MOTOR RESPOXSES IN INVERTEBRATES 611 



c. Color Vision. — Lubbock (129) clearly demonstrated that ants are 

 sensitive to ultra-violet. This has by various methods been confirmed 

 for a considerable number of different insects (Hess, 104; Kiihn and 

 Pohl, 122; Lutz and Richtmyer, 136; Lutz, 134, 135; Kuhn, 119). The 

 results obtained by these investigators are, however, qualitative. Bert- 

 holf (16, 17) ascertained quantitatively the distribution of stimulating 

 efficiency in the spectrum for the honey bee and for the fruit fly, Dro- 

 sophila. He presented to the insects two translucent glass plates of the 

 same size, one illuminated with white and the other with monochromatic 

 light produced by means of a quartz prism, and then changed the inten- 

 sity of the white light until its effect on their photopositive response was 

 equal to that of the monochromatic light in different regions of the 

 spectrum. He came to the following conclusions: 



As the wave-length decreases, the stimulating efficiency for the honey 



o 



bee increases from zero at about wave-length 6450 A to a maximum of 

 100 arbitrary units at 5550 A, then decreases to 10 at 4350 A, after which 

 it rapidly increases to a second maximum of 450 at 3650 A and then 

 rapidly decreases to zero at about 2800 A. The distribution of stimulat- 

 ing efficiency for the honey bee is, in the visible spectrum, nearly the 

 same as it is for man, but in the ultra-violet the maximum is 41-2 times as 

 high as the maximum in the visible spectrum, whereas it is zero for man. 

 The distribution of stimulating efficiency for Drosophila is essentially 

 the same as it is for the honey bee, except that the limits of the spectrum 

 and the maximum in the visible are shifted considerably toward the 

 ultra-violet and the maximum in the ultra-violet is more than 5^^, in 

 place of 43^, times as high as the maximum in the visible spectrum. 



Sander (193), in observations on the honey bee obtained results which 

 differ greatly from those obtained by Bertholf . He agrees with Bertholf 

 in the contention that there are two maxima; but he holds that the one is 

 at 5700 A in place of at 5550 A, and the other at 4700 A (blue), in place 

 of at 3650 A (ultra-violet), and that the efficiency at the former is slightly 

 greater than at the latter, in place of only one-fourth as great. He con- 

 sequently does not agree with Bertholf in the contention that the stimu- 

 lating efficiency of ultra-violet is comparatively very high. Bertholf's 

 contention is, however, strongly supported by results obtained by Lutz 

 (134, 135). 



Sander's methods were similar to those used by Bertholf, except that 

 he ascertained the frequency of selection of each of a number of lights 

 which differed in wave-length but were equal in energy; while Bertholf 

 ascertained the intensity of white light that resulted in the same fre- 

 quency of selection as each of a series of monochromatic lights obtained 

 from different regions of the prismatic spectrum, i.e., lights which differed 

 both in wave-length and in energy. Sander maintains that the relation 

 between stimulating efficiency and energy varies with wave-length, and 



