GEOLOGICAL OUSKUVATIONS IN FIJI. 81 



the floor of an old crater. The hi<;h peaks at the western side of the 

 bay are tlie edges of flows dipping away from tlie crater. Asli-beds 

 which form the hills to the east of the bay also show quaquaversal 

 dips. 



Looking from the Beacon, the highest summit in Mbengha, towards 

 the bay, several minor hills may be seen rising within the depression. 

 As seen from the north, these hills appear to enclose the southern side 

 of the bay, but actually the depression encircles the hills and a very 

 slight subsidence would cut off the eastern side of the island and 

 transform the hills into islands. 



Other irregularities in the coast are believed to have an origin 

 similar to that of Malumu Bay. They are irregularities, formed by 

 the rapid erosion of weak ash-beds, and were not formed by the em- 

 bay ment of river valleys. The streams entering these bays are 

 torrential, with boulder-covered beds within a hundred yards of their 

 mouths. The evidences of embayment do not warrant the assump- 

 tion that the island has subsided to any great extent; certainly not 

 enough to support the view that the subsidence of Mbengha has con- 

 verted a former fringing reef into the present barrier reef. 



It has been shown, in the chapter on Viti Levu, that the larger 

 island has experienced a number of earth movements, during which 

 a lagoon floor was produced extending westward to the Yasawa 

 islands. The Yasawa islands have been built up on this platform. 

 It is thought probable that Mbengha also is a Post-Pleistocene vol- 

 cano built on the ruins of an older one, which had been greatly eroded, 

 then submerged and levelled by sedimentation and wave-cutting 

 during a period when corals did not flourish. 



The sub-mature topography of the island and the slight lateritization 

 of its surface rocks show that the island is not very old. It appears 

 to be of the same age as Munia and the more recent part of Kanathea. 



The absence of signs of extensive submergence suggests that the 

 younger INIbengha volcano, constituting the existing island, became 

 active near the close of the Glacial epoch. Since that time the tropical 

 climate has caused deep erosion, especially on the windward side. 



Kandavu. 



The group of islands of which Kandavu is the largest lies within 

 the Great x\strolabe reef, some 40 miles south of Viti Levu. Kandavu 

 is the southernmost of the group and extends for 15 to 18 miles in an 

 east-west direction. A few miles off its northeastern shore is the 



