90 THE BIOLOGY OF INSECTS 



presumably less than that of a vertebrate's because the 

 retinal structure of an insect's eye is distinctly coarser. 

 And it has already been mentioned that the lack of any 

 provision for accommodation results in the insect being 

 very short-sighted. *' We know," remarks Eltringham, 

 " how readily one white butterfly will pursue and investigate 

 another to see if it is a suitable mate, but I have never seen 

 this kind of flirtation begin from a distance of more than a 

 few feet." The compound eyes of insects, therefore, while 

 giving their possessors a wide range of vision as regards 

 direction, are ineffectual in perceiving objects at a distance, 

 and in some cases only can they receive from objects close 

 at hand any definite indications of form. 



A subject of great interest in connection with the sight 

 of insects is their appreciation of colour. Information as to 

 this cannot be obtained by microscopic examination of the 

 eye structure, but approximate certainty has been reached 

 by means of careful observ^ations and experiments. 

 Reference has been made above to the fact that a butterfly 

 often pursues one of its own species ; such action is due to 

 visual recognition, because the attraction is also exercised 

 by a dead dried specimen or even by a coloured model, as 

 Eltringham has demonstrated. He found that a common 

 " fritillary " (Brenthis euphrosyne) dipped directly to a spot 

 on the ground where was lying a wing accidentally broken 

 off from an insect of its own kind long dead. This observa- 

 tion affords convincing evidence that the butterfly could 

 recognise the characteristic colour of the wing, for any 

 specific scent must have been for a long period absent from 

 such a dried fragment. Many investigators have concluded 

 that insects of various kinds distinguish the colours of 

 flowers on account of the apparent preference which they 

 show for certain hues. Eltringham watching Vanessid 

 butterflies on a bed of asters, white, pink, and purple, found 

 that of 427 visits, 47 were to white, 135 to pink, and 245 to 

 purple flowers, though the purple blossoms were only three- 

 quarters as many as the pink. H. Mliller (1878) was long 

 ago led to the opinion that the colours of flowers serve as 



