Sir D. Brewster on the Colours of Mixed Plates. 195 



we now incline the incident ray to the plate of nacrite more 

 than 90°, the faint red band next the yellow gradually be- 

 comes brighter, while the other bands become fainter ; and at 

 the boundary of light and darkness all the other bands dis- 

 appear except this red one, which is the comjolementari/ colour 

 to the green, (produced by the union of the blue, green and 

 yelloio bands,) and the colour which is seen upon the dark 

 space next the edge of the flame, as described by Dr. Young. 

 If we, on the other hand, incline the incident ray in an oppo- 

 site direction, so that it forms with the plane of the plate a 

 less angle than 90^j the red band next the blue will now be- 

 come brighter; and at the boundary of light and darkness, 

 when all the other bands have disappeared, the i-ed band will 

 afford the complementary colour to the green. 



As the edge of the plate of nacrite is rough and unpolished, 

 and accurately perpendicular to the parallel faces, there are 

 no reflected nor refracted pencils, whose combinations with 

 one another, or with the direct rays, can be employed to ac- 

 count for the complementary colours. The phaenomena of 

 mixed plates, indeed, are cases of diffraction when the light is 

 obstructed by the edge of very thin transparent plates placed 

 in a medium of different refractive power. If the plate were 

 opake, the fringes would be exactly those which have been so 

 often described, and explained by the principle of interference. 

 But owing to the transparency of the plate, fringes are pro- 

 duced within its shadow ; and owing to the thinness of the 

 plate the light transmitted through it and retarded, interferes 

 with the partial waves which pass through the plate and with 

 those which pass beyond the diffi'acting edge with undimi- 

 nished velocity, and modifies the usual system of fringes in 

 the manner which we have described. 



As the plate of nacrite diminishes in thickness, or as the fluid 

 in which it is immersed approaches to it in refractive density, 

 the central coloured bands, whose union constitutes the direct 

 tint, will diminish in number, and descending gradually in 

 the scale will finally disappear when the retardation produced 

 by the plate does not perceptibly alter the phase of the raj'. 

 When the plate, on the other hand, increases in thickness, or 

 the fluid diminishes in refractive power, the central bands will 

 become closer and more numerous, and will finally resemble 

 the fringes within the shadow of the ordinary system. 



When the plate of nacrite is thicker at one place than an- 

 other by the partial removal of a parallel film, the edge where 

 tlie increase of thickness takes place produces exactly the same 

 pha.'nomena as the edge of the film that is removed, or of the 

 film that is elevated above the general surface, and hence we 

 02 



