1898.] Funafuti, Rotuma and Fiji. 485 



be deposited, rather than in the passages themselves in which the 

 absence of corals is more probably to be accounted for by the 

 inability of the larvae to fix themselves. The ship's entrances at 

 Fulanga and Lakemba in the Lau Group of Fiji, where tidal 

 currents of 4 — 5 knots an hour are met with, are very free from 

 corals, which yet live in great luxuriance off their entrances as 

 deep down as the bottom can be seen. At Funafuti, iu the ship's 

 entrances, the genus Millepora is extremely abundant, but true 

 corals are rare. 



With regard to ocean currents, which, in some groups of islands, 

 are very powerful in certain positions, I think that they probably 

 modify considerably the form of the whole reef, by preventing the 

 proper fixation of the larvse both of corals and nullipores, and by 

 the distribution of the talus on the reef slopes. The form of the 

 channel, between Wakaya and Ovalau, is probably due to this, the 

 slope being very sharp down to 40 fathoms off each island, where the 

 steep is reached, the reefs on either side running also very evenly, 

 and with few projecting points. Such currents as cannot pass this 

 way must pass to the south of Viti Levu, and all along the coast 

 here the reef lies close to the land. The current between Taviuni 

 and Vanua Levu is not altogether tidal, but probably has had a 

 still greater effect on the reefs. I cannot entirely agree with 

 Semper 1 , when he states his opinion that currents are one of the 

 principal causes which have given the reefs their very remarkable 

 forms. Given a bank at a suitable depth, I think that it would 

 require a very powerful current, a current such as is only found in 

 narrow straits between islands, to prevent a reef from forming on 

 it in suitable latitudes. 



Where currents are not too strong, they assist no doubt very 

 materially in the formation of the reefs in some way. Most atolls 

 have their land to windward and their passages to leeward ; 

 Cliarybdis reef to the north of Viti Levu, a case in point, is well 

 protected against the wind, but its reef is, apparently owing to the 

 current, especially well formed to the east but quite open to the 

 west. In the Gilbert Group the current and wind both come from 

 the east, on which side also the land generally lies. In the Marshall 

 Group, the current sets east and the prevailing wind is east-north- 

 east, the two being thus opposed to one another ; the atolls have 

 therefore land both on their windward and leeward sides. 



At Rotuma corals are found in greatest abundance in the boat 

 channel at points intermediate to passages in the reef. Before I 

 observed this, however, I removed specimens of a number of 

 different species down the coast for about half a mile to a position 

 almost opposite to a big passage — from an area very rich in corals 



1 Animal Life. By Carl Semper, 1881, p. 229. 



