1898.] Funafuti, Rotunia and Fiji. 475 



islands to windward seem to have been formed by an aggregation 

 of boulders, and their subsequent consolidation underneath by the 

 deposition of carbonate of lime, caused by the evaporation of the 

 spray and salt water at low tide. The large island to leeward 

 appears to be a large sand cay on the reef, formed partially by 

 fragments of reef-rock, but principally by sand washed up from 

 the lagoon and formed into a bank by the assistance of the wind. 

 The lagoon has a depth of about 10 fathoms, and is stated by the 

 natives to be of a basin-like form, rather than a flat with pre- 

 cipitous walls. As the tide falls, the water rushes out of the 

 lagoon by a few narrow passages, which, however, would seem to 

 be insufficient to keep its water at the same level as that of the 

 external ocean ; hence it is probable that (as the natives stated) 

 there are tunnels from the lagoon, which assist in carrying away 

 the water. It is interesting to note that on the western island of 

 Nukulailai there is a spring of pure fresh water, which, according 

 to the natives, is continually running. This probably indicates 

 that the foundation of this atoll is situated at no great depth. I 

 was informed by Capt. Wilson that Arecifos, or Providence Island, 

 to the east of the Marshall Group, has also a spring of clear fresh 

 water in the coral-rock, and that this and Nukulailai are the only 

 coral islands in the Central Pacific, where such are found. 



The appearance outside a reef is that of a number of more or 

 less overhanging and anastomosing masses of rock with pools 

 between, having sandy bottoms. The same appearance is retained 

 on protected reefs, until they become awash, when by the growth 

 of corals and nullipores the masses fuse, and a smooth platform is 

 formed. The inner part of the Great Barrier Reef of Fiji consists, 

 where it is single, of deep pools, separated from one another at the 

 surface but joined below. Inside the more solid outer part of the 

 reef here mushroom-shaped heads of coral rock grow up, anasto- 

 mosing with one another at the surface to form an almost solid 

 platform, resting on so many large pillars. I have obtained, in 

 pools on the inner parts of the Great Barrier Reef and of the reef 

 to the west of Wakaya, 5 — 6 fathoms of water, and at the same 

 time noticed that the tide sets strongly through them, indicating 

 a free connection with the ocean beneath the surface. Outside 

 reefs, where they are exposed to the ocean, there seems to be the 

 same process as within the Great Barrier Reef, but the joining 

 organisms are almost entirely nullipores ; in the account of the 

 outer reef of Funafuti I have shown how it is going on at the 

 surface, and it certainly seems to be taking place outside the 

 rim. At greater depths the same formation also, I believe, 

 occurs ; the corals and nullipores cannot grow upwards with 

 such vigour as inside the reef on account of the strong back- 

 wash of the waves which break on the reef, and so grow relatively 



