124 MEMOIR OF C. F. BEAUTEMPS-BEAUPRE. 



Dumont d'Urville, passed in the corvette la Coquille, 2d and 3d August, at 

 about halt' a degree to the W.SW. of the island. Strong eastwardly winds 

 prevented him from approaching nearer, but he took numerous bearings which 

 seryed to rectify the position of the island, and then obeyed without thought 

 the wind which bore him away from it, having himself no reason for supposing 

 that this obscure spot presented any trace of the expedition of la Perouse. 



Yet the veil was about to be withdrawn. Four years after, in December, 1827, 

 and January, 1828, M. Dumont d'Urvillo was lying with the Astrolabe in the 

 port of Hobarttowu, situated in those parts of Van Diemen's Land which MM. 

 d'Entrecasteaux and Beautemps-Beaupre had surveyed with so much care while 

 they were still desert. Here reports reached him, vague indeed, and even con- 

 tradictory, respecting a surprising discovery made by Captain Dillon, com- 

 manding an English vessel, engaged in commerce. This mariner, it was said, 

 had acquired authentic information relative to the shipwreck of la Perouse, and 

 had even brought aAvay the handle of a sword which he claimed to have belonged 

 to that celebrated navigator. 



Notwithstanding the slight authority for these reports, M. Dumont d'Urville 

 thought himself justified in modifying the route which his instructions traced 

 for him. He touched, February 10, at Tikopia, where he found among the 

 natives a lascar named Joe, a sailor and native of Calcutta, who was the same 

 that had sold the sword-handle to Captain Dillon. This man, after a little 

 hesitation, acknowledged that some years before he had gone to the Vanikoro 

 isles, which are no other than the group of la Recherche, where he had seen 

 many objects belonging to the vessels of la Perouse ; that he had been then 

 told that two very aged whites were still alive, but he himself had not seen 

 them. 



The next day, February 11, 1828, the Astrolabe sailed for the Vanikoro 

 islands, situated, according to the natives, about forty leagues W.NW. from 

 Tikopia. The vessel came to anchor, February 14, at the place of its 

 destination, and remained till the 17th of March. M. Dumont d'Urville, being 

 quite seriously indisposed, could not quit the corvette, Avhich, besides, was, in 

 more than one respect, not considered in entire safety ; but, after having inter- 

 rogated the natives, he despatched in succession several parties commanded by 

 responsible officers, with whom he associated his faithful surgeon, M. Gaimard, 

 whose recent death has been a new occasion of sorrow to the friends of science. 



The chain of reefs which, at a distance of two or three miles, forms an im- 

 mense girdle around Vanikoro, closely approaches the southern coast near Paiou, 

 in front of a place called Ambi. Here it is but a mile off, and it was here that, 

 on a first visit, the native who preceded M. Jacquinot stopped his canoe in an 

 opening between the breakers, and made a sign to the Frenchmen to look be- 

 neath the water. There; at a depth of twelve or fifteen feet, were clearly 

 distinguishable, scattered here and there, and imbedded in corals, anchors, 

 cannons, bullets, and divers other objects, especially numerous sheets of lead ; 

 the wood had entirely disappeared. The position of the anchors seemed to in- 

 dicate that four of them had sunk with the ship, while two others had probably 

 been let go. On a second visit M. Guilbert succeeded in withdrawing from the 

 reefs the following objects : An anchor of about eighteen Imndred pounds weight, 

 without a stock, much rusted and covered with a crust of corals apparently from 

 one to two inches in thickness ; a cast cannon, likewise covered with corals, and 

 so much oxydized that the metal readily yielded under the hammer ; a small 

 swivel of brass and a blunderbuss of copper in much better preservation, one 

 bearing on its trunnions 548 as its number, and 144 as its weight ; the other 

 286 and 94 for its number and weight respectively, with no other marks ; a pig 

 of lead and large sheet of the same metal, together with some fragments of por- 

 celain. The remains of a kettle had been previously procured at Nama, a vil- 

 lage of the coast. 



