FUNDAMENTAL FORMULATIONS OF ELECTRODYNAMICS. 233 



and it has been shown on a previous occasion* that the only form of specification of 

 the energy distribution which is consistent with our usual ideas on these matters is 

 that which makes the density of the magnetic energy equal to B 2 /8ir, and as the 

 present discussion shows that our only hope of discrimination lies in that direction, we 

 may assume that the evidence in favour of this special form is conclusive, at least for 

 the present ; it is besides the only form in which the most general case is completely 

 representative of the distribution in any volume of the field without requiring the 

 introduction of boundary terms involving surface distributions. 



11. We next turn to a consideration of the expression for the forcive of electro- 

 magnetic origin acting on the polarised media in the field. We have seen that the 

 mechanical forcive on the dielectrically polarised media is such that its .T-component 

 per unit volume at any place is of the form 



or as it first appears in the analysis 



This result is in complete agreement with that derived by LA.RMOH in the electron 

 theory,! but the present derivation indicates clearly the origin of the different terms 

 in it. The expression 



is that corresponding to the expression derived in the statical theory from energy 

 considerations and corresponds to MAXWELL'S magnetic expression ; the second t j rm. 



viz., 



-[P Curl E], 



is'one of the terms arising as a result of the convection of the media, and this is tin- 

 term which is effective in reducing the electric part of the forcive to the form 



(PV) E, 



which is the result derived in the elementary theory by regarding the forcive as the 

 resultant of the forces on the elementary bi-poles. 



* ' Phil. Mag.,' vol. 34 (1917), p. 385. Of. also ' Phil. Mag.,' vol. 32 (1916), p. 1G2. 

 t ' JEther and Matter,' p. 104. 



2 K 2 



