Refracted and Diffracted Light. 349 



fringes are complimentary to each other; but, at first view, 

 if we suppose them to be connected, we are as much at a 

 loss to account for their great distance from each other, 

 which includes the whole field of view, as the early ex- 

 perimentalists in galvanism were to account for the ap- 

 pearance of oxygen and hydrogen, in the proportions which 

 form water, in distant parts of the apparatus. A system of 

 mutual intermediate compensations, satisfactorily accounts 

 for both these appearances : the colours which are deficient 

 in the fringes on one side are not transferred by the prism 

 to the fringes on the opposite side, but to the white light 

 immediately adjoining, which surrenders equal portions of 

 the same colours to form another adjoining surface of white 

 light, until, by a succession of such transfers, the compli- 

 mentary colours appear at the opposite side, where there 

 is no white light from which these colours can be com- 

 pensated. 



When we look at the sky, or any other white object, 

 with the naked eye, we see three images of it, which are 

 superposed so as to form one white image ; and this image 

 appears in the proper position of the object; but when we 

 look at it through a prism, we see the three images in three 

 different positions, neither of which is its true position ; 

 the deviation being less in the red image than in the green, 

 and less in the green than in the violet image. 



If, then, under this new arrangement, we interpose an 

 object, either before or behind the prism, we do not inter- 

 cept the same part of the different images, as we do when 

 these images are formed of parallel rays, but equal portions 

 of different parts of them ; so that, if the deviation of the 

 images from each other be equal to the breadth of the in- 

 tercepting object, we still see the whole of the sky, or 

 whatever other object we are looking at, in the different 

 colours : for the part of the object, not seen in one colour, 

 is seen in the others ; and it is these separated parts of the 

 three images which form the fringes observed in light re- 

 fracted by the prism. 



It clearly appears from these and other experiments, 

 that parallel rays are not essential to the formation of white 

 light; and that, if the colours are properly proportioned, 

 the intersection of the rays may be made at any angle, 

 without impairing its purity. 



