THOUGHTS ON THE NATURE AND ORIGIN OF FORCE. 255 



In the peculiar phenomena of magnetism, we have reason to believe 

 that the attraction (as in the cases of gravitation, cohesion, and affinity) 

 is resident in the ultimate molecules, if not in the constituent atoms. 

 However minutely we may subdivide a magnetic bar of steel or nickel, 

 (cobalt or manganese,) we cannot separate the poles — the attractive from 

 the repellant. However small these fissiparous segments, each is itself 

 a perfect magnet, with equal opposite polarity. 



It is evident that the arrangement of the molecules in a magnetic bar, 

 all axially accordant in direction, must be one of constraint, since 

 although a single line of such particles would find a stable equilibrium 

 in that uniform direction of polarity which should satisfy the attractions 

 of all their contacts, yet a second adjacent line, similarly directed, would 

 constantly present to the first line the antagonism of connatural or 

 homogeneous poles. Hence, to secure a general stability of equilibrium, 

 the supposed lines of molecular magnets should be alternately reversed. 

 But such an arrangement would result in a complete neutralization of 

 the attractive and repulsive forces, and the mass or bar of metal thus 

 constituted, possessing no internal differential resultant, would exhibit 

 no traces of integral polarity or magnetism. 



Assuming (as we have reason from various considerations to believe) 

 that in all solids the constituent particles, however firmly locked by 

 their cohesive forces in their fixed relations of distance and position, 

 are yet capable of rotation within their magic spheres, without affecting 

 their solidarity, we can easily' perceive how, under the attractive induc- 

 tion of a i)owerful external magnet, a bar of iron constituted as just 

 suggested would have at once its constituent molecules polarized in 

 direction, and thus become a sensible magnet, while it would as quickly 

 relapse from this condition of restraint on the removal of the controlling 

 agency. The actual behavior of iron is exactly conformable with such 

 an organization. The entire efficiency of the electro-magnet for the 

 purpose of telegraphy depends upon this property of facile molecular 

 rotation in the iron magnet, whereby it instantly becomes spontaneously 

 (Zepolarized on the suspension of the electric current. 



Under the quaquaversal disposition of the molecular axes, thus shown 

 to be the normal condition of equilibrium, it follows that the magnetism 

 of matter, however powerful, should remain forever latent, and all its 

 manifestations in abeyance, did not some potent interference disturb 

 this neutrality. 



If, however, we further assume that in very rare cases, as in that of 

 nickel or of hardened steel, a considerable amount of adhesion (the 

 nature of Vv^hich we are unable to comprehend) exists between the con- 

 stituent particles, preventing their easy rotation, then we can perceive 

 that tlie presentation of a x)owerful external magnet to such a bar 

 would not easily or at once induce a sensible magnetization. If, however, 

 the solicitation of the attracting magnet were sufficiently powerful, and 

 sufficiently continued, to overcome the resistance of this molecular 



