COLOUR SENSE OF INSECTS 363 



necessary, and in some experiments a smaller number was 

 used. His honey bees were first drilled to visit a piece of 

 blue or of yellow paper placed among the grey papers, 

 but with a watch-glass of sugar-water on it to act as an 

 attraction. The drilling lasted two days. After that the 

 bees were offered the series of papers arranged at random, 

 and without any sugar on the coloured slip. To avoid any 

 influence that a difference in smell between the coloured and 

 grey papers might occasion, the whole was covered with a 

 glass sheet. The bees settled only over the coloured paper, 

 and never on any of the greys. This held for blue and for 

 yellow. For red it was found that the bees settled on black 

 or on dark grey as readily as on the colour ; for blue-green it 

 was found that the bees settled with equal readiness on a 

 medium dark grey. It was further found that the bee could 

 not distinguish between yellow, orange-reds, and yellow- 

 greens, or between blues and purples. If bees were drilled 

 to a green or orange-red, they subsequently settled rather 

 on yellow. Drilled to a purple-red paper, they would settle 

 on a blue, but not on a red. 



Von Frisch concludes that the bee can distinguish the 

 colours yellow and blue, but not red or green. When 

 it is presented with yellow-greens or orange-reds it sees in 

 these only the yellow constituent ; when it is presented 

 with purples it sees only the blue constituent. Its colour 

 sense therefore approaches that of a red-green colour-blind 

 person (the commonest kind of colour-blindness), though 

 it is not exactly the same. 



The bee can therefore see yellows, blues, and purples, 

 and the colour of the flower affects its senses. We have 

 here an experimental explanation of the facts that bee 

 flowers are characteristically blue or purple in colour, and 

 that yellow is also frequent. When we say that these 

 colours attract the bee, we do not mean that the bee likes 

 these colours ; we mean that it can distinguish them, and 

 that it comes to associate them with nectar, so that it uses 

 them in the search in nature exactly as it does in the 

 experiment. 



