346 0. Bar us — Inversion in the Wave Mechanism 



tion of two waves might be an uncomfortable consideration. 

 In fact only if we accompany the electric current with a 

 counter-clockwise magnetic whirl, and the magnetic flux with 

 a clockwise electric whirl— only by reversing the relations of 

 electric and magnetic field, which everybody is in the habit of 

 associating with electromagnetic phenomena, can we make the 

 wayward wave run right. 



3. To arrive at an insight into the conditions here presented 

 one may note that if the Y forces were constant, there would 

 be no magnetic field in the dielectric due to the occurrence of 

 the electric excitation. Since the magnetic forces a, j3, y, are 

 in this case partial differential coefficients of a potential func- 

 tion, there may be a magnetic field due to extraneous causes. 

 This however has nothing to do with the immediate problem. 

 In the diagram Y, Y', . . . would alone occur unaccompanied 

 by magnetic whirls. In the second place if b Y/bt is constant 

 a magnetic field must be evoked in the dielectric ; but this 

 field, apart from the manner of its origin, will during the 

 period of constant bY/bt be invariable in intensity and posi- 

 tion. It will be a stationary magnetic field, the lines of which 

 are fixed both as to their geometry and their number. As such 

 it is entirely unable to evoke displacement currents. In the 

 above diagram this would mean that, although the forces Y, 

 Y, . . . the whirls a, a, . . . and in consequence the forces 

 7, y' exist, the electric whirls e, e . . . and c, c . * .do not 

 exist. 



Finally let b 2 Y/bf have a significant value, positive or nega- 

 tive. The values b Y/bt will then be variable. Displacement 

 currents will therefore be evoked in the dielectric, and the 

 electric whirls c, c . . . and e, e . . . , the resultant electric 

 forces Y ± , Y\, . . . and Y 2 , Y\, . . . , the magnetic whirls 

 b, b . . . and d, d, . . . etc., will in general all have to be 

 entered in the diagram. One may conceive the disturbance to 

 be transferred from PP in both directions indefinitely. It 

 need not however be a harmonic disturbance. 



If however the time variation of Y, Y\ ... is simple har- 

 monic, then 6 2 Y/bf, etc., are essentially negative and nega- 

 tively increasing of the Y's themselves are essentially positive 

 and increasing ; and vice versa. Hence 7 is essentially decreas- 

 ing if Y is essentially increasing in the lapse of time. The 

 magnetic polarization is therefore falling off. The effect of an 

 increasing Y is thus virtually to superpose on any existing 

 magnetic field, new forces 7 and 7' which have directions 

 respectively opposite to those given in the diagram. These 

 new forces 7 and 7' then constitute a counter-clockwise whirl 

 of magnetic forces a, a ... , around the electric forces Y, 

 Y\ ... j thus wholly reversing the directions of the parts 

 a, a . . . , y and y' of the diagram. 



