336 ON LIGHT. 



than half of the second (discordant), still less of the third, 

 &c., so that on the whole there will be a preponderance 

 of the concordant undulations so introduced, and p will 

 be more strongly illuminated than before. When re- 

 moved one step farther however, since the newly intro- 

 duced half of (B) the second zone almost exactly coun- 

 teracts that of (A), the effect of the change will have its 

 character decided by the proportional magnitudes of the 

 segments of (c), (D), &c., disclosed, among which the 

 preponderance is evidently in favour of (c), that is, of 

 discordant undulation, so that by this removal of the 

 shading obstacle the illumination of P will be diminished; 

 and so on alternately. Now at each stage of these re- 

 movals of the shading body, the edge of the geometrical 

 shadow retreats farther and farther from p, or (which is 

 the same thing) p is successively farther and farther out- 

 side of the edge of the shadow, becoming alternately 

 more and less illuminated than at the actual edge. 

 Here then we see the origin of the bright and dark ex- 

 ternal fringes exhibited in homogeneous light ; and there- 

 fore by the very same reasoning, of the coloured ones 

 produced by the successive overlapping of those formed 

 by several coloured rays to each of which corresponds a 

 different breadth of fringe ; that for the red being broad- 

 est and for the violet narrowest. 



(117.) The twinkling or scintillation of the stars partakes 

 so far of the nature of a phenomenon of diffraction, as 

 that it depends for its origin on the mutual interference 

 of discordant rays arriving at one instant, but by different 

 routes, on the same point of the retina of the eye j and 



