2IO SUPPLEMENT TO THE INTRODUCTION 



near, but would be able to perceive it at a greater distance, and would then 

 turn aside. 



The observation of Delpino, mentioned on p. 141 of this volume, can likewise 

 only be explained by reference to the visual power of insects, and I therefore repeat 

 his description. — On a meadow in Vallombrosa there were numerous plants of 

 Bellis perennis and Anemone nemorosa, equally mixed, and distributed at about 

 equal distances from one another. Delpino saw a bee collecting pollen zealously 

 from Anemone. When flpng from one flower to another it repeatedly made a 

 mistake and went to flowers of Bellis, though on reaching these recognized its error, 

 and at once flew on again. 



Plateau has given a very one-sided interpretation to his experiments, without 

 regard to the earlier observations of other investigators. For instance, he quite over- 

 looks the experiments of Forel, who proved that blinded insects could not recognize 

 the part of a flower on which they wished to settle, while others from which he had 

 removed the antennae — which bear the olfactory organs — flew with certainty from 

 flower to flower. 



Plateau further entirely ignores the results of the observations of Hermann 

 Miiller, which have been fuUy confirmed by the statistical investigations of 

 E. Loew, J. MacLeod, and myself. By these the following conclusions have been 

 established : — 



1. Other things being equal, a flower is visited by insects in proportion to its 

 conspicuousness. Among nearly allied species which agree closely in the form and 

 colour of their flowers, and naturally also agree in their floral mechanisms, those 

 which are most conspicuous receive the most numerous visits, while those which are 

 least conspicuous have the smallest number of visitors '. 



2. In a number of cases odour has more to do with the attraction of insects 

 than the size and colour of the corolla. ' The richly scented flowers of Convolvulus 

 arvensis,' says Hermann Miiller ('Fertilisation,' p. 572), 'are far more abundantly 

 visited than the larger and more conspicuous but scentless flowers of C. sepium ; 

 the sweet-scented violet is much more visited than the larger, brightly coloured, but 

 scentless pansy ; the small, insignificant, but strongly perfumed flowers of Lepidium 

 sativum surpass in the abundance of their visitors the other more conspicuous but 

 scentless Crucifers.' 



3. Dull yellow flowers (Bupleurum, Anethum, Pastinaca, Alchemilla, and 

 others) are not visited as a rule by beedes, while nearly related flowers that are 

 white, or of some other conspicuous colour, attract these insects even when nectar- 

 less (e.g. Helianthemum, Papaver, Genista). Reddish blue or violet flowers are 

 preferred by bees, butterflies, and hover-flies, which are highly specialized visitors, 

 while the insects that appear most frequently on white or yellow flowers have a short 

 proboscis, and are unskilled visitors. Bees with a long proboscis (humble-bees) 

 appear to be least dependent upon the colour of flowers. As Hermann Miiller 

 expresses himself ('Alpenblumen,' p. 496), ' these — the most intelligent of flower 



' H. Miiller (' Fertilisation," p. 570) gives as examples species of Ranunculus, Geranium, Malva, 

 Polygonum, Stellaria, Cerastium, Epilobium, Rosa, Rubus, Veronica, Carduus, Hieracium, and the 

 various flower-forms of Euphrasia officinalis, Rhinanthus Crista-galli, and Lysimachia vulgaris. 



