i.] EFFECT OF INSECTS ON FLOWERS. 13 



paper. No one who saw my bee at that moment could 

 have had the slightest doubt of her power of distinguish- 

 ing blue from orange. 



Again, having accustomed a bee to come to honey on 

 blue paper, I ranged in a row other supplies of honey 

 on glass slips placed over paper of other colours, yellow, 

 orange, red, green, black, and white. Then I continu- 

 ally transposed the coloured paper, leaving the honey on 

 the same spots ; but the bee always flew to the blue 

 paper, wherever it might be. Bees appear fortunately 

 to prefer the same colours as we do. On the contrary, 

 flowers of a livid, yellow, or fleshy colour are most 

 attractive to flies ; and moreover while bees are at- 

 tracted by odours which are also agreeable to us, flies, 

 as might naturally be expected from the habits of 

 their larvae, prefer some which to us seem anything 

 but pleasant. 



Among other obvious evidences that the beauty of 

 flowers is useful in consequence of its attracting insects, 

 we may adduce those cases in which the transference of 

 the pollen is effected in different manners in nearly allied 

 plants, sometimes even in the same genus. 



Thus, as Dr. H. Miiller has pointed out, Malva sylves- 

 . tris (Fig. 9) and Malva rotundifolia (Fig. 10), which grow 

 in the same localities, and therefore must come into com- 

 petition, are nevertheless nearly equally common. 



In Malva sylvestris, however (Fig. 11), where the 

 branches of the stigma are so arranged that the plant 

 cannot fertilize itself, the petals are large and conspi- 

 cuous, so that the plant is visited by numerous insects ; 

 while in Malva rotundifolia the flowers of which are 

 comparatively small and rarely visited by insects, the 



