Fisher on the Variation of the Compass. 103 



ttiagnetic force iu this country, in accelerating the balance of a 

 chronometer by means of a magnet. The chronometer used 

 for this purpose was made by Arnold, which, besides the usual 

 compensation for temperature in the balance, had a steel rim at- 

 tached to it ; and when the south pole of a magnet, fixed in the 

 plane of the balance, was presented to it, an immediate accele- 

 ration of about 32" in twenty-four hours was observed. The 

 following are the rates I have been able to obtain by transit — 



Rate in 24 hours M.S. time. Barometer. Intervals. 



+ 13."6 29.702 3 days 



+ 14.7 29.789 3 — 



+ 16.9 29.991 4 — 



+ 14.8 29.940 3 — 



+ 16.8 30.061 5 — . 



+ 17.3 30.074 3 — 



+ 14.6 29.883 5 — 



The chronometer had a very steady rate of — IS'. 5, before the 

 application of the magnet. These experiments, and likewise the 

 effects produced on the acceleration by the violet ray, together 

 with a comparison of the forces accelerating the oscillations of 

 a dipping needle at Spitzbergen, Shetland, and London, I will 

 give in another paper. 



It appears now from the observations, that the afternoon vari- 

 ation, taken between noon and midnight, exceeds the morning 

 variation taken between midnight and the following noon by 

 2r.20". This excess is even apparent in the observations taken 

 ■with the azimuth compasses on the ice, particularly with Kater's 

 compasses, which, in point of lightness and delicacy of con- 

 struction, far exceed the others. It appears from them by a 

 mean of a great number of observations, made in the summer 

 of 1818, that in lat. 80°. and long. 11°. E., the morning varia- 

 tion was 23°.49'. W,, and the evening 24°.16'., making a dif- 

 ference of 27 '. : the mean variation, therefore, m thai latitude 

 is 24°.2',30''. W. 



The following arc the results of the observations made for 



