18981 '^^HE BROADENING OE A TOLL-ISLETS 175 



higli bank of coral shingle facing the waves, then a gentler, longer 

 slope inland, after this another shingle bank, less steep and less 

 high, followed by a similar slope, tlien another and perhaps still 

 another bank, each lower than its predecessor, the last scarcely 

 perceptible. The coral blocks of which these banks are built vary 

 according to their situation ; on the seaward side they are pebbles 

 and boulders, mostly so ground and worn as to be scarcely recog- 

 nisable as corals, but on the banks further inland they are rough 

 and much etched by the weatlier, there they are also brittle from 

 ilecay. 



The history of these successive banks of coral debris seems 

 obviously to be, that each was piled up by the ocean waves and was 

 afterwards shut in by its eastern neighbour, built up to windward 

 at a later date. The lower present level of the inner banks must be 

 the result of collapse of its material, comparable to the collapse that 

 occurs in an unused heap of road metal, while the more advanced 

 state of decay of that material, previously noted, likewise points to 

 its superior antiquity. While the upper parts of the hurricane 

 beaches thus change, their lowest stratum may by solution and 

 deposition be welded into breccia, as a heap of hail may be 

 turned by slight melting and freezing into a solid mass. 



These wave-like rows of shingle appear to be an ordinary 

 phenomenon of atoll structure and have been noted by many , 

 observers. Guppy ^ saw on the windward angle of Keeling Island 

 a massive looking slope of large blocks of coral forming a kind of 

 glacis. On North Keeling Island he was informed that the shore 

 on the landing place on the west coast had advanced some fifteen 

 or twenty paces during the previous ten or fifteen years, the old line 

 being indicated by the overhanging cocoanut palms removed that 

 distance from the sea ; it was largely the work of a single night, 

 a huge pile of coral blocks being piled on the beach during a 

 cyclone. 



Speaking of the atoll of Peru, one of the (xilbert or Kingsmill 

 Archipelago, Whitmee ^ says, " The Island itself is formed of suc- 

 cessive ridges of sand, broken coral, and shells. These ridges are 

 most of them from thirty to fifty feet across, and the hollows 

 formed between them are generally from four to six feet in depth. 

 For some distance, at that end of the island which I examined, they 

 run across, and in the middle they run parallel with the sides of 

 the island. The whole extent examined presented the same 

 appearance, and the ridges were so regular that they gave one 

 the idea of being artificially formed. The waves must exert a 

 mighty force during heavy weather to form these extensive ridges. 



1 Guppy, Scot. Geo(jr. Mag., vol. v., 1889, pp. 296 and 468. 



- Whitinee, "A Missionary Cruise in the South Pacific," 1871, p. 35. 



