Sensory Discrimination: Vision 157 



experiments and could not confirm his results. Using col- 

 ored lights of measured intensities and studying their effect 

 in causing the bees to collect under them, he compared 

 these effects with the influence of the various colors on the 

 reflex contraction of the pupil in human beings. He found 

 that the smallest differences in intensity to which the 

 bees reacted were those just perceptible to the human eye, 

 and that the relative effect of different colors was like their 

 relative effect on a color-blind human being. He reports 

 similar results with butterflies. Hess thinks this method, 

 which deals with reflexes, superior to any method which, 

 like Von Frisch's, involves learning on the part of the ani- 

 mals. But again we may remind ourselves that it does not 

 follow that because a human being who finds the yellow- 

 green, rather than the yellow, the brightest spectral region, 

 is totally color-blind, therefore an animal, especially an 

 invertebrate animal, the chemical substances .in whose 

 eye may have no resemblance to those in the human eye, 

 is color-blind if it shows these reactions to the differ- 

 ent regions of the spectrum. Hess's method is defective 

 just because it deals with reflexes whose stimuli are inten- 

 sity differences. If an animal is capable of distinguishing 

 both intensity differences and color differences, the use of 

 reflexes that depend on the former is a poor way of study- 

 ing the capacity to discriminate the latter. 



We have already noted the dispute as to how far visual 

 sensations in general are involved in the reactions of bees to 

 flowers, and have seen that Plateau maintains their relative 

 unimportance in this connection, as compared to smell. 

 Besides the experiments which we have quoted on pp. 104 f ., 

 he adduces the facts that he could never persuade in- 

 sects to alight upon artificial flowers, though these were 

 not distinguishable by human eyes from real ones (600- 



