132 -MR. A. E. FINCKH. 



The islets on the two sides of the atoll have themselves their characteristic 

 difFereuces. Iii hoth cases they are accumulations of debris on the rim, and it is this 

 debris whicli enables us to classify them into leeward and windward islets. The 

 former agree in their constitution entirely witli their present biological surroundings, 

 being large accumulations of the fragments of the branching LitJiotli amnion. Inter- 

 mixed with this is a quantity of foraminiferal sand, as well as fragments of the frailer 

 corals at present growing in the neighbourhood, together witli Hcdimeda remains and 

 shell fragments, the whole forming a coarse sand. These large heaps of sand begin 

 quite abruptly on the platform witliout the intervention of a breccia"'^ zone, and this 

 accounts for the fact that the shapes of the islets change from time to time, as stated 

 by the native inhabitants. 



The windward islets are, as already stated, also accumulations of debris, l)ut this is 

 not, as it is in the case of the leeward islets, the direct product of living organisms. 

 The boulders and pebbles heaped up on tlie H\u-ricane Bank indicate that they were 

 once part of the general LithutJiamnion rock. Sand on these islets is the exception ; 

 it consists, where present, mainly of Hcdimeda remains together with a large 

 percentage of foraminifera. 



The slopes lagoon wards are in all localities sandy ; purely Ilalimeda to the east, 

 and containing on the west besides the Halimeda plentiful remains of LitJiothamnion. 

 The bottom of the lagoon within a very short distance of the rim, in every instance 

 where dredgings were carried on, was found to consist of almost pure Halimeda remains, 

 with a small admixture of foraminifera, as already described by Mr. F. Chapman. 



The examination of the ocean slojies is a somewhat difficult matter. Iji four diii'erent 

 localities, however, satisfactory information was obtained— at the island of Funafuti 

 near the 1897 bore ; off the rim between tlie islets of Funafuti and Fatato, at the 

 islet of Pava to the noi'th, and off Fualo])a on tlie western side of the atoll. 



The observations at the first of these localities have been discussed in the preceding 

 sub-section. The other localities differ only in minor details from it. Between 

 Fmiafuti and Fatato, where tlie reef rim, as described above, is very narrow and 

 consists of a "flat" from lagoon to ocean face, there is no slope oceanwards from the 

 rim surface, but the reef flat with its irregular edges, its buttress-like projections and 

 its intermediate channels, ends in vertical walls one or two fathoms deep, at which 

 depth the general submarine slope of tlie atoll begins. Considerable quantities of 

 isolated coralla of Pocillopora (?s^a) exist here. A few specimens of the mushroom 

 coral, Montipora incognita, are present, as well as the only exam})le of a " lirain coral " 

 (about 4 feet in diameter) noticed on tlie atoll. There is no sand jiresent, liut pebbles 

 lie in the hollows of the uneven slope. As the depth increases, the ground becomes more 

 even and shows many small isolated growing corals, with here and there a patch of 

 Halimeda and non-calcareous green seaweeds. 



* Sandstone formation and breccia (the latter only lagouiiwaids) are occasionally met with, but they are 

 quite the exception on the western rim of the atoll. 



