of the Electro-magnetic Theory of Light. 347 



The same result may be reached somewhat differently : con- 

 stant b Y/bt is equivalent to a fixed magnetic polarization in 

 circuit a, the value of which may be either large or small but 

 which is clockwise around Y. Retarded variation of bY/bt, 

 such for instance as occurs in simple harmonic motion, is thus 

 equivalent to a waning of the clockwise magnetic polarization ; 

 or from another point of view, it is equivalent to an increase 

 of counter-clockwise magnetic polarization around Y, hence to 

 a series of counter-clockwise whirls of magnetic force around 

 Y, Y . . . yielding 7 and 7' reversed as resultants. 



4. Turning now to the values of 7 and 7' which by the 

 present inferences are opposite in direction to those given in 

 the diagram, and are increasing in value in this reversed direc- 

 tion at the retarded rate demanded by the simple harmonic 

 equation, it will be expedient to carry forward the argument 

 in the step for step fashion of the preceding paragraph. If 7 

 were constant it would be unaccompanied by an electrical 

 effect, if we except the case that X r Y, Z, may in such a case 

 have a potential due to extraneous causes of no relevant interest. 

 If 7 is variable but by/bt constant, the result would be an 

 electromotive force curled counter-clockwise around 7. There 

 would be a curled electric polarization of the dielectric but no 

 displacement current around the axis of magnetic force. 

 Finally if b 2 y/bf is significant displacement currents must 

 occur and be maintained. 



But for the case of an inherent simple harmonic law origi- 

 nating in Y and transferred to 7, the time variation of 7 is 

 always retarded from an absolute point of view. Hence the 

 counter-clockwise electric polarization around 7, reversed, is fall- 

 ing off ; in other words, clockwise electric polarization around 7 

 reversed is increasing. Virtually therefore a clockwise dis- 

 placement current is running around 7 reversed, i. e. a counter- 

 clockwise current must encircle the 7 as drawn in the diagram. 

 Hence both the whirls c, c . . . and 0, e . . . , are correctly 

 drawn in the diagram for positive increasing Y, Y' . . . and 

 the new electric excitations (0, e) tend both to weaken Y, and 

 to evoke the new wave front Y x and Y 2 as shown. 



These forces Y l and Y n are then to be treated each in its 

 turn in the same way as Y, reproducing the same results indefi- 

 nitely. Thus the following paradoxical rule relative to the 

 construction of the mechanism of a simple plane electric wave 

 would seem to follow : Provide the original electric flux (J") 

 with a counter-clockwise magnetic curl (a) ; provide the result- 

 ing magnetic fluxes (7, 7') each with a clockwise electric curl 

 (c and e) ; and so forth indefinitely. For such a wave the elec- 

 tric and magnetic forces and the advance of the wave will be 

 in correct directional relations (§ 1, end). The wave would 



