ARTICLES 439 



give excellent views of the tapering points and re-entrant em- 

 bayments of the coast. It may be safely assumed that both 

 the sea-level and the elevated reefs are unconformable on the 

 granitic spurs that slope down between the embayments ; for 

 granite is not a surface-made rock ; and even if it were, its 

 present spur and valley form could not be its original form ; 

 but no description of the island that I have found makes explicit 

 mention of these significant inferences. 



When the deduced consequences of various coral-reef theo- 

 ries are confronted with the facts regarding the Seychelles reefs, 

 it becomes plain that no theory can be accepted that does not 

 involve submergence preceding or during reef formation, as 

 well as long-continued erosion before submergence. Whether 

 the sea-level reef has been formed directly after the uplift by 

 which the elevated reef gained its present altitude, or whether 

 the uplift has been followed by a new subsidence cannot be 

 determined from the scanty records now available ; but in any 

 case strong submergence, explicable only by subsidence, must 

 have preceded the formation of the elevated reef. Hence all 

 so-called still-stand theories are excluded from application here. 



Furthermore, the features of the Seychelles include three 

 conditioning factors : the extraordinary area of the vast sub- 

 marine bank, the great volume and long duration of erosion 

 implied by the unconformity of the reefs on their sloping 

 granitic foundations, and the special changes of level involved 

 in the formation of the vast bank and the two fringing reefs. 

 The changes of level thus demanded are manifestly unlike the 

 changes recorded on other reef-encircled islands in the Indian 

 and Pacific Oceans ; hence these factors cannot be satisfied 

 by any theory of coral reefs which postulates stationary reef 

 foundations in an ocean that rises and falls everywhere by the 

 same amount. The vast area of the submarine bank, and the 

 great volume of erosion before the sea-level and the elevated 

 fringing reefs were formed, both demand a longer period of 

 emergence than can be provided by the Glacial-control theory ; 

 moreover, the granitic islands of the Seychelles group are not 

 cliff-rimmed, as they should be if the submarine bank were the 

 work of abrasion by the lowered and chilled ocean, as the 

 Glacial-control theory assumes. 



Again, the narrow sea-level fringing reef around Mane* is 

 so unlike the three-mile reef plain adjoining Rodriguez island 



