#F 



" Experiments and Observations on ike Properties of Light." 5 

 The simple facts affirmed in Prop. III. Exp. 1 and 3, when di- 

 vested of all theoretical language, -pj 4_ 



appear to be,, that if three edges, 



E, F, G, be placed at successive di- 

 stances from the origin in the order 



of the letters, E and G being on ^ 



one side of the ray and F on the 



other ; then if E and F alone give 



fringes as at o (fig. 4), and G be 



then made to act upon them, or if 



F and G alone give fringes and E 



be made to act upon them, in either 



case the fringes will be shifted to p 



towards the side on which E and G G 



lie, and become broader; and the 



conclusion which the author chiefly 



insists upon is that all three edges 



act in producing the ultimate result : 



the same thing being further con- — 



firmed by exp. 3, in which a curved llllllil ll ll IMIjI 



form given to the edge E, is still ex- P " P' 



hibited in the form of the fringes after the action of F and G. 



That all three edges should be in some degree effective in pro- 

 ducing the ultimate character of the fringes, would, on a general 

 view, be obviously consistent with the wave theory; since, on 

 that theory, a new set of waves originates at each edge, all of 

 which conspire to produce the ultimate result ; though antecedent 

 to exact calculation, it would be impossible to say what would be 

 the precise action of each. 



On repeating the experiment, however, in regard to the parti- 

 cular appearances described by the author, I have found consider- 

 able difficulty : consistently with the conditions before remarked, 

 if the edges £ and F form a narrow aperture so as to give a white 

 centre, and within such limits of distance along the ray as to 

 produce fringes on each side (as in fig. 2), then if G be also 

 within the same limits, and be advanced so near to the ray late- 

 rally as to make a still narrower aperture, the fringes on each side 

 will expand further into tlie shadow. If the edges be beijoyid 

 those limits (which seems to be implied in the author's description, 

 since he speaks of only one set of fringes), then E and F will give 

 a white centre with fringes on the side towards E, as at o; and 

 when G is introduced it will narrow the aperture and give new 

 fringes on the side towards F, at //, that is, just the opposite rvaij 

 to that which the author describes. In repeating the experiment 

 a great number of times at very diffiirent distances, and under 

 varied conditions, I have never been able to obtain any other re- 



