96 Mr. Cooper on the Colours that enter into the 



would no doubt be the case with a single ray of each, but 

 with a breadth of light, such as would be required to make 

 the experiment, the divergence would, in all probability, be 

 too small to effect the complete separation, except at a con- 

 siderable distance, the greater refrangibility of the violet 

 light being compensated by the greater refracting angle of 

 the green. Hence, the blue formed by the junction of the 

 two rays, cannot be separated by methods which succeed 

 when the rays intersect each other at a greater angle. 



If we now refer to the violet ray after six reflexions, v, we 

 may observe that it is parallel to the ray I v ; and, there- 

 fore, as the space between v and 1 v, if the figure were 

 correctly drawn, would be less than the breadth of light 

 necessary to make the experiment, the two images must be 

 partly superposed, and, being formed of parallel rays, must 

 continue so at all distances : the same will hold good of 

 images formed after twelve reflexions, or any other multiple 

 of six ; and the whole will emerge at the same angle, one 

 above the other, and form one lengthened image only. 



Hence the extension of the spectrum in strong light, a 

 circumstance not accounted for in the Newtonian theory; 

 and hence, probably, the chemical rays, when the reflexions 

 are too weak to be visible. 



By tracing the reflexions of the red ray we shall find 

 every circumstance attending it giving support to the same 

 views. 



The white light E F, (Fig. 5) is dispersed by the refrac- 

 tion of the surface A C, into violet v, green g, and red 1 r ; 

 the green ray being, as in the former figure, parallel to the 

 base of the prism A B : the red ray, after three reflexions, 

 emerges at 4 r, at a greater angle of refraction than the 

 direct ray 1 r, and, supposing the angle of dispersion formed 

 by the violet and the red rays with the green ray equal, 

 parallel to the violet ray ; consequently, and for the reasons 

 assigned with regard to violet light, the ray, after three 

 reflexions, will form an image in the spectrum above the 

 red and below the violet images ; it must, therefore, either 

 be a distinct image, or be blended with the green ; and, if 

 with the green, with the lower part of it ; for, although, to 

 render the subject more simple, we have assumed the dis- 

 persion of the violet and red rays, from the mean ray, equal, 



