46 DISPERSION. 



(56) The dispersive power of a substance is measured, 

 not by the absolute dispersion, which varies in general with 



the angle of refraction, but by the ratio which that quantity 



g _ g 

 bears to the total deviation, or by -. But, in the case of 



Ci 



a ray which passes nearly perpendicularly through a thin 

 prism, this ratio is constant ; for, dividing the third of the 

 equations of the preceding article by the first, 



The dispersive power, therefore, is measured by the difference 

 of the refractive indices of the extreme red and violet rays, 

 divided by the refractive index of the former minus unity. 



(57) Newton supposed that the dispersive powers of all 

 substances were the same. He was led to this erroneous con- 

 clusion by observing that when a prism of glass was inclosed 

 in a prism of water with a variable angle, their refracting an- 

 gles being turned in opposite directions, the emergent ray was 

 coloured when it was inclined to its original direction ; while, 

 on the other hand, it was colourless whenever, by varying the 

 angle of the water prism, the refractions of the two prisms 

 were made to compensate each other, or the ray to emerge 

 parallel to the incident ray. Hence he concluded that the 

 dispersion was always proportional to the total deviation ; 

 and that refraction could never take place without a separa- 

 tion of the refracted ray into its coloured elements. 



When Newton's experiment with the two prisms was re- 

 peated a long time after, by Dollond, he found that the 

 results were exactly the opposite to those stated by New- 

 ton ; that, in fact, the emergent ray was colour -ed, when the 

 deviation was nothing, or the ray parallel to its original di- 

 rection ; and that, on the other hand, when the dispersions 



