CORALS FROM MURRAY, COCOS-KEELING, AND FANNING ISLANDS. 205 



Although it is to be expected that either this or closely related species occur 

 generally in the tropical Pacific, I have seen no named species to which it can 

 positively be referred. It is distinct from any of Dana's species represented by 

 types in the U. S. National Museum, but is close to P. cylindrica, which is the next 

 species to be described. Dana's type of P. levis is not in the National Museum. 

 Bernard identifies his Porites Fiji Islands (24)1 with P. levis {op. cit., p. 43, plate 

 2, fig. 2). This is almost certainly what I am here calling P. andrewsi. The same 

 species is represented by Bernard's Porites Tonga Islands (10)8 and (io)9- Dana's 

 figure of the calices of P. levis (plate 54, fig. $c) represents septal granules different 

 from those of P. andrewsi, i. e., there are not the usual two clear-cut rings of granules 

 between the pali and the thecal summits; and the growth-form of P. levis is different. 

 It is probable that Bernard is correct, but unless it is more positively shown that 

 the two actually intergrade, I prefer to consider P. andrewsi as distinct from P. levis. 



Distribution. — Great Barrier Reef; Fiji Islands; Tonga Islands. 



16. Porites cylindrica Dana. 



Plate 92, figures 3, ]a, Dana's type of the species. 

 1846. Porites cylindrica Dana, U. S. Exp. Exped., Zooph., p. 559, plate 54, fig. 4. 



This species is so very close to P. andrewsi that for some time I was undecided 

 as to whether they should be separated. The growth-form is similar enough for them 

 to belong to the same species, and the fundamental skeletal plan is the same in both, 

 but the septal granules inthetwo forms are constantly different accordmgtothe speci- 

 mens available for study. They are smaller in P. cylindrica, less regularly developed, 

 and not arranged in conspicuous rings separated by circular depressions as in P. 

 andrewsi. However, it does not appear improbable that P. levis, P. cylindrica, 

 and P. andrewsi may all belong to the same species. 



Type: No. 708, U. S. Nat. Mus. 



Locality. — Fiji Islands? 



1 7. Porites nigrescens Dana. 



Plate 91, figures 3, 3a, Dana's type of the species; plate 92, figures I, la, lb, specimen from Cocos-Keeling 

 Islands; figures 2, 2a, probably a variant of the species, from Cocos-Keeling Islands. 



1846. Porites nigrescens Dana, U. S. Expl. Exped., Zooph., p. 557, plate 54, figs. I, la, lb. 

 1905. Porites fidjiensis octava Bernard, Cat. Porites Indo-Pacific, p. 51. 



Dana's type is in the U. S. National Museum, No. 691. As good descriptions 

 of the form have been given by both Dana and Bernard, that feature of the corallum 

 need not be treated here. The following is a description of other features of Dana's 

 type: 



The intercorallite walls are wide and flaky, usually without a well-defined wall-ridge, 

 which, when present, is represented by a slightly elevated broken row of curly flakes. The 

 calicular cavities are subcircular in outline and are slightly sunken below the flaky reticulum. 

 Distance between thecal summits about 2 mm.; diameter of calicular openings about 1 mm. 



The septa are thick, triangular in horizontal outline, outer ends wide, flaky, with a 

 curled denticle near the wall; sides granulate; interseptal loculi narrow; number 12, formula 

 complete, members of triplet ending separately. Typically 8 palar knots, those before the 

 lateral pairs somewhat the larger. Palar ring of synapticulse irregularly developed. 



Columellar tubercle a small, twisted knot. Columellar tangle irregularly developed, 

 as the ring of palar synapticuloe is not complete. 



In the specimens from Cocos-Keeling Island this species is represented by 

 pieces of three colonies. Those from two colonies differ from the type principally 

 by the branches being more terete in cross-section (see plate 92, figs. 1, la, lb), 

 not anastomosing, and by delicate, curly, lace-like tracery along the mural summits. 

 The latter difference may be due to the state of preservation of the specimens. 



