108 Au Account of the Repeating Circle, 



rations to a distance on both sides of the meridian to obtain 

 the requisite number of repetitions, incurring the error arising 

 from uncertamty of time; or keeping close to the meridian, 

 put up with a Umited range, and render his work equally un- 

 satisfactory through that circumstance. It was before stated 

 that, to avoid hurried observation in the use of A, stars at 

 nearly equal distances of right ascension should be chosen ; 

 and it should be remarked, that it is almost equally important 

 to have them also at nearly equal distances of declination ; be- 

 cause the more regularly they are dispersed over the whole 

 range of arc employed, the more perfect will be the correction 

 of erroneous dividing : for two stars having the same declina- 

 tion will evidently be affected by the same error, and jf they 

 differ only a little, will probably partake of the same. When 

 A is used in the observation of many stars, tlie errors bolh ol 

 observation and graduation are reduced by taking the mean. 

 With II, the errors of many observations on one star are di- 

 minished precisely in the same degree, but as the errors of the 

 whole intermediate arc are passed over and do not enter at all 

 into the account, only those of the zero reading and the final 

 one are charged upon the general mean. This distinction, cer- 

 tainly in favour of R, may be gathered from what has been said 

 belbre, and is the little all of W. It is here brouglit forward 

 again, for the use of those numerous observers who have not 

 seen through the blind process of repetition, and who have 

 attributed to this instrument powers of approximating towards 

 the inaccessible point of truth little short of the miraculous. 



I will now subjoin in detail the observations of some zenith 

 distances of a fixed mark made about ten years ago with an 18 

 inch repeating circle at St. John's College Cambridge, by a 

 gentleman who ranks high in science at that distinguished uni- 

 versity, and who is a most worthy member of this society. 



Double Obs. Zenith dist. Zenith dist. 



liy repetition. by successive pairs, 



1 = 92 4-5 3^2 3,2 



2 = 4,8 6,3 



3 = 6,3 9,4- 



4 = 6,8 8,3 



5 = 7,5 10,4. 



6 = 6,9 4,0 



7 = 7,2 8,7 



8 = 6,9 5,2 



9 = 7,1 8,7 



10 = 7,0 5,9 



11 = 7,2 8,7 



12 6,8 3,1 



Since 



