776 LIGHT AND LIFE 



Heintz (28) placed bees in a box, and using different wavelengths 

 measured the number that crawled over an illuminated slit in a 

 three-minute period as a function of intensity. The curve second 

 from the bottom in Fig. 1 is the reciprocal of the intensity required 

 for a constant effect — an action spectrum. Again the major peak is 

 in the near ultraviolet, regardless of whether a small (solid curve) 

 or a large (broken curve) effect is selected. These curves will not be 

 found in Heintz's paper, as he chose to present his data in another 

 form. 



Sander's (46) action spectra for phototaxis of the bee are some- 

 what anomalous in that they show nothing of a peak in the ultra- 

 violet. There is no ready explanation for this discrepancy. [It has 

 been reported (62) for some beetles that when the test lights are 

 sufficiently bright, the peak in the green is higher than the peak in the 

 ultraviolet. Presumably Sander's three curves differ from each other 

 because they were obtained at different levels of intensity; however, 

 this point is not made very clear.] Somewhat perturbed by Sander's 

 data, I have tested the relative phototactic effectiveness of green (546 

 m^u,) and ultraviolet (365 m/^) lights of equal energies (20). The re- 

 sults of this experiment, involving just two wavelengths, support the 

 conclusions of Bertholf (5) , Daumer (9) , Heintz (28) , and Weiss, 

 Soraci, and McCoy (62) in indicating that for the bee ultraviolet light 

 is more effective than green light in stimulating positive phototaxis. 



Except for the work of Sander, the conclusion is consistent that 

 for all those species studied the near ultraviolet wavelengths are the 

 most effective in exciting phototaxis and similar reactions. The same 

 conclusion was reached by Peterson and Haeussler (42) and Weiss 

 et al. (60, 61, 62, 63, 64) , who studied the responses of insects to 

 broad bands of wavelengths of equal energy. The taxonomic occur- 

 rence of this phenomenon, in so far as it is known, is summarized in 

 Table 2. Unfortunately there are few comparative data. 



With extracellular electrodes it is an easy matter to record relatively 

 large (ca. 10 mv or more) retinal action potentials from the eyes of 

 insects, and recently measurements of spectral sensitivity have been 

 made in the ultraviolet region of the spectrum by using this electrical 

 response. Because the retinula cells lie close to neurons in the optic 

 ganglion with which they synapse, electroretinograms (ERG) do not 

 reflect the activities of receptor cells exclusively. This is a situation 

 to be borne in mind, ))articularly in those instances where more than 

 one kind of color receptor appears to be contributing to the retinal 

 action potential. There are as yet no records of the activity of single 

 units in the compound eyes or optic ganglia of insects. 



