254 ON LIGHT. 



duced by different media different for the same mean 

 amount of refraction, but within those lengths the dis- 

 tribution of the several colours differs, the spaces occu- 

 pied by the several tints differing very considerably in 

 proportion to each other and to the whole. Thus, in 

 the spectrum formed by flint-glass, and most other of 

 the highly dispersive media, the green is situated nearer 

 to the red than to the violet end of the spectrum, while 

 in that formed by muriatic ("hydrochloric") acid the 

 reverse is the case. 



(38.) By the reunion of all the coloured prismatic 

 rays (winch may be effected by an equal and contrary 

 refraction of tlie whole spectrum through a prism of the 

 same material reversely placed), white light is repro- 

 duced. And hence we conclude that colour is not a 

 siLperinduced but an inhei'ejit quality of the Iwrwious rays. 

 Again, if we exclude from this reunion any portion of 

 the spectrum, the reconstituted beam is coloured : and if 

 the rays so excluded be not extinguished^ but diverted 

 aside, and themselves collected and reunited into 

 another and separate beam (which may easily be 

 effected, with a little management, by one skilled in ex- 

 perimental optics), this will also be coloured, but with a 

 tint couiph7nentary to that of the first. Between the 

 tints so arising is always found to prevail that beautiful 

 and, so to speak, har7nonious contrast which is so effec- 

 tive in the ornamental arts, where one colour is said to 

 set-off another, or show it to the greatest advantage. 

 Thus, crimson or pink is complementary to green, 

 scarlet or orange to blue, yellow to purple, Uq.. The 



