Magnetized Non-crystalline Substances. 249 



trough, close to the point where it is pressed upon by the 

 body within. Using small glass balls (which, when empty, 

 exhibit no sensible effects of the influence of the magnet), the 

 magnetic conditions of different liquids filling them may be 

 easily tested. Faraday's beautiful experiments on the rela- 

 tive magnetic capacities of solutions of sulphate of iron of dif- 

 ferent strengths, or rather, other experiments to illustrate the 

 same principles, may be performed in an extremely convenient 

 manner, by filling a glass ball of this kind with a solution, 

 hanging it from one end of the arm, and, by a suitable ad- 

 justment of the weight at the other, immersing it below the 

 surface of another solution contained in the trough. I have 

 found that whenever the difference of the strengths of the two 

 solutions was considerable, the ball immersed was attracted 

 or repelled by the external magnet, according as the solution 

 contained in the ball was stronger or weaker than the solu- 

 tion surrounding it. 



On the stability of small inductively magnetized bodies 

 in positions of equilibrium. 

 20. In the paper published in the Mathematical Journal 

 (referred to above), I pointed out that a small ball of either 

 ferromagnetic or diamagnetic substance placed in the neigh- 

 bourhood of a magnet, and not acted upon by any non-mag- 

 netic force, is in equilibrium if it be in a situation where the 

 " resultant force " (that which was denoted by R) is either a 

 maximum or minimum, or " stationary" in value; that a dia- 

 magnetic ball is in stable equilibrium if, and not in stable 

 equilibrium unless, it be situated where the force R is a mini- 

 mum in absolute value; and that "if there be any point 

 external to the magnet, at which the resultant force has a 

 maximum value, it would be a position of stable equilibrium 

 for a small bar of soft iron, and any other position is essen- 

 tially unstable." Shortly after the publication of that paper, 

 I succeeded in proving that the resultant force cannot be an 

 absolute maximum at any point external to a magnet, and 

 consequently that no position of stable equilibrium for a ferro- 

 magnetic ball, perfectly free from all constraint, can exist. I 

 have very recently found that there may be points where the 

 resultant force is an absolute minimum without being zero ; 

 and therefore there may be positions of stable equilibrium for 

 a diamagnetic ball not included in the case of the force va- 

 nishing, noticed in the previous paper. This case however 

 affords the simplest illustration that can be given of that most 

 extraordinary fact, that a solid body may be repelled by a 

 magnet, or magnets, into a position of stable equilibrium. If, 



