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



the position of high water. The height of this point of intersection above the zero of 

 the level was taken as the height of the object above liigli water, and as the eye 

 f)f the o])server was seldom more than a few yards distant from the levelling stafP, 

 the results were approximately accurate, and the method was certainly expeditious. 

 In some cases the levels Avere taken by means of the spirit levels of our clinometers. 

 Where special importance attached to the levels, as in the case of the raised 

 Ueliopora cctrulea reefs in the mangrove swamps of the main island (see Section 9 

 of Plate 9) the levels were checked several times and may be considered to be fairly 

 correct. As the details of the geological survey were filled in mostly by pacing, 

 errors of several yards no doubt occur in places. The general shape of the islands, 

 however, is fairly accurate as shown, or at all events sufficiently so to form a record 

 for future reference. 



As regards the method of showing the outer l)oundary line of the rock marked 

 O.L. 10, on Plate 17, the living Lithothamnion zone, the following plan was adopted. 

 Mr. A. E. FiNCKH having accurately measured, with a tajje, the details of the reef 

 jolatform down to low-water springs, opposite the main diamond-drill camp, this 

 plate was drawn from his traverses, and the outline shown on it was then repeated 

 over and over again to represent the characteristic outline of the outer shore of the reef 

 at low tide, but conspicuous indents in the shore line are shown in their true position 

 by indentations of the boundary line between O.L. 10 and 0.3 as, for example, near 

 S^ to the right of section-line 9 on Plate 9. In the same way the pattern for 

 the boundary of the reef lagoon-wards at low tide has been taken from Mr. Finckh's 

 traverse of the lagoon platfoi'm of the reef to the north of the Mission Church (see 

 Plate 17). 



The general shape and submarine contour of the atoll may be seen from the two 

 figures on Plate B. These have been made from a plaster model in the Australian 

 Museum, reproduced from a wooden model carved by Captain E. C. Hore, of the 

 London Missionary Society's steam yacht "John Williams." Captain Hore carved 

 Ins wooden model to true scale, taking as his basis the British Admiralty Chart, 

 No. 2983. In general outline the atoll resembles a caricature of a human head, 

 facing the west, the positions of the mouth and eyebrow being indicated by two dark 

 streaks marking deep passages into the lagoon fiom the ocean. The greatest 

 diameter of the atoll lies in a direction N. 20" E. (true). Kramer* has argued that 

 this elongation of the islands of the Ellice group, in common with that of numerous 

 other coral islands of the Pacific, is due to the direction taken by the prevalent ocean 

 currents and winds. In the case of Funafuti, however, the prevalent winds from 

 about March to November are the S.E. or E.S.E. trade winds, whereas from December 

 to February north-westerly monsoonal winds prevail. Thus the long axis of Funafuti 

 is almost at right angles to the prevalent winds. As regards its shape in relation to 



* " Ueber den Bau der KorallenriflFe uud die Planktoiivertheilung an den Samoanischen Kiisten," 

 pp. 88-90. 



