Gravity and Magnetism. 53 



strated, the influence of the sun upon terrestrial magnetism ; 

 Secchi ascertained that " the diurnal excursion of the needle is 

 the sum of two distinct excursions, of which the first depends 

 solely on a horary angle, and the second depends, besides, on 

 the sun's declination""*, and that " all the phenomena hitherto 

 known of the diurnal magnetic variations may be explained by 

 supposing that the sun acts upon the earth as a very powerful 

 magnet at a great distance "f. 



This hypothesis has been objected to, on the ground that it is 

 difficult to understand how any conceivable intensity of solar 

 magnetism, by its simple induction, could produce so great a 

 disturbance as is daily observed. Therefore it will probably 

 follow the fate of the earlier ones, which attributed terrestrial 

 magnetism to one or more powerful magnets lying nearly in 

 the line of the earth's axis, while Barlow's idea that the mag- 

 netism is superficial and in some manner induced {, will still re- 

 main in the ascendant. Secchi's conclusions are, however, none 

 the less interesting, and from the fact that magnetism is, like 

 gravity, a central force, varying inversely as the square of the 

 distance, they lend encouragement to those who are endeavouring 

 to find new evidences of the unity of force. 



My own experiments and researches have led me to the belief 

 that all magnetism is a simple reaction against a force which dis- 

 turbs molecular equilibrium, that the numerical equivalent of 

 the magnetic force is therefore equal and opposite to that of the 

 disturbing force, (+ M= +D), and that all the phenomena of 

 terrestrial magnetism result from tidal and thermal changes in 

 terrestrial gravitation. 



Sullivan § and Reinsch || have pointed out the effect of musical 

 vibrations upon the magnetic needle; and I have shown the con- 

 trolling influence of a purely mechanical polarity ^[. A careful 

 examination of the polarizing thermal and rotation currents** 

 will show that the spirals, which they have a tendency to pro- 

 duce, are quasi-horizontal cyclones — one set flowing in a nearly 

 constant direction along the magnetic meridian, and the other 

 toward the momentarily shifting solar meridian. From an in- 

 vestigation of these currents and a comparison of various obser- 

 vations, I have deduced the following theses : — 



I. The daily magnetic variations, though subject to great dis- 

 turbances at different hours, show an average approximation to 

 the differences of the gravitation-tidal currents. 



* Phil. Mag. S. 4. vol. viii. p. 396. f Ibid. vol. ix. p. 452. 



% Phil. Trans. 1831. § See De la Rive's 'Electricity,' vol.ii.p. 635. 



|| Phil. Mag. S. 4. vol. xiii. p. 222. 



IF Proc. Amer. Phil. Soc. vol. ix. p. 359. ** Ibid. p. 367 et seq. 



