78 OBSERVATIONS OF THE MAGNETIC INTENSITY 



needles since they came into my possession. The needles were always kept 

 separate from each other; while travelling, they were carried about my person, 

 and, when stopping for any considerable length of time, were deposited as far 

 from iron as was necessary to their safety. From the experiments made with 

 these needles, both of which have been magnetised several years, and which 

 have been kept carefully for more than six years, I should be disposed, in fu- 

 ture, to adopt the plan of procuring needles of as nearly equal force as possible,* 

 and keeping them in pairs, which renders them much more convenient to 

 carry. It is certain that permanence of force has not resulted, in these nee- 

 dles, from the opposite plan, and that the labour of observation and calculation 

 are much increased by the necessity of ascertaining and applying a correction 

 for the loss. In comparing two sets of experiments at a distant date, to ascer- 

 tain the loss of magnetism by the needle, the results are affected by the change 

 of dip which has taken place in the interval, and as it is not probable that this 

 change is produced by an alteration in the total intensity, a correction is to be 

 applied, which, however, except in the longest interval of my series, was 

 scarcely appreciable. 



The magnetic dip was observed in the usual way, the poles of the needles 

 being reversed in each series. The bar magnets for reversing the poles were 

 placed in the top of the box containing the dipping circle, each pair of opposite 

 poles being connected by a keeper. Notwithstanding this arrangement, the 

 bars lost much of their strength, probably from the percussion resulting from a 

 slight play which was allowed in the bed where they were placed; and, on my 

 arrival at Berlin, their magnetism was so much diminished that the dipping- 

 needles could no longer be charged by them to saturation. Since that time I 

 have always taken the precaution, after changing the poles, to oscillate the 

 needles within determinate arcs, and when resting on the agate planes, to ascer- 

 tain, by the time of oscillation, that they are charged at least nearly to satura- 

 tion. The statement of this precaution may be of service to others, since, with 

 a diminished force in the needles, the liability to take up some other position 

 than that corresponding to the true dip is increased, and the error cannot, ne- 

 cessarily, be detected by discrepancies in the several readings. Of the two 



* The importance of attending to these conditions appears very strikingly from the experiments 

 of Mr. Airy, with large magnetic bars. Royal Society's Transactions, Part I., for 1839, pp. 196, 

 197. 



