138 



SECTION IV. 

 ABSOLUTE DECLINATION RESULTS. 



The observation for Declination consists of two operations : 



1. The determination of the direction of the magnetic axis of the suspended magnet relative to the 



zero axis of the instrument. 



2. The determination of the azimuth of a fixed mark, to which the direction of the zero axis of the 



instrument is referred. 



Shortly, Part 2 is known as the determination of the azimuth of the mark. 



Observations to determine the azimuth of the mark were made on ten occasions, during the years 1902 

 and 1903. 



The results of the observations, in some cases, show considerable inconsistency, and in order to form an 

 opinion as to the best value, the change in the sun's bearing, corresponding to the interval of time between 

 the two successive transits of each observation, was calculated independently from the observations 

 themselves. 



The observed change of bearing, as derived from the difference of the readings of the unifilar circle, 

 represents : 



The true change of bearing twice the amount of instrumental error due to the axis of the mirror 

 bearings not being at right angles to the line of the telescope (or to the plane of the mirror not being 

 parallel to the axis of its bearings). 



Thus, by comparing the calculated and observed change of bearing, we have the means of determining 

 the error due to the above causes. 



In the following table are given the calculated change of bearing, the change of bearing observed, and the 

 amount of instrumental error derived from each observation. 



TABLE III. 



In the observation of date October 21, 1903, Mr. BEKNACCHI apparently observed with the mirror in 

 the reversed position first ; there is no objection to this, but it accounts for the observed change of bearing 

 being only 17'. 



It is evident from the instrumental errors deduced above that the instrument was in fair adjustment up 

 to and including the observation of date January 27, 1903; equally is it evident that subsequent to that 

 date the instrument was considerably out of adjustment so far as the mirror is concerned. 



The error of observation due to this cause is eliminated by taking the mean of the readings observed 

 with the mirror erect and reversed, but the evidence of the mirror being out of adjustment, together with 

 the inconsistency of the results of the observations taken on two consecutive days, October 20 and 21, 

 1903, seems to suggest the probability that the instrument was out of adjustment in other respects. 



