§ 46. NOMENCLATURE OF COLOURS. 65 



proportions of the colours of the rainbow — supposing the 

 whole to form 100 — red 11, orange 8, yellow 14, green 17, 

 blue 17, indigo (or purple) 11, violet 22. This last division 

 I shall consider in noticing the secondary colours. 



A. The first division of the three primaries is the most 

 simple and intelligible, though still it is necessary to deter- 

 mine exactly what are blue, red, and yellow, since each colour 

 is composed of different hues and tones ; and the mere name 

 of a colour is otherwise indefinite. 



By blue should properly be understood (as by the other 

 two) that colour which appears in the prism, when light is 

 decomposed by it ; but it is necessary to describe these and 

 other colours, as the names of most of them are very con- 

 ventional. 



Blue may be considered equivalent to that of the deepest 

 coloured sky (or to lapis lazuli, or a French blue) ; not what 

 ive call sky blue, but the colour of the sky in those southern 

 climates where the atmosphere is clear, and where it appears 

 to the eye an intense bright blue. What it is in an ex- 

 ceptional case, when the atmosphere is foggy, it is unimportant 

 to consider ; nor is it necessary to examine the question of 

 the sky being really white ; nor even to inquire into the reason 

 of its blueness from the reflection of the blue ray, which takes 

 place so readily in meeting with a medium of a different 

 density ; nor why some shadows are blue instead of black. 

 These have no bearing on arrangement of colours, nor even 

 on their nomenclature ; and the decomposition of light, and 

 various optical phenomena, interesting as they are, have no 

 connexion with the question now before us. 



It is not always easy to determine what the exact tone or 

 even hue is, when we mention some colours ; and it is there- 

 fore necessary to agree as to our meaning when speaking of 

 any one. a. The blue of the sky, then, is the one to which 

 the name blue most properly applies. It was evidently that 



F 



