Inteoduction. xix 



The following are the corrections which have been applied to the observations 



XT 



of declination, on account of the estimated torsions ; the first value of ^ applying 

 only to the observations before May 23*^ 20**. 



Period. 



Torsion. 



Correction 





o 



Sc. Div. 



May 16<i 23i^— 23<i 5^, 



180 



+ 16-3 



May 23-1 201^ 30^ 20^, 



249 



+ 19-5 



May 30'i 231', 



160 



+ 125 



May 31<i 2^ 5\ 



40 



+ 3-2 



May 31'! 20i'— June 2<i 5'', 



37 



4- 2-9 



June 2d 201^ June 3^ 5\ 



9 



- 0.7 



June 6d 20ii. Two fibres found broken ; they were cut off and the torsion removed. 



June 20d 21^. Torsion tried ; found to be 61° 30'. 



June 26^ 201*. Fibre broken ; removed the torsion which was 173°. 



June 28"^. The torsion is always determined and removed during the period of 

 dip observations. For these determinations, see notes to the Daily Observations. 



July 18^ 20^. Fibre broken ; the torsion was removed as nearly as possible for 

 the day's observations. 



July 19*^ 5''. The thread, prepared May 26^, and having a brass weight sus- 

 pended since then, was now inserted, the other having become weak. This thread 

 was composed of 20 fibres. The torsion was completely eliminated this evening, 



19. Errors from accidental sources. 



It is believed that a small magnet intended to be placed in the brass bar to 

 facilitate the elimination of torsion from the suspension thread of the declinometer, 

 but which was never used, had been lying in the writing-desk between July 1841 

 and March 29. 1843. The position of the desk was to the east of the reading tele- 

 scope of the declinometer, except on term-days, when, for convenience, it was moved 

 to a position nearly midway between the piers of the Declination and Balance 

 magnetometers . 



In the usual position of the desk, the greatest effect of the small magnet on the 

 declination might be from + 0'-2 to — 0'-2 ; and during term-days from + I'-l to — I'-l, 

 It is probable that the magnet remained in the same position in the desk for a long 

 period. 



The effect would be constant for each term, and from term-day till term-day. 



Absolute Declination. 



20. The absolute declination is determined in the following manner: — 



The middle wire of the theodolite telescope is brought to coincide with the 



