184 



MEASUREMENT OF AN ARC ON THE 



taken on the other part of the arc, by turning tlie 

 instrument half round on its vertical axis. The 

 mean of these two was therefore the true observed 

 zenith distance, and half the difference was^he error 

 of collimation. For applying these to the Jvinpose 

 in question, the mean of the zenith distances\ being- 

 corrected for refraction, the declination of tl\e star 

 for each of these nights, was corrected for nuti.tion, 

 aberration, &c. to the time of observation, andy the 

 mean of the two taken for determining the latiti^de. 



In this manner has the whole series of observa- 

 tions been continued, by turning the sector half 

 round every night, for the purpose of observing on 

 opposite parts of the arc, and each compared with 

 its preceding and succeeding one. In pursuing this 

 method, it was unnecessary to notice the error of 

 collimation for any other purpose than as a test to 

 the regularity of the observations ; for until they 

 became uniform, no notice was taken of the zenith 

 distances, concluding that there had been some 

 mismanagement, or some defect in the adjustment. 



The following tables contain the observations by 

 the star Aldebaran, for determining the length of 

 the arc. 



