PHYSICAL PROPERTIES OF LIGHT 487 



the following appearance. Visibility usually begins at about 8000-7800 A.U., the 

 rays of longest wavelength being red. As the wavelength becomes shorter the colour 

 gradually changes to orange, the. transition being at 6500 A.U. nearly. From orange 

 the colour changes to yellow, at 6000 A.U. nearly. From yellow to green at 5500 A.U., 

 to blue-green at 5000 A.U., to blue at 4500 A.U., and to violet at 4000 A.U. The 

 violet extends to 3800 A.U., where visibility ceases. The spectrum exhibits therefore 

 a gradual change of colour with wavelength. Above the red is the invisible region 

 occupied by the infra-red or heat rays, and below the violet the invisible ultra-violet 

 or actinic rays, as explained above. 



The colours of the spectrum have important properties which form the foundation 

 of the science of colour mixture. If the spectrum produced from white light is caused 

 to fold up again, it is found that white light is reformed. But white light is produced 

 if certain pairs of colours only are caused to combine in the correct proportion. Thus 

 red (6562 A.U.) and blue-green (4921 A.U.) when mixed correctly form white light, 

 so also do yellow (5636 A.U.) and violet (4330 A.U.).' Such pairs are called comple- 

 mentary colours. But since there is in the spectrum a gradual transition from one 

 colour to the next, so there are between red and yellow an infinite number of rays 

 of different wavelength, each of which has its complementary colour, between blue- 

 green and violet. If therefore from white light we remove one of a pair of complemen- 

 tary colours the other member of the pair will be left unneutralised, and the light there- 

 fore becomes tinted with its colour. Green rays do not possess a complementary in 

 the spectrum ; but it is found by experiment that, by combining red and violet to form 

 purple, the required colour may be produced. If we include purple with the spectral 

 colours, we can imagine these colours to form a closed ring. Each colour will then have 

 its complementary opposite to it. 



THE SPECTRUM COLOURS have another important property, for if red and 

 yellow are caused to combine, they are found to produce orange, the intermediate colour. 

 If red and green are mixed, then again the intermediate colour, yellow, may be produced. 

 It is found that by varying the intensities of the two components, it is possible to produce 

 orange, or yellow-green, or in fact any other intermediate colour at will. Careful 

 experiment shows that the intermediate colour thus formed is no mere approximation 

 but an exact match. If red and green are thus able to combine to form the intermediate 

 colour, while red and blue-green are complementaries producing white by their mixture, 

 the question arises as to the effect produced by mixing red with a colour intermediate 

 between green and blue-green. Experiment shows that a range of colours is produced 

 containing an amount of white light, which varies with the intensities and wavelengths 

 of the combining colours. Colours diluted with white light are spoken of as unsaturated. 

 In order that the colours produced by a mixture of red and green rays shall be fully 

 saturated, and thus be able to match the colours of the spectrum exactly, the green 

 must not be shorter in wavelength than 5400 A.U. Similar phenomena are to be found 

 at the other end of the spectrum ; green and violet, when mixed in various proportions, 

 form colours which match the intermediate spectral colours. With red, green and violet, 

 it is therefore possible to match the whole spectrum. But since red and violet, when 

 mixed, form the intermediate purples, with the three coloured rays it is possible to 

 imitate the whole range of colours. Now the purple formed from red and violet is, as we 

 have seen, the complementary colour to green ; by means of these three colours it is 

 thus possible to produce white light. It should therefore be possible to match 

 an unsaturated colour as easily as a saturated one. Experiment shows that such is 

 the case. The third property beside colour and saturation, is intensity, which depends 

 on the amplitude of the waves. The intensity of the mixture formed by red, green 

 and violet, can therefore be readily adjusted by varying the intensity of each of the 

 three component rays. We may summarise the above facts by stating that by varying 

 the intensities of the red, green and violet rays it is possible to match every shade and 

 colour. This statement has been put to the test by Maxwell, Abney and other observers, 

 and has been found to hold good in all cases but one, spectral blue being slightly more 



