1898.] Funafuti, Rotwnia and Fiji. 495 



Islands in particular, considered the fact of great importance, 

 and discussed very fully its bearing on the theory of subsidence 

 joined with the complicated system of up and down movements, 

 that would have been required under this theory to explain their 

 formation. For the Fiji Group Moore 1 and Murray' 2 have already 

 dealt with Kandavu. Wakaya has to the west off its highest 

 summit (590 feet) a narrow fringing reef, but elsewhere a broad 

 barrier reef; the land has cliffs in accordance with this to the 

 west, but toward the east near the sea there is a gradual slope 

 of about 10°. Assuming the slope to be only 5°, Dana's more 

 moderate estimate, the thickness of this reef would be 1840 feet, 

 and the cliff to the west, if we accept his theory of subsidence, 

 must have been 2345 feet high. Mbatiki, Ngau, Nairai, Ma- 

 kongai, Moala and Matuku, all of which are volcanic islands, slope 

 to the sea fairly evenly on all sides, and show every variety of 

 reef from a narrow fringing to a well-formed barrier. Lakemba 

 slopes evenly from its two central peaks ; its reef to the west 

 fringes, and to the east is a barrier, inclosing a lagoon 5 miles 

 broad. The greater part of the coast of Taviuni has no reef at 

 all or only a narrow fringing one, yet a barrier reef lies off Vuna 

 to the south, and to the north-east, including also the islands of 

 Ngamia and Lauthala ; northwards there is a regular line of shoals 

 enclosing an area with a depth of 40 — 50 fathoms. The slope 

 from the hills of Taviuni is in all directions very uniform, and it is 

 inconceivable that a subsidence of this kind of 300 feet (calculating 

 it only from the depth of the water inside the reef to the north), 

 has left no trace of any reef along the greater part of its coast. 

 There is no difference between the appearance of the hills north 

 and south of Vanua Levu, or in their rock ; there is, however, a 

 far distant barrier reef to the north and along the greater part 

 of the south coast, a narrow fringing reef. 



I submit that the above observations afford strong reasons 

 against Darwin and Dana's theories of the structure and forma- 

 tion of coral reefs in general and of those of Fiji in particular. 

 As Dana spent three months in the Fiji Group, and had a very 

 unique experience of coral islands, I have given his account 

 more weight than that of Darwin, which has the priority. 

 Although it is very unlikely that subsidence has been one of 

 the determining features in the formation of the reefs of Fiji, 

 or of most of the other reefs of the Pacific, I consider that, if a 

 slow subsidence as is suggested, or a series of rapid subsidences, 

 as would be more probable, took place, barrier and atoll-reefs 

 would undoubtedly result, not, however, of the internal structure 

 as suggested by Dana. My opinion is that the growth of the 

 debris and sand is not so great, as Dana has suggested, and 

 1 Nature, xl., 1889, p. 203. 2 Nature, xl., 1889, p. 222. 



39—2 



