CORALS FROM MURRAY, COCOS-KEELING, AND FANNING ISLANDS. I97 



wall; along its summit are projecting trabecular, which, however, are radially com- 

 pressed. The wall is straight, not zigzag. Where these conditions occur there 

 is usually a well-developed mural shelf below the upper edge of the wall. As in 

 places there is an intercorallite reticulum, it is not practicable to recognize two 

 species. The specimens from 1,400 and 1,600 feet from shore show a great develop- 

 ment of the reticulum which may well up in the angles between calices or form low, 

 curved ridges. The close similarity of P. mayeri and P. viridis is discussed on 

 page 201. Larger suites of specimens may show that the two forms are really 

 only variants of one species. 



Distribution. — Torres Strait. 



7. Porites haddoni, new species. 

 Plate 87, figures I, la, lb, specimen from Murray Island. 

 1905. Porites queenslandia tenia et vicesima Bernard, Cat. Porites Indo-Pacific, p. 127, plate 16, fig. 4; 



plate 21, fig. 13. 

 1905. Porites queenslandia tricesima Bernard, Cat. Porites Indo-Pacific, p. 134, plate 17, fig. 6; plate 21, 

 fig. 22. 



The following is a description of this species: 



Growth-form, rising from a small base, expanding and massive above, with radiating 

 rounded ridges and intervening depressions. Type specimen about the size of a man's fist. 



Calices polygonal, shallow; usual diameter on upper surfaces 1 to 1.5 mm., 8 to 10 

 calices to 1 cm.; in depressions, diameter slightly less than 1 mm. 



Walls, straight, narrow, membranous, continuous, but with perforations; denticles along 

 summit indefinite, often compressed parallel to the mural plane, about twice as numerous 

 as the septa. Surfaces delicately spinulose. 



Septa 12 in number, fairly thick; as thick as, or somewhat thicker or thinner than the 

 interseptal loculi. Surfaces closely beset with granulations which restrict the interspaces. 

 Two synapticular rings within each calice. The outer, which is near but usually detached 

 from the wall, varies greatly in completeness on different parts of the corallum. On the 

 lower part of the corallum it may be highly developed and surmounted by pronounced outer 

 denticles, between which and the wall is a definite excavated ring. The outer ends of the 

 septa between the denticles and the wall are frequently bifurcate. On other parts of 

 the corallum, although the outer synapticular ring is represented, it is only slightly 

 developed and does not form a mural shelf. Here the condition is similar to Bernard s 

 plate 17, figure 6. The inner synapticular ring corresponds in position to the pah. The 

 palar formula may be complete, i. e., one on the solitary directive, one on each member 

 of the triplet, and one at the fusion of each lateral pair; or the pali may be suppressed 

 on the laterals of the triplet, sometimes suppressed on the ventrical directives, and fre- 

 quently there are 6 pali, when trident formation may be obvious. Those before the 

 lateral pairs are the more prominent, reaching to or almost to the level of the upper edge of 

 the wall; they are slender and closely granulate. The other pali are shorter and not so 

 thick, but also densely granulate. Sometimes, but not invariably, there is a denticle 

 between the outer denticle and the palus. .... , ... 



The columella is an erect, narrow, densely granulate plate, joined to the septa by radii, 

 between which the spaces are narrow, but usually open, except near the lower edge of the 

 living tissue. 



Stations, Murray Island. — Southeast reef, line I: 



400 feet from shore; water about 5 inches deep. 



600 to 1,000 feet from shore; water about 10 inches deep. 



650 feet from shore; water about 1 1 inches deep. 



675-720 feet from shore, water about 10 inches deep; bottom sandy, with coral rock (type). 



This species differs from P. australiensis by its membranous wall, the more 

 delicate frosting of the mural denticles, the more definite detachment of the septal 

 denticles from the wall, the presence of trident formation in the triplet, and its 

 lobate growth-form. Bernard pointed out that his P. Great Barrier Reef (42)23 

 and (42)30 might belong to the same species. 



