134 Prof. Tyndall on the relation of Diamagnetic Polarity 



have illustrated by thirty-seven cases in the Bakerian Lecture 

 for 1855. The effects, it will be remembered, consisted of the 

 turning of elongated paramagnetic bodies suspended between 

 jjoiivted poles from the axial to the equatorial position, and of 

 elongated diamagnetic bodies, from the equatorial to the axial 

 position, when the distance between the suspended body and 

 the influencing poles was augmented. I know this to be a 

 subject of considerable difficulty to many, and I therefore 

 claim the indulgence of those who have paid more than ordinary 

 attention to it, if in my explanation I should appear to presume 

 too far on the reader's want of acquaintance with the question. 

 Let us then suppose an elongated crystal of tourmaline, stauro- 

 lite, ferrocyanide of potassium, or beryl, to be suspended be- 

 tween the conical poles N, S, fig. 9, of an electro-magnet ; sup- 

 posing the position between the poles to be the oblique one 

 shown in the figure, let us inquire what are the forces acting 



Fig. 9. 



upon the ciystal in this position. In the case of all paramag- 

 netic crystals which exhibit the phsenomenon of rotation, it will 

 be borne in mind that the line of most intense magnetization is 

 at right angles to the length of the crystal. Let sn be any 

 transverse line near the end of the crystal ; fixing our attention for 

 the present on the action of the pole N, we find that a friendly 

 pole is excited at s and a hostile pole at n : let us suppose s 

 and n to be the points of application of the polar force, and, for 

 the sake of simplicity, let us assume the distances from the 

 point of the pole N to s and from s to n, to be equal to one 

 another. We will further suppose the action of the pole to 

 be that of a magnetic point, to which, in reality, it approxi- 

 mates ; then, inasmuch as the quantities of north and south mag- 

 netism are equal, we have simply to apply the law of inverse 

 squares to find the difference between the two forces. Calling 

 that acting on s unity, that acting on n will be \. Opposed to 

 this difference of the absolute forces is the difference of their 

 moments of rotation ; the force acting on n is applied at a 

 greater distance from the axis of rotation, but it is manifest 

 that to counterbalance the advantage enjoyed by s, on account 



