DESCRIPTION OF SMALL ISLANDS OF THP] ATOLL. 109 



8 inches by 7 inches by G inches, was obtained in a shallow pool which allowed sufficient water to cover 

 it at low tide. The central part of the lagoon beach is occupied in part by recent gravel and shingle 

 from the breccia with fragments of coral. The Hurricane Banks surmount the breccia sheet, and for 

 most of its length incline sharplj^ from it to a central depressed area in which are small spaces where 

 solution is taking place. The general surface of the middle of the island is composed of fragments of 

 breccia and coral Avith but little sand, though large masses of coral breccia exist in parts of its southern 

 half which is undergoing solution. These occur as rugged masses, between which pits and caverns arc 

 formed and forming ; they afford a suitable home for the large Puka-Vai trees (Caloj)hi/Uiim inopJujUum) 

 growing here. These trees give shelter to large numbers of birds, as do the Fetau trees at Amatuku. 



On the ocean beach occur several outliers of breccia, often well out from the Hurricane Bank and 

 above the general surface of the breccia sheet, some containing in their upper parts small corals in the 

 position of growth filled in with smaller fragments cementing all together. Their lower parts often 

 exhibit worn blocks of coarse coral with a suggestion of almost horizontal stratification. These breccia 

 masses are really the most oceanward part of the breccia sheet, which extends to and across the island, 

 but is now corroded and worn by the waves till only the low pinnacles and these larger isolated masses 

 remain to mark the original boundary. Behind these outposts at the Hurricane Bank the breccia sheet 

 is nearly obscured by the drifted dSris forming the bank, but it still continues far into and at both 

 ends at least quite across the island. Near the southern end, on the lagoon side, occur beds of a finer 

 breccia with a distinct dip, north 40° east, at 10° reposing on the platform (O.L.I). The southern end of 

 the ocean platform is of a rusty brown hue from the presence of an alga which covers almost everything 

 from near low water almost up to the beach line. 



FUNANGONGA. 



The platform between the last island and this (Plate 8) is much constricted, but widens again as it 

 approaches. Though subject to some variation it maintains on the whole about the average width on the 

 ocean side of the island, and is fairly regular on the lagoon side. A striking featirre of the rough zone 

 on the ocean side is the very irregular outline of the breccia sheet, which like the southern part of Funafara 

 presents a rampart-like outline with projecting buttresses and short sloping beaches between them, 

 appearing at high tide as a headland and bay outline. These headlands or buttresses are sometimes con- 

 tinuous with the breccia sheet behind them and under the Hurricane Bank on into the island. This 

 breccia sheet is also seen at each end of the island to extend quite across and on to the lagoon platform 

 where it forms the base of the island for some distance, itself resting on the platform O.L.I. Over the 

 corroded zone, which is also a portion of the same breccia sheet, the waves sM^eep at half tide and onwards, 

 nrshing up between the buttresses and carrying quantities of coral and breccia fragments, which thus become 

 worn and rounded. Here the bulk of this beach debris consisted of fragments up to 8 inches long, of two 

 species of coral, both of which retained the bright fresh colours and bloom-like surface common to them 

 during, and for a comparatively short period after life. This shows them to be in a living and flourishing- 

 condition, near by and almost certainly, therefore, on the ocean reef beyond. They are Heliopora and 

 the yellow Millepora. The fragments of the former included robust specimens, many of which were gathered 

 and some brought away ; the blue of the Heliopora and the bright A^ellow of the Millepora were striking for 

 their clearness. 



The erosion caused by the attrition of this material has smoothed the breccia in j^laces and is wearing 

 away the inner sides of the buttresses, some of which are quite severed from the main breccia sheet and 

 standalone ^s outliers. When they are higher, as is sometimes the case, they appear at first in their 

 isolated position to have no connection with the main mass, but here, as at a few other places, their 

 relationship is in part disclosed. On the upper portion of some of them corals are seen in the position of 

 growth, with the spaces between them filled in vdih coral fragments and rJe'hrii( and all cemented together. 



