202 



in South America, at the bottom of a valley densely covered 

 with a blue flowering Papilionaceous plant, I obser'ved some 

 large Pierids {Catopsilia philea) rapidly cross the valley at 

 a height of from 30 to 36 ft. above the ground. Between the 

 blue flowering shrubs there were scattered also a few isolated 

 flowers of a very brilliant red, resembling somewhat our 

 geranium ; upon these the butterflies would occasionally pre- 

 cipitate themselves from their considerable height, remaining 

 for a short time to sip the nectar. Once attracted from above, 

 they would continue visiting the neighbouring blossoms, es- 

 pecially the blue ones, which they seemed to favour, probably 

 because they constituted the food plant of the caterpillar. It 

 seems obvious that the butterflies, notwithstanding their 

 predilection for these blue flowers, had not seen them from the 

 great height at which they were flying, but, being induced by 

 the more conspicuous red flowers to descend to within i or 

 2 metres (3 to 6 ft.), they were well able to recognise them 

 distinctly. Thus it seems that the eyes of insects also perceive 

 the red colour at a much greater distance than the blue. Any 

 one studying the behaviour of flower-loving insects will be easily 

 convinced that the gorgeous colour and fragrance which we 

 often notice in flowers, as well as their honey, principally serve 

 to attract insects for the purpose of fertilisation, and indeed we 

 are acquainted with some species in which the fact that fer- 

 tilisation has taken place and the production of honey has 

 ceased is indicated by an immediate change of colour, the object 

 evidently being to keep the insects from wasting their time. 

 But such a colour-signal would be useless if the insects were 

 unable to perceive and interpret its meaning. 



There still remains another question to be answered, namely, 

 whether the insects not only distinguish between the different 

 colours, but whether they perceive them in the sarrie way as 

 is the case with the eyes of vertebrates. 



As we know that there exist rays, e.g. the ultra-violet ones, 

 which are invisible to the human eye, it may be thought possible 

 that the eyes of insects might in this respect be more perfect 

 than the vertebrate eye. Such a view is supported by a well- 

 known experiment, consisting in aflowing sunhght to pass 

 through a prism upon the bottom of a box containing a number 



