THE GEOLOttY OF FUNAFUTI. 73 



on the main island a probable index of tlie amount of material added to tlie latter 

 by silting since the deposition of the conglomerate, the latter being perhaps nearly- 

 contemporaneous with, though a little newer than, the old breccia sheet. As will be 

 seen by referring to Plate 9, the island has been extended westwards for a distance 

 of about 270 yards since the time this conglomerate was formed. The rock of type 

 L.2.A. jDasses insensibly into L.3. L.3 is defined on the index as hard foraminiferal 

 Lithothamnion sandstone, passing in places into fine conglomerate. It is of various 

 ages and has ])een formed chiefly on the lagoon side of the reef, but exception- 

 ally it occurs on the ocean side of the reef, as shown on Plate 10, where Section 

 line 3 crosses the island. It is usually bedded, haying a thickness of from 3 to 

 5 feet. It is found now chiefly .between the levels of high and low tide. In places, 

 as at the islet of Mulitefala, it extends up to fully the level of high tide. Its 

 general appearance is seen on Plate E (upper figure), the photograph being taken near 

 the N.W. end of the islet of Amatuku. It is formed of more or less water- worn 

 fragments of Lithothamnion, Halimeda and coral, together with numerous fora- 

 miniferal tests, especially those of Orhitolites and Tinoporus, and fragments of 

 moUuscan shells, the whole being bound together by a clear calcareous cement. 



It appeared to us that the most im|Dortant questions about this sandstone are 

 (1) as to its mode of origin, with special reference to the nature of the cement, 

 and (2) as to its age. With regard to the second question it is clear that its age 

 is variable. In certain areas it is indubitably of some antiquity, as, for example, at 

 Mulitefala,"^ where it is evident, from the direction of dip of the calcareous sandstone 

 (or sandy limestone), that since its formation the whole islet has been driven in 

 lagoonwards, so that this sandstone, originally formed at the back of the island, 

 is now situated in front of it. A similar case occurs at Fualifeke Isletf and at 

 Tefala. On the beach at Fuafatu, on the other hand, a fine coral rubble is even now 

 being cemented into a fine conglomerate. An examination of specimens from the latter 

 in thin sections shows the cement to be a fibrous radial calcite. In the case of the 

 calcareous sandstones, sections examined under the microscope show that they too 

 owe their cement to a secondary fibrous carbonate of lime devoid of organic structure. 

 At first we inclined to the opinion that no cementation of loose material was 

 taking place between the tide marks, but that it had all formed at or below the 

 level of low tide. At the same time a careful examination of this sandstone showed 

 that incoherent layers of sand were frequently interstratified with it and occurred in 

 beds parallel to those of the sandstone. 



The extensive corrosion so noticeable towards the centres of the islets shows that 

 carbonate of lime is being removed extensively by the agency of humic acid from 

 these areas. Probably carbonate of lime is being removed at an equally rapid rate 

 from the banks of sand and rubble on either side of these depressed central areas, 



* See Plate 12, Section 2, of Mulitefala. 



t See Plate 13, Section 1, of Fualifeke. 



L 



