354 PLtJCKER ON THE THEORY OF DIAMAGNETISM. 



lies between both. The curve of the indifferent points will be 

 in no way affected, whether we assume //, as constant or as vari- 

 able; we can always compare the results of calculation with 

 those of experiment. In the assumption that fi is variable, 

 certain modifications of the developments in 23 are necessary, 

 and the values of the limited integrals <J> and ^ do not remain 

 the same. 



27. According to the foregoing, the deportment of a prism of 

 tourmaline suspended between the poles is no longer a mechanical 

 paradox. It is only necessary to assume that its ultimate par- 

 ticles become magnetic under the action of the magnet, but so 

 that the straight line which unites the two poles of every par- 

 ticle is perpendicular to the optic axis. To imitate the tour- 

 maline, it is only necessary to construct a prism of thick brass 

 wire, or any other feebly magnetic or non-magnetic substance, 

 and through the prism, at right angles to its axis, to introduce 

 bits of iron wire in all possible directions. Thin iron plates, 

 perpendicular to the axis, and whose centres are in the axis, 

 might be substituted for the wires. 



28. To explain the phaenomena exhibited by crystals of oxide 

 of tin, described above, it is only necessary to assume that in this 

 case the permanent polarity, or that excited by induction, always 

 takes place in the direction of the optic axis. To imitate these 

 phaenomena, we can introduce through a mass, which is but 

 feebly acted upon by the magnet, a number of iron wires having 

 a common direction ; this direction will represent the optic axis 

 of the crystal. 



29. Finally, if instead of such crystals as we have regarded 

 in the two last paragraphs, we take those whose substance is 

 diamagnetic, the phaenomena are similarly explained, when we 

 assume that the diamagnetic induction is opposed to the mag- 

 netic. We thus make the assumption, that in calc-spar the 

 directions in which the polarity takes place are all at right angles 

 to the optic axis, while in the case of ice these directions all 

 coincide with the optic axis. 



30. According to the theory of Ampt^re, the hypothesis that 

 we have made in 27 to 29, reduces itself to the assumption, 

 that in those crystals which do not belong to the regular system 



