THE COLOR VISION OF BIRDS 3 



like conditions, however, and the results obtained are not com- 

 parable. Katz and Rev6sz ('09) advanced evidence for color 

 fatigue in the fowl after exposure to red which seems to prove 

 some differential action of wave-length. Yerkes ('15) showed 

 the ability of the ring-dove to distinguish spectral red from 

 green but did not control the brightness of the lights. 



Finally the study of the action currents of the retina gives 

 evidence of differential sensitivity to wave-length in day birds, 

 Kohlrausch and Brossa ('14) obtaining characteristic types of 

 galvanometer deflection for different wave-lengths irrespective 

 of intensity. The correlation between action currents and color 

 vision has not been made however. 



All this evidence offers good ground for the presumption of 

 a true color vision in day birds, yet the results are in no case, 

 absolutely conclusive. The question of the nature of the bird's 

 spectrum, the number of primary colors, has not been approached 

 experimentally. Hess describes the color vision of the fowl 

 variously, as like that of a normal man, and as like that of a 

 man looking through a piece of orange glass. He takes the 

 latter view because of the supposed shortening of the spectrum 

 at the violet end. Watson's data upon the threshold of the 

 fowl make this view untenable. Beyond the probability that 

 the bird is capable of some sort of reaction to wave-length, no 

 definite conclusion can be drawn from the studies reviewed here. 

 The details of spectral sensitivity, contrast, after images, color 

 adaptation, and the role of color vision in the normal activities 

 of birds are still open problems, promising to tax objective 

 methods to the utmost. 



The present paper offers further evidence for the existence of 

 color vision in the fowl, in the form of data upon the relative 

 stimulating effect of different wave-lengths upon the light and 

 dark adapted eye, the ability of the fowl to react upon the basis 

 of wave-length, and the appearance of relatively abrupt changes 

 in the stimulating value of different parts of the spectrum. 

 The experiments were begun in 1912, but, owing to delays 

 incident to reorganizing the apparatus and to pressure of other 

 work, they have progressed but slowly. The chief problem and 

 the method of attack I owe to the suggestion of Professor J. 

 B. Watson who has also given generously of his time in the 

 readjustment and control of the apparatus. I am indebted, 



