342 



motion of a magnet, which, independently of gravity, would rest in 

 the direction of the magnetic force. 



It has been customary, however, to give theories of the cause of 

 magnetical variations, with reference solely to the diurnal variations 

 of the magnetic declination (and not unfrequently with a very indif- 

 ferent knowledge of the facts with respect even to that element). I 

 venture to say, that it will only be from a careful comparison of the 

 whole facts relating to the motions of a freely suspended dipping 

 needle, not for one place, but for different and distant portions of the 

 earth's surface, that a satisfactory theory will be obtained. The 

 attempt to deduce one from a consideration of the declination varia- 

 tions alone, can only be likened to a similar attempt with reference 

 to planetary motions, the apparent position of the planet being stu- 

 died without any relation to the direction or rate of motion of the 

 place of observation. 



Dr Lloyd, who has done so much for magnetical science, has 

 lately brought forward a discussion of his declination observations, 

 which he considers strongly in favour of the theory that the diurnal 

 variations of magnetic declination are due to the sun's heating effect 

 upon the earth, in opposition to the atmosphere. ,1 venture also to 

 offer my guess, founded upon a consideration of various meteorologi- 

 cal facts, that it is in the atmosphere, and not the earth, that we shall 

 find seated the secondary causes of magnetic variations. In the mean- 

 time, it is facts that are wanted. 



It may be noticed, chiefly with reference to the months from 

 March to October, that a line passing through the positions of noon 

 and midnight, also passes through, or nearly through, the mean 

 position, or the centre of gravity, each hour having equal weight : 

 also a line passing through the positions, about four hours before, 

 and four hours after noon, passes nearly through the centre of gravity 

 of the perimeters ; the former of these lines lies nearly in the direction 

 of the minor axis, the latter nearly in that of the major axis of the 

 rude ellipses for the midsummer months. The horizontal line pass- 

 ing through the centre of gravity also passes nearly thi'ough the 

 positions of 1** a.m. and 1^ p.m., which, therefore are the epochs of 

 mean dip. 



Angular Distances between the Hourly Positions from the Mean 

 of all, and from the Undisturbed Days. — In order to render the 

 following result intelligible, it must be stated that, after a careful 



