84 Mr. R. Moon on the Undulatory Theory of Interference. 



form of the first and third must be more rapid than in any two 

 immediately consecutive waves, it follows that the divergence 

 from the geometrical shadow must be greater in the former 

 case than in the latter. Other lines of interference occurring 

 in the same manner, we are led to the conclusion that light of 

 one colour, as it actually presents itself to our observation, is 

 not, in fact, homogeneous, but consists of the same constantly 

 recurring series of waves of the same length, but otherwise so 

 far differing from each other that each individual of the series 

 changes its form after diffraction less rapidly than its imme- 

 diate predecessor*. 



The explanation upon the above principles of the fringes 

 within the shadow of a narrow body illuminated from a single 

 point is so obvious as to require no particular comment. 



Recurring to our original example, it remains to say a few 

 words on the effect produced by intercepting the diffracted 

 rays by a transparent plate. It is well known that the velo- 

 city of the wave is retarded in traversing the transparent me- 

 dium, and it seems very natural to suppose that the same 

 thing should obtain with reference to the lateral extension of 

 the wave after diffraction. Now if the lateral extension were 

 retarded, the relative change of form of the consecutive waves 

 would likewise be retarded ; and by necessary consequence 

 the divergence of the line of interference from the geometrical 

 shadow would be diminished, or in other words, "the fringes 

 will move within the shadow." Also, as the relative change 

 of form takes place more slowly in the fixed medium, it is clear 

 that the interposition of the transparent plate must produce the 

 same effect as if the screen op which the diffracted light is re- 

 ceived were removed to a distance from the diffracting body ; 

 that is, the fringes would become broader and more faint, and 

 gradually overlapping each other would ultimately disappear 

 altogether. 



The principles above laid down, if correct, may be applied 

 to explain all the ordinary phaenomena of interference. We 

 shall merely observe, in conclusion, that the resolution of light 

 of one colour into waves of different diffrangibility, which we 

 have thus endeavoured to establish, may be subjected to the 

 same test as the Newtonian resolution of the solar ray into 

 rays of various colours. To ascertain whether diffracted rays 

 are further diffrangible, would require experiments of great 

 delicacy, but may not be impossible. 



* If the velocity were supposed to increase instead of diminishing after 

 the diffraction, each succeeding wave must change its form more rapidly 

 than the one preceding it. 



