44 4 Sir G. C. Haughton on the Common Nature of 



or shock. I now proceed to an important step in these dis- 

 coveries. 



On the relative dynamic xmlucs of the degrees of the compass, 

 and on the cause of the needle resting in the magnetic meri- 

 dian. 



Shortly after I had begun to measure the angles of devia- 

 tion, I saw that it was a matter of the first necessity to ascer- 

 tain whether the degrees of the compass after 90° ijicreased in 

 strength, or went on diminishing in some given ratio, as it 

 appeared to me that the connexions were made with increased 

 facility in proportion as the N. pole of the needle approached 

 nearer to 180°. For the purpose of settling this question, I 

 made a number of most careful measurements, after drawing 

 proper lines upon a sheet of paper, which all crossed one 

 another in acommon centre, by ascertaining the distances at dif- 

 ferent azimuths at which a bar-magnet, placed with its axis di- 

 rected to the centre of a magnetic compass-needle (the pivot 

 of suspension of which was exactly over the common centre 

 of the lines of which 1 have spoken), caused the needle to as- 

 sume a position in which its axis was in the same line with 

 that of the magnet, I found that these points of distance form 

 a peculiar curve, which was nearest to the axis of the magnet 

 at the east and west azimuths, and receded from it as it ad- 

 vanced to the north and south ; and was twice the distance 

 from it at the north that it was at the south azimuth. Such 

 is the sum total of the results which I communicated to the 

 Royal Society in June 1846, and of which a short account has 

 been given in the Abstracts of the Papers read before the So- 

 ciety, and republished in this Journal for last November*. It 

 since however occurred to me, that the longest distance from the 

 east and west azimuths would be better ascertained by operating 

 by the means of repulsion than by attraction, as I had done; for 

 it is evident that repulsion will sooner discover a minute devia- 

 tion of the needle when in its natural position at N. than can 

 be effected by attraction, because if the axis of the bar-magnet 

 deviates in the slightest degree from the line, a false value will 

 be obtained ; and there always therefore remains a doubt in 

 proceeding by attraction, respecting the point where the reci- 

 procal action of the needle and the magnet ceases, in conse- 

 quence of the identity of their axis. Now in proceeding by 

 repulsion, the reverse takes place ; and the minutest diflference, 

 except so far as the friction of the needle on its pivot is con- 

 cerned, is immediately detected. I was surprised therefore 

 to find that where I had only been able to discover a devia- 

 * Vol. xxix. p. 405. 



