104 APPENDIX I. 



ill dtu, but that seen ut Nukusavalivali ami the much larger specimen here make this still more evident. 

 A ie\Y _Forite$ are seen above the worn or eroded breccia sheet. This sheet has apparently been 

 continuous from Telele to Funafara, as it still is at the similar easternmost corner of Fongafale (the 

 principal village on Funafuti), each of which angles have many striking resemblances to each other.* 



Here the breccia sheet has been broken through ; there it has liecome very narrow and will probably 

 be also breached at no very distant date. At Fongafale, again, a soft sandstone covers a considerable 

 area inside the Hurricane Bank at the swamp ; at Tefota there are several beds of soft sandstone in 

 a curve, so that the parts now remaining on the west breach form portions of a basin of sandstone 

 dipping from every side to its centre, lying inside the remains of the breccia sheet and now being 

 more or less eroded. There is also a small outcrop of incipient sandstone on the cast side of this 

 channel not far from the entrance to the tiny bay there, also dipping to the passage and being eroded. 

 The two sand points on either of the inner ends of this shallow passage are extending lagoonwards, 

 but the east side is undoubtedly l)eing eroded, as is proved by the low undermined cliff without a 

 storm bank of any kind whatever ; with the cocoanut trees falling and the matted roots projecting 

 as they grew before the breach occurred. This has been a portion of land very similar — apart from 

 its size — to that at the main village, and is now being slowlj' removed by erosion, accelerated by 

 the breach or passages here. 



As the lagoon near this islet of Tefota is very narroAv, it has silted up much more rapidly here 

 than at the main island. There a swamp exists inside the Hurricane Bank ; here there is little doubt 

 that the inner part of this wide l)reach and the site of the soft sandstones and present tiny bay to 

 the north were not very long ago so occupied, and the present swamp extending from the tinj^ bay 

 along the inside of the Hurricane Bank is but a northward extension of this swamp to the island of 

 Funafara. 



Bounding the northern side of the breach eastwards of Tefota there is a Hurricane Bank which 

 looks much like a deflection of that which bounds the ocean side of Fimafara, once contiiuiing across 

 the breach to Tefota l)Ut now diiven back, like a gate pushed open. There can I think l)e no doubt 

 but that this is practically what has happened ; first a breach has been formed through the old 

 Hurricane Bank, then erosion of the breccia has followed, and some of this removed material has been 

 driven in and piled up to a small extent along this side of the channel which, as it widened, rolled 

 back the Hurricane Bank, adding more to it and increasing its height and bulk up to the present time. 



FUNAFARA. 



This island (Plate 5) is now the largest on this end of the atoll and the second largest of all. It is 

 formed at its southern end by a wide area of flat sandy land which continues to the northern end, though 

 of reduced width and with an admixture in varying proportions of coral fragments. Near the south 

 end of this island is the site of the village of Funafara, next in importance to Fongafele. 



At the widest portion of its south end the breccia sheet does not extend through to the lagoon side, 

 being undoubtedly masked by the sand accumulation there, Mhich near the point is unmistakable. As the 

 island narrows, the eroded breccia is exposed in places, and sometimes forms the creiuilated outline, as it 

 does northwaids of the village where, for about one-thii-d of the length of the island, it extends in a series of 

 thin sloping tongue-like projections similar to those on Telele, l)etween which are sometimes a little beach 

 sand and dchrii. North of this the breccia maintains a moderately even line till towards the end, when it 

 is observed to be overlain by the newer breccia. 



* Thus at both, the lagoon platform is very wide — uncommonly so; at both, the breccia sheet has extended 

 quite round the corner; at both, silting up has taken place in the lagoon, while extensive sand accumulations 

 have greatly extended the island lagoonwards. 



