all known Substances to the Magnetic Injluence, Sfc. 105 



In all my previous experiments, which amounted to][_from 

 fifteen to twenty series, the tangent of the angle of deviation, 

 at the distance of the first length of the bar, was always consi- 

 derably less than the mean of the column of ratios, because of 

 the assumed length between the foci of the magnet being too 

 great ; but in this last series, for the first time, the tangent be- 

 longing to one focal-length's distance is somewhat too great, 

 which might seem to indicate that the assumed length was now 

 too small. Where, however, all the ratios at other distances are 

 found to coincide so very closely, the small difference in the first 

 tangent must be referred to a peculiar cause, which, the greater 

 discrepancies with shorter magnets eventually enabled me to 

 detect. 



In the foregoing investigations on the magnetic forces acting 

 upon a compass, the attractive and repulsive actions of the same 

 pole of the magnet on the different poles of the needle have been 

 considered as a simple and not a compound action, because in 

 most cases, excepting at short distances, they so combine as to pro- 

 duce almost exactly double the effect of either influence separate- 

 ly. This double effect, therefore, which for simplifying the in- 

 vestigation may generally be considered as the result of a single 

 force, cannot, in the case of a short distance, and with a large 

 compass-needle be so considered, without being the occasion of 

 a very perceptible error. For although the north pole of a 

 magnet, when placed at a coasiderable distance from a coifipass, 

 in the direction of its east or w^est point, will attract the south 

 pole of the compass, and repel the north pole with equal energy, 

 the same magnet, if placed very near the compass, so as to pro- 

 duce a great deviation, will then have a difference of energy, as 

 to its attractive and repulsive influences, because of the attracted 

 pole of the needle being much nearer to the magnet than the 

 pole which is repelled. Neither will the mean action of these 

 two forces be the same as that of the force belonging to the dis- 

 tance as measured from the focus of attraction to the centre of 

 the compass. As, for example : 



