APPENDIX. 



*' of eighty degrees it will caufe an error of twenty-three 

 " feconds. As we generally took feveral fucceffive ob- 

 " fervations, any error in the obfervation itfelf will be 

 " generally independent of the reft; and as I have calcu- 

 " lated each feparately,- the conclufions will fhew which 

 " are erroneous, by their differing much from the mean 

 " of all, which cannot but be very near the truth. 



" In calculating thefe obfervations, I found by the 

 " logboard how much we had altered our latitude {ince 

 " the lafl: obfervation; and fometimes, when we had 

 ** an obfervation the noon following the obfervation 

 " for the time, the latitude of the fhip at the time 

 " the altitudes were taken was inferred from it. As mod 

 **" of our altitudes were obferved when the fun was near 

 *' the prime vertical, a fmall error in the latitude will 

 " not produce any conGderable change in the time; 

 *' indeed, if it is exadlly in the prime vertical, it will not. 

 *' make any change at alh 



*' To find the Longitude from thefe obfervations: to 

 "the apparent time found by calculation, apply the 

 *' equation of time according to its fign, which will 

 ** give the mean time ; the difference between which and 

 " that marked by the watch, will flievv how much it is . 

 " too flow or too faft for mean time. 



" Captain 



3^3 



