494 Mr Gardiner, The Coral Reefs of [Mar. 7, 



Where land is formed on the reefs of lagoons, other agencies, 

 such as the deposition of sand, drift, etc. come into play, and 

 the whole lagoon may be finally converted into land. The Ellice 

 Islands show the stages of this conversion, which depended largely 

 on the size of the atolls, and the number and depth of the passages 

 through their rims. Funafuti is the largest, and has six deep 

 passages, one of 24 fathoms, and a general depth of 25 fathoms 

 in the lagoon, with only a deposition of sand in one place off the 

 large island to windward. Nukufetau has only one passage, the 

 depth of its lagoon is about 22 fathoms, with a sandy bottom in 

 most parts. Nukulailai has more land, broader reefs, no passages, 

 and a depth of 10 fathoms with sandy bottom. Oaitupu is nearly 

 all land, there being only a small lagoon with a depth of 1 — 2 

 fathoms and a narrow channel to the sea at its north end ; this 

 island is completely covered by sand, except along the coast. 

 Niutau, another sandy island, has in its centre a small mangrove 

 swamp, which is dry at low tide. Gente Hermosa, one of the 

 Union Islands, is of interest, having in its centre a lagoon a few 

 hundred yards across with precipitous sides, a depth of 5 — 6 

 fathoms and a small island in its centre ; there is also, slightly 

 to the north, another smaller lagoon with a similar depth ; unlike 

 the other islands which are covered with sand, the solid rock is 

 here continued from the reef to the lagoons, and it is probable 

 that the atoll owes its land to a different mode of formation 1 . 

 There is no evidence that the channels of the atoll and barrier 

 reefs are filling up in Fiji, but if they are doing so, it is certainly 

 not by the accumulation of debris in them, but rather by the 

 growth upwards and inwards of corals and nullipores on their 

 bottoms and sides. 



Lastly, the waves " accumulate on the reef material for beaches 

 and dry land." Certainly in the boulder zone at Rotuma an 

 accumulation of this kind is going on ; it is also especially notice- 

 able on the reefs to the south-east of Viti Levu and on the 

 Mumbolitha reef to the south of Ngau. Most of this debris comes, 

 according to Dana, from the rim and outer portion of the reef, but 

 even narrow fringing reefs have these parts, and yet the land they 

 surround does not show much debris piled up, or any extension 

 outwards by its means. Such an accumulation may take place 

 but, in my opinion, both Darwin and Dana have greatly over- 

 estimated the amount of the debris thrown up by the waves. 



Semper, Murray, and other observers, lay great stress on the 

 presence in the same group off different sides of the same islands 

 of fringing and barrier reefs. Semper 2 , dealing with the Pelew 



1 I am indebted for this information and much of that about the Ellice Group 

 to Capt. Wilson. 



2 Animal Life. By Carl Semper. 1881. Chap. vm. 



