ORIGIN OF THE COLOUR-SENSE. 247 



destroy the seeds of the larger fruits and nuts which 

 they devour, and which are not usually coloured ; but the 

 irritating effect of bright colours on some of them does 

 not support this view. It seems most probable there- 

 fore that man's perception of colour in the time of 

 Homer was little if any inferior to what it is now, but 

 that, owing to a variety of causes, no precise nomencla- 

 ture of colours had become established. One of these 

 causes probably was, that the colours of the objects of 

 most importance, and those which were most frequently 

 referred to in songs and poems, were uncertain and 

 subject to variation. Blood was light or dark red, or 

 when dry, blackish ; iron was grey or dark or rusty ; 

 bronze was shining or dull ; foliage was of all shades of 

 yellow, green, or brown ; and horses or cattle had no 

 one distinctive colour. Other objects, as the sea, the 

 sky, and wine, changed in tint according to the light, 

 the time of day, and the mode of viewing them ; and 

 thus colour, indicated at first by reference to certain 

 coloured objects, had no fixity. Things which had more 

 definite and purer colours as certain species of flowers, 

 birds, and insects were probably too insignificant or 

 too much despised to serve as colour-terms ; and even 

 these often vary, either in the same or in allied species, 

 in a manner which would render their use unsuitable. 

 Colour-names, being abstractions, must always have 

 been a late development in language, and their com- 

 parative unimportance in an early state of society and 

 of the arts would still further retard their appearance ; 

 and this seems quite in accordance with the various 

 facts set forth by Mr. Gladstone and the other writers 

 referred to. The fact that colour-blindness is so pre- 



