OF THE TONGA ISLANDS. 



603 



In ordinary weather the streams flow beneath the limestone rock, 

 and it is only after heavy rains tliat runninjj; water is found in the 

 ravines whicli traverse the terrace. ^loroover, the sides of these 

 slope abruptly down from the level of the terrace, and their course 

 is for the most j)art quite indepojident of the direction of tlie depres- 

 sion. I'inally, it may be asked, A\ hy should the drainage from a flat 

 area excavate the interior and leave the margin comparatively 

 untouched along almost the entire length of the terrace? It appears, 

 then, tluit the terrace had much the same general conformation 

 which it now presents when it was upraised from the sea ; in other 

 words, that the elevated border constituted a small barrier-reef 

 lying nearly a mile from the shore of the island as it then existed, 

 with a break in it at one point, and enclosing a lagoon with a depth, 

 of 6 fathoms over a considerable area. At its northern and southern 

 ends the reef joined the shore and became fringing. A similar 

 relation of reef to shore occurs in the reefs now forming the harbour 

 of Suva 'in Yiti Levu, the capital of Piji, where the barrier which 

 protects the anchorage is continued into a fringing reef along the 

 southern shore of the island. 



As above described, there is a narrow limestone-terrace running 

 along the eastern side of the island at about the same level as 

 terrace a on the western, and interrupted in the middle by a deep 

 gap, the head of which is formed by the cliffs of nearly the highest 

 part of the island. On the north of the gap the terrace ends 

 abruptly in a cliff", facing south, and standing out beyond the cliff 

 is an isolated pinnacle of limestone rock, which I estimated as 

 measuring about 300 feet from top to bottom. The rock of the base 

 of the pinnacle presents to the naked eye a compact homogeneous 

 structure, except that in one specimen there is a cavity which was 

 occupied by a univalve of the genus Ceritliium. Dr. Murray has 

 examined sections of the rock * and finds that it is made up prin- 

 cipally of calcareous algoe, together with fragments of echinoderms, 

 molluscs, and a number of f oraminifera f. 



The limestones of terrace h, which stands at an elevation of 

 about 600 feet, are also for the most part homogeneous compact 

 rocks. A section from the edge of the terrace t is found by Dr. 

 Murray to be " chiefly made up of calcareous organisms. Fragments 

 of molluscs, echinoderms, polyzoa, and calcareous alga), together 



* Shdes 1269 and 1270. 

 t Viz. :— 



Spiroloculina, sp. 

 Bulimina (cf. compressa). 

 BoHvina textilarioides (Reuss). 

 Textularia trochus (d'Orb.). 

 Cristellaria, sp. 

 Globigerina {rubra ?). 

 Truncatulina, sp. 

 Planorbulina (cf. larvata). 



Voliftrcma miniaceiim (Linn.). 

 liotalia, sp. 

 Tinoporns bacculaftos. 

 Cymbalopura Voeiji (d'Orb.). 

 Gjipsina inhcBrens (Schiiltze). 

 CyclocIijj)eus Carpenteri (Brady). 

 Hderostegina dcprcasa (d'Orb.). 

 Hammulites Camivyii (Carpt.). 



\ Slide No. 1272. 



