Ch. L south AMERICA. 15 



between the magnetic and true north of the world; but 

 alfo as, by repeated obfervations of this kind, the lon- 

 gitude may be found, and we may know within a de- 

 gree, or a degree and a half, the real place of the ihip ; 

 and this is the neareft approximation to which this has 

 been carried by thofe who revived it at the beginning 

 of this century. Among thcle ihe chief was that cele- 

 brated Engliihman, Dr. Edmund Halley : in emula-- 

 tion of whom, many others of the (¿une nation, as alfo 

 feveral Frenchmen, applied themfclves to the improve- 

 ment of it. We already enjoy the iruits of their la- 

 bours in the variation charts lately publiihed, tliough 

 they are principally ufeful only in long voyages; where 

 the difFerence of two or of even three degrees is not 

 accounted a conliderablc error, when there is a cer- 

 tainty that it cannot exceed that number. This fyilem, 

 though new with regard to the ufe it is now applied 

 to, is lar from being fo among the Spaniards and Por- 

 t'jgiicfe, very plain veitiges of it remaining in their old 

 treaiifes of navigation. ?daniel dc Figueyredo, cofmo- 

 grapher to the king of Portugal, in his Hydrographia, 

 or Examin de Pilotos, printed at Lilbon in 1608, 

 chap. ix. and x. propotes a method for finding, from 

 the variation of the needle, the diiiance run in failing, 

 cail and weft. And Don Lázaro de Flores, in his Arte 

 de Navegar, printed in 1672, chap. i. part i ¡.quotes 

 this author, as an authority to conGrm the fatne re- 

 mark made by himfelf; adding (chap, ix.) that tfjc 

 Portusucie, in all their reixulalions conccrninT navi- 

 gation, recommend it ns a certain method. It mufl- 

 however be acknowledged, that tiiote ancient writers 

 have not handled this point with the penetration and 

 accuracy of the Englifh and French, affifted by a 

 greater number of more recent obfervations. And that 

 the obfervations made in this voyage may be of the 

 moft general ufe, Í íhall infert tliem in the tv/o follow- 

 ing tables ; previoully informing the reader, that the 

 If.mgitudcs correiponding' with each arc true, the 



error 



