s 



126 Prof. Tyndall on the relation of Diamagnetic Polarity 



cases of real difficulty which have been actually encountered by 

 those who have worked experimentally at the subject. 



To liberate the thoughts from all effects except those which 

 are purely magnecrystallic, we will for the present operate with 

 spheres. Let a sphere of carbonate of lime be suspended before 

 the pole S, fig. 1, of an electro-magnet, so that the axis of the 

 crystal shall be hori- 

 zontal. Let the line Fig. 1. 

 ab mark any position 

 of the axis inclined 

 to the direction of the i\/\ 



force emanating from .^-— L..:^-...-\—- 



S ; and let the dotted z^LlS^ 



line dc make an equal 

 angle with the direc- 

 tion of the force at 



the other side. As the sphere is diamagnetic, the face of it which 

 is turned towards S will be hostile to S, while that turned from 

 S will be friendly to S, according to the principles established in 

 the paper above referred to ; and, if the sphere were homoge- 

 neous, the tendency to set ab at right angles to the direction of 

 the force W'Ould be exactly neutralized by the tendency to set cd 

 in the same position : the sphere would consequently stand still. 

 But the case is otherwise if the intensity of diamagnetization along 

 ab be greater than along cd, which I have elsewhere shown to be 

 the fact*. If we suppose the sphere to vanish, with the excep- 

 tion of two thin needles taken along the lines mentioned, the 

 hostile pole at a will be stronger than that at c, and the friendly 

 pole at b will be stronger than that at d; hence the ends a and b 

 being acted upon by a mechanical couple of superior power, the 

 line ab w"ill recede from its inclined position, and finally set itself 

 at right angles to the direction of the force. Whatever be the 

 inclination of the line ab to the magnetic axis, this superiority 

 will belong to its couple ; it is therefore manifest that the entire 

 sphere will tui'n in the manner here indicated, and finally set 

 with the axis of the crystal equatorial, which is the result esta- 

 blished by experiment. 



For the diamagnetic calcium, contained in this crystal, let the 

 magnetic element, iron, be substituted. Each molecule of the 

 crystal becomes thereby magnetic ; we have carbonate of iron in 

 place of cai'bonate of lime ; and the line which, in the latter sub- 

 stance is that of maximum repulsion, is that of maximum attraction 

 in the former. This, I think, is one of the most suggestive points f 



* Phil. Mag. S. 4. vol. ii. p. 1/6. 



t I'or its bearing upon the question of a magnetic medium see Phil, 

 ilag. vol. i.\. p. 208. 



