Conductivity of Tourmaline Crystals. 371 
during changes of intensity of the current, work would be done 
against or by the springs. If the current passed in such a 
direction that the majority of the magnets were set so as to 
remain unchanged, there would be less work done by changes of 
intensity than if the current were in the opposite direction, and 
this would give rise to an apparent unilateral conductivity. I 
say “apparent” because the weakening of the current is due to 
an inverse electromotive force, and not to a true increase of the 
resistance. The same effect would be produced by supposing a 
majority of the magnets turned in the same direction along the 
magnet, and kept in position by two springs, one on each side, 
but one stronger than the other, when of course a current would 
have to do more work in turning them to one side than to the 
other, so that in this case also there would be apparent unilateral 
conductivity during variations of the current. 
Now, suppose that the proportion of polarised magnets or their 
strength depended on the temperature of the system. It is then 
evident that during changes of temperature there would be 
changes in the numbers or strengths of the polarised magnets ; 
either would produce an electromotive force in the wire during 
the change. Hence the phenomena of pyro-electricity would be 
manifested by such a system. I put these forward merely as 
illustrations, not supposing that the structure of tourmaline is 
necessarily at all like either of them, but there are generally 
great analogies between different systems exhibiting the same 
phenomena, and an illustration gives us a concrete stepping-stone 
to found our conceptions on during the difficult transit to the 
abstract. 
The passage of a current through an iron wire is accompanied 
by the production of a series of magnetic elements round it, and 
the effect of this has been noticed as causing an apparent change 
of resistance during changes of the current, but as there is no 
want of symmetry in the wire there is no apparent unilateral 
conductivity. 
Scien. Proc., R.D.S. Vou. 1, Pr. v. 2c2 
