120 THE CORAL REEFS OF THE MALDIVES. 



growth of an elongated ring of corals enclosing an interior lagoon. The 

 great land rims of the Paumotus, however, have been cut down from ledges 

 once of a greater height, upon the outer faces of which corals have grown ; 

 while in the Maldives the flats are made up of recent corals which have 

 grown up from the base forming the underlying plateau of the Maldives 

 and have only been slightly elevated. 



As we go south from Mulaku the Maldivian character typical of the 

 island groups gradually disappears more and more. A mere glance at the 

 chart shows this conclusively (PI. 1). Neither Kolumadulu nor Haddum- 

 mati would be chosen as typical representatives of Maldive groups with 

 their characteristic rings and faros ; they combine Maldivian features with 

 extensive narrow reef flats, characteristic of Pacific atoll-lagoon reefs. 

 The northeast horn of Haddummati reminds us of the eastern horn of 

 Arhno in the Marshall Islands. The Maldivian type is still less apparent in 

 Suvadiva and Addu, an atoll on the southernmost plateau of the group, 

 such as might be found either in the Paumotus or the Ellice, Gilbert, and 

 Marshall Islands. 



It is true that within the area of Mulaku there are more than one 

 hundred flats and rings, but they are small and do not compare with the 

 characteristic larger, circular, or elliptical, faros, often several miles in length, 

 crowded in North and South Male, Ari, and within the basins of other north- 

 ern groups. The absence of flats and faros within the area of the southern 

 groups is striking in Kolumadulu, still more in Haddummati and Suvadiva. 

 The aspect of the southern enclosed basins (lagoons) reminds us of such 

 large lagoons as Fakarava and others of the Paumotus, where there are 

 either numerous small islets or only shoals to be seen in the great ex- 

 panse enclosed within the outer land rim. 



The western face of the western faros of Mulaku is edged by stretches of 

 small coral boulders and coarse shingle set in transverse dikes running at 

 right angles across the outer rim of the faros, they are more prominently 

 heaped up on the two sides of the passes separating the faros (Pis. 64, 

 fig. 1 ; 65, fig. 1). These stretches of boulders and shingle are largest on 

 the western face of Kureli Faro ; they are usually mainly masses of Pontes 

 tailing into long sand spits towards the east. When seen entering the pass 



