404 



Transactions of the Society. 



say, the point ri x . That being the position of affairs, assume now 

 that the glow at e is extinguished and that the point 77 begins to 

 glow in its turn, under the following conditions. It is to give out 

 light exactly like that which it received, that is to say, identical 

 in colour, attuned in phase and proportioned in intensity, so that 

 the emitted light would upon the surface of the point 77, and, 

 therefore, at every surface along the optical system, balance and 

 neutralise the light received from e. I will, with your permission, 

 call this supposed beam of light which thus reacts to and cancels 

 the original — I will call it the reverted beam. It is not a reflected 

 beam because the individual rays are supposed to travel back along 

 the same paths by which they came, not along paths inclined to 

 the optical axis at equal and opposite angles, and the several wave- 

 fronts do not carry on the original order of phases, they reverse it, 



Fig. 91. 



and therefore they cancel the arriving wave-fronts at all points of 

 the path and not only at periodic distances of half a wave- 

 length. 



Consider now what will happen when the reverted beam re- 

 enters the aperture. There being no radiation from the whilom 

 object-point e to quench the reverted beam, it will fill the aperture 

 and the region round about it with wave-fronts exactly like the 

 original wave-fronts in intensity, phase, frequency and form, but 

 travelling in the opposite direction. These wave-fronts will, of 

 course, give rise to diffracted wave-fronts at the aperture; and 

 although the diffracted beams of light will now be propagated 

 towards the object plane instead of being propagated away from 

 it, we can estimate their direction and magnitude by the aid of the 

 original observation. For the diffraction in this case must be 

 equal and opposite to the original diffraction. 



