PACIFIC AND BEERING'S STIIAIT. 249 



islands, and to the men having been a long time on C ynx' 

 a reduced allowance of salt provisions. ^-^^ 



The islands which were visited between Bow i826. 

 Island and Otaheite were all of the same character 

 and formation as those already described, and furnish- 

 ed us with no additional information beyond the cor- 

 rect determination of their size and position ; which, 

 with some remarks that may be useful to navigation, 

 are given in the Appendix to the 4to. ed. Among the 

 number there were two which were previously un- 

 known ; the largest of these, which was also the most 

 extensive of our discoveries in the archipelago, I 

 named Melville Island, in honour of the first lord 

 of the Admiralty ; and the other, Croker Island, in 

 compliment to the right honourable secretary. 



The discoveries of Cook and Wallis in this track 

 are relatively correctly placed ; but those of the lat- 

 ter are as much as forty miles in error in longitude, 

 and several miles in latitude, which has occasioned 

 two of them to be mistaken for each other by Bel- 

 linghausen, and one to be considered as a new dis- 

 covery by Captain Duperrey. It would not have 

 been easy to detect these errors, had we not visited the 

 discoveries of Wallis in succession, beginning with 

 Whitsunday and. Queen Charlotte's Islands, which 

 are so situated that no mistake in them could possi- 

 bly occur. Moreover, we always searched the vici- 

 nity narrowly for the existence of other islands. 



The mistakes have arisen from placing too much 

 confidence in the longitude of the early navigator. 

 The true place of Cumberland Island lying much 

 nearer the alleged position of Wallis's Prince Wil- 

 liam-Henry Island than any other, has occasioned 

 Bellinghausen's mistake ; and the true position of 



