GEOLOGICAL OBSERVATIONS IN FIJI. 55 



of an atoll. Some of the higlu>r volcanic peaks of Vanna Mhalavu 

 may have risen above the surface of the bank, but the indications are 

 that the andesitic hills were entirely covered by limestone. The 

 submerged surface of this bank may or may not have been as flat as 

 the present lagoon floor. The irregular surface of the sunken land 

 woidd, initially at least, invohe considerable variations in the depth 

 of water on the shoal. 



Subsequent elevation brought great masses of the coralliferous 

 limestone above the sea. Much of this limestone has been sv/ept 

 away by atmospheric solution and wave erosion, as is shown by the 

 isolated islets of limestone dotting the present lagoon. The depths 

 within the lagoon and the distribution of the limestone remnants 

 suggest the possibility that the present andesitic hills represent only 

 the high points of the older topography; that the uplift only served 

 to shallow the depths over the bank and not in all parts to elevate 

 its surface above sea-level ; and that the subsequent basaltic eruptions 

 and long continued erosion have furnished material with which the 

 lagoon floor has been smoothed. 



During uplift the coral would continue to grow about the edge of 

 the elevated limestone mass. When uplift ceased, the destruction of 

 the limestone by atmospheric solution and wave cutting would go on 

 apace. If these limestones were not protected by a reef during the 

 Glacial epoch, they must have been rapidly eroded. Certain it is 

 that the elevated coralliferous limestones have disappeared from the 

 edge of the present reef and only isolated remnants remain within 

 the lagoon. 



The depths between the isolated patches of elevated limestone are 

 exceptionally great when compared to the average lagoon depths. 

 They reach 90 to 100 fathoms at the eastern border of the lagoon, 

 depths too great to have been produced by wave-cutting. Submarine 

 solution cannot be postulated for reasons which are well expressed 

 elsewhere, (see Vaughan, 1914, p. 27; Daly, 1915, p. 231; Davis, 1914, 

 p. 576). Submergence furnishes the only valid explanation. 



Direct evidence is lacking on the question whether this submer- 

 gence was due to actual subsidence or to a rise of the ocean level after 

 the Glacial period. The re-distribution of the loose volcanic products 

 emitted during the basaltic period of eruptivity, and the erosion of the 

 basaltic cones to maturity have furnished material sufficient to level 

 the lagoon floor to a considerable degree. Even today the lagoon 

 floor has an exceptionally irregular topography but these irregularities 

 may be residuals of the eroded topography of the andesites. Portions 



