ON THE NUMERICAL RELATIONS OF GRAVITY AND MAGNETISM. 



119 



ulterior results, and it may be reasonably supposed that such a ratio betokens a natural 

 connection between the forces themselves. This hypothesis is greatly strengthened in the 

 case now under consideration by the experimental illustrations that I have given of the 

 influence of mechanical vibrations upon the magnetic needle (Proc. A. P. S., vol. ix, p. 

 359), and by the evident tendency of simple known causes to produce spiral eddies in the 

 air and aether analogous to those which arc constantly circulating around permanent mag- 

 nets. (Ibid., pp. 367, sqq) * Having the elements of a mechanical polarity thus given, 

 we might reasonably suppose that any excess of magnetic influence in one portion of the 

 day would be balanced by a deficiency in some other portion, so that the average hourly 

 differences would assume a tidal form, as stated in Prop. I. That such is the case is 

 shown by 



TABLE I. DAILY DIFFERENCES OF MAGNETIC FORCK.f 



DIFPHRBNOBS IN HUNDRED-THOUSANDTHS OF TOTAL FOR0B, 



RATIOS. 



Astronomical Time. 



lh. 



2 h. 



3 h. 



1 h. 



2h. 



8h. 



Before 8 h. 



After 8 li. 

 Before 9 h. 

 After 9 h. 

 Before 15 h. 

 After 15 h. 

 Before 21 li. 

 After 21 li. 

 Means, 



20 



20 



5 



4 



2 



1 



31 



29 



14 



42 

 39 

 15 



4 



5 



2 

 54 

 51 

 26.5 



55 



55 

 26 



2 



7 



4 

 66 

 61 

 34.5 



.864 



.864 



.192 



2.000 



,286 



.250 

 .470 

 .475 

 .400 



.764 

 .709 

 .577 

 2.000 

 .714 

 .500 

 .818 

 .836 

 .761 



1 

 1 

 1 

 1 

 1 

 1 

 1 

 1 

 1 



Mean of Ratios, 







.503 



.865 



1 



'■'- 'Hie .simplest of the three experiments referred to is performed with an ordinary binnacle compass. Holding 

 the gimbals so as to allow motion in only one direction, and causing the box to swing like a pendulum, the needle 

 will tend towards the line of oscillation, thus showing that simple mechanical vibrations can produce a species of 

 polarity. If the combined action of convection and rotation upon the solar meridional currents, and on the 

 temperature currents between the equator and the poles of greatest cold, be carefully studied, it will be seen 

 that they furnish the elements of two sets of spiral eddies, or quasi-horizontal cyclones, in the air and eether, one 

 flowing in a nearly constant direction along the magnetic meridian, and the other towards the momentarily shifting 

 solar meridian. As the air has a certain degree of specific magnetism, the polarity which these eddies produce, 

 may explain the directive energy of the compass needle. 



-\ All of these tables are compiled front the St. Helena Observations and General Sabine's Discussions. 



