COLOUR OF FLOWERS 361 



is white or yellow, often of a dirty tinge, or dull red or 

 green ; in class AB yellows and whites predominate ; in 

 class B blues and purples are also found ; in class H, the 

 bee flowers, blue and purple are predominant, though 

 yellows are common. In the butterfly and moth flowers of 

 class F, the colours are varied, though pale tints of pink or 

 purple predominate. The fly flowers of D are frequently 

 lurid red or dirty green. There is a very distinct tendency 

 for the less highly specialised flowers to be white or yellow, 

 for bee flowers to be blue or purple, for pronounced fly 

 flowers to be reddish-green, for moth flowers to be pale. 

 Pure reds are extremely rare. It is generally assumed 

 that the blue tints have appeared at a later stage of evolution 

 than the yellow and red. The very fact that flowers related 

 to different classes of insects have different types of colouring 

 should be a warning not to generalise from the behaviour 

 of one insect to that of others, and still less from our own 

 colour-sense to that of the insect. 



The importance of colour in the entomophilous flower 

 is also emphasised by its absence from anemophilous 

 flowers ; in these pigment, if it is present, is a red anthocyan, 

 as in the larch cones or stigmas of the hazel. It has nothing 

 to do with polHnation, and it is doubtful if it has any signifi- 

 cance at all. It is the colour conspicuous by its absence 

 from the insect-visited flower. In the floras of Oceanic 

 islands bright flowers are scarce, and this is related to the 

 absence of insect life. Wallace, in his Tropical Nature, 

 writes, "... the Galapagos Islands, which . . . with a 

 tolerably luxuriant vegetation in the damp mountain zone 

 yet produce hardly a conspicuously coloured flower ; and 

 this is correlated with, and no doubt dependent on, an 

 extreme poverty of insect Ufe, not one bee and only a single 

 butterfly having been found there." Finally, we may repeat 

 the statement that the great advances in the evolution of 

 insects and of brilHant flowers were contemporaneous. 



Colour Sense of Insect. — Such facts make it certain that 

 the bright colouring of the flower is related to the insect 

 visit, but thev do not decide whether colour is more 



