96 



theories, which start from the second point of view. However, if 

 we should wish to describe the electro-magnetic phenomena in 

 certain cases by the motion of ions or electrons, which has many 

 and well-known advantages, - - the first standpoint is certainly 

 more convenient. These motions then, and the moving ions or 

 electrons themselves can be looked upon as reflected in a plane, 

 and it might be imagined that electric charges are attributed to 

 the "reflected ions" or electrons with the same algebraic sign as 

 they have in the original electro-magnetic field. In this way a 

 description of the phenomena in the "mirror-image" will be possible 

 just in the same way as if we were dealing with the original field; 

 and the mirror-image is thus in truth a "possible" electro-magnetic 

 system, fulfilling the above mentioned condition of the preservation 

 of the general relations between the electric and magnetic parameters. 



From this it will now be clear that the symmetries attributed 

 to physical phenomena are really relative symmetries, determined 

 by the general relations between the different natural phenomena 

 themselves, and by the particular choice of the symmetry primarily 

 given to a certain phenomenon which is considered as the starting- 

 point for the definition of the others related to it. 



12. The symmetry of the "image" /in any point P of a phy- 

 sical system determines the maximum symmetry compatible with the 

 occurrence of the phenomenon considered in P. The phenomenon, 

 namely, can occur in a medium, if its symmetry be the same, or if 

 it be that of a sub-group of the symmetry characteristic for the 

 phenomenon in question. 



If we have a crystal of turmaline whose symmetry with respect 

 to the cohesion-phenomena (which are closely related to its internal 

 structure), is that of the group C V 3 , and if this crystal be heated 

 uniformily to a certain temperature, the symmetry of the crystal 

 is of course by this scalar change altered in no respect; it remains, 

 as before, C V 3 . But C V 3 is a sub-group of C ; and therefore the 

 possibility exists that a dielectric polarisation, the symmetry of which 

 is precisely C , will occur in the heated crystal. Nevertheless nothing 

 has yet been said about the true magnitude of the expected pheno- 

 menon, nor about the real necessity of its occurrence. It is possible 

 that the effect is for instance so extremely small, that it cannot be 

 tested by any experimental method now available 1 ). 



') It is a curious fact, for instance, that the theory of Stokes on the con- 

 ductivity of heat in certain crystals, as scheelite, etc., could not be verified 



