56 FOYE. 



of the lagoon are fairly flat and the question arises whether this flat- 

 ness has been induced by sedimentation, by wave-cutting, or by both. 

 The lack of pronounced embayment of the basaltic islands, suggests 

 that most of the period during w^hich the basaltic cones were eroded to 

 maturity has been one of comparative still-stand, yet there is no 

 evidence that this period has been sufficiently long to level the lagoon 

 floor. 



The outstanding fact remains that the present barrier reef is de- 

 veloping on the remnants of a bank of coral limestone once elevated 

 and eroded, and now submerged. 



Tuvulhd. 



The general features of Tuvutha, about 20 miles south of the Ex- 

 ploring Group, are similar to those of Thikombia-i-lau. Precipitous 

 limestone clilTs surround the island on nearly all sides. The accom- 

 panying map and cross-section (Figure 25) show the distribution of 

 limestones and volcanic rocks, and their relation to one another. 



Only for a short distance along the northwestern side of the island 

 are volcanic rocks known to outcrop at sea-level. The highest peaks 

 on the island are of limestone. The greater portion of the interior is 

 a rough, heavily vegetated, limestone country eroded into weirtl forms 

 by solution. 



Descending the steep limestone cliifs by tortuous ravines covered 

 with thick undergrowth, one is suddenly confronted by a barren 

 andesitic ridge, lying in the center of the island. The ridge is half a 

 mile wide and about a mile and a half long. It is deeply lateritized 

 and, at the time it was visited, bore no vegetation because of drought. 

 The ridge, 540 feet high, is surrounded by a circular moat. Outside 

 the moat rises a continuous cliff of limestones. Limestone patches 

 also o^•erlie the andesite of the central ridge. 



Tuvutha, like Tliikombia-i-lau, was formed by the mature erosion 

 of an isolated volcanic peak. It was then submerged, overlain by 

 limestone, and uplifted. The occurrence of olivine-basalts in a 

 conglomerate near the sea-coast, and the irregularity of certain lime- 

 stone-andesite contacts suggest that a second period of volcanic 

 activity aft'ected the island after its uplift, but full evidence for this 

 conclusion is not forthcoming. 



The contact of the limestone with the volcanic rocks of the interior 

 shows the surface of the andesite to umlulate in low, rolhug hills. 



