i7i different Latillides. 97 



equator is circular, might be represented very exactly, by the 

 action of two magnetic centres placed at a small distance one 

 from the other, near the centre ot the earth. Botli M. Hum- 

 boldt and myself have been led to this result in the work which 

 I have mentioned aljove, and our Memoir was already published 

 when I learned that the celebrated astronomer Mayer had also 

 arrived at the same conclusion, in discussing the dips known in 

 his time : and that he even used it to represent the variations, in 

 a Memoir read to the Society of Gottingen, but never printed*. 

 The son of this great astronomer having been so obliging as to 

 send me an extract from it, I have been enabled to convince 

 mvself of this identity; and I also found that Mayer had disco- 

 vered, in an experimental manner, that the law of magnetic at- 

 tractions is reciprocally as the square of the distance. 



Tbi^ consequence, which we have l)oth deduced from elements 

 so different, appears to indicate something more than a merely 

 empiric law. It is therefore necessary to examine it a little 

 closer. First, it is easy to see that a single magnet placed at 

 the centre of the earth could not produce all the phenomena; 

 for then the magnetic equator would be a great circle perpendi- 

 cular to the right line drawn by the two centres of action, and 

 there would not result from it the inflexion which we have ob- 

 served in the South Sea. In other respects, such a magnet, in 

 whatever way it might be placed, would necessarily give symme- 

 tric phrenomeaa on both sides the plane drawn by its two centres 

 and by the centre of the earth, a synunetry which is no way con- 

 formable to the facts observed, especially in the South Sea and 

 the continent of Asia. 



Not being able to adopt this simple idea, let us endeavour to 

 depart from it as little as possible ; and since we have found 

 that it represents the observations made in Europe and in the 

 Atlantic Ocean sufficiently well, let us try to make such a modi- 

 fication that it may be scarcely sensible in this part of the 

 globe, and vet that it shall be pretty large in the opposite part, 

 where the magnetic eciuator suddenly experiences its inflexion. 

 W'c shall arrive at this by placing a second excentric magnet near 

 this ])oint, whose position and relative energy shall be determined 

 in such a way as to agree with the observations. Now, by making 

 the calculation, we find it sufficient to give a very small force to 



* This cirninistancc is nicntionod by Professor R()()isoi), at the end of 

 the article Mnfinelism, in the Supplrineiit to the tliird edition of the En- 

 cyclofailia Biitdv/iicu. He also jiivcs tl)C results of Mayer's calrulntions 

 as they were first |)iil.li>hL(l hv Lichttiihiirg in his edition of l'".xliben's 

 Elements of Natural I'liilosopliy, 178-1. The article here referred to, and 

 that of Varialitm in the hody of the work, contain, jjerhaps, the most com- 

 plete Theory of Maunetisni that has hitherto hrcn pnhlished. 



Vol.49.No.22(». Fe/'. 1817. <' this 



