84 PROFESSOR T. W. EDGEWORTH DAVID AND MR. G. SWEET. 



currents. Since the determination of the broad outhnes of the island its shape has 

 been modified chiefly by :■ — - 



(a) Organic growth of the reef-forming plants and animals, notably Ilalimeda and 

 Lithotliamnion among the former, and foraminifera, corals and Millepora among the 

 latter. 



(b) By prevalent winds, which are respectively the S.E. to E.S.E. trade-winds and 

 the N.W. monsoons. 



(c) The ocean and tidal currents. 



The convexity of the atoll on the eastern side and the concavities in its outline on 

 the western side are thought to be due respectively to the work of the trade-winds 

 being less destructive than that of the monsoons. The monsoons are frequently more 

 violent than the trade-winds and tend consequently to keep open breaches in the 

 reef, whereas the trade-winds are steadier and less violent, and thus have a less 

 effect than the former in sweeping debris off the reef platform into the lagoon. 

 Another likely reason for the atoll being more open on the western than on the 

 eastern side is that if, as the magnetic survey seems to indicate, there is a nucleus of 

 magnetic rock near the eastern side, that portion may be the older, and the reef may 

 have started to form there earlier than it did on the west side. The two deep 

 passages tlirough the western reef rim, 1 e Ava Tebuka and Te Ava Fuagea, each 

 about 25 fathoms deep, are evidently of considerable antiquity, as is shown by the 

 great extent of the reef growth which flanks their sides for a distance of over a mile. 

 At the point where the prolongation oceanwards of Te Ava Fuagea meets the 100- 

 fathom line the soundings show evidence of a deposit about 300 feet in thickness. 

 This is also in fjivour of that passage being of considerable antiquity. The lowest 

 visible foundation rock of the reef platform along the whole of the northern, 

 eastern and southern portions of the atoll is Heliopora ccerulea, with occasional 

 Porites. On the western rim Porites reefs abound, but none of the old Heliopora 

 ccerulea reef was observed there in situ, though it probably underlies the present 

 horizon of the dead Porites, a view rendered highly probable by the occurrence of 

 fragments of Heliopora ccBndea at the base of the breccia sheet of the western rim. 



Since the time that the Heliopora ccpridea reef flourished just west of the Man- 

 grove Swamp in the main island, a downward movement of the shore line has taken 

 place of at least G^ feet, while in the case of the raised Porites reefs of the western 

 side of the atoll, there must have been either a land-elevation or a sea-sinking of at 

 least 9 to 10 feet. Prior, however, to this final downward movement (which is probably 

 still in progress) there was a slight similar movement exposing the Heliopora and 

 Porites reefs to the full denuding force of the waves, followed by a contrary move- 

 ment of fully 8 feet, for we find that the surface of the old Heliopora ccBndea reefs 

 was for the most part covered with coral fragments, in a manner which could only 

 have been accomplished by the denuding and disintegrating action of the ocean 

 between tide marks, and it is necessary to assume a subsequent rise of the sea of 



