TIMOTHY H. GOLDSMITH 713 



The Extent of the Spectrum Visible to Insects and the 

 Regions of Maximum Sensitivity 



There have been numerous attempts to measure the spectral sensi- 

 tivities of the eyes of insects, using both behavioral (4, 5, 6, 7, 9, 14, 

 28, 46, 49, 50, 60, 61, 62, 63, 64, 65) and electrophysiological (1, 2, 

 11, 19, 20, 23, 33, 56, 57, 58, 59) techniques. These experiments vary 

 widely in sophistication, and in the following paragraphs the terms 

 "action spectrum" and "spectral sensitivity" will be used only where 

 the quantity measured was the relative energy for a constant effect. 

 In addition, the term "spectral sensitivity" will be further restricted to 

 electrophysiological measurements of the eye, as distinct from action 

 spectra for behavioral responses of the whole organism. 



Almost all attempts to measure the relative effectiveness of various 

 wavelengths in stimulating insects, electrophysiological as well as 

 behavioral, indicate a broad maximum in the blue-green or green 

 regions of the spectrum and gradually diminishing effectiveness in the 

 red. The fly Calliphora, however, possesses a second peak of sensi- 

 tivity in the red (2, 58, 59) (see Fig. 5) . What is in many respects 

 the most interesting feature of these observations is that for at least 

 some insects the visible spectrum extends into the near ultraviolet. 



The Question of Sensitivity to Near Ultraviolet Light 

 In 1886 Sir John Lubbock (38) showed that sunlight from which 

 the ultraviolet had been filtered was much less effective than normal 

 sunlight in stimulating ants to remove their pupae to the dark. The 

 relative efficiencies of different wavelengths in eliciting behavioral 

 responses such as phototaxis have been determined by several in- 

 vestigators, and where the data permit quantitative comparison, it 

 seems that the near ultraviolet is the most effective region of the 

 spectrum (Fig. 1). 



The data reproduced in Fig. 1 require further comment. Bertholf 

 (6) reported the "relative stimulative efficiency" of various wave- 

 lengths in stimulating the positive phototactic response of the fruit 

 fly Drosophila. Ihifortunately, his method was somewhat indirect; 

 monochromatic lights were matched in stimulative effect by a w^hite 

 light of variable intensity. Owing to the fact that the test wavelengths 

 varied in energy, the reciprocal of the intensity of white light re- 

 quired for an equal effect — the "relative stimulative effect" — has little 

 meaning. Bertholf attempted to correct for this defect by dividing 

 the observed values of "relative stimulative effect" by the relative 

 energies of the monochromatic test lights, and it is this quantity, the 



