5l6 ACCOUNT OP ASTRONOMICAL OBSERVATIONS 



which quantity o".13l is entered on the IX. column of 

 Table II. and so of the rest. 



6. I WAS at first doubtful respecting the best mode 

 ©f obtaining a very accurate mean latitude for the Ob- 

 servatory, and hesitated between making a selection of 

 a certain set of stars whose declination was determined 

 ^t Greenwich after the same method, and with the same 

 instruments ; or taking the whole mass of my observa- 

 tions without adverting to the catalogues either English^ 

 French^ or German, from which I had taken the decli- 

 nations. 



7. Had the whole of the stars given in Table III. 

 been computed when I began writing this paper, I 

 might 'have been induced to think the separate cata- 

 logue in Table II, unnecessary, since the two means ^ 

 only differ by 0''.36j. However, I was, at the time, 

 determined by an opinion that the results of a few very 

 accurate operations were always preferable to the mean 

 of a great number of indifferent ones, and chose there- 

 fore twelve principal stars (six on each side of the ze- 

 nith) the declinations of which are given in Dr. Mas- 

 kelyne's catalogue for January 1802. With these 

 I constructed Table II. to which I particularly wish to 

 calV the attention of the reader, as every thing that I 

 shall say hereaf*'er is grounded on the mean latitude 

 which is derived from it. 



8. In this catalogue, the maximum of deviation in 

 the respective latitudes is only 4''.551, and their gra- 

 dual decrease as the stars become more southerly indi- 

 cates that this difference is not solely to be attributed 

 to inaccuracy in the observation ; for it is to be ob- 

 served that the regularity of this decrement (which is 

 scarcely interrupted) cannot altogether be ascribed to 



