VOYAGE TO THi: PACIFIC OCEAN. 17 



navigated it completely, and ascertained its extent 

 and division into two islands.* Repeated visits since 

 that have perfected this important discovery, which, 

 though now known to be no part of a Southern con- 

 tinent, will, probably, in all future charts of the 

 world, be distinguished as the largest islands that 

 exist in that part of the Southern hemisphere. 



5. Whether New Holland did or did not join to 

 New Guinea, was a question involved in much doubt 

 and uncertainty, before Captain Cook's sailing be- 

 tween them, through Endeavour Strait, decided it* 

 We will not hesitate to call this an important acqui- 

 sition to geography. For though the great sagacity 

 and extensive reading of Mr. Dalrymple had dis- 

 covered some traces of such a passage having been 

 found before t, yet these traces were so obscure, 

 and so little known in the present age, that they had 

 not generally regulated the construction of our charts; 

 the President De Brasses t, who wrote in 17^6, and 

 was well versed in geographical researches, had not 

 been able to satisfy himself about them ; and Mons. 

 de Bougainville, in I768, who had ventured to fall 

 in with the south coast of New Guinea, near ninety 

 leagues to the westward of its south-east point, chose 

 rather to work those ninety leagues directly to wind- 

 ward, at a time when his people were in such distress 

 for provisions as to eat the seal-skins from off the 

 yards and rigging, than to run the risk of finding a 

 passage, of the existence of which he entertained the 

 strongest doubts, by persevering in his westerly 



* Its southern extremity nearly in latitude 4-7, and its northern 

 in S5h. See Captain Cook's chart, in Hawkesworth, vol. ii. 

 p. 28U 



f See the track of Torre, in one of Quiros's ships, in 1606, be- 

 tween New Holland and New Guinea, upon Mr. Dalrymple's Chart 

 of Discoveries in the South Pacific Ocean, before 1764-. 



J M. de Brosses says of New Guinea: " C'est une longue isle, 

 " ou presqu' isle, si elle touche k la Nouvelle Hollande.' Navi- 

 gations aux Terres Auatrales, torn. i. p. 434. 

 VOL. V. e 



