82 Mr W. Swan on a Simple Variation Compass. 



I find that the most consistent readings are obtained, not 

 by waiting until the magnet comes to rest, but by causing the 

 theodolite wire to bisect the arc of vibration of the magnet, 

 by estimation, as soon as that arc is reduced to about 8' or 10'. 

 If the magnet has come to rest, it is easy to make it vibrate 

 again in a small arc, by cautiously approaching to it a magnet 

 or a piece of iron, which is again removed to a sufficient dis- 

 tance before making the observation. 



As an example of the performance of the instrument, I 

 select the last observation of magnetic declination I have 

 made.* 



Greenwich Mean Time. 



Observed Azimuth of line of 



Collimation of Magnet : mean of 



two Verniers. 



1854, April. 

 10 d It 30™ 

 34 

 38 



Before reversal of magnet. 



77° 46' 10" 

 46 10 

 46 5 



l h 45 m 

 48 

 51 



After reversal of magnet. 



77° 43' 37" 

 43 45 

 42 25 



s,:l ih 4im 



77° 44' 42" 



Azimuth of the magnetic axis of the magnet = 77° 44' 42" 

 Azimuth of true north . • . =102 48 44 



Variation of needle 



= 25° 4' 



The Kater's circle, by means of which these observations 

 were made, has both its vertical and horizontal limbs 6-5 inches 

 in diameter, each furnished with two verniers, reading 10". 

 The azimuth of the true north was deduced from transits of 

 the sun, taken near the meridian, in the following manner : — 

 The vertical circle being placed approximately in the meridian, 

 the sun's transit over the five diaphragm wires was observed ; 

 and the error and rate of the chronometer used were ascer- 

 tained by comparison with the Edinburgh time-ball within 



* This paper was read on 19th June 1854. 



