[)0 APPENDIX I. 



No dense sub-base to this island overlying the platform was anywhere observable ; but there does 

 exist, much nearer the ocean margin than the present island, in addition to the rorifcg bosses, in a few 

 places over a small area, a very thin breccia-like covering scarcely distinguishable from the platform 

 itself. 



The whole surioundings of this island indicate that, as already stated, its existence is due to the 

 exceptional width there of the platform and to the slight protection of the few Voritcs and i-emaining low 

 bi'eccia elevations, though even this is not sufficient to insure its stability. The few masses of Pontes 

 with the imtches of breccia, while they mark the site, also represent the noM' almost obliterated foundation 

 of the former island which has been eroded, the finer material being driven further and further lagoon- 

 wards till it has reached its present position, which serves to mark the distance it has shifted from its 

 earlier site. It also suggests that any extension of this earlier island or of others existing either north 

 or south, or both, would probably have fared similarl}' : Init to have removed so fai' as this would have 

 carried them beyond the platform, into the lagoon, which is undoubtedly what has happened. Now it is 

 to be observed that nearly all along the western rim and southwards, the Admiralty Soundings indicate 

 unusually shallow water from close up luider the lagoon reef to a varying distance out from it, thus 

 accounting for the material removed from the islands which at one time almost certainly occujned 

 positions on this reef. 



Except in the case of channels, the spaces lietween the islands are occupied by a continuation of the 

 reef platform on which thev rest, while heie and there along the middle or outer portion of these platforms 

 on the west side of the atoll, lines of low scarps exist coriesponding with those on the wide platform 

 north of Mulitefala, with pinnacles and a thin worn and rugged breccia sheet, in some ])laces not yet all 

 removed. Beside this, there are blocks of blackened coral to be seen, at intervals all along, now 

 numerous, now rare ; these, the so-called " negro-heads " of some writers, can be examined at many points. 

 On this side they ajjpear to be chiefly ruritc^, which are generally /;; situ ; and while present at other 

 parts of the atoll, their j^i'esence is often conspicuous on the western reef platform, from the compara- 

 tive absence of land. Here only a few of them were examined, but those so testeil were, as usual 

 on this side, found to l>e big bosses of rviitcs firndy embedded in and projecting through the platform 

 as well as through the breccia also where it is still present. 



There can, I think, be no doubt that these low scarps and pinnacles with the J'uritrs bosses antl stumps 

 are the lingering remains of the foundations of former islands. It will be noticed that only where these 

 rocks are nimierous or when there is a consideral^le mass of the breccia sheet not yet eroded away, will 

 there be an island or a sand cay, each of which rarely exists without one or other of these defences, or 

 else a wide expanse of reef platform between it and the ocean. 



TEFALA. 



This islet, like Fuagea, stands well up abo\e the M-aves, its highest point being about 9 feet abo\e high 

 water. It supports a fair number of cocoanut palms, though it is conspicuous for its Fandanus (Fala) 

 trees, whence its name. It is free from fallen trees on the maigin of its surface, and is not so stdjject 

 to variations of oxitline as Fuagea, besides being more sheltered than it fiom rough seas (Plate i). 



The material of its upper portion is throughout like that of the last islet, sand and small fragments, 

 with the addition of some larger old coral fragments scattered on the surface, and accumulated 

 in some quantity on a part of the S.W. beach. Its surface is even moi'e fiat than in the last, fewer 

 undulations 1)eing present, though three or four are prominent at the east and S.E. jjortion. This pait 

 is exposed to the force of easterly gales across the lagoon — as my experiences there, limited as they weie, 

 sufficiently proved to me. IStich gales as can exert their full force on the island, across the lagoon and 

 over the narrowest part of the lagoon reef on to its beach, aie, however, not common, but that they 

 do o:cur is proved by the undulations on the upper mai'giii of the island, and the sectional outline 



