AFRO-INDIAN REGIONS. 



333 



island ; and described it, as " covered with woods on all sides," and as 

 " reported to contain forty towns." 



On the 26th, returning towards Viti-levu with a fine breeze, the 

 Peacock passed Vatu-lele ; an elevated coral-island, already alluded 

 to in the account of Metia. It is situated on the outskirts of the 

 Feejeean Group, but quite near the main island. As examined with 

 a glass from the deck of the Peacock ; Vatu-lele, for the greater part 

 of its length, appeared to be only twenty to thirty feet high, and 

 nowhere more than fifty feet. Besides vertical fissures, some of them 

 wide enough to interrupt the land-surface, the face of the cliff seemed 

 everywhere inaccessible and extremely ragged ; presenting open- 

 mouthed caves lined with stalactites, and long groove-like horizontal 

 excavations, as though scooped out by the surf prior to the change in 

 level. 



Leeward portion of the group. The Peacock having reached 

 the Western side of Viti-levu, we w^ere surprised at the total change 

 in the aspect of the country ; above the coast-line of mangroves, every- 

 where barren and brown, with forest-growth only in deep ravines, and 

 on declivities around mountain summits; a result of the above- 

 described difference in climate. 



The immense tract of coral-reef that fills the ocean on the Leeward 

 side of the group next came in sight; with Meloloand the other rocky 

 AsAUA Islets projecting in the midst. This tract of coral-reef is 

 separated from the rocky shore of the two main Feejee islands by a 

 narrow channel, continuous throughout for some hundreds of miles; 

 and the Peacock having entered, we were confined to the quiet 

 water close along the land during the remaining weeks of our 

 stay. The navigation, however, requires daylight; and the Peacock 

 anchored. 



On the 27th, the ship proceeded, with a light wind, threading her 

 way in the narrow passage behind the coral-tract ; but in spite of 

 every precaution, struck on a lump of coral, about 5 p. m. ; and get- 

 ting clear shortly afterwards, was anchored. — On the 28th, I landed 

 with others for about twenty minutes on the exposed surface of the 

 coral-tract; and occupied myself with marine productions. 



The ship proceeded, and on the 30th, I landed with others on the 

 Northwest or leeward side of Viti-levu ; a few miles from the town of 

 Mba, which was concealed by mari(jroves. Several interesting plants 

 were met with; including a remarkable Oippaiis?, a woody vine 



84 



