94 .< .Mr, Powell on tke Communication [FebJ 



r«so&> obtained by placing the wire ^t such a distance from the 

 pole of the needle as to make it deviate as nearly as possible 

 2C)° 34', and the same distance was kept with the other wires. 

 By examining the tables, it will be seen that the tan. of 26° 34' 

 is very nearly the half of radius, or sin. 90°, and the tangent* 

 which are the nearest to the halves of the other sines o-iven 

 above are those of 



26° 13'; 25° 10'; 23° 25'; 20° 58'; 17° 49'; 14° 2'; 9" 43'; 4° 58'; 



and this set of arcs differs from the mean of the above experi- 

 ments by quantities, which are covered by the unavoidable uncer- 

 tainties both of experiment and observation. 



Should the law which I have proposed be considered suffi- 

 ciently established, I conceive it affords a strong confirmation 

 of the truth of the ingenious theory, proposed by Mr. Christie 

 respecting the nature of magnetic action (of which he has given 

 an account in the Cambridge Philosophical Transactions Part I. 

 and in the Edinburgh Phil. Journ. No. 10), when combined (as 

 Mr. C. admits it may easily be) with the theory of M. Ampere 

 concerning the magnetic or electrical currents. 



Let us suppose, according to Mr. 

 Christie's idea, magnetic currents in 

 the direction of the dip x a d ; then dj' 

 at right angles to this line is the magne- 

 tic equator. Let d c a e represent a por- 

 tion of one of the wires magnetized, as in 

 my experiments, of which the thickness 

 is a by and the inclination < c df. Then 

 also the current x a d\^ supposed to be 

 composed of currents perpendicular to its 

 axis ; let a c be the direction of one of 



these. The wire is magnetized by imbibing these magnetic 

 currents, being put into a state fit for imbibing them by the 

 torsion ; but when magnetized, it also possesses currents perpen- 

 dicular to its axis ; therefore, if it imbibe a current in the direction 

 a c, this must be resolved into a b and b c, of which a h is alone 

 effective in producing magnetism. Then it is obvious that c a : 

 a b :: rad. : sin. inc. and consequently the intensity of magnetism 

 which is measured by the tan. of the deviation produced on the 

 needle, varies as sin. I, which is the law I have deduced from 

 experiment. 



This coincidence appears to afford a strong presumption in 

 favour of the actual existence of magnetic currents, or rather 

 systems of magnetic currents (to adopt M. Ampere's idea) in 

 the atmosphere ; and for leading us to consider them as the 

 causes by which magnetism is communicated, as in these experi- 

 ments, or generally by position. 



* When pieces of iron are placed at different inclinations, it has 

 bfeen found that by mere position they may, after a time, imbibe 



