food, which in time producing vegetation, may continually accumulate until the whole mass become a solid stratum of earth. But this does not solve the present difficulty, for, on the islands I am describing, large pieces of madrepore are found, disposed in regular layers, full twenty feet above the level of high water mark, and for this circumstance no satisfactory reason, in my opinion, can be assigned, but the supposition of the sea having retired since they have been so deposited.
The small island, which I have mentioned as different from the rest, consists of a solid rock of calcareous stone, through which run veins of calcedony. On the east side of it is a large cave, used by the masters of dows, frequenting this bay, as a store-house, for laying up their goods; and from this circumstance, as we could not ascertain its native appellation, we named it Safety Island.
The shores to windward of these islands are, in general, steep; and, when the weather is foul, difficult of approach, owing to the encroachments of the sea, which have undermined the rocks, leaving in many places singular-shaped pillars and hollow caves, bearing a strong resemblance to works of art. On the leeward side, a grove of rack trees is commonly found, particularly convenient for supplying fire-wood to vessels, and the natives will permit any quantity to be cut down for a few dollars.
One of the islands named Kutto, appears at some distant time to have been inhabited, as the ruins of stone houses and a fort plainly shew; the latter was evidently intended to command the passage leading into an inner harbour, adjacent to the village of Duroro. There also exists in the centre of the same island, a connected set of four large cisterns, excavated in the shape of a cross, each of which is thirty feet long, nine broad, and seven high, all of which are lined with chunam; these, when filled, would hold, at a moderate computation, one hundred and twenty thousand gallons of water. These cisterns seem to have been constructed by the same people who formed those which I have formerly described on the Island of Dahalac. A tradition current among the natives ascribes this undertaking to the Pharsees, or Per-
