ON LIGHT. 333 



ally verified by M. Fresnel (to whom its suggestion is 

 due), and more recently by M. Billet himself, who by 

 merely interposing (concentrically) between the luminous 

 point and the centre of the screen, a small opake an- 

 nulus exactly corresponding to the calculated dimensions 

 (for red rays and using red light) of the first even ring (B) 

 obtained an illumination at P estimated at five times 

 that when no obstacle was interposed. 



(115.) By way of showing the kind of explanation 

 these principles afford of some of the simplest and 

 easiest cases of diffraction (for their calculation is for the 

 most part very complicated in its details, though simple 

 enough in its principles) ; let us suppose first the case of 

 a screen illuminated by a minute radiant point o through 

 a small circular aperture, and consider only the illumina- 

 tion of the central point of projection on the screen, or 

 of P in our figure. Suppose P to approach the screen 

 from a very great distance so great that the difference of 

 its distance from the centre and either edge of the aperture 

 shall be less than a semi-undulation of the light con- 

 sidered (say loojoooth of an inch). Then the undula- 

 tions from every part of the aperture will reach P in 

 phases more or less accordant with each other, and P 

 will therefore be more or less illuminated : and, P still 

 approaching, its illumination will increase till it attains 

 such a distance that the difference in question exactly 

 equals a semi-undulation. In this case the portion of 

 the wave transmitted corresponds precisely to the whole 

 of the central circle (A) of our system of wave-zones 

 above discussed, and we have here the greatest possible 



