TIMOTHY H. GOLDSMITH 119 



The "preference" of an insect for one wavelength or color may be 

 different at different times in the life cycle. Use has provided a clear 

 example. Feeding cabbage butterflies (Pieris brassicae) tend to alight 

 on blue or yellow flowers and can be attracted to blue or yellow paper 

 models; green, on the other hand, is not visited (30). Females of the 

 same species when laying eggs, however, seek out green stems and 

 leaves. When paper models are again employed, the insects go to blue- 

 green and green while ignoring red, yellow, blue, and all shades of 

 gray from black to white (31) . 



The sensitivity of insects to ultraviolet light clearly does not depend 

 on indirect excitation of a single blue- or green-sensitive receptor 

 through fluorescence of the ocular media. There are several reasons 

 for rejecting this idea. (1) It is difficult to see how the sequence ab- 

 sorption-fluorescence-absorption could be several times more efficient 

 than direct excitation of the visual pigment of a blue or green re- 

 ceptor; yet if there is neither a special ultraviolet receptor nor a 

 single receptor maximally sensitive in the ultraviolet, this is what is 

 implied by the phototactic response curves reproduced in Fig. 1. (2) 

 The repeated success of a number of investigators in training honey- 

 bees to search for food at ultraviolet-illuminated feeding dishes sug- 

 gests that the wavelengths shorter than 400 m^u, are a distinct color for 

 bees (e.g., 35, 29, 9) . (3) The fact that the wave forms of the electro- 

 retinograms are dependent on the wavelength of the stimulus 

 means that different regions of the spectrum exert qualitatively 

 different effects at the level of the receptor process (56, 57, 20, 

 23). (4) Where electrophysiological techniques reveal the presence of 

 two peaks in the spectral sensitivity function — in the green and the 

 near ultraviolet — it is possible to alter their relative heights by selec- 

 tively adapting the eye with lights of different wave lengths (58, 59, 

 20) (see also Fig. 2) . 



So far this paper has considered the occurrence of color vision 

 among insects and the question of vision in the ultraviolet. In doing 

 so, several different kinds of evidence were introduced. It will be the 

 task of the following two sections to examine several of these experi- 

 ments in more detail and to summarize what is known about the visual 

 systems of two species of insect. 



The Color Vision of the Honeybee, Apis mellijera 



Color vision has been studied more extensively in the honeybee 

 than in any other insect. Von Frisch (16) showed that bees could be 

 trained to collect food on either blue or yellow cards surrounded by 



