plCcker on the theory of diamagnetism. 343 



opposite faces are laid successively upon the pole of the electro- 

 magnet. When the vertical faces are perpendicular to the axis 

 the action must be a minimum, and when the vertical faces are 

 parallel to the axis the action must be a maximum. I have 

 proved, by exact measurements, that this does not take place ; or 

 rather, that the difference of attraction is so small that it cannot 

 be the cause of the change of position on the part of the crystal. 

 (I am now convinced that this assertion is incorrect. It was 

 founded on experiments which were not made under altogether 

 favourable circumstances. The substance experimented with 

 was a cube of the protoxide of iron, which stood immediately 

 upon the approximated semi-armatures. Such experiments 

 have been conducted by Tyndall with great exactitude, and he 

 has corrected my assertion.) 



10. If we suppose the prism of tourmaline to be divided longi- 

 tudinally into two equal portions, then at a suitable distance 

 of the poles the centre of gravity of each half recedes from the 

 poles. The same takes place with the entire tourmaline when 

 it is so suspended that it can move freely in a horizontal plane 

 round a point which lies in the production of its axis. 



11. These observations, based upon manifold experiments, have 

 caused me to reject the above idea (8), or at least to regard it 

 entirely as an indefinite one, which first needs a closer explanation, 

 and thus finally I have arrived at the conviction that the problem 

 before us is to explain, by the general laws of mechanics, how it 

 is possible, in a freely suspended crystal, a direction can be re- 

 pelled by the magnetic action of the pole of a magnet, and on 

 the other side a direction attracted by the diamagnetic action of 

 the same pole ; that, in fine, the question is to shoio by calcula- 

 tion and to confirm by experiment, that under certain conditions a 

 piece of soft iron is repelled by the poles of a magnet and a piece 

 of uncrystallized bismuth attracted"^. 



* The following is literally copied from the original memoir : — 

 " In order to obtain a clear conception of our view that a crystallized body is 

 more ready to assume magnetic polarity in one direction than in another, 

 the familiar thought presents itself of making use of thin bars of iron, which 

 easily assume magnetic polarity at their ends, but with difficulty at their sides, 

 and to unite these bars in a maimer conformable with the above view and 

 expose them to the action of a magnet. I will here describe a few such ex- 



