OBSERVATIONS WITH THE ZENITH TELESCOPE. 



301 



turning it slightly in azimuth after each reading. The individual readings would prob- 

 ably never agree exactly, even if the performance of the level were perfect, owing to 

 want of precision in the vertical axis. Large deviations from the first reading were, 

 however, regarded as sufficient ground for rejecting such observations. 



It is perhaps an open question whether anything was gained by this procedure, as 

 there are obvious difficulties in formulating a criterion for rejection founded on the 

 evidence so furnished. 



The work of 1885-86 is in all essential respects a repetition of that of 1876-78. 

 The same star list was used with the exception of three pairs which precession had 

 carried beyond the range of the Micrometer. 



The series of 1888 was intended to be a repetition of the same procedure, with some 

 improvements which it was hoped would add materially to the accuracy of the results. 

 The Micrometer was thoroughly repaired and the old springs replaced with new ones. 

 A level by Stackpole & Brother took the place of that of Kahler, which had been in 

 use since 1870. A level-trier was also provided, with which it was proposed to make 

 frequent tests of the sensitiveness of the tube. 



Although the new level proved to be much better than the old one, it failed to meet 

 the requirements in respect to precision, and was returned to the maker after a trial of 

 six weeks. The difficulties in finding a tube of the required delicacy proved greater than 

 was anticipated; a delay of nearly a year was the consequence. 



Meanwhile, in the future prosecution of the work a more elaborate program had 

 been decided upon. This included a reobservation of the old list, with the exception of 

 the three pairs which were no longer available. It is with this part of the program that 

 we are now concerned ; that relating to the new List prepared at this time being treated in 

 another section. 



The distribution of the observations on the old star list, originally comprising 60 

 pairs, was as follows : 



Tear. 



Number 



of Observations. 



Tear. 



Number of Observations 



1876 





187 



1886 





167 



1877 





178 



1888 





121 



1878 





12 



1889 





17 



1885 





115 



1890 



Tola!, 



397 

 1144 



A number of the older observations, in which the level corrections were large or 

 where the readings appeared to be fictitious, were excluded for reasons already given. 



A. P. 8. VOL. XX. MM. 



