xviii Introduction. 



nation are not given, on this account, till noon, when the thread was wound up. 

 The observations after are probably affected by a considerable torsion-force. 



January 19*^ 2^. 1842. Two fibres of the suspension-thread found broken ; they 

 were removed, but the plane of detorsion had evidently varied to a considerable 

 extent. During the following night and day (term-day), the thread broke away 

 fibre by fibre ; and though readings of the instrument were continued, they cannot 

 be considered of much value. They are given from January 19^ 10^ till 20'' 6% and 

 are corrected by — 13'.4, an approximate reduction for torsion obtained from a com- 

 parison with the usual daily observations. 



January 21. 1842. A new thread prepared ; allowed to stretch for several days, 

 when the magnet was suspended. The torsion eliminated, the circle reading 280°. 



May 24. 1842. A fibre found broken, removed by cutting near the top and 

 bottom ties. 



May 26. A new thread prepared, formed of 20 fibres. A weight was attached 

 to it, and suspended till required. 



May 30'^ 20''. No observations of torsion were made lately. On removing 

 the declination magnet, in order that the dip observation might be made, the brass 

 bar which was inserted was generally caught by a wooden block ; it was allowed to 

 move freely to-day, when the bar went round about 400°, the motion opposite to that 

 of the sun. The torsion-circle read, vernier A. 146°. The torsion was supposed to be 

 about 180°. The arm of the torsion-circle was turned through nearly 90°, by mistake, 

 instead of 180° ; torsion-circle reading, B. 235°. The magnet was replaced till the 

 observation at 23'', after which the brass bar was again inserted, and ultimately the 

 torsion-circle left reading B. 354° 55'. After b^ the brass bar was suspended until 



May 31^ 19" 30"", when the arm of the torsion-circle was turned to B. 358° 4'. 



June 2'' 5'' + . Torsion again tried, and the circle left reading B. 44° 1.3'. 



June 3*^ 5'' + . It was found, when the torsion-circle read B. 35° 0', that the tor- 

 sion was removed as nearly as possible. It is believed that, from various causes, 

 the torsion was not completely eliminated till now. Upon examining the daily ob- 

 servations, it was evident that a marked change had taken place in the declination 

 readings between May 16* 20'' and 23'', the period of the dip observation. There was 

 no doubt but that the torsion had been induced at this time. It is almost certain 

 that the brass bar must have gone round half a revolution before being checked by 

 the wooden block; or, that being inserted in the revei^se position, it had been supposed 

 to have gone round and altered accordingly. 



The breaking of a fibre May 24*^, and the probable existence of torsion previ- 

 ously, will account for the diff'erence from 180°. 



TT 



From the observations for the value of ^ March 3*^ and June 6^, a torsion of 180° 



corresponds to a deflection of the magnet of 16-3 and 14'12 scale divisions, re- 

 spectively. 



