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



do show traces of a hardening on the surface, but the raised coral 

 reefs in Viti Levu, closely examined by me, are all overlain by 

 soapstone, a rock of such density and of such a nature that it 

 is very doubtful if it would allow water to percolate through it at 

 all. Considering the large size of the delta of the Rewa, it is 

 obvious that the eastern half of Viti Levu has been upheaved for 

 a considerable period of time, and hence it is only reasonable to 

 suppose that the upheavals which formed the Lau Group must 

 have been extremely ancient. 



In Lau it has been seen that there are many perfect specimens 

 of raised atoll-reefs, of the same form as existing atolls, save only 

 that the slope is not gradual below their summits for some 

 distance and then precipitous, but at once an almost perpendicular 

 cliff. Naiau and Tuvutha are such raised atolls with perfect rims, 

 and lagoons apparently seldom more than 30 fathoms deep ; 

 Mango has the rim broken and is deeper, probably from denuda- 

 tion. Kambara, once a typical atoll, has a broad flat rim, remains 

 of numerous shoals in a flat even lagoon, and many passages in 

 the original reef. Round Tuvutha are cliffs of over 100 fathoms, 

 and these are approached or exceeded by those of many of the other 

 islands. The cliffs are limestone ; in no place except Lakemba, 

 where the raised reef is of the barrier type, is the foundation 

 visible. Considering the known sections of atoll-reefs, the pre- 

 cipitous cliffs to the sea can only be accounted for by the splitting 

 off of great masses of the rock, owing to undermining, and their 

 subsequent solution (or the formation on them of the fringing 

 reef). The reef sections of Funafuti 1 and of Sydney and Canton 

 atolls in the Phoenix Group 2 show that it is only at a distance of 

 150 — 200 yards from the edge of the reef that the steep com- 

 mences. Yet there is no trace of any distance approaching this 

 from the summits to the cliffs in such islands as Fulanga, Ongea, 

 Namuka and Kambara ; indeed the edge of the cliff is often the 

 actual summit of the island. Hence the action of the sea under- 

 mining the cliffs must have been extremely extensive. As the 

 breadth of the fringing reef, supposing it to have been largely 

 built up on the fallen fragments, is in many places not comparable 

 to such an amount of undermining, it necessarily follows that the 

 solution under water of the fallen fragments must for some time 

 at any rate have kept pace with the breakup away of the land. 

 The higher elevations probably represent the highest points of 

 the reef, parts awash or low islands on the rim of the atoll ; when 

 first upheaved, such would naturally be the most consolidated, 

 and hence least liable to solution. 



Denudation owing to climatic causes cannot, I consider, have 



1 Nature, vol. lv. 1897, p. 375. 2 Vide Admiralty Charts. 



