THE COLOR VISION OF BIRDS 1? 



of the lights and experiment box the monochromatic lights were 

 arranged so that both could be projected upon the same stim- 

 ulus patch. They were exposed in irregular order with the 

 other patch illuminated with the intense white light. Under 

 these conditions Chick E chose — 



Red 6 White 6 when r and w were exposed together. 



Green 1 White 5 when g and w were exposed together. 



Chick D chose — 



Red 2 White 4 



Green 5 White under the same conditions. 



The white light was immediately reduced to threshold intensity 



and the tests repeated. 



Chick E chose — 



Red 8 White 1 



Green 3 White 8 



Chick D chose — 



Red 2 White 5 



Green 5 White 



The change in the intensity of the white light with the resultant 

 change in the dilution of the monochromatic points to a sen- 

 sitivity to wave-length. The complete change in the technique 

 eliminates a reaction to secondary criteria. 



The conclusions from the third set of tests are based upon 

 the time of reaction to a single monochromatic light: the time 

 required for the chick to go from the door of the starting box 

 to the food dish, a distance of 3 feet. The results of the tests 

 with the three chicks are given in table 5. During these tests 

 the intensity of the lights was varied from 1/18 to the full energy 

 of the spectrum. The birds, although instinctively photo- 

 positive, showed a marked delay in advancing to the negative 

 light, irrespective of its intensity. Unless the brightness differ- 

 ences of the monochromatic lights for the chicks are extremely 

 great this experiment is sufficient to eliminate reaction to the 

 absolute intensity of the light. 



These tests are given in some detail as illustrating a fairly 

 satisfactory technique when the energy and brightness values 

 of the stimulus lights are unknown. Such methods have the 

 disadvantage of introducing many disturbing elements; in such 

 tests the animals ultimately become confused and give con- 

 flicting results. 



