SAIL.ING DIRF,?TIONS. ?L?7 The GEOOR?r?E REEFs extend for more ?o,? twelve A. miles, and perhaps are joined to the !and. Their southern Sect. IV', parts dry at low water. The (?eog?aphe salied throu? lq. West them, so that it is probable they are de_?*?e?d in *?,*_merous reefs. At FORESTIER ISLANDS we saw the coastsgain. The nutin is here very low, 'but from the shcalneas of the water we wer? not able to penetrate behind Depuch Island. It is very uncertain whether the coast line that is laid down upon the ch.art is correct: it was scarcely visible from the deck, and was so low that it might have merely been the dry parts of extensive reefs. The high land retires for fifteen or twenty miles, and forms an amphitheatre or deep -bay, with some bills of considerable elevation in the ,distancoo �All the islands of this group are low and sandy, excepting Dsx, ucu, which is high, and of a very peculiar formation; 4t is described in the first volume, at page 14?;. We did not !and upon it, but on its north-east side there appeared to be a bey, on which the French found a streem of water� Betwee? DEduce ISLAND and CAre LAMBERT the coast is very shoal. Towards the latter the hills approach the .sea, and the bottom is deeper. BzZOUT ISLAND iS connected to the cape by a reef, on which there are severer dry rocks; we passed close FOUnd its north-east edge, and had eleven .fathoms. To.the westward of (?ape Lambert, in latitude 20 end 1on?itnde 1 ! 7 � there are two deep openings, which appeared to be merely beys, but their bottom was not dis- .tinctly seen. On the top of the hill of the pr. ojectiug point that separates them, there are three remarkable rocky sum*
�
