i8 A VOYAGE TO Book I. 



Juan ; and as ih'is officer continued his obfervations 

 to tiie end of the voyage with the lame needle, the 

 difference, which at firii was negative, on. the varia- 

 tion changing its denomination became poiitive ; and 

 from my changing initruments, the difference on my 

 fide continued always negative. The reaibn of this is, 

 that the difference of the five firft obfervations pro- 

 ceeded lets from a real diíFerence i-n the variation, than 

 from the poles of the needle, which was {o far from 

 anfwering exailly v/ith tlie meridian-line on the com- 

 pafs-card, that it inclined towards the N. W. ; the 

 contrary happened in tiie Iccond compafs made ufe of, 

 its inclination being towards the N. E. ; confequcntly, 

 whatever the angle of that inclination was, it occa- 

 iioned a proportionate diminution in the variation of 

 a contrary fpecies. 



These obfervations, thus compared, ffiew the errors 

 to which navigators are liable, for want of attention in 

 making choiccí of proper needles, which they fhould be 

 caieful to procure, not only well made and exa6i, but 

 alfo ftri(5tly tried with regard to their inclination to the 

 true tneridian, before they venture to depend upon 

 ihem in any voyage. In this point Spain is guilty of a 

 notorious negleit, notwitliltanding it is evidently the 

 fource of a thouland dangerous errors; for a pilot, in 

 corre6ting thecourfe he has fleered, in making ufe of a 

 compafs whole variation is different from the true, will 

 conlequentlv find a diff'creuce between the latitude by 

 account and the latitude obierved ; and to make the 

 neceflary equation according to the rules commonly re- 

 ceived in failing on points near the meridian, he muff 

 cither increale or diminiih the difiancc, till it agrees 

 with the latitude, whereas in this cale the principal 

 error proceeded from the rhomb. The fame thing hap-' 

 pens in parts where it isapprehended'tlieremay be cur- 

 rents; which often occur in failing when thelatitudeby 

 nccount, and that by obfervation, difagree ; though 

 in reality the water has no motion, the differevice })ro- 

 cecding entirely from making ufe of another varia- 



' • tiott 



