vii EETINAL EXCITATION 375 



the flame of A, will be red, while shadow a, which is illuminated 

 only by Hume B and should therefore appear white, will look 

 green by contrast. 



This experiment may be varied by substituting daylight for 

 one of the candle flames. In this case it is not necessary to use 

 coloured glass, because daylight differs from candlelight in being 

 white, not yellowish. Daylight casts a grey shadow, but on 

 lighting the candle the grey shadow turns yellow, and the other 

 shadow which is due to daylight appears blue by contrast. 



In Ragona-Scina's experiment the effect of simultaneous 

 contrast is obtained by reflection. A black spot is observed 

 through coloured glass, at an angle of 45, and the glass is so 

 arranged that the reflected image of another black spot is seen 

 beside it ; seen separately, the first spot will be the colour of the 

 glass, the second black ; viewed together, the 

 second assumes the complementary colour of a _ i _ 



the first. ~~\ 7~ 



Colour-contrasts are more easily obtained \ / 



with pale than with saturated hues. This *j c 



is readily demonstrated by the following ex- ,' \ 



periment of H. Meyer. If a square of grey 

 paper is placed on a sheet of coloured paper, 

 its tint does not alter perceptibly ; but if / 



the whole is covered with a sheet of semi- A 

 transparent tissue-paper the small square & p] 

 assumes the complementary hue of the back- 



j -j ii_ i j -i. PIG. 178. Experiment on 



ground : it the latter is red, it appears coloured shadows. 

 greenish ; if yellow, bluish ; if blue, yellow- 



ish ; and so on. This striking effect proves that contrast-effects 

 are much more pronounced when the colours are rendered less 

 saturated by the addition of white. 



Other analogous contrast-effects can be obtained by means of 

 revolving discs. If black sectors with serrated edges are rotated 

 on a white disc (Fig. 179), a series of concentric rings appear, 

 which look darker from the periphery to the centre of the disc. 

 The amount of black in each of these rings is constant, but owing 

 to contrast each ring appears brighter at the inner part, next to 

 a darker ring, and darker at the outer part, next to a lighter 

 riii- Masson's experiment). 



If instead of black sectors on a white ground two different 

 colours are taken, then while the disc is rotating each ring shows 

 dit't'erent colours at its two edges, although the objective colour 

 is uniform throughout each ring. If, e.g., blue and yellow are 

 substituted for black and white, the resulting rings are of different 

 shades of grey, but each ring shows an inner yellow border, as 

 contrasted with the preceding ring, which is bluer, and an outer 

 blue border in contrast to the next and more strongly yellow ring. 



