CORALS FROM MURRAY, COCOS-KEELING, AND FANNING ISLANDS. I37 



larger septa extend to the edge of the columella fossa, where their margins curve, drop per- 

 pendicularly, and fuse to the columella down in the calices. 



Columella usually well developed as a solid, more or less compressed, stout, prominent 

 style, but sometimes represented only by fused septal ends. 



Habitat, etc., Cocos-Keeling Islands. — Dr. Wood Jones says regarding this 

 specimen: 



"From east end of Pulu Tikus, on lagoon rocks, in water I fathom deep at low tide. 

 This colony was the only one found, nor were any dead fragments picked up. The growth- 

 form is as upright, thin lamina?, which were pale but bright. Color yellow. The colony 

 was very conspicuous beneath the water. When dived tui it was broken with great diffi- 

 culty with a large hammer, as it was very hard." 



As the original of Dana's Pavonia boletiformis (non Lamarck) =type of Milne 

 Edwards's Lophoseris danai is in the U. S. National Museum, No. 135, it is figured 

 (plate 56, figs 2, 2a), and the following notes are based on it: 



The form and size of the type are sufficiently indicated by the figures. The fronds are 

 dissected, the lobes curled or twisted, edges sharp, and carina; are present. Calicinal series 

 are only fairly obvious. There may be no calices within 10 mm. of the edge; farther down 

 on the frond the distance from the calicinal centers in one series to those immediately above 

 in the next series is from 2.5 to 3 mm.; the distance apart in the same series is from 2 to 2.5 

 mm. The ambulacra are crossed by strongly alternating septo-costae, between which 

 synapticulae are obvious. The costal edges are sharp. From 6 to 9 prominent septo-costa; 

 extend as septa to the columella axis, and alternate with much smaller septa. Near the 

 edge of the frond the columella is absent or obscure, but in older calices it is well developed 

 either as a flat calcareous plug or a lamellate style, compressed in the direction of the axis 

 of the series. 



A comparison of the description of Dana's type with that of the specimen from 

 Cocos-Keeling shows that their characters are fundamentally similar, and that the 

 principal difference consists in the apparent absence of carinae on the latter. Dana's 

 type is from Sulu Sea. Fortunately the U. S. National Museum possesses three 

 good specimens and several fragments of P. danai, collected by J. B. Steere in 

 Zamboanga, Philippines. These specimens (which are as nearly typical as can be 

 imagined, one being almost a duplicate of the type, except that it is a more nearly 

 perfect specimen) are young colonies; but one shows widening of the fronds by 

 peripheral fusion of the lobes, and the carinae become obscured with the thickening 

 of the lower part of the frond. Two other specimens (one from Zamboanga and one 

 from Mariveles, Luzon), which are older than the three colonies mentioned, show- 

 widening of the fronds with increasing size and nearly bridge the gap between the 

 type of P. danai and the specimen from Cocos-Keeling Islands. The gap is com- 

 pletely bridged by an excellent specimen collected by Professor W. A. Bryan in the 

 Caroline Islands. 



Pavona laxa Klz. 1 is almost certainly this species. Klunzinger's Pavonia 

 angularis is surely the same. Ridley referred to this coral as Pavonia lata Dana 

 in his list of corals collected by Forbes in the Cocos-Keeling Islands. 2 The species 

 is closely related to both P. decussata and P. lata, the principal difference being in 

 the greater lobation of the plates. Therefore both Ridley and Bedot put the species 

 in the proper place in the genus. 



Distribution. — Red Sea; eastern Indian Ocean; Amboina (Bedot); Sulu Sea 

 (Dana's type); southern Philippines (J. B. Steere); Luzon, Philippines (A. M. 

 Reese); Caroline Islands (W. A. Bryan). 



>Korall. Roth. Meer., pt. 3, p. 73, 1879. 



2 Forbes, H. O., A naturalist's wanderings in the Eastern Archipelago, p. 47, 1885. 



