602 MAGNETIC FIELD. 



at any rate partially, by using an uneven number of magnets, 3 

 or 5. If the ratios of the results obtained in two different stations 

 are the same as for the three needles, it is probable that they are 

 exact; when one of them disagrees with the two others it is to be 

 assumed that its magnetism has undergone an accidental variation, 

 and acquired a new value, which will serve for the following ob- 

 servations. In this way, provided that two at least of the needles 

 have not been modified in the interval of the two stations, the 

 series of observations is never interrupted whatever changes may 

 have taken place in the state of the needles. 



1174. POISSON'S METHOD. Poisson* was the first to point out 

 how the horizontal component in a place may be determined in 

 absolute measure independent of the magnetic moment of the 

 magnet. 



The oscillations which a bar makes under the influence of the 

 earth determine the directing couple MH = A. On the other rrand, 

 the magnetic field of the bar is proportional to M, and the com- 

 parison of this field with the terrestrial field determines the ratio 



= B. Two numbers thus formed by experiment enable us to 

 H 



calculate separately the values of M and H: 



M 2 = AB, 



(21) H2 = A. 



B 



Taking into account the induced magnetisation, the observation 

 of oscillations gives 



The bar is then made to act on a needle which is parallel to it, 

 and at a distance R in the principal position in respect of the bar 



for instance, in its prolongations. By fresh oscillations (1140) the 



F 

 ratio = B' of the field of the magnet is thus determined. 



H 



The bar being parallel to the magnetic meridian and in the 

 same direction, its magnetic moment, increased by the induced 



* POISSON. Connaissance des Temps pour 1828, p. 113. 



