S2 



CATÎPMAEL AND McLEOD ON THE 



TABLE VI. 



The single night's work being considered insufficient For the accurate determination 

 of the personal eqviatiou of the observers, the following equations of condition from the 

 results of exchange on the individual nights of time determination have been used, with 

 the result of the special determination, for obtaining a better value of the personal equa- 

 tion ; the numbers set opposite the individual equations are the weights used, and are 

 such that the probable error of each equation so weighted should be one second of time. 



It may be noted that the probable error of a single weighted equation, as obtained 

 from the equations themselves, is 1.85 sec, which is considerably in excess of the amount 

 (1.00 sec.) deduced from the probable errors of the time determinations on the various 

 nights. This may be due to several causes ; — 



1. Errors in the determination of the levels ; the amount, however, due to this cause 

 would probably be very small. 



2. To refraction in Azimuth, due to the planes of equal refraction not being horizonal ; 

 this cause is also likely to give rise to but small errors. 



