66 PEOFESSOE T. W. EDGEWOETH DAVID AND ME. G. SWEET. 



and the 3000-fathom line at rather more than double that distance. The soundings 

 taken during the survey by H.M.S. " Penguin" show that there is an absence of any 

 connecting ridge between Funafuti and its nearest neighbours in the EUice Group, 

 Nukufetau and Nukulailai. Funafuti is thus an isolated conical mountain rising 

 16,000 to 18,000 feet above the general level of the surrounding ocean floor, and 

 crowned with sand and reef 



The two deepest channels into the lagoon respectively, compared to the mouth 

 and eyebrow of the head, are, we think, of considerable geological antiquity. It may 

 be something more than a coincidence that they converge lagoonwards on the spot 

 where Captain Creak has located the chief focus of magnetic disturbance. Possibly 

 they may indicate the site of submerged valleys or silted up inlets. 



Symbols used on Maps and Sections. 



As regards the meaning of the symbols and of the numbers employed an index 

 is given on Plate 2, but it will be as well to add some further explanation of them here. 

 The symbol O. means that the rock so denoted was formed chiefly on the ocean 

 side of the reef, the symbol L. that it was formed chiefly on the lagoon side, and the 

 symbol O.L. that it was formed on both the ocean and lagoon sides of the reef The 

 higher numbers denote the newer rocks, the lower numbers the older. The rock 

 marked O.L.I, and denoted by the blue colour on the geological maps, is the oldest 

 and by far the most important of those forming the hard rock of the atoll ring. It 

 is the substratum of the principal islets, probably of all the islets, at all events of 

 those on the north, east and south sides of the atoll. It is composed chiefly of 

 Heliopora, ccerulea, with occasional clusters of Porites, which usually accompanies 

 it among the reefs of the eastern side. On the western side of the atoll, however, 

 though Porites was present, and in the south-western part was much more abundant 

 than on the eastern rim, very little Heliopora ccerulea was observed, and even that 

 was not in situ, but present as fragments in the reef-breccia sheet. On the western 

 side of the atoll, however, Porites greatly predominates, no old reef-rock formed of 

 Heliopora ccerulea being there visible, in spite of the fact that there is an abundance 

 of Heliopora coBvidea living inside the lagoon on its western side, but it must be 

 noted that no excavations were made deep enough on this rim to reach the general 

 horizon of the Heliopora. 



The Heliopora ccpridea occurs in stocks from a few feet up to 24 feet in diameter. 

 These are best seen at the Mangrove Swamp, (See Plate 18,) 



As stated on the plate, the shape and size of the coral heads figured are drawn 

 to actual scale from measurement, whereas the branching is merely sketched with 

 only approximate accuracy. An interesting feature in connection with this 

 development of Heliopora cinrnlea is the fact that at the centre of nearly every one 

 of the larger heads a cushion-like mass of Porites is found ?n situ, in each of which 



