THE GE0L0C4Y OF FUNAFUTI. 81 



together with Pocillopora grandis. It will be noticed that pebbles of reef rock 

 strew the bottoms of some of these channels, but further seaw^ard, according to his 

 observations, the Lithothamnion rock is bare of detritus until the zone of live 

 Halimeda is reached. The latter ranges on the ocean side of the reef, from 5 fathoms 

 down to 45 fathoms. The structure of the reef below this level is described in 

 Section VII. 



In regard to the intricate succession of channels and piers forming the boundary 

 of O.L. 10, at low tide, we must inquire whether the former indicate channels of 

 marine erosion, the jiiers representing incipient outliers, and the mounds beyond 

 complete outliers, or whether the piers represent areas of more, the channels of less, 

 active growth. If the latter hypothesis be considered, it may be remarked that 

 when once re-entering angles are developed at the edge of the reef platform, through 

 some portions of it growing out more rapidly than others, the heavy surf, funnelled 

 in between the piers, would tend to erode back the bay, or at all events to check 

 temporarily the further outgrowth of the reef at such points. Meanwhile as the piers 

 are pushed further outwards owing to accretions of Lithothamnion, &c., the tendency 

 is for these channels to be roofed over near the level of low tide, which is the zone of 

 most vigorous growth of Carpenteria, Polytrema and Lithothamnion, The spaces 

 left open further inland form blow-holes. Meanwhile, the sea-water surging to 

 and fro in these channels, and sweeping leef debris backwards and forwards, continues 

 to erode the bottoms of the channels, so long as the holes remain open, and thus admit 

 of an ebb and flow of the water. 



We thus hold the view, as the result of the study of the O.L. 10 zone, that a very 

 slow peripheral enlargement of the atoll is still taking place.* On the eastern rim of 

 Funafuti this growth is dependent almost entirely on Lithothamnion, Polytrema, and 

 Carpenteria. Mr. Finckh's experiments proved that the Lithothamnion can spread 

 horizontally at the rate of 1 inch in 6*56 months, but the layer so formed w^as so 

 thin as to be almost inappreciable, probably between 1/50 and 1/100 of an inch. 

 As, therefore, it may take fifty years to form a layer 1 inch thick of the encrusting 

 type of Lithothamnion, it is obvious that portions of this Lithothamnion rock which 

 have a thickness of several inches must be of considerable antiquity, and the exten- 

 sion outwards of the atoll rim by this agent must be proportionately slow. Unfor- 

 tunately, the presence of Polytrema and Carpentaria amongst the Lithothamnion was 

 not detected at the time the experiments on rate of growth of reef-forming organisms 

 were being conducted, so that more data are needed. 



We may now pass on to consider the development of O.L. 10 around the edge of 

 the reef platform on the lagoon side. A remarkable feature in the lagoon reef 

 platform is the low cliff, from 2 to 3 fathoms deep, in which it generally terminates. 

 The edge of this cliff, together with a small portion of the upper surface of the reef 

 platform, where it is awash at low tide, is encrusted with living Lithothamnion. 



* ' Natural Science,' vol. 12 (1898), p. 174, "The Broadening of Coral Islets." By C. HedlEY. 



M 



