APPENDIX. 



to remove every ciicumftance which might be fuppofed to 

 create an accidental error; the obfervations being taken 

 fometimes by different people with the fame compafs, in 

 the fame and diflerent places; fometimes with different 

 compaffes, changing the places and the obfervers re- 

 peatedly, to try whether there was any error to be im- 

 puted to local attradion, or the different mode of obfer- 

 vation by different perfons. I have fince my return tried 

 the compaffes by a meridian as well as by taking azimuths, 

 and find them to agree with one another, though the fama 

 compafs fometimes differs from itfelf a degree in fuccefliva 

 obfervations. 



That every perfon may (as far as is pofflble without 

 having been prefent at the time) be enabled to judge cf 

 the degree of accuracy to be expected in fuch obfervations,, 

 as well as the degree of attention paid to thofe made by 

 us, I have fet down every circumftance that I thought 

 material, giving every part of each obfervation, with eacli: 

 feparate refult, and the mean of every fet, with the 

 weather at the time. Whenever I mention its blowing; 

 frefh, it was only comparatively with refped to the reft 

 of the voyage, no obfervation having been made in any 

 weather which might not generally fpeaking be called 

 fine^ 



Having faid fo much of the inaccuracy of the inftrur 

 ment, I muft add, that I think fome general and rather 

 curious inferences may fafely be drawn from thefe 



S 3 obfervations^ 



iC9> 



