1898.] Funafuti, Rotuma and Fiji. 471 



been of sufficient importance to have greatly affected the position 

 of the summits, or it should have left traces of itself in water 

 channels, etc. in the cliffs. It no doubt has been considerable, but 

 I think that the heights of the highest points are not far, not 

 more perhaps than 40 — 50 feet, below their original elevations, 

 when first upheaved. Crosby, in the case of El Yunque in Cuba, 

 has come to the conclusion that " erosion must have diminished 

 the height by as much as 200 feet 1 ," but such an amount is 

 impossible for the Lau Group and seems to me very extreme for 

 that mountain. 



That solution in the sea is taking place very rapidly admits of 

 little doubt from the presence of small lagoons within many of the 

 raised atolls. The beginning is seen in the pools in the centres 

 of Tuvutha, Naiau and other islands; in Mango these have further 

 dissolved themselves out a channel to the coast. Wangava is said 

 to have a tidal lake in the centre, while in Namuka and the two 

 Ongea, the water has broken through the reefs of the former atolls, 

 and opened lagoonlets to the sea ; in their channels the water is 

 deep and within shoals gradually, denoting the extent to which 

 solution has taken place. Fulanga is an example of a later stage, 

 where a perfect lagoon, 10 fathoms deep, has been formed by this 

 means. The islands of its rim are high and protected from all 

 rapid solution oceanwards by the living reef, which now surrounds 

 them. The same action is, however, still going on within the 

 lagoon, the rocks of which, the shoals of the former atoll, are seen 

 to be disappearing one by one, and its ring of land to be lessening. 

 Only on the seaward face is any increase of the land visible, and 

 there is due merely to the growth of sand flats at the base of the 

 cliffs ; such flats, however, I only saw on this one limestone island 

 in the Lau Group. Should the Yangasa Group, as I am led to 

 believe from the appearance of their hills, be the remains . of a 

 raised atoll, they would form another stage, only a few isolated 

 peaks being left. Within the barrier Navatura and Yavutha are 

 connected by a reef awash, while the terraces on Yangasa Levu 

 I regard as caused by denudation in a channel through the 

 original reef; if they represent progressive steps in elevation, they 

 should rather have been separated by perpendicular cliffs. Finally, 

 some of the neighbouring atoll-reefs may possibly represent the 

 last stage, where all the raised reef has gone, and only the living 

 reef is left. In support of this I may here note that there are, in 

 addition, a number of large barrier reefs in the Lau Group, sur- 

 rounding very small islands, or rocks of a similar limestone to that 

 of these raised atolls. 



The thickest cliffs, which I had the chance of closely examining, 



1 On the Elevated Coral Reefs of Cuba. Proc. Boston Soc. of Nat. Hist. Vol. xxn. 

 1882-3. 



