DESCRIPTION OF SMALL ISLANDS OF THE ATOLL. L07 



of that material occur, causing the platform near the breccia and the Ijeach to resemble closely in general 

 character and appearance some of the less rough of the ocean beaches. 



The middJe portion of the island, however, is convex in outline on this side ; here erosion has been 

 much less severe, and so the island is correspondingly wider, the Hurricane Bank scarcely exists, and what 

 there is, is more sandy. As the northern end of the island is approached, the erosion has again been 

 active, the Hurricane Bank reappears though less in height, and the beach is formed of pebbles, which 

 are swept up and down it, with great force even in moderate weather, by the back-wash caused by the 

 adjacent channel which separates this island from that of Falefatu. By that action this end of the lagoon 

 side of the island has been worn back, so that it has become very narrow. Near it a very large block, 

 8 feet by 7 feet by 4 feet, of single coral of the species allied to Goniasircea favistella occurs in situ projecting 

 through the beach above high water. This is the first deep or wide passage that Ave have passed in our 

 journey round the islets since that of Fuagea on the western reef of the atoll. 



FALEFATU. 



This island (Plate 7) occupies a unique position on this atoll, situated as it is between two of the chief 

 passages — that of Mateika and Te Bua-Bua (the former being the widest of any on the atoW, i.e., IJ miles) — 

 both of which are sufficiently deep to give safe passage to ships entering the lagoon, and are the two greatest 

 passages for the waters of the lagoon to enter and escape with every rise and fall of the tides. The shape 

 of its breccia sheet, though a considerable distance back from the edge of the reef, yet strikingly conforms 

 to it. It would appear to be also regulated largely by the direction of the chief winds and currents. 

 This island is one of a few where the effect of their force in first forming, and then removing, under 

 changed conditions, the breccia sheet as well as the island upon it, may be best studied. The maps will 

 show that the distance of the breccia sheet from the outer edge is increased wherever it crosses the direct 

 line of the general direction of the winds and currents, and is diminished wherever this strikes the outline 

 of the island in an ol^lique direction. The existence of the island itself in this position is probably due 

 to a considerable projection oceanwards of the 4 and 100 fathom lines of submarine reef, which presents 

 at this place a keel-like outline the result of which, as the waters strike and divide on it and so lose some 

 of their force, is a great width of reef rim or platform. 



Similar submarine projections oceanwards would appear to favour the development and continuance of 

 islands in the areas sheltered behind such projections, inasmuch as at each position where these conditions 

 obtain, even though they may appear to be the most exposed, there are yet found considerable widths of 

 upper reef rim and islets, as at Fuafatu, Tefota, Falefatu and even at Fongafale. I may add that in this 

 position there is little doubt that this island would not have remained till now, but for the submarine 

 projection, which favours the formation of breaches and passages. 



, Falefatu consists of a sheet of breccia of very irregular outline lying on a base of tidal reef or 

 platform, which appears nearly all around the island. It is in a more or less eroded condition, except 

 at the central portion of the concave lagoon side, where pebbles and foraminiferal sand have accumulated 

 and now obscure the breccia sheet for about half the length of this side. Here there is a considerable 

 forward wash which throws ujj the sand and shingle in a steep bank, while the roar caused by the 

 rushing forward and backward of the pebbles of coral and coral breccia, especially near the northern 

 end where the island is narrow, was so great at three-quarter tide, with but a moderate breeze, that 

 one could scarcely hear one's own voice. Extensive erosion has taken place on the ocean side as 

 indicated by the remnants of breccia still existing on the tidal platform. This has been so far invaded 

 that the island near its centre is almost severed, there being but a few paces connecting the north 

 and south ends together. Very little Hurricane Bank or coarse material exists anywhere as the southern 

 end is approached. 



This, the widest portion of the island, is composed chiefly of sand and finer material ; no approacti 



P 2 



