Faraday-Maxwell Mechanical Stress. 653 



B 2 : A 2 . Such a conclusion seems hard to reconcile with 

 the simple linear character of the electromagnetic equations 

 of the aether. 



'25. Another aspect of the question is this : By the change 

 of parameter from A to B the total energy of wave-motion 

 within a given region at a given instant is changed in the 

 ratio B 2 : A 2 , while according to our assumption the trans- 

 lational energy of the sether within the same region is 

 changed in the ratio B 4 : A 4 . This implies that, notwith- 

 standing the absence of any change of type in the wave- 

 train considered, there has heen a change in the partition of 

 energy in the train ; the translational energy of the aether 

 having become a greater or a less proportion of the whole 

 energy according as B is greater than or less than A. 



26. If the difficulties enumerated in the last eight para- 

 graphs are to be regarded as insurmountable, the conclusion 

 must be that the Poynting vector is not associated with 

 translational aetherial velocity (in free aether) *, and this 

 conclusion also seems, on other grounds, to be a reasonable 

 one. For a steady electrostatic field and a steady magnetic 

 field are, so far as we know, entirely without interaction, and 

 where fields of these two kinds coexist, the resulting energy 

 is simply the sum of the separate energies of the two fields 

 in question, without product terms such as we should expect 

 to appear if the sether were indeed in motion with velocity 

 proportional to the vector product of the electric and magnetic 

 vectors. 



27. To sum up some of the objections which may be urged 

 against the doctrine that a flow of aether takes place in the 

 direction of the Poynting vector. This doctrine leads to the 

 conclusion that the translational motion of the sether in a 

 progressive wave-train is longitudinal, resembling the motion 

 of air which is transmitting sound-waves, with a general 

 bodily drift in the direction of propagation superposed. In 

 the absence of special assumptions which might conceivably 

 be devised to remove the difficulty, it leads also to the con- 

 clusion that radiation is accompanied by a continuous creation 

 of aether. It assumes the existence of a translational energv- 

 term of which the uncontested expressions for total electro- 

 magnetic energy (say in free sether) give no indication ; this 

 assumed term constituting in fact a violation of the linear 

 character which is an outstanding feature of the electro- 

 magnetic equations of the aether. 



* The arguments above given are of course, mutatis mutandis, equallv 

 valid against the view that the Poynting vector with sign reversed 

 corresponds to translational velocity of the tether. 



