ON THE NUMERICAL RELATIONS OF GRAVITY AND MAGNETISM. 



131 



terms of Horizontal Force) for each Year, from 1841 to 1857, including only those Days 

 of Great Magnetic Disturbance, in which Records were made by the three Instruments." 



Theoretical value of M, 0011o7 



Observed " « 001150 



The " Westerly Force" I regard as representing the permanent magnetic force produced 

 by the constant flow of currents between each point of the isothermal equator and the 

 poles of greatest cold; the "Northerly" and "Nadir" forces as the tendencies towards the 

 solar meridian, and towards or from the earth's centre, produced by the diurnal solar wave. 

 This assignment appears to be justified by Mr. Airy's observation, that " the Aggregate 

 for the Westerly Force .... (taken in comparison with that for the Northerly Force), 

 appears to show that, on the whole, the direction of the Disturbing Force is 10° to the 

 east of south ;" p. 628. This indicates a line of mean disturbance about midway between 

 the magnetic meridian (which, at London, is about N. 24° W.), and the solar meridian, 

 or midway between the meridians of decussation in the two sets of principal spirals, 'o 

 which I have referred. 



The law of varying attraction suggests a plausible explanation for the approximate mean 

 proportionality of the barometric to the tidal and magnetic variations. For the ratios of 

 attraction of any planet when in solar conjunction, at quadrature, and in opposition, vary 

 as (n~\ L)'\ ri\ and (n — l) 2 , respectively, the attraction at the mean distance being nearly a 

 mean proportional between the maximum and minimum attractions. The barometric 

 fluctuations are occasioned by variations in the gravitation of the air towards the earth's 

 centre,— the tidal motions, by the influence of distant heavenly bodies, — and the mag- 

 netic, according to my hypothesis, by the oscillations of the air and author in their efforts 

 to restore the unsettled equilibrium. The three disturbances, therefore, must evidently 

 have nearly the same mutual relations as if they were produced by three forces, one cen- 

 tripetal, and the other two centrifugal, the two latter being nearly equal in amount but 

 diametrically opposed in direction. 



In our discussion of Prop. V, we found some indications of a lagging of about 13' in the 

 time of the magnetic mean disturbance, which is about one-fourth of the average baro- 

 metric retardation. This corresponds very nearly with the ratio of the lunar-barometric 

 to the lunar-magnetic disturbance (4.384), and with Mr. Welsh's determination of the 

 moment of magnetic inertia (4.4696; Phil. Trans., v. 153, p. 297). 



The foregoing comparisons have been based almost exclusively on Major-General 

 Sabine's discussions of the St. Helena records. It would be desirable, if it were possible, 

 to confirm them by observations at other stations near the equator ; but the need of such 

 confirmation is in great measure obviated, by the variety of ways in which I have shown the 



