Observations and Experiments on Light. 127 



respectively to the parallel sides of the opening, and consequently- 

 crossing each other at angles equal to those of the opening. But 

 a part of the light would pass through the centre of the opening 

 unbent, and would form upon the screen a white image at the in- 

 tersecting point of the two rows of colored fringes. Thus it 

 will be seen, that a beam of light passing through a single open- 

 ing of the kind above described, would be divided into nine parts, 

 four being produced by the inflection of the four sides, four more 

 by the deflection of the same, and one being the remains of the 

 beam that pass on unmodified. Now let us suppose that a beam 

 of light, instead of passing through a single opening, passes 

 through an extremely minute lattice, containing an indefinite 

 number of such openings, as in the case of the feather. As all 

 the bars of the lattice are parallel respectively to those which sur- 

 round each individual opening, it is evident that the general ef- 

 fect upon the beam will be the same as that of a single opening, 

 with this difference, that the range within which interference 

 would take place, would be greatly enlarged, by enabling the in- 

 flected and deflected rays from different openings to interfere with 

 each other ; and thus the fringes, which are scarcely perceptible, 

 when formed by a single opening or a single edge, become bril- 

 liant spectra, when a beam of light is passed through a lattice of 

 the kind described. All this is realized in the experiments with 

 the feather. It is proper to remark, however, that the central 

 white image is probably not formed entirely of unmodified light, 

 but is partly produced by light slightly inflected by the opposite 

 edges of the bars of the lattice, and corresponding with the in- 

 ternal fringes, first explained by Dr. Young upon the principle of 

 interference. It is not improbable, that some of the deflected 

 rays fall within the central white image and add to its brightness. 

 The faint spectra in the angular spaces may be explained by sup- 

 posing that they are formed by light, which has undergone two 

 inflections or two deflections, or one inflection and one deflection, 

 by two contiguous bars of the lattice. 



It should be noticed here, that all the colored spectra, as well as 

 the central white one, are considerably elongated in a direction 

 perpendicular to the barbs of the feather. With a very delicate 

 feather and a small luminous object, the eye can easily distinguish 

 a row of colored spectra arranged in the same direction. This is 

 what might have been expected, and gives us some idea of the ef- 



