THEORY OF COLORS. 



resented in fig. 3. This fan-like form is produced by the fact that some of the 

 < rays which compose the beam are more strongly refracted by the prism than 

 others, and the divergence of the fan depends upon the difference between the 

 extent of the deflection of the most refrangible, and the least refrangible rays. 

 The angle of divergence of the fan has been called the dispersion of the origi- 

 nal beam by the prism. 



When the rays, thus dispersed, in virtue of their different susceptibility of 

 refraction, are received upon a white screen, they exhibit a streak of surface 

 illuminated by a series of different tints of color, which, in their general char- 

 acter, are conformable to the distinction assigned to them by Newton ; but ac- 

 curate examination shows that there are no distinguishable boundaries between 

 the successive tints ; that throughout the limits of the red the degree of red- 

 ness varies, that it insensibly melts away into the beginning of the orange, 

 which, increasing to a point where its intensity is greatest, again gradually 

 melts away insensibly into the yellow, and so on, the successive colors and tints 

 of color fading imperceptibly into each other. Now there is nothing in these 

 circumstances to afford any rigid justification of the seven elementary colors 

 assigned by Newton, and when we consider, what is not disputed by Newton 

 himself, that the commingling or blending together of lights of different colors 

 will produce intermediate tints, it follows that there are an infinite variety of 

 ways in which the constituent colors of light might be imagined to be arranged 

 which would equally produce the phenomenon of the prismatic spectrum. 



This problem has, accordingly, been taken up in our own times by Sir Da- 

 vid Brewster, with all the advantages which the increased knowledge and experi- 

 ence of the age, and improved methods of inquiry, could afford. He has shown, 

 by innumerable experiments on the transmission of light through colored me- 

 dia, and on artificial lights, produced by combustion, of various circumstances, 

 that the pure and elementary simple lights are one or other of the three 

 colors, red, yellow, and blue ; that the light of each of these colors, respect- 

 ively, is composed of constituent rays which are differently refrangible, so that 

 if a beam of any one of these lights were transmitted through a prism, an ob- 

 long spectrum would be produced, of one uniform color, corresponding to that 

 of the light itself. .Thus if we suppose a beam of red light transmitted through 

 a prism in the same manner as the original beam of white light, fig. 3, was 

 transmitted, then we should obtain an oblong spectrum, similar in form and 

 length to that which we originally obtained, but all of one tint. It would be 

 all red, although the redness would be greatest at one particular point, and 

 would decrease from that point toward each extremity, and gradually fade 

 away. These circumstances may be represented by the diagram, fig. 4. 



Let L M represent the screen, and let L represent the lower and M the up- 

 per end of the spectrum ; let N be the point at which the redness is most in- 

 tense, it will gradually diminish from N to M and from N to L. Let us sup- 

 pose that we draw a curved line, L P' P P" M, so that the lines or distances 

 N P', N P, N P", &c., shall, respectively, represent the intensities of the 

 light at the several points N' N 7 ', &c. Such a figure will exhibit, geomet- 

 rically, the gradation of tints from the point N, where the red is brightest, up- 

 ward and downward to the points where it fades away. It is found by ex- 

 periment that the point where it is brightest is near the lower extremity of the 

 spectrum. 



In like manner, if a beam of pure yellow light be transmitted through the 

 prism, a similar yellow spectrum will be produced, which may be represented 

 in -I similar manner, the point of greatest brightness, however, being at a high- 

 er point in the spectrum, represented in figure 5, by similar letters. 



Finally, let us suppose a beam of blue light transmitted through the prism in ( 



