94 AITEXDTX T. 



the islet. Tlio I'drifrs blocks, though fewer in numlier liere, appcir through the lieach sand on the lagoon 

 side, while the sand seems to retard gieatly, even if it does not prevent, the process of solution. On the 

 west or ocean side of this islet is a small hut rugged clinker field whose outer boundary is also the limit of 

 vegetation. Beyond this is the Hurricane Bank, which is succeeded by a depression along its side many 

 yards wide and enclosed by an outer and smaller ridge forming a loop. The enclosed depression is carved 

 out of the very hard and dense breccia sheet, and contains no Poriti'^ e.xcept as low stumps ; it also 

 follows exactly the shape of this side of the islet with which it is conterminous. 



Apart from the central solution area in this islet and a great accumulation of sand at the lagoon side of 

 the northern end of this islet {riil" Avalau), there is only one feature in its historj' disti.nguishing it from 

 the islets above described on the western side — namely, that of the depression on the ocean side of the 

 islet. Probably this represents the site of an earlier solution area equally or even more intense than the 

 present one in the centre of the islet. If so, the area affected by solution is receding from the ocean side, 

 and its Hurricane Bank has been driven lagoonwards, so that an earlier position of this is marked by what 

 is now the outer low ridge. 



TENGASU AND TEAFOAFOU. 



The general characters obseived in the platform at Tc Falaoingo and Tutanga are continued on to 

 Tengasu (Plate 4) and Teafoafou (Plate 5). 



Immediately to the south of Tutanga, and therefore between that and Tengasu, is a good l)oat cliannel ; 

 south of this there is a gradual but marked reduction in the numbei', and especially in the size of the 

 PorHi'ri. Passing still further southwards, they continue to diminish both in number and size, till verj' 

 many are only al)out 2 feet in diameter, and less than that in height. They are, however, similarly 

 situated in I'clation to tlic lircccia. The effect of tlie reduced olistructions to strong westerly winds, 

 caused 1iy these modifications in the conditions, is shown in the character of the islets. There can 

 scarcely be said to l)e any Huri'icane Bank on either side of them, unless the ))elt of fragmcntal 

 material can l>e so regarded. This, however, is not higher tlian the islet itself, the platform rising 

 gradually to the highest level of tlie land, with but a sliglit step upwards where the loose fragmental 

 material connncnees. On the lagoon side, the width of the platform varies considerably, though it 

 bears some little relation to the fcjrm of that side of the island; while the sand flats invade and 

 cover a portion of the breccia floor on this side. The hardy Xgie trees stand in places well out from 

 the island on the eroded breccia towards the ocean platform, with theii- loots firmly grasping the 

 breccia floor, resisting every advance of the invading Avaters, and marking an caidier lioimdary of the 

 now receding islet. They also serve to break the force of the waves and surf, and furnish a temporary 

 shelter for loose fragments of coral tliat may chance to l)e driven landwards. The btiffeting of the 

 waves seems to increase rather than to reduce the vitality of this tree, till some unusual force — 

 pei'haps a large worn mass of l)i-eccia or coral driven by the M'aves — breaks it oft" short at the roots, 

 when it is swept away. During high tides the IjiUows roll up to and over some parts of the island 

 of Teafoafou, making surface breaches, from which the Aegetation has been in part or wholly removed : 

 the soil being reduced at nearly eveiy such attack faster than it can be replaced. Then the highei- 

 vegetation, receiving insufficient support and nourishment, gradually sickens and dies, oi' is carrie<l away, 

 soil and all, l)y an mnisually fierce gale. Thus it is l)eing slowly divided into several parts, just as it has 

 l>een separated from Tengasu. Near each end of Teafoafou it is flat-topped, with much pumice nearly 

 covering small areas, on which vegetation grows more lai-ge and healthy. This, with the luxuriant 

 vegetation of Tutanga, and of one part of Tengasu, bears evidence of what it probably was throughout, 

 before the soil was subject to the incursions of the sea. A small area in the north end of Teafoafou, 

 from which the material of this end of the islet had appai-ently been almost, if not quite, removed by the 



