Royal Society. 389 



observed when we compare the abnormal paramagnetic bar with the 

 abnormal diamagnetic one. The former, in the magnetic field, sets 

 equatorially, the latter axially. The former sets parallel to an elec- 

 tric current, the latter perpendicular to the same. If the deflection 

 of the former be from right to left, the deflection of the latter, under 

 like conditions, is from left to right. Finally, the position of equi- 

 librium of the former is always at right angles to that of the latter. 



But if the deportment of the normal paramagnetic bar be com- 

 pared with that of the abnormal diamagnetic one, it will be found 

 that they are in all cases identical ; and the same identity of deport- 

 ment is exhibited when the abnormal paramagnetic bar is compared 

 with the normal diamagnetic one. The necessity of paying atten- 

 tion to structure in experiments of this nature, could not, it is ima- 

 gined, be more strikingly exhibited. 



It is proved by these experiments that the simple substitution of 

 an attractive force for a repulsive one would completely convert the 

 phenomena exhibited by paramagnetic bodies into those exhibited 

 by diamagnetic ones ; that if that which Gauss has railed the ideal 

 distribution of magnetism in an iron bar be reversed, we have a 

 distribution which would produce the phenomena of a bismuth bar 

 of the same dimensions. All the phenomena of diamagnetic bodies 

 become equally intelligible with those of paramagnetic ones, when 

 we assume that the former class possess a polarity the same in kind, 

 but the opposite in direction, to that of the latter. 



It is well known that a bar of iron surrounded by a helix in which 

 a voltaic current circulates is converted into a magnet, and exhibits 

 that twoncss of action, — those phenomena of attraction and repulsion 

 at its two ends — to which we give the name of polarity. The 

 present paper contains an account of experiments made with the 

 view of ascertaining whether similar phenomena were exhibited by 

 a bar of pure bismuth. A cylinder of the latter substance, 6^ inches 

 long and 0'4 of an inch in diameter, was suspended by a suitable 

 device within a helix of covered copper wire, so that it could vibrate 

 freely from side to side. The ends of two electro-magnetic cores 

 were brought to bear upon the two ends of the bismuth bar, and it 

 was so arranged that the two magnetic poles acting upon the bar 

 might be of the same or of opposite qualities. The helix being first 

 excited by a strong current, a current of considerably less power was 

 sent round the electro- magnetic cores, and their action upon the 

 bismuth bar was observed : by means of a gyrotrope the current in 

 the helix was reversed, and the effect of the reversion noted ; per- 

 mitting the current through the helix to flow in its last direction, 

 by means of a second gyrotrope, the current which excited the cores 

 was reversed ; and finally, while the last magnetic polarity remained 

 unaltered, the direction of the current in the helix was once more 

 changed, and the consequent deportment of the bismuth bar was noted. 



In making these experiments and exercising some judgment in 

 the choice of the relation between the strength of the magnets and 

 the strength of the current in the helix, the most complete mastery 

 was attained over the motions of the bar. With one disposition of 



