to Magnecrystallic Action. 135 



of its greater proximity, the distance x z would require to be 

 four times that of x y. Taking the figure as the correct sketch 

 of the poles and crystal, it is plain that this condition is not 

 fulfilled, and that hence the end of the crystal will be drawn 

 towards the pole N. What we have said of the pole N is equally 

 applicable to the pole S, so that such a crystal suspended be- 

 tween two such poles, in the manner here indicated, will set its 

 length along the line which unites them. 



While the crystal retains the position which it occupied in 

 fig. 9, let the poles be removed further apart, say to ten times 

 their former distance. The ratio of the two forces acting on the 

 two points of application s and n will be now as the square of 

 11 to the square of 10, or as 6 : 5 nearly. Taking fig. 10, as 

 in the former case, to be the exact sketch of the crystal, it is 



Fig. 10. 



manifest that the ratio of xz to x y is greater than that of 

 6 to 5 ; the advantage, on account of greater leverage, possessed 

 by the force acting on n is therefore greater than that which 

 greater proximity gives to s, and the consequence is that the 

 crystal will recede from the pole, and its position of rest between 

 two poles placed at this distance apart will be at right angles to 

 the line which joins them. It is needless for me to go over the 

 reasoning in the case of a diamagnetic body whose line of 

 strongest diamagnetization is perpendicular to its length. Re- 

 versing the direction of the arrows in the last two figures, we 

 should have the graphic representation of the forces acting upon 

 such a body; and a precisely analogous mode of reasoning would 

 lead us to the conclusion, that when the polar points are near the 

 crystal, the latter will be driven towards the equatorial position, 

 while where they are distant, the crystal will be drawn into the 

 axial position. In this way the law of action laid down empiri- 

 cally in the Bakerian Lecture for 1855 is deduced a priori from 

 the polar character of both the magnetic and diamagnetic forces. 

 The most complicated effects of magnecrystallic action are thus 

 reduced to mechanical problems of extreme simplicity; and, 

 inasmuch as these actions are perfectly inexplicable except on 

 the assumption of diamagnetic polarity, they add their evidence 



