hgs. 5; 

 plate 29, figs. 1-4; plate 35, fig. 2. 



