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



other atolls, can properly exist. Considerable denudation and 

 wearing awa}^ by tidal action has certainly gone on, and I think 

 that its amount is shown by the extent of the rough zone, the 

 pinnacles and buttresses left being the remains of harder, viz. 

 coral-formed parts of the reef. Sand enters considerably into 

 the formation of the islands to leeward, and atoll reefs generally, 

 such as I suppose the atoll of Funafuti to have been before 

 it was elevated, are not so high to leeward as to windward, so that 

 I see no difficulty in the width of this zone there. From the 

 position of the pinnacles it is necessary to suppose that the 

 animals, or plants with the hardest and most durable skeletons, 

 i.e. the corals, were then living on the extreme edge of the reef in 

 the full force of the breakers, as they are declared by Darwin to 

 have been doing on the reefs of Keeling Atoll, when he visited 

 them. The formation of the present reef would of course be 

 subsequent to this elevation. 



The evidence from the facts, bearing on the theory of formation 

 of coral reefs, collected at Funafuti admit of four conclusions : — 

 (1) The atoll has been elevated about 10 feet. (2) The land is 

 now being washed away, the extent of the erosion being shown to 

 some degree by the breadth of the rough zone. (3) The reef 

 seems to have been mainly formed by the growth of nullipores, 

 which are now building up masses outside the rim and adding 

 them on to the reef, causing its extension seawards. (4) The 

 lagoon does not seem to be filling up to any appreciable extent 

 by the growth of shoals, etc., and solution is indicated by the 

 characters of its slopes and floor. 



PART II. 



ROTUMA. 



The island of Rotuma, situated in lat. 12° 30' S. and long. 

 177° 15' E. is about 260 miles distant from the Great Barrier Reef 

 of Fiji, the islands of which group form its nearest high land. 

 It rests on a plateau 1500 fathoms in depth with the Fijian and 

 Tongan islands, extending to the west to the Santa Cruz Group. 

 The island is volcanic with hills ranging up to 900 feet in height, 

 and there are no less than seven craters, the appearance of some 

 of which leads me to believe that they have not been long inactive 1 . 

 From these hills broad lava streams can be traced to the sea and 

 along the beach, except where they are covered by a deposit of 



1 Vide " The Geology of Eotuma," Quar. Jour. Geo. Soc. liv. pp. 1 — 11. 





