EXPERIMENTAL FLORAL BIOLOGY 367 



which has reached the flower to find the nectar ; it does 

 show, however, that the impression which guides the insect 

 to the flower is that of a combination of white circle and 

 violet ground. It also demonstrates the colour sense of the 

 moth for blue. This moth also visits the flowers of Linaria 

 vulgaris, and individuals which frequent this flower are 

 found to react to yellow instead of blue. Like the bee, the 

 moth distinguishes a yellow group of colours and a blue 

 group ; but an individual which frequents a particular 

 flower becomes " drilled " to the colour of that flower 

 (exactly like the bees in the experiment) and always reacts 

 to that colour. The flower-fly {Bombylius fuliginosus) has 

 the same colour sense, 

 distinguishing yellow 

 and blue. In both 

 these insects the sense 

 of smell is of very sub- 

 ordinate importance. 



Special Cases. — A 

 puzzling case may be 

 mentioned here. There 

 are three orchids, 

 Ophrys aptfera, O. imis- 

 cifera, O. aramferaythe 

 flowers of which, to 

 our eyes, bear a close 



resemblance to a bee settled on a rose flower, to a fly, and to 

 a spider respectively. It has been supposed that these 

 resemblances frighten away visiting bees, which perceive that 

 the flower is already occupied, and that the bee is an un- 

 welcome visitor — an explanation rejected by Darwin (1862). 

 Detto (1905) has shown that when one of these flowers is 

 pinned into the flower or inflorescence of another plant 

 habitually visited by bees — e.g. the peony — bees approach, 

 and then, apparently perceiving that the flower is already 

 occupied, swerve off, just as they do if a real insect is 

 already in possession. Detto inclines to believe in the 

 unwelcome guest theory. But there is no evidence that 



Fig. 52. — Knoll's artificial flower; a 

 black dot and white circle on a purple 

 ground ; the small spots are the traces 

 of the tongue of a visiting moth. (After 

 Knoll.) 



