1898.] Funafuti, Rotuma and Fiji. 441 



in all essential respects to that off Sororoa, but its breadth is 

 about 50 yards and its channel has a depth of 6 — 9 feet. The 

 platform attains a breadth of about 12 yards, and on it an outer 

 rather higher part can be distinctly recognised. No fissures 

 reach through the platform and, as each breaker drives a large 

 amount of water over it, there results a strong current to passages 

 to the east and south. A few Prionastrea and Madrepora grow 

 on the platform among incrusting nullipores, but there are none 

 in the channel and few nullipores, most of the surface being 

 covered with low green algae. 



As the edge of the shallow reef lies further from the shore, 

 the platform gets broader, as is well seen off Lopta. A distinction 

 into two parts is now marked, an outer 5 — 6 yards broad about 

 9 inches above and an inner at the low tide level. The division 

 is still more obvious in the reef between Solkopi and Pepji, where 

 the outer is about 8 yards broad and the inner 30 yards. These 

 parts, as they seem homologous to the rim and reef-flat at 

 Funafuti, are accordingly so named here. As the reef- flat 

 broadens, fragments, torn from the rim or the reef outside, are 

 thrown on its inner side so that it is itself divided into two parts, 

 the outer of which is equivalent to the reef-flat proper of an atoll, 

 and the inner to the piled-up masses of reef-rock, which in an 

 atoll might be supposed to form the land ; the latter is termed the 

 boidder zone. A section (Fig. 4) then of the reef to the north 

 of the island or off Noatau, where it is about 500 yards broad, 

 presents four zones which with their characteristics are as 

 follows : — 



(i) A boat channel (Fig. 4 bb) extending out for about 

 250 yards from the shore with a greatest depth of 8 feet at low 

 tide. Its bottom is covered by a thin layer of sand and mud over 

 a hard well-consolidated rock, formed of nullipores, corals and 

 sand. The sand is found principally near the beach and is the 

 same as the beach sand, while the mud, which lies towards the 

 outside, is much finer, being formed mainly by boring worms and 

 molluscs. Halimeda is very abundant, but massive nullipores are 

 not found ; corals do not grow near the beach, but the remainder 

 of the channel is studded with large clumps up to 20 feet in 

 diameter. Of these Heliopora and a variety of Millepora are very 

 common in clusters 5 — 6 feet across, consisting of parallel lamellae, 

 arising from broad and firm bases. Symphyllia, Coeloria, Favia, 

 Astvaeopora, Porites and Montipora form large masses, which 

 after they have reached the surface die in the centre, but extend 

 at the periphery so as to overhang for several feet. They are 

 then hollowed out by various boring animals, but their edges 

 extend evenly outwards a few inches below the low tide level, 

 so that even at spring tides the living polyps are not exposed. 



