FADIFOLU. 401 



islands do not reach to within a considerable distance of the seaward edge of the reef, on wliich side 

 they vary greatly in shape. To the W., liowever, their shores form a line 500 — 600 yards behind 

 the lagoon edge of their reef. Belonging to both the series are the islands Hudufuri and Difuri. 

 The southernmost of tliese, Hudufuri, has a belt of rock riglit on the outer side of the line of rocky 

 islands, joined by a narrow isthmus — formed by two deep bays to tlie N. and S. of the island — to 

 a stretch of sand}' land as decidedly belonging to the inner line. Difuri is less regularly divided, 

 but a bay to the S. with for the most part rock outside and sand within indicates the two lines. 



As to the changes going on there is to the E. of the rocky belt a marked washing away, which, 

 continuing for some distance round the ends of the islands, has formed the deep bays in Hudufuri 

 and Difuri'. There is also some slight washing away to the E. of the islands of the sandy line. 

 To the W. of these same islands there is practically no change at all, but the ends of all these 

 islands are growing out along the reef towards one another. Whether the islands themselves will 

 ever meet is uncertain. They are at present separated by channels of 4 — 6 feet of water with very 

 strong currents, a depth so unusual for such a position that it appears probable that the dividing 

 channels increase in depth with their restriction in breadth, the two processes going on until equi- 

 librium is established. Tiie above, however, is not quite true for Difuri, which, while perhaps a little 

 growing at its ends, shows a distinct washing away along its whole lagoon side. 



From the N. end of Hudufuri to the Olivelifuri clump of islands a velu extends along the W. 

 side of the reef. It is separated by a sand flat of about 90 yards in breadth from the inner line of 

 islands, but opposite the channels between them this ilat broadens considerably. Its depth varies 

 from 3 fathoms to the N. to 6 fathoms a little south of Deru. As to its character it is quite open, 

 and has a soft, sandy bottom. Life, both free-living and sedentary, is in it very scarce, the latter 

 being only represented by a few coral knolls on the W. side and some small coral colonies opposite 

 the passages between the islands. To the N., almost between the islands of Difuri and Hudufuri, is 

 a separate small pool of 1| fathoms. 



The lagoon reef is a flat about 50 yards broad — narrower to the S. — part of which seems to be 

 awash at low spring tides. The whole way along it appears to be so even that a man could wade 

 upon it at low tide from Difuri to Olivelifuri, and probably from the latter again to Aligau. Along 

 the W. of Difuri it extends for about the half of the length of that island, thence further N. 

 gradually merging into an area, where there is no true reef, but merely a number of isolated per- 

 pendicularly-walled reef patches with sandy ground — the slope of tlie Difuri shore — between. Against 

 the lagoon the reef has no edge, but passes into an area covered with large, massive Madrepora 

 and Porites arenosa. This again opposite the islands falls precipitously to the atoll lagoon, but in 

 the open part to the N. of the Olivelifuri set of islands we found an area with 5 — 10 fathoms of 

 water, covered with loose, dead coral masses, the dredges and swabs practically failing to secure any 

 fixed life at all. To investigate this small bank would have required a residence for a considerably 

 longer time in the atoll than we had to spare. We found no area elsewhere which could be in any 

 way compared with it, and I am inclined to think that it is a formation entirely due to local causes. 

 The currents of the N.E. monsoon would naturally sweep along the reef by the islands and across in 

 the open space, clearing off any coral masses from the W. reef and depositing them inside. Further, 

 the condition would be almost that to the W. of the long main island of Funafuti, where there is 

 in the centre an extensive and increasing deposition of sand^. 



Off Olivelifuri I was unable to localize any fixed point of the reef to the S.E., the contour 

 appearing quite rounded. The island is of sand alone, and no part approaches within 200 yards of 

 the reef edge. All sides of the land except perhaps a sand spit to the S. show washing away. This 



1 These are inhabited by immense numbers of stinging - Vide "The Coral Reefs of Funafuti, Rotuma, etc.," 



rays, a fish which we saw nowhere else in the Maldive Group. Proc. Camb. Phil. Soc. vol. ix. p. 435, 1898. 



