126 Observations and Experiments on Light. 



and a deflected ray which have diverged from the same point, 

 and is always parallel to and passes through the unmodified beam 

 of light. When the difracting edge is a straight line, the plane 

 of difraction is always perpendicular to a plane passing through 

 the difracting edge and the corresponding outline of its shadow. 

 In an irregular or curved difracting edge the same law will hold 

 with regard to any indefinitely small portions of it, which may 

 be assumed as straight lines. 



I am aware that the terms inflection and difraction are used as 

 synonymous by many who have written upon the subject of light. 

 But without the definitions and limitations, which I have just 

 indicated, I should be compelled to resort to circumlocutions, 

 which might render ambiguous the explanations which I am 

 about to give of the phenomena of the feather. Again, lam not 

 aware that the law which regulates the position of the plane of 

 difraction has been stated by any other writer, although it is fairly 

 inferrible from the facts which they have brought forward, as 

 well as from experiments performed by myself, and which I hope 

 to notice more fully in a subsequent communication. It will be 

 seen in the sequel, that the law which regulates the position of 

 the plane of difraction determines the angle, which the two rows 

 of colored spectra make with each other. 



Let us now turn our attention to the lattice-work formed by 

 the crossing of the barbules of the feather, and inquire how the 

 light passing through a single opening would be affected. The 

 openings of the lattice are of course one of the four varieties of 

 the parallelogram. The angles of these openings difler in the 

 feathers of different birds, and in diff"erent feathers of the same 

 bird. Let a 6 c c? represent one of these openings; and let us sup- 

 j pose a beam of light passing through it perpendicular to the 

 plane of the paper. It is evident that each of the sides of 

 the opening will be a difracting edge ; and if we take any 



two opposite sides ah, dc, the inflected rays of one side will be bent 

 in the same direction as the deflected rays of the other, and will 

 be liable to interfere with each other, and produce colored fringes 

 upon a screen placed to receive the difracted light, and these frin- 

 ges would extend on each side of the opening in a line perpen- 

 dicular to the two sides in question. The same will be true of 

 the other two sides ad, be, and thus we should have two rows of 

 colored fringes, whose hues of direction would be perpendicular 



