including an Examination of the Magnetic Field. 343 



friend ; " either you must refer tlie weakening of a quadrant to 

 magnetic interference, or you must conclude, that that induced 

 state, whatever it be, which causes the bismuth to be repelled by the 

 magnet, causes it to be attracted by the coil, the resultant being 

 the difference of both forces. In the same manner the strengthen- 

 ing of a quadrant is accounted for by the fact, that here the in- 

 duced state which causes the bismuth to be repelled by the 

 magnet Causes it to be repelled by the coil also, the resultant 

 being the sum of both forces. The matter may be stated still 

 more distinctly by reference to Eeich^s experiments*. He 

 found that when a bundle of magnet-bars was brought to bear 

 upon a diamagnetic ball suspended to the end of a torsion 

 balance, when similar poles were presented to the body, there was 

 a very distinct repulsion ; but if one half of the poles were north 

 and the other half south, there was no repulsion. Let us imagine 

 the two halves to be brought to bear upon the ball consecutively; 

 the first half will cause it to recede to a certain distance ; if the 

 second unlike half be now brought near, the ball will approach again, 

 and take up its original position. The question therefore appears 

 to concentrate itself into the following : — Is this ' approach ' 

 due to the fact that the magnetic forces of the two halves annul 

 each other before they reach the ball, or is it the result of a 

 compensation of inductions in the diamagnetic body itself ? If 

 a sphere of soft iron be suspended from a thread, the north pole 

 of a magnet will draw it from the plumb-line ; if the south pole 

 of an exactly equal magnet be brought close to the said north 

 pole, the sphere will recede to the plumb-line. Is this recession 

 due to a compensation of inductions in the sphere itself, or is it 

 not ? If the former, then, by all parity of reasoning, we must 

 assume a similar compensation on the part of the bismuth.^^ 



26. That bismuth, and diamagnetic bodies generally, suffer in- 

 duction, will, I think, appear evident from the following consi- 

 derations. The power of a magnet is practically ascertained by the 

 mechanical effect which it is able to produce upon a body pos- 

 sessing a constant amount of- magnetism, — a hard steel needle, 

 for instance. The action of a magnet in pulling such a needle 

 from the magnetic meridian may be expressed by a weight which 

 acts at the end of a lever of a certain length. By easy practical 

 rules we can ascertain when the pull of one magnet is twice or 

 half the pull of another, and in such a case we should say that 

 the former possesses twice or half the strength of the latter. If^ 

 however, these two magnets, with their powers thus fixed, be 

 brought to bear upon a sphere of soft iron, the attraction of the 

 one will be four times or a quarter that of the other. The 

 strengths of the magnets being, however, in the ratio of 1 : 2, this 

 attraction of 1 : 4 can only be explained by taking into account 

 * Phil. Mag., S. 3. vol. xxxiv. p. 127. 



