l6 PAPERS FROM THE DEPARTMENT OF MARINE BIOLOGY. 



appears to be composed of these veneering organisms, but in reality it con- 

 sists of dead corals which have been protected from erosion by a thin coating 

 of organic material. The dead-coral element is the dominant one, although 

 hidden from sight, while the veneer which is so apparent is merely superficial, 

 covering the dead surfaces of the corals and filling the interspaces between 

 them in the manner of a cement. 



There is no lithothamnion ridge along the outer edge of the reef which 

 fringes the northwest side of the island, for breakers appear on this shore 

 only during the occasional storms of the season of northwesterly winds from 

 November to March and the incrusting nullipores thrive only where the 

 breakers dash continually over the reefs. 



The northwest reef-flat slopes gradually seaward, the water being about 

 3 feet deep along its precipitous outer edge at spring tide. Near shore one 

 finds extensive grass-flats composed of Posidonia australis, among which 

 are large areas of Monti pora ramosa, which are laid bare by even moderately 

 low tides. The outer edges of this reef-flat are covered with well-developed 

 stocks of Acropora hebes, A. pulchra var. alveolata, etc., the stems of which 

 may project 6 inches or more above the level of low spring tide. Here, as 

 elsewhere around the island, the seaward edge of the reef-flat ends in a steep 

 slope extending abruptly downwards to a depth of 1 5 to 30 fathoms. Fragile 

 corals, such as Acropora and Montipora, which require pure, cool, relatively 

 quiet water, grow upon the wall of this precipice; but the southeast trade 

 wind blew so constantly during September and October that no detailed 

 study of this outer slope could be made. 



On the southeast side of the island, where the reef-flat is widest, wave- 

 worn lava boulders are strewn along the shore and extend outward fully 200 

 feet from mean low-tide line. These boulders have been used by the natives 

 to construct the walls of the numerous fish-traps which extend outward about 

 350 to 390 feet from mean high-tide line. (See plate 5 a and plate 6 b.) ■ 

 The steep lava bluff fronting the whole northeasterly and the more east- 

 erly part of the southeast side of the island is about 20 to 30 feet high, and 

 this (together with the wave-worn masses of lava boulders at its feet) leads 

 one to believe that it has been eroded and encroached upon by the sea. 

 Judging, indeed, from the trend of the fairly regular slope above the shore- 

 bluff (see figs. 2 and 3) the shore was originally about 400 to 600 feet farther 

 out than at present. If this be true the reef-flat has extended seaward about 

 1,200 feet beyond the old shore-line, but in default of a series of borings 

 through the reef this is at best only a matter of conjecture. 



The entire visible reef belongs to the recent period subsequent to the 

 cessation of volcanic activity and has evidently grown seaward over its own 

 talus, so as to widen the reef-platform surrounding the island, and (as is 

 commonly the case in the Pacific) the sea front of the reef-precipice is densely 

 covered with corals, those in relatively quiet water at depths of a fathom or 

 more being mainly foliated, fragile, rapidly growing forms, such as Montipora. 



