Coloitr. 6 r 



depend on the quality of the illumination. This was illustrated 

 by holding different coloured papers in different parts of the 

 spectrum, and also in the pure yellow light of a sodium flame. 

 All the various tints and shades could be got from the separate 

 spectrum colours by altering either their intensity or their 

 purity, with the exception of purple tints, for which it was 

 necessary to mix the extreme colours of the spectrum. In very 

 bright light colours tend to become yellowish ; m. dim light, 

 such as moonlight, the bluish colours are alone visible. 



The laws of mixture of colours were then explained and 

 illustrated experimentally by making coloured patches on the 

 screen overlap, and by whirling rapidly parti-coloured discs. 

 The results obtained had to be distinguished from those got by 

 mixing pigments, and led to the assumption of red, green and 

 violet as the primary colour sensations. The treatment of the 

 rather complicated facts of colour-mixture was rendered easy 

 by the use of a colour diagram in the form of a triangle, with 

 the primary colours at its angles. Attention was drawn to the 

 unique position occupied by green, and the artistic consequences 

 of this. Specially important from the artistic point of view 

 was the grouping of colours into complimentary pairs, which 

 combine to give white or grey — e.g.^ blue and yellow, green 

 and purple, red and greenish-blue. The phenomena of con- 

 trast depended directly on these groupings. They might be 

 explained as an error of judgment on the part of the eye, the 

 standard of white being affected by the prevailing colour. 

 Thus in a prevailing yellow illumination the standard of white 

 would be displaced towards a yellowish tint, and surfaces which 

 were really grey would look to have the complementary colour 

 blue. This was illustrated by the well-known effects of 

 coloured shadows, seen, for instance, in a room lit by both 

 ordinary and incandescent gas burners. In general, coloured 

 objects seen on an extended coloured background had their 

 true colours mixed with that complementary to the back- 

 ground. A number of instances were shown in which two 

 rings, cut from the same coloured paper, but pasted on back- 



